Contribution to M2M architectures for Smart Grid

Document technical information

Format pdf
Size 6.9 MB
First found May 22, 2018

Document content analysis

Category Also themed
Language
English
Type
not defined
Concepts
no text concepts found

Persons

Henri Matisse
Henri Matisse

wikipedia, lookup

Organizations

Places

Transcript

UNIVERSIDAD CARLOS III DE MADRID
TESIS DOCTORAL
Contribution to Machine-to-Machine
Architectures for Smart Grids
Autor:
Gregorio Ignacio López López
Director:
Dr. José Ignacio Moreno Novella
DEPARTAMENTO DE INGENIERÍA TELEMÁTICA
Leganés, Enero 2014
- ii -
TESIS DOCTORAL
Contribution to Machine-to-Machine
Architectures for Smart Grids
Autor:
Gregorio Ignacio López López
Director:
Dr. José Ignacio Moreno Novella
Firma del Tribunal Calificador:
Firma
Presidente:
Jürgen Jähnert
Vocal:
David Fernández Cambronero
Secretario:
Francisco Valera Pintor
Calificación:
Leganés,
- iii -
de
de 2014
- iv -
A María y a Emma
A mis padres y hermanos
-v-
- vi -
Acknowledgments
En primer lugar, me gustaría dar las gracias al principal responsable de que esté
escribiendo estas líneas, al que considero como un padre a nivel profesional, José Ignacio
Moreno. Gracias por darme la oportunidad de trabajar contigo. Gracias por todo lo que me has
enseñado, tanto a nivel profesional como personal, a lo largo de estos años. Gracias por tu
dedicación y esfuerzo para que esta tesis sea lo mejor posible tanto en contenido como en forma
y por tu apoyo incondicional, ánimo y comprensión en los momentos difíciles que he afrontado
a lo largo de ella. Gracias por ser, además de mi director de tesis, un buen amigo.
En segundo lugar, quiero darle las gracias a los buenos amigos con los que he tenido el
placer de trabajar en nuestro grupo de investigación, de los que tantas cosas buenas he
aprendido: a Juan Pablo, que me acompañó durante la primera etapa, a Paco, con el que tuve la
gran suerte de volverme a encontrar durante la última, y especialmente a Víctor, que me ha
acompañado a lo largo de todo el proceso. No podría haber elegido mejor compañía para este
viaje. Gran compañero, mejor amigo. Parte de esta tesis es también tuya.
Eu também quero agradecer ao Aníbal de Almeida por me oferecer a oportunidade de
investigar no ISR-UC e por ser tão bom anfitrião. Mas quero agradecer especialmente ao Pedro
Moura, o meu outro companheiro nesta viagem, com quem tenho compartilhado tão bons
momentos. Eu nunca poderia agradecer suficientemente o que tu me ensinaste sobre energia e
sobre a vida, nem todo o teu apoio incondicional. Estiveste a ajudar-me até ao último momento.
Parte de esta tese também é tua, amigo.
Também quero ter algumas palavras de agradecimento para o Cláudio Lima. Conhecemonos por casualidade num ónibus em Sophia-Antipolis em abril de 2011. Desde então não deixou
de me ajudar em tudo o que precisei e sempre teve palavras de apoio para me animar. Obrigado,
Cláudio.
También me gustaría darle las gracias a Cayetano Lluch, por ayudarme en todo lo que ha
podido y por involucrarme a través del COIT en actividades de estandarización de AENOR
relacionadas con Smart Grids.
I would like to thank my colleagues from the ENERsip project. I learned so much working
side by side you guys and we had such a good time. Gracias. Obrigado. Dêkuji. Toda. Dank.
Gracias a todos los compañeros y amigos de la universidad y del Departamento de
Ingeniería Telemática que me habéis ayudado a lo largo de estos años y me habéis animado en
esta última etapa. Spasibo. Sas efcharistó. Grazie. Multumesc. Choukran.
Gracias a mis amigos del UC3Marathon y del dual bike, con los que intento poner en
práctica eso del mens sana in corpore sano en un ambiente inmejorable.
- vii -
Gracias a todos mis amigos, en general, que afortunadamente tengo muchos, por los
buenos momentos que me hacéis pasar y por hacer que cuando nos vemos parezca que no haya
pasado el tiempo ni nos hayamos movido de lugar. Obrigado. Danke. Grazie. Thank you.
Gracias a mis suegros, Tomasi y Antonio, por la gran ayuda y todo el cariño que me dais
día a día, a mis cuñados, Miryam, Dani, Nuria y Miki, por vuestro cariño y apoyo, y a mis
sobrinos, Ainara y Álvaro, por regalarme siempre una sonrisa.
Gracias a mis hermanos, Álvaro y Claudio, con los que he pasado los mejores 31 años de
mi vida, por estar siempre a mi lado, no importa la distancia que nos separe, y por hacer que
para nosotros no pase el tiempo y nos lo sigamos pasando como niños.
Gracias a mis padres, Antonia y Gregorio, por su fe ciega en mí, por su amor, por su
compresión y por estar siempre dispuestos a ayudarme. No hay mejor premio que saber que
estáis orgullosos de mí.
Gracias a mi pareja, a mi amor, a mi mejor amiga, María. Sin ti no habría sido capaz de
llegar hasta aquí. Gracias por ser como eres. Gracias por haberme hecho pasar los mejores años
de mi vida. Pero, sobre todo, gracias por haberme hecho la persona más feliz del mundo, gracias
por haberme regalado la vida. Emma, hija mía, esta tesis, como todo lo bueno que haga en esta
vida, te lo dedico a ti, para que nunca dejes de sonreir.
- viii -
“Knowing a great deal is not the same as being smart; intelligence is not information
alone but also judgment, the manner in which information is coordinated and used.”
Dr. Carl Sagan, Cosmos
- ix -
-x-
Abstract
The electrical grid is a huge and complex system which represents a critical infrastructure.
Due to this fact, the electric power industry has traditionally adopted a conservative attitude
regarding changes. As a result of that, the electrical grid has experienced very few
breakthroughs for decades and currently is not prepared to face novel challenges, such as
properly integrating DERs (Distributed Energy Resources) or proactively controlling the energy
demand by means of the so-called DR (Demand Response) programs, which mainly derive from
nowadays society concerns on global warming and climate change.
Upgrading traditional electrical grid to the so-called Smart Grid represents one of the most
complex engineering projects ever and will certainly drive the next wave of research and
innovation in both the energy and the ICT (Information and Communications Technology)
sectors. The road towards the Smart Grid will mean an unprecedented revolution especially at
the power distribution and customer domains, since the unpredictable and uncontrollable nature
of renewables will impose the coordination of generation and consumption points in almost real
time.
M2M (Machine-to-Machine) communications allow networked devices to communicate
between them without further human intervention. What in the very beginning seemed to be a
tailored solution for telemetry applications, has become a communications paradigm itself,
addressing the myriad of applications existing and yet to be in the wide context of the Internet
of Things. As a matter of fact, M2M communications represent one of the main pillars of the
Smart Grid in that they will enable the bidirectional real-time exchange of information between
the consumption and generation facilities to be monitored and controlled, and the information
systems where the optimization processes run.
There is a plethora of communications technologies and protocols available within the
scope of M2M communications for the Smart Grid. Hence, research is needed in two directions.
On the one side, it is required to evaluate how different communications architectures and
technologies meet the specific requirements of the Smart Grid before undertaking the important
investments needed to deploy this kind of infrastructures on a large scale. On the other side, it is
crucial to develop common data models which serve as reference to future horizontal or wide
scope protocols which expand across different domains or areas.
This thesis aims to tackle these issues. The main goal of the thesis is to contribute to the
area of M2M communications architectures tailored to the power distribution and customer
domains of the Smart Grid. In order to achieve this overall objective, first we carry out a survey
on the most relevant standardization activities developed in parallel to this thesis and on the
most outstanding technological and research trends within the Smart Grid area, identifying gaps
and challenges.
- xi -
Second, we propose a novel M2M communications architecture to support energy
efficiency and optimum coordination of DER (Distributed Energy Resources) within the socalled energy-positive neighborhoods, which are neighborhoods which ensure a substantial part
of their consumption by local generation based on renewables. The proposed architecture
comprises three network segments, for the sake of flexibility and scalability, and combines
different communications technologies to meet the specific communications requirements of
each of them.
Next, we formally model the domain of knowledge of energy efficiency platforms for
energy-positive neighborhoods by means of an ontology developed in OWL (Ontology Web
Language), with the aim that it becomes a reference data model for the application of M2M
communications to this context. Thus, this ontology has been made public through the EC
(European Commission) eeBuildings Data Model community, so that other researches can reuse it and extend it.
We also propose a methodology that can be applied, in general, to characterize any
communications overlay deployed on top of an infrastructure devoted to any purpose. Following
this methodology, we model the traffic of the proposed M2M communications architecture in
realistic large-scale scenarios. The main goal of this model is to ensure that potential works
based on it actually mean and bring value to the interested parties. Although the model is
tailored to the Portuguese power distribution grid, since it is based on actual data provided by
EDP (Energias de Portugal), it can be easily adapted to other scenarios by suitably tuning the
appropriate parameters.
Taking this model as reference, we finally evaluate the core of the proposed M2M
communications architecture twofold. On the one side, we analyze the impact of using IPSec
(Internet Protocol Security) or TLS/SSL (Transport Layer Security/Secure Socket Layer) as
VPN (Virtual Private Network) technologies on the operational costs of a potential energy
efficiency platform which relies on the proposed M2M communications architecture. To the
author’s best knowledge, no similar studies are available in the state of the art. The main
conclusion of this analysis is that using TLS/SSL along with data aggregation is the best option
to minimize operational costs at neighborhood level.
On the other side, we evaluate by means of simulations the performance of IEEE 802.11b,
using as metric the goodput (i.e., throughput at the application layer), and GPRS (General
Packet Radio Service), using as metric the transmission time. The first conclusion of these
simulations is along the line that IEEE 802.11b meets the requirements in terms of goodput of
the NAN (Neighborhood Area Networks), which is of special interest to the Smart Grid
community taking into account the low cost and wide adoption of this technology. The second
conclusion of such simulations is that GPRS meets the requirements in terms of bandwidth of
the backhaul network, thus confirming that it represents a very attractive technology considering
that it is the most mature and widely deployed cellular technology in Europe.
- xii -
Keywords
Demand Response; Distributed Generation; Electric Vehicle; Energy-positive
neighborhoods; Home Area Networks; Home Energy Management Systems; Information and
Communications Technologies for Energy Efficiency; Internet of Things; Machine-to-Machine
communications; Neighborhood Area Networks; Modeling; Simulation; Smart Grid
- xiii -
- xiv -
Contents
Chapter 1 - Introduction ................................................................................................ 1
1.1 A new and novel electrical grid paradigm: The Smart Grid....................................... 1
1.1.1 Smart Grid drivers ............................................................................................... 3
1.1.2 Smart Grid evolution ........................................................................................... 8
1.2 Thesis motivation ..................................................................................................... 10
1.3 Thesis objectives ...................................................................................................... 12
1.4 Thesis organization ................................................................................................... 12
Chapter 2 – State of the art ..........................................................................................15
2.1 Introduction ...............................................................................................................15
2.2 Overview of standardization activities ......................................................................16
2.2.1 NIST Smart Grid Interoperability Panel .............................................................16
2.2.2 IEEE 2030 ..........................................................................................................19
2.2.3 European Standardization Organizations ...........................................................21
2.3 Overview of related technologies ..............................................................................24
2.3.1 Smart meters .......................................................................................................24
2.3.2. Smart appliances ................................................................................................25
2.3.3 Monitoring and control systems .........................................................................26
2.3.4 Communications architectures, technologies, and protocols .............................30
2.4 Main trends on standardization and research ............................................................36
2.5 Overview of related projects .....................................................................................37
2.7 Conclusions ...............................................................................................................41
Chapter 3 – Network Architecture ............................................................................. 45
3.1 Introduction .............................................................................................................. 45
3.2 System description.................................................................................................... 46
3.2.1 Building domain ................................................................................................ 47
3.2.2 User domain ....................................................................................................... 49
3.2.3 Information System domain .............................................................................. 50
3.2.4 Neighborhood domain ....................................................................................... 50
3.3 Relation to standardization activities........................................................................ 51
3.4 Communications technologies .................................................................................. 53
3.4.1 Home Area Network.......................................................................................... 54
3.4.2 Neighborhood Area Network ............................................................................ 54
3.4.3 Backhaul ............................................................................................................ 55
3.5 Security ..................................................................................................................... 55
- xv -
3.5.1 Home Area Network.......................................................................................... 55
3.5.2 Neighborhood Area Network ............................................................................ 57
3.5.3 Backhaul ............................................................................................................ 58
3.6 Address management and end-to-end addressability ............................................... 58
3.6.1 ID distribution and management ....................................................................... 60
3.6.2 Routing .............................................................................................................. 63
3.7 Conclusions .............................................................................................................. 64
Chapter 4 – Formal modeling.......................................................................................65
4.1 Introduction ...............................................................................................................65
4.2 Ontology description .................................................................................................66
4.2.1 Classes ................................................................................................................67
4.2.2 Properties ............................................................................................................73
4.3 Use Case ....................................................................................................................75
4.4 Conclusions ...............................................................................................................78
Chapter 5 – Practical Modeling ...................................................................................81
5.1 Introduction ...............................................................................................................81
5.2 Methodology..............................................................................................................82
5.3 Communications infrastructure modeling .................................................................83
5.3.1 Context and scope...............................................................................................83
5.3.2 Characterized scenarios ......................................................................................84
5.4 Conclusions ...............................................................................................................88
Chapter 6 - Evaluation ..................................................................................................91
6.1 Introduction ...............................................................................................................91
6.2 General considerations and assumptions ...................................................................92
6.3 Evaluation of end-to-end security protocols ..............................................................93
6.3.1 Technical comparison .........................................................................................95
6.3.2 Economic comparison ........................................................................................96
6.4 Evaluation of communications infrastructure performance ....................................100
6.4.1 Simulation setup ...............................................................................................100
6.4.2 Performance evaluation ....................................................................................104
6.5 Conclusions .............................................................................................................105
Chapter 7 - Conclusions ..............................................................................................107
7.1 Introduction .............................................................................................................107
7.2 Conclusions .............................................................................................................107
7.3 Thesis impact ...........................................................................................................109
7.4 Future work .............................................................................................................112
Table of Acronyms ..................................................................................................... 113
Table of Symbols......................................................................................................... 117
References.................................................................................................................... 119
- xvi -
List of Figures
Figure 1-1 – Main subsystems of the traditional electrical grid .................................................... 2
Figure 1-2 – Role of communications in the Smart Grid [IEA2011] ............................................ 3
Figure 1-3 – Typical photovoltaic generation and demand profiles in a residential building
[Moura2012] ................................................................................................................................. 4
Figure 1-4 – Electricity consumption trends in the EU Industry, Transport, and Buildings sectors
until 2010 [Eurostat2013].............................................................................................................. 5
Figure 1-5 – Load profile of an average EU household [De Almeida2011] ................................. 6
Figure 1-6 – Deployment of EVs and PHEVs [IEA2011] ............................................................ 7
Figure 1-7 – The evolution of the Smart Grid [Farhangi2010] ..................................................... 8
Figure 1-8 – Smart Grid technology areas [IEA2011] .................................................................. 9
Figure 1-9 – Smart Grid generations [Carvallo2011] ................................................................. 10
Figure 2-1 - SGIP organizational structure ................................................................................. 16
Figure 2-2 – NIST Smart Grid conceptual model [NIST2012] ................................................... 17
Figure 2-3 – NIST conceptual reference diagram for Smart Grid communications networks
[NIST2012] ................................................................................................................................. 18
Figure 2-4 – Evolution and scope of IEEE 2030 standardization process [IEEE2011] .............. 19
Figure 2-5 – IEEE 2030 Smart Grid communications architecture [IEEE2011] ........................ 20
Figure 2-6 –CEN/CENELEC/ETSI Smart Grid conceptual model ............................................ 21
Figure 2-7 – CEN/CENELEC/ETSI SGAM [SGCG2012a] ....................................................... 22
Figure 2-8 – Main domains of the ETSI M2M reference architecture ........................................ 23
Figure 2-9 – Mapping of ETSI M2M main domains onto the Smart Grid main layers
[ETSI2012].................................................................................................................................. 23
Figure 2-10 –Application of ETSI M2M architecture to Smart Grid sub-systems [Lu2012] ..... 24
Figure 2-11 –Example of whole-house power monitoring system [OWL2014] ......................... 26
Figure 2-12 – Example of power outlet monitor [Efergy2014] .................................................. 27
Figure 2-13 – Example of remote controlled sockets [Efergy2014] ........................................... 27
Figure 2-14 – Example of residential monitoring and control system for electricity consumption
[Cloogy2014] .............................................................................................................................. 28
Figure 2-15 – Example of monitoring system for residential micro-generation [Enlighten2014]
..................................................................................................................................................... 28
Figure 2-16 – Example of residential monitoring and control system for energy consumption
and generation [SHM2014] ......................................................................................................... 29
Figure 2-17 – Hierarchical and heterogeneous M2M communications architecture for the power
distribution domain of the Smart Grid [Fadlullah2011] .............................................................. 31
Figure 2-18 – OSI layer placement of most of the analysed standards [Moura2013b], [De
Craemer2010] .............................................................................................................................. 36
Figure 2-19 – Timeline of the most relevant milestones of this thesis along with the most
relevant standardization and research milestones ....................................................................... 43
Figure 3-1 – Overall system architecture .................................................................................... 46
Figure 3-2 – Prototype network for the monitoring and control of I-BECI [Carreiro2011] ....... 49
- xvii -
Figure 3-3 – Prototype network for the monitoring and control of I-BEGIs [López2013a] ....... 49
Figure 3-4 – Mapping of the scope of the target platform onto the NIST Smart Grid conceptual
model ........................................................................................................................................... 51
Figure 3-5 – Mapping of the proposed M2M communications architecture onto the IEEE 2030
SGIRM ........................................................................................................................................ 52
Figure 3-6 – Mapping of the proposed M2M communications architecture onto the ETSI M2M
domains ....................................................................................................................................... 52
Figure 3-7 – Mapping of the proposed M2M communications architecture onto the ETSI M2M
architecture applied to the Smart Grid [Lu2012] ........................................................................ 53
Figure 3-8 – Zigbee PRO Standard Security Mode [Gislason2008] ........................................... 56
Figure 3-9 – Zigbee PRO High Security Mode [Gislason2008] ................................................. 57
Figure 3-10 – E2E addressability in the proposed addressing solution....................................... 60
Figure 3-11 – M2M Platforms ID distribution and management ................................................ 60
Figure 3-12 – CNTRs ID distribution and management ............................................................. 61
Figure 3-13 – ADR EPs ID distribution and management .......................................................... 62
Figure 3-14 – Device ID distribution and management .............................................................. 63
Figure 3-15 – Forwarding of data in the uplink .......................................................................... 63
Figure 3-16 – Routing of commands in the downlink................................................................. 64
Figure 4-1 – Taxonomy of an energy efficiency platform for energy-positive neighborhoods .. 68
Figure 4-2 – Overall picture of the use case................................................................................ 76
Figure 4-3 – Instances of the I-BECI and I-BEGI in the use case .............................................. 77
Figure 4-4 – Use case interaction at different levels of abstraction ............................................ 78
Figure 5-1 – Typical electricity distribution infrastructure ......................................................... 83
Figure 5-2 – Mapping of the designed M2M communications architecture onto the power
distribution infrastructure ............................................................................................................ 84
Figure 6-1 – Details of the model considered in chapter 6 ......................................................... 93
Figure 6-2 – (a) NxM direct secure tunnels from the ADR EPs to the M2M GW; (b) M secure
tunnels from the CNTRs to the M2M GW; (c) NxM secure tunnels from the ADR EPs to the
CNTRs + M secure tunnels from the CNTRs to the M2M GW .................................................. 94
Figure 6-3 – Protocol stack at CNTR and M2M GW for: (a) IPSec &Aggr. (b) TLS/SSL &
Aggr. (c) IPSec & FF. (d) TLS/SSL & FF .................................................................................. 96
Figure 6-4 – Annual savings per district between implementing Fast Forwarding or
Aggregation at CNTRs for each security protocol in each scenario ........................................... 98
Figure 6-5 –Annual savings per district between implementing IPSec and TLS/SSL at CNTRs
in each scenario depending on whether Fast Forwarding or Aggregation is used ..................... 99
Figure 6-6 – Network used to simulate the NAN...................................................................... 100
Figure 6-7 – Internal structure of the module MobileHost ........................................................ 101
Figure 6-8 – Network used to simulate the Backhaul ............................................................... 102
Figure 6-9 – Internal structure of the module WirelessHostSimplified ..................................... 103
Figure 6-10 – NAN results for each scenario for a simulation time of 3600 s.......................... 105
- xviii -
List of Tables
Table 2-1 - Domains and Actors in the NIST Smart Grid conceptual model ............................. 17
Table 2-2 - Communications networks defined in the IEEE 2030 CT-IAP of special interest to
this thesis ..................................................................................................................................... 20
Table 2-3 - Summary of some relevant commercially available HEMSs ................................... 30
Table 2-4 - Summary of some of the communications technologies likely to be deployed in
HAN and NAN in different countries [Lo2012] ......................................................................... 32
Table 2-5 - Summary of communications technologies and their scope..................................... 32
Table 2-6 - Summary of PLC communications technologies ..................................................... 34
Table 2-7 - Summary of higher layers standards and protocols .................................................. 35
Table 2-8 - Summary and comparison of considered R&D projects .......................................... 40
Table 4-1 - Summary of the properties defined in our ontology ................................................. 73
Table 4-2 - Mapping of the Services onto the Stakeholders they are addressed to ..................... 75
Table 5-1 - Main parameters for Urban and Rural scenarios ...................................................... 85
Table 5-2 - Main parameters for Short-term and Long-term scenarios....................................... 87
Table 5-3 - Main parameters in Low usage, Medium usage and High usage scenarios ............. 87
Table 5-4 - Scenarios considered in the model presented in this chapter ................................... 88
Table 6-1 - Summary of the parameters relevant to the scenarios considered in chapter 6 ........ 93
Table 6-2 - Summary of IPSec and TLS/SSL technical comparison .......................................... 95
Table 6-3 - Overhead introduced by IPSec and TLS/SSL [Alshamsi2005] ................................ 96
Table 6-4 - Summary of the results of the analysis of the impact on the operational costs of
using IPSec or TLS/SSL ............................................................................................................. 97
Table 6-5 - Difference in terms of cost (in €) per CNTR and per district during one year between
using Fast Forwarding and using Aggregation in each scenario ................................................ 98
Table 6-6 - Difference in terms of cost (in €) per CNTR and per district during one year between
using IPSec and TLS/SSL in each scenario ................................................................................ 99
Table 6-7 - Summary of Linksys WRT160NL datasheet.......................................................... 101
Table 6-8 - Summary of minimum transmission powers required for our target application ... 102
Table 6-9 - Summary of the most important parameters for each communications technology
................................................................................................................................................... 104
Table 6-10 - Summary of NAN results (Goodput in bps) ......................................................... 104
Table 6-11 - Summary of Backhaul network results (Transmission Time in s) ....................... 105
Table 7-1 - R&D projects related to this thesis ......................................................................... 111
- xix -
- xx -
Chapter 1
Introduction
1.1 A new and novel electrical grid paradigm:
The Smart Grid
Nowadays we live in such a globalized World driven by such a digital society where
everything gets outdated so fast that the Darwinian motto could be rephrased as “update or die”.
The electrical grid, however, has remained unchanged, without major architectural
improvements, for decades. Is then the electrical grid an exception to the aforementioned motto?
The answer is definitely no. The reason why the electrical grid is more resilient to changes has
to do with the fact that it is a huge and complex system which represents a key critical
infrastructure, so stability and security come first to everything else. However, for some time
now, the electrical grid is undergoing slowly but surely its inexorable metamorphosis under the
new paradigm of the so-called Smart Grid.
The basic topology of the traditional electrical grid is strictly hierarchical, with clear
demarcations between its generation, transmission, and distribution subsystems, as shown in
Figure 1-1. The traditional electrical grid is also unidirectional in nature, with an energy flow
from the power plant to the customer without any real-time share of information between the
consumption and generation points [Farhangi2010]. As a result, it presents quite a few
inefficiencies and problems, e.g., the generation capacity is over dimensioned to meet peak
demands, which makes the whole system inefficient, or the lack of real-time monitoring and
control of critical processes and assets causes blackouts and failures, which is unacceptable for
Chapter 1 – Introduction
such a critical infrastructure and implies high costs to operators (e.g., to the so-called utilities,
which operate at power distribution level). In addition, the traditional electrical grid is not
prepared to meet the new requirements and features that nowadays society demands from it.
Industrial
Customer
Primary
Substation
Generation
Transmission
Secondary
Substation
Commercial
Customer
HV (High Voltage) lines (> 70 kV)
MV (Medium Voltage) lines (2kV – 70 kV)
Residential
Customer
LV (Low Voltage) lines (220V – 1kV)
Distribution
Customer
Figure 1-1 – Main subsystems of the traditional electrical grid
During the last decades, social concern about global warming and climate change has
dramatically increased. As a token of that, governments worldwide are making great efforts to
reduce GHG (Greenhouse Gas) emissions through the reduction of energy consumption and the
use of renewable energy. The “Kyoto Protocol”1 agreed and developed by the UNFCCC (United
Nations Framework Convention on Climate Change) [UNKP1997] or the 20-20-20 target of the
EU (European Union) climate and energy package2 [EUCR2008] are two representative
examples of the commitments made by developed countries to fight against climate change.
However, the defined scenarios for decarbonizing the energy sector will lead to important
impacts on the electrical grid, namely due to the high penetration of intermittent renewable
technologies, with more reliance on DG (Distributed Generation), and to the proactive control
of the demand of electricity [EPIA2012].
The increasing share of variable renewable energy resources (mainly solar photovoltaics
and wind power) will lead to major challenges to the electricity system’s stability, security and
reliability [EURELECTRIC2011]. Moreover, such new generation resources are deployed
mostly as DG, creating a more decentralized system, where the figure of the “prosumer”3
emerges [Toffler1980]. As a result, the scalability of the grid is improved and the energy flows,
and so the losses, are reduced, but the complexity of managing such infrastructures increases
dramatically.
The increasing electricity consumption in most countries must be also properly addressed
by the electrical grid, which must have new tools to optimize it and to promote energy
efficiency both in the supply and in the demand sides, such as the monitoring and control
infrastructures based on SANs (Sensor and Actuator Networks) or the DR (Demand Response)
1
The “Kyoto Protocol” is an international treaty wherefore the industrialized countries commit to reduce
their GHG emissions.
2
The 20-20-20 target of the EU climate and energy package sets that the following goals must be
achieved in the EU by 2020: 20% reduction of GHG emissions compared to 1990 levels; raising the share
of EU energy consumption produced from renewable resources to 20%; 20% of improvement related to
energy efficiency.
3
In the electricity market, the term “prosumer” refers to such a stakeholder that both consumes electricity
and produces it.
-2-
Chapter 1 – Introduction
programs. The electricity consumption will increase even more dramatically with the foreseen
wide adoption of EVs (Electric Vehicles) [Faria2012]. However, EVs can be also an important
tool to the grid management with the possible control of their charging cycles and the use of
their batteries as local energy storage equipment. Thus, the monitoring and control of electricity
consumption together with the use of energy storage resources will provide the needed
flexibility to ensure the generation/consumption matching.
Therefore, the Smart Grids is expected to address the major shortcomings of the existing
electrical grid and to exploit the new resources. Thus, the Smart Grid will no longer be
unidirectional, since energy will flow in both directions between the grid and the customers’
facilities, and the assumption that the demand for electricity dictates the amount of electricity
produced will no longer hold. Although the Smart Grid will still partly rely on large scale
generation, the presence of energy storage and renewable energy generation facilities (the so
called DERs – Distributed Energy Resources) at the different grid levels will increase
dramatically. Furthermore, the Smart Grid will provide the operators of electrical infrastructures
(e.g., utilities, aggregators) with enhanced sensory and control capabilities to ensure the full
visibility and pervasive control over their assets and services [EPRI2011], [IEA2011].
Hence, the Smart Grid will facilitate the integration of diverse supply-side resources,
support the integration of distributed and on-site generation on the customer side, promote more
active engagement of demand-side resources and participation of customers’ loads, and allow
the widespread permeation of dynamic pricing to beyond-the-meter applications
[Sioshansi2012].
As Figure 1-2 illustrates, ICT (Information and Communications Technologies) are crucial
to upgrade current electrical grid to the Smart Grid, since they will be responsible for providing
both the required real-time bidirectional communications among the huge volume of devices
involved and the required real-time massive data handling tools.
Figure 1-2 – Role of communications in the Smart Grid [IEA2011]
1.1.1 Smart Grid drivers
The increasing penetration of DG, the growing demand of electricity, the large-scale
deployment of EVs, and the availability of new energy storage technologies, are the main
drivers of the Smart Grid. They do not only introduce new challenges which cannot be solved
by the traditional grid, but they do also offer new resources which can be part of the solution.
-3-
Chapter 1 – Introduction
1.1.1.1 Distributed generation
The already mentioned efforts to reduce GHG emissions related to electricity generation
have been leading to a fast increase in the deployment of renewable generation, in particular
photovoltaic and wind power. Such renewable energy sources are being deployed not only as
bulk generation facilities but also as distributed local generation facilities. These distributed
local generation facilities can be connected directly to the power distribution network, as in the
so-called VPPs (Virtual Power Plants) [Hernández2013], or to private consumption
infrastructures [Claudy2011], enabling self-consumption and leading to the so-called energypositive neighborhoods4.
DG has the potential to provide site-specific reliability improvement, as well as
transmission and distribution benefits, although it also presents some drawbacks and introduces
new challenges.
Electric power systems have been designed for more than a century ago in a top-down
perspective, based on highly predictable bulk generation power plants following demand
variations. However, the renewable generation resources have characteristics that differ from
conventional energy sources. The output of wind and solar power is determined by random
meteorological processes, outside the control of the generators or the system operators.
Therefore, unlike conventional capacity, wind and solar generated electricity cannot be reliably
dispatched or perfectly forecasted and exhibits significant temporal variability [Moura2010a]. In
addition, as Figure 1-3 illustrates, renewable energy generation does not typically match the
consumption profile of residential buildings and households [Moura2012].
Figure 1-3 – Typical photovoltaic generation and demand profiles in a residential building [Moura2012]
In the EU, the variable generation could represent up to 65% by 2050 according to the EC
(European Commission) Energy Roadmap [EC2011a]. As penetration rates of variable
generation increase over levels of 15% to 20%, and depending on the electricity system in
question, it can become increasingly difficult to ensure the reliable and stable management of
electricity systems relying solely on conventional grid architectures with limited flexibility
[IEA2011]. Therefore, new monitoring and control tools are crucial to increase the system
flexibility and maintain stability and balance at the power distribution grid level.
4
It should be noted that energy-positive neighborhood is considered to be a neighborhood which ensures
a substantial part of its consumption by local generation and not necessarily a neighborhood with more
generation than consumption.
-4-
Chapter 1 – Introduction
1.1.1.2 Electricity demand
Over the last few decades, worldwide energy demand has increased due to industrial
development and global economic growth, being the electricity the fastest growing component
of the overall energy demand. In the EU, it is expected that electricity will double its share in
the final energy demand from the current levels up to 36-39% by 2050, mainly due to its
increasing use in transports and buildings [EC2011a].
Buildings, in particular, are the largest electricity consuming sector in the World and
account for over one-third of total final energy consumption and an equally important source of
GHG emissions [IEA2013]. As Figure 1-4 shows, in the EU in particular, during the last few
years the electricity consumption in the industry and transport sectors has kept bounded,
whereas the electricity consumption in the building sector has been steadily increasing,
becoming a major problem for governments, utilities, customers, and the environment.
TWh
2000
1800
1600
1400
1200
1000
800
600
400
200
0
2001
2002
2003
2004
Industry
2005
2006
Transport
2007
2008
2009
2010
Buildings
Figure 1-4 – Electricity consumption trends in the EU Industry, Transport, and Buildings sectors until
2010 [Eurostat2013]
The main reasons for these different trends are twofold. On the one hand, electricity
consumption has been traditionally a well-identified problem in the industrial sector, since it
translates into higher costs. Therefore, huge investments have been devoted to develop EMSs
(Energy Management Systems) that reduce dramatically such consumption and, in turn, such
costs. On the other hand, electricity consumption in households is not individually very
significant; its true impact arises when it is summed up over millions of homes. In addition, the
widespread utilization of new types of loads and the requirement of higher levels of comfort and
services have also driven such an increase in the electricity consumption in the residential sector
[Firth2008].
Indeed, the electricity consumption breakdown in the EU households was recently
characterized [De Almeida2011], showing clearly the increasing importance of electronic loads,
which represent more than 21% of the overall consumption. Such loads are mostly
entertainment and ICT appliances with standby consumptions that represent about 7% of the
total annual electricity consumption per household. HVAC (Heating Ventilation and Air
Conditioning) loads also show high consumption and an increasing penetration rate. Given such
-5-
Chapter 1 – Introduction
increasing consumptions and the difficulty on identifying the major contributors, in-house
monitoring and control systems, with sub-metering capabilities, are needed to make information
about unwanted consumptions available to end-users and to give them tools to efficiently
control these loads, thus enabling energy efficiency [Moura2013b].
The baseline consumption in an average EU household is fairly high (near 200 W) mostly
due to the cold appliances and HVAC loads, as Figure 1-5 shows. However, if properly
controlled, such loads can be used as a DR resource. Washing and drying appliances consume
more than 16% of the electricity with high consumption at peak hours, while they might be
shifted to other periods. Thus, the washing and drying appliances might be rescheduled to
periods of lower energy consumption or higher energy production. The cold appliances, HVAC
and water heating loads might be also interrupted during short periods of time, without major
reduction of service quality, to avoid the most unbalanced situations between generation and
consumption.
Figure 1-5 – Load profile of an average EU household [De Almeida2011]
In the past, the electrical system was planned and operated under the assumption that the
supply system must meet customers’ demand of electricity. However, that supposition starts to
change in the 1980s with the emergence of a new approach to control a resource traditionally
uncontrollable: the load [Gellings2009]. Through the proper application of DSM (Demand-Side
Management), and providing incentives to the consumers, it is possible to control the
consumption so that it matches shortages in the conventional generation capacity or the
uncontrollable dips and peaks of renewable generation.
Thus, using DR programs it is possible to directly or indirectly force a consumption
reduction in critical situations in a short period of time, as long as the required communications
infrastructure is in place, as it is the case in the Smart Grid. Traditionally DR technologies were
typically used to attend upon economic concerns related to balance supply and demand,
involving industrial customers which present very high electricity consumption. However,
nowadays they can be used to improve the system reliability, reducing or increasing
instantaneously the electricity consumption to avoid the problems that result from the
intermittence of renewable generation [Moura2010b], and exploiting the great potential of the
residential sector [Moura2013a].
-6-
Chapter 1 – Introduction
1.1.1.3 Electric vehicle and energy storage
The electrification of transport will be responsible for a large increase in the electricity
consumption. Europe has a target of 10% share of EV by 2020, which need to be charged by
means of the electrical grid [EURELECTRIC2011]. As Figure 1-6 shows, the IEA
(International Energy Agency) estimates that the transport sector will make up to 10% of overall
electricity consumption by 2050 due to a significant increase in EV and PHEV (Plug-in Hybrid
Electric Vehicles) [IEA2011].
LDV: Light Duty Vehicle
Figure 1-6 – Deployment of EVs and PHEVs [IEA2011]
The impact of EV on the distribution network load diagram will require a new approach in
load control, since if the vehicle charging is not managed intelligently, it could increase peak
load and require major infrastructure investments to ensure the system reliability [Mets2012].
However, the EVs, due to their storage capacity, have a great potential as controllable
loads, drawing power and storing energy when not in use (e.g., when they are parked at home or
at work, what represents the main part of the day/night) [Pang2012]. Based on these singular
characteristics, EV can support the integration of unpredictable intermittent renewable sources
and contribute to the system stability, namely by provisioning ancillary services [Goebel2013].
The Smart Grid will allow the smart charging of EV during periods of low demand and/or
high generation, thanks to its real-time communications capabilities [López2013b].
Furthermore, over the long term, when V2G (Vehicle-to-Grid) technologies are implemented, it
could also allow EVs to feed the electricity stored in their batteries back into the system when
needed [Clement2011], [Ma2012]. Additionally, other energy storage technologies, such as
batteries and supercapacitors are already available and present increasing performance and
decreasing costs.
The need of new energy storage technologies is also due to the alternative planar structure
of the Smart Grid. The traditional grid already has energy flexibility (e.g., hydropower dams),
but it is in the same place and works in the same manner as the generation, i.e., centralized from
top to bottom, which limits the storage capacity. However, the new storage equipment does not
need to be located near to the power plants and can be installed in any point of the grid. That
choice enables supporting the integration of intermittent energy and the mitigation of
congestion.
-7-
Chapter 1 – Introduction
Therefore, the new energy storage technologies associated with the Smart Grid can ensure
the matching between generation and consumption at different grid levels, not only at largescale but also at neighborhood or building levels.
1.1.2 Smart Grid evolution
Upgrading current electricity grid to the so-called Smart Grid represents one of the major
engineering challenges ever. As a result, the road towards the Smart Grid will be long and needs
to be paved gradually, certainly driving the next wave of research and innovation in both the
energy and the ICT sectors.
The metering side of the distribution system has been the focus of the Smart Grid
investments so far, with the initial introduction of the AMR (Automated Meter Reading)
systems to read meter data (Figure 1-7). However, AMR systems do not allow the transition to
the Smart Grid, due to the absence of control capabilities. In a second stage, the AMI
(Advanced Metering Infrastructures) have been used to provide a two-way communication
system to the meter as well as the ability to ensure load management and revenue protection.
The next step is the leverage of AMI to implement distributed command and control strategies
with pervasive control and intelligence across all geographies, components, and functions of the
system [Farhangi2010].
Figure 1-7 – The evolution of the Smart Grid [Farhangi2010]
The evolution of the Smart Grid will then be ensured by the introduction of new layers to
enable more advanced functionalities. The basic layer is the already described monitoring and
automated infrastructure. The second layer is the connection between participants, integrating
customers and energy service providers into the grid. Then, a layer to sense and response is
needed to share information, analysing and acting upon it to balance all the resources in realtime. Finally, the layer to analyse and optimize will manage the network using rules, constrains
and intelligent agents [IBM2011]. Such layers will enable the development of the different
Smart Grid technological areas shown in Figure 1-8, which spread across all the electrical grid
domains [IEA2011].
-8-
Chapter 1 – Introduction
Figure 1-8 – Smart Grid technology areas [IEA2011]
Therefore, the evolution towards the Smart Grid can be divided into the three generations
shown in Figure 1-9 [Carvallo2011]. The Smart Grid 1.0 (today’s first steps) will be focused on
increasing the visibility and awareness of the status of the power distribution network. Thus,
throughout this first generation, the existing services will be transformed using advanced
communication to offer pre-paid metering, in-home displays, intelligent disconnect, fine-grain
load control, advanced outage management, bi-directional metering, and DR.
The Smart Grid 2.0 (grid resident intelligence) will enable future services and foster
innovation with applications such as micro-grids and DG, intelligent street lighting, V2G,
storage/distribution of renewable generation, fault prediction/outage prevention, energy asset
management, and automatic DR.
Finally, the Smart Grid 3.0 (grid leveraged applications) will make the grid completely
manageable, extending the applications to take full advantage of the grid resident intelligence.
Thus, the Smart Grid of the future will become not just a way to deliver electricity more
efficiently, but also an entirely new social and transactional platform with new business models,
applications, services, and relationships.
-9-
Chapter 1 – Introduction
Figure 1-9 – Smart Grid generations [Carvallo2011]
1.2 Thesis motivation
The Smart Grid especially represents a revolution at the distribution and customer domains
(notably, at commercial and residential level). The electric power industry has traditionally
devoted more attention and resources to power transmission networks and primary power
distribution networks (i.e., power networks responsible for transforming high voltage levels to
medium voltage levels) rather than to secondary power distribution networks (i.e., power
networks responsible for transforming medium voltage levels to the low voltage levels required
by commercial and residential customers), since the former ones allowed keeping generation
and consumption balanced under the traditional approach.
Therefore, bulk generation plants, power transmission systems and – to some extent primary power distribution systems, have been traditionally monitored using legacy
communications networks which allow a certain level of centralized coordination; whereas
secondary power distribution systems have been traditionally passive systems with no or limited
communications capabilities. However, as it has been shown throughout this chapter, the impact
of the main drivers of the Smart Grid is especially relevant at distribution and commercial and
residential customer level (which are connected through secondary power distribution
networks), so it is right there where the major breakthroughs are required.
ICT will play a key role on making the Smart Grid a reality at distribution and customer
level. On the one side, M2M (Machine-to-Machine) communications will enable the required
real-time bidirectional communications between the huge number of devices to be monitored
and controlled and the information systems where the optimization processes run. On the other
side, advances on data mining, data analytics, the emerging Big Data, and cloud computing, will
allow managing, processing and making decisions based on such a huge amount of information
[Motamedi2012], [Kezunovic2011], [Miller2013], [Sakr2011].
There is a plethora of communications technologies available for being used in such M2M
communications infrastructures for the Smart Grid [Güngör2011], which in practice slows the
wide deployment of the Smart Grid down, since this situation introduces uncertainty on the
- 10 -
Chapter 1 – Introduction
market and so hampers the required investments. Wireless communications technologies are of
special interest to the distribution and customer domains of the Smart Grid [Aravinthan2011].
As a token of that, the NIST (National Institute of Standards and Technologies) has launched a
specific Working Group within the PAP2 (Priority Action Plan 2) to tackle the challenges and
opportunities of wireless communications in this emerging application domain [NIST2011].
Moreover, communications for the Smart Grid present specific requirements from both the
technical and economic perspectives, such as [Güngör2013], [Yan2013], [Liu2012]:

QoS (Quality of Service). The communications infrastructure must provide a given level of
QoS that fits the target application. Notably, QoS policies are mainly oriented to traffic
prioritization and resource allocation to face congestion situations. Some parameters which
are widely used to quantify such QoS level are:
o Latency. It can be defined as the E2E (End-to-End) delay of the data.
o Bandwidth. The communications infrastructure must provide an aggregated data rate as
high as to carry the traffic associated to the target application. In general, this will
depend on the volume of devices as well as on the size of the exchanged packets and the
traffic pattern.
o Reliability. The communications infrastructure must guarantee that it will work
correctly during a given percentage of time throughout a year. The more critical the
application is, the higher such a percentage will be.

Interoperability. The communications infrastructure must allow equipment from different
manufacturers to interact seamlessly. In order to achieve this goal, the main functional
blocks which compose the communications infrastructure as well as the interfaces among
them must be defined and standardized. Standardization is crucial to effectively achieve this
goal, which eventually fosters competition and thus yields more reliable products at lower
cost.

Scalability. The communications infrastructure must ensure scalability from both the
technical and economic perspectives. On the one side, taking into account the huge number
of devices this kind of systems involves, the selected communications technologies must
minimize the deployment, maintenance and operational costs. On the other side, the
communications architecture must be able to incorporate new devices and to accommodate
new services.

Security and privacy. Due to the fact that Smart Grid applications handle sensitive data,
security (both physical and cyber-security) and privacy represent key factors for their wide
deployment and adoption. If privacy is not guaranteed, many users will not embrace many
of the new services. If security is not guaranteed, many service providers will not
implement or rely on many of such new services. However, since these two features and
cost are usually directly proportional, a trade-off is required in order to obtain feasible
solutions.
As a result, it is crucial to evaluate how different communications architectures and
technologies meet such requirements before undertaking the important investments needed to
deploy this kind of infrastructures on a large scale. Simulations represent a powerful, costeffective and flexible solution to achieve this goal, although the relevance of their results tightly
depends on how accurately the model behind such simulations fits real World scenarios.
Therefore, a proper characterization of the communications requirements of the target
application is of capital importance in order to obtain meaningful results [López2012a],
[ETSI2012], [Khan2013].
- 11 -
Chapter 1 – Introduction
At higher levels of the communications stack, there is also a myriad of protocols for every
specific application within the Smart Grid, which hampers interoperability both within a given
domain or area (e.g., at distribution domain, at customer domain) and across them. In this
regard, it is crucial to develop common data models which serve as reference to future
horizontal or wider-scope protocols.
1.3 Thesis objectives
This thesis addresses some of the hottest topics - from the communications perspective introduced by the already presented Smart Grid drivers at distribution and commercial and
residential customer level, representing a remarkable contribution to the development of the
aforementioned Smart Grid 1.0 and 2.0. Notably, the overall objective of this thesis is to
contribute to the area of M2M communications architectures tailored to the power distribution
and customer domains of the Smart Grid. Next a breakdown of the specific objectives of this
thesis goes:

Analyze the most relevant standardization activities and research trends within this area,
identifying gaps and challenges.

Design a novel M2M communications architecture for energy-positive neighborhoods that
promotes energy efficiency and consumption and generation matching at neighborhood
level.

Formally model the domain of knowledge of the energy efficiency platforms for energypositive neighborhoods to foster re-using and extending our work, thus increasing its
potential impact.

Model the traffic carried by the proposed M2M communications architecture in realistic
large-scale scenarios in order to maximize the impact of potential simulations based on
them.

Evaluate the performance of the proposed M2M communications architecture by means of
simulations and draws conclusions that can be valid as guidelines for potential deployments.

Assess the impact of using different security protocols on the operational costs of a potential
energy efficiency platform which relies on the proposed M2M communications architecture.
1.4 Thesis organization
The remainder of this dissertation is organized as follows. Chapter 2 provides an overview
of standardization activities, research activities and potential gaps within the Smart Grid area.
Chapter 3 describes the M2M communications architecture designed to support energy
efficiency and proper integration of local and distributed micro-generation within energypositive neighborhoods, including the definition of the required functional blocks and the
interfaces among them, as well as proposing the communication technologies to be used and
an application-layer solution for address management and end-to-end addressability. Chapter 4
presents the ontology that formally defines the vocabulary and taxonomy and captures the
engineering and business semantics of the domain of knowledge of the energy efficiency
platforms for energy-positive neighborhoods, highlighting how future work can make the
most out of it. Chapter 5 maps the proposed M2M communications architecture onto the power
distribution network and characterizes the communications requirements and features of our
specific target application. Chapter 6 evaluates, from different perspectives, the core of the
proposed M2M communications architecture taking Chapter 5 as baseline. Notably, Chapter 6
- 12 -
Chapter 1 – Introduction
assesses the operational costs of using different security solutions to establish VPN (Virtual
Private Networks) and the performance of the selected communications technologies based on
different metrics. Finally, Chapter 7 concludes this dissertation.
The research conducted during this thesis and presented throughout this dissertation has
been gradually disseminated and validated in the research community and appropriate
standardization organizations, leading to the following publications:
1. G. López, V. Custodio, F. J. Herrera, J. I. Moreno, “Machine-to-Machine Communications
Infrastructure for Smart Electric Vehicle Charging in Private Parking Lots”, International
Journal of Communication System, Wiley, November 2013.
2. P. Moura, G. López, J. I. Moreno, A. de Almeida, “The role of Smart Grids to foster energy
efficiency”, Energy Efficiency, Volume 6, Issue 4, Pages 621-639, November 2013.
3. G. López, J. I. Moreno, “PRICE Project: M2M communications architecture for large-scale
AMI deployment”, 4th ETSI M2M Workshop, Mandelieu-la-Napoule, France, November
2013.
4. E. El achab, G. López. J. I. Moreno, “Evaluación de mecanismos de seguridad en entornos
de Smart Grid”, JITEL 2013: XI Jornadas de Ingeniería Telemática, Granada, Spain,
October 2013.
5. P. Moura, G. López, J. I. Moreno, A. de Almeida, “Impact of Residential Demand Response
on the Integration of Intermittent Renewable Generation into the Smart Grid”,
EEDAL2013: 7th International Conference on Energy Efficiency in Domestic Appliances
and Lighting, Coimbra, Portugal, September 2013.
6. G. López, J. Moreno, P. Moura, A. de Almeida, M. Perez, L. Blanco, “Monitoring System
for the Local Distributed Generation Infrastructures of the Smart Grid”, CIRED 2013: 22nd
European Conference and Exhibition on Electricity Distribution, Stockholm, Sweden, June
2013.
7. G. López, P. Moura, B. Kantsepolsky, M. Sikora, J. I. Moreno, A. de Almeida, “European
FP7 Project ENERsip: Bringing ICT and Energy Together”, Global Communications
Newsletter, November 2012.
8. P. Moura, G. López, A. Carreiro, J. I. Moreno, A. de Almeida, “Evaluation Methodologies
and Regulatory Issues in Smart Grid Projects with Local Generation-Consumption
Matching”, EEMSW2012: International Workshop on Energy Efficiency for a More
Sustainable World, São Miguel, Portugal, September 2012.
9. G. López, P. Moura, V. Custodio, J. I. Moreno, “Modeling the Neighborhood Area
Networks of the Smart Grid”, IEEE ICC 2012, Ottawa, Canada, June 2012.
10. A. Carreiro, G. López, P. Moura, J. I. Moreno, A. de Almeida, J. Malaquias, “In-house
monitoring and control network for the Smart Grid of the future”, ISGT Europe 2011: 2nd
IEEE PES International Conference and Exhibition on Innovative Smart Grid Technologies,
Manchester, UK, December 2011.
11. G. López, P. Moura, M. Sikora, J. I. Moreno, “Comprehensive validation of an ICT
platform to support energy efficiency in future smart grid scenarios”, IEEE SMFG 2011:
IEEE International Conference on Smart Measurements for Future Grids, Bologna, Italy,
November 2011.
- 13 -
Chapter 1 – Introduction
12. G. López, P. Moura, J. I. Moreno, A. de Almeida, “ENERsip: M2M-based platform to
enable energy efficiency within energy-positive neighborhoods”, IEEE INFOCOM 2011
Workshop on M2M Communications and Networking, Shanghai, China, April 2011.
13. G. López, J. I. Moreno, “Smart Energy-positive Neighbourhoods for the Smart Grid.
Architecture, Communications Technologies and Address Management for the ENERsip
platform”, 2011 ETSI Workshop on “Standards: An Architecture for the Smart Grid”.
Sophia Antipolis, France, April 2011.
- 14 -
Chapter 2
State of the art
2.1 Introduction
Upgrading current electricity infrastructure to the so-called Smart Grid is one of the most
complex engineering projects ever. As a result, there are many challenges to face and problems
to overcome, which will certainly drive the next wave of research and innovation in both
the energy and the ICT (Information and Communications Technology) sectors.
This chapter provides an overview of the most outstanding standardization and R&D
(Research and Development) activities within the distribution and customer domains of the
Smart Grid, identifying the gaps and challenges tackled in this thesis. The Smart Grid involves a
wide range of technologies and a myriad of standards. Therefore, this chapter does not aim to be
exhaustive, but to focus on the most relevant work related to this thesis.
The remainder of the chapter is organized as follows. Section 2.2 outlines the main
standardization activities related to this thesis developed by the NIST (National Institute of
Standards and Technology), the IEEE (Institute of Electrical and Electronics Engineering), and
the ESOs (European Standardization Organizations). Section 2.3 presents the most relevant
Smart Grid technologies related to this thesis, paying special attention to communications
architectures, technologies, and protocols. Section 2.4 summarizes the main trends in
standardization and research within the Smart Grid area from the ICT perspective. Section 2.5
summarizes some relevant European and national R&D projects which tackle the issues
Chapter 2 – State of the art
previously presented throughout the chapter. Finally, section 2.6 draws conclusions and
highlights the gaps and challenges addressed in this thesis.
2.2 Overview of standardization activities
2.2.1 NIST Smart Grid Interoperability Panel
The NIST SGIP (Smart Grid Interoperability Panel) is a private/public partnership funded
by different industry stakeholders in cooperation with the US (United States) Federal
Government that is aimed at the development of a framework for coordinating all Smart Grid
stakeholders in an effort to accelerate standards harmonization and advance in the
interoperability of Smart Grid devices and systems.
The SGIP was established by the NIST in late 2009 as part of a broader plan aimed at the
coordination of a standards development process for the Smart Grid, in fulfillment of the
responsibilities it had been assigned by the EISA (Energy Independence and Security Act of
2007) US law. In 2012, the SGIP was composed of over 780 member organizations representing
22 stakeholder categories, including international organizations, federal agencies, as well as
state and local regulators. In January 2013, the SGIP entered a new phase becoming a selfsustaining entity with the majority of funding coming from industry stakeholders, despite NIST
still maintains an active role.
The SGIP organizational structure is illustrated in Figure 2-1. The Technical Committees
deals with transversal issues and establish guidelines. More specific tasks are carried out by
temporary Working Groups belonging either to the DEWGs (Domain Expert Working Groups)
or to the PAPs (Priority Action Plans) categories. There are two Technical Committees and
a Working Group performing activities in a permanent basis, namely the SGAC (Smart
Grid Architecture Committee), the SGTCC (Smart Grid Testing and Certification
Committee), and the CSWG (Cyber Security Working Group).
Figure 2-1 - SGIP organizational structure
As a first step towards the harmonization of Smart Grid standards to fully support
interoperability, the NIST SGIP developed the Smart Grid conceptual model shown in Figure 2-
- 16 -
Chapter 2 – State of the art
2. The first version of this conceptual model was published in January 2010 [NIST2010a] and it
was reviewed and updated in February 2012 [NIST2012].
Figure 2-2 – NIST Smart Grid conceptual model [NIST2012]
As Figure 2-2 graphically shows, the NIST Smart Grid conceptual model defines seven
domains as well as the electrical and communications flows among them. It can be seen that
electrical flows involve the traditional subsystems of the electrical grid; whereas
communications flows almost create a mesh topology between every domain, which illustrates
the importance of communications in the Smart Grid.
Each domain - and its sub-domains - encompasses Smart Grid actors and applications.
Actors include devices, systems, programs, and stakeholders that make decisions and exchange
information. Applications are tasks performed by one or more actors within a domain (e.g.,
home automation). All these pieces can be orchestrated to obtain useful use cases (i.e., select a
given domain, a given application, its specific requirements, the actors involve in this
application, and describe how they interact). Table 2-1 summarizes the main actors involved in
each domain.
Table 2-1 – Domains and Actors in the NIST Smart Grid conceptual model
Domain
Bulk Generation
Transmission
Distribution
Customers
Operations
Markets
Service Providers
Actor
Generators of electricity in bulk quantities
Carriers of bulk electricity over long distances (the so-called TSOs –
Transmission System Operators)
Distributors of electricity to and from customers (the so-called DSOrs Distribution System Operators)
End users of electricity. They may also generate, store, and manage the use of
energy. Traditionally, three customer types are considered: home,
commercial/building, and industrial
Managers of the movement of electricity
Operators and participants in the electricity market
Organizations providing services to electrical customers and utilities (e.g.,
aggregators, retailers, ESCOs – Energy Services Companies)
- 17 -
Chapter 2 – State of the art
Figure 2-3 zooms in every domain and illustrates the main actors involved in each domain
as well as the relationships between them from the communications perspective.
Figure 2-3 – NIST conceptual reference diagram for Smart Grid communications networks [NIST2012]
Although other standardization bodies have also defined their own Smart Grid conceptual
models, the NIST Smart Grid conceptual model is the most widely accepted and so it is taken as
reference in the remainder of this thesis.
Beside the Smart Grid conceptual model, one of the main outcomes of the SGIP activity is
the elaboration of a compendium of standards, practices, and guidelines that allow the
development and deployment of a robust and interoperable Smart Grid. As a result, in May
2011 the SGIP Governing Board established the so-called CoS (Catalog of Standards) and the
first six standards to be included were approved by the SGIP Plenary in July 2011. This CoS is
available on-line through the NIST Smart Grid Collaboration wiki [NIST2014]. As of today, the
CoS comprises 20 individual standards and 5 series of standards that in turn contain 36
additional standards, which accounts for a total of 56 standards. The CoS list available in
[NIST2014] includes for each standard:

A brief description.

Mapping onto functional areas.

Date in which the standard was included in the CoS.

Reviews from the SGAC (Smart Grid Architecture Committee), the SGTCC (Smart Grid
Testing and Certification Committee), and the CSWG (Cyber Security Working Group).

Mapping onto the Smart Grid conceptual model domains.
- 18 -
Chapter 2 – State of the art
2.2.2 IEEE 2030
Once a conceptual model of the Smart Grid is defined, a reference architecture which
works such a conceptual model out by defining functional blocks and interfaces, thus bringing it
closer to implementation and so to developers, is required.
The IEEE project 2030 was pioneer on developing such reference architecture, leading to
the so-called SGIRM (Smart Grid Interoperability Reference Model). The SGIRM extends the
NIST Smart Grid conceptual model and defines three IAPs (Interoperability Architectural
Perspectives), which represent the main areas of expertise involved in the Smart Grid
[IEEE2011]:

Power Systems (PS-IAP);

Communications Technologies (CT-IAP);

Information Technology (IT-IAP).
Each IAP defines the main functional blocks required in each domain of the NIST Smart
Grid conceptual model from the appropriate perspective, as well as the interfaces between
functional blocks (intra-domain interfaces), and the interfaces between domains (inter-domain
interfaces). The defined IAPs are further particularized for the most important applications in
the Smart Grid area, such as AMI (Advanced Metering Infrastructure) or PEV (Plug-in Electric
Vehicle). Figure 2-4 illustrates the IEEE 2030 standardization process and overall reference
architecture.
Figure 2-4 – Evolution and scope of IEEE 2030 standardization process1 [IEEE2011]
1
It should be noted that dates refer to when the work was developed, but do not represent when the work
was published.
- 19 -
Chapter 2 – State of the art
The most relevant IAPs to this thesis are the IT-IAP and the CT-IAP. The IT-IAP deals –
among other topics – with data modeling. The CT-IAP defines the communications networks
that may be used in every domain. Figure 2-5 illustrates the defined communications networks
or segments as well as the interfaces among them. Table 2-2 describes briefly the
communications networks defined within distribution and customer domains which are of
special interest to this thesis.
Figure 2-5 – IEEE 2030 Smart Grid communications architecture [IEEE2011]
Table 2-2 – Communications networks defined in the IEEE 2030 CT-IAP of special interest to this thesis
Communications network
xAN/ESIs
NAN
Backhaul
Description
xAN represents HAN (Home Area Network), BAN (Building
Area Network), and IAN (Industrial Area Network), which
encompass all the IEDs (Intelligent Electronic Devices) that
allow monitoring and controlling energy status and patterns
within each context. ESIs (Energy Services Interfaces) represent
logical gateways
NAN (Neighborhood Area Network) is a last mile
communications network that connects ESIs and smart meters as
well as DER s(Distributed Energy Resources) and microgrids to
the utility control and operation center through the backhaul
network
Backhaul network provides connectivity between the utility
control and operation center and any communications network
within the distribution and customer domains
Based on this reference architecture, the IEEE has sorted their own catalogue of standards
and has identified the functional blocks and interfaces where they can be applied, as well as the
standardization gaps where new standards are required [IEEE2014].
- 20 -
Chapter 2 – State of the art
2.2.3 European Standardization Organizations
The three main ESOs (European Standardization organizations) – namely CEN (European
Committee for Standardization), CENELEC (European Committee for Electrotechnical
Standardization), and ETSI (European Telecommunications Standards Institute) - are working
together on the Smart Grid standardization process. This collaborative work is being driven by
the following EC mandates:

Standardization mandate M/441 to develop an open architecture for utility meters involving
communications protocols enabling interoperability (March 2009) [EC2009].

Standardization mandate M/468 concerning the charging of EVs (June 2010) [EC2010].

Standardization mandate M/490 to support European Smart Grid deployment (March 2011)
[EC2011b].
In response to such mandates, three working groups, involving the participation of the three
ESOS, have been created:

SM-CG (Smart Metering - Coordination Group) in response to M/441.

Focus Group on European Electro-Mobility in response to M/468.

SG-CG (Smart Grid – Coordination Group) in response to M/490.
Through the EC standardization mandate M/490 and the corresponding SG-CG, the
CEN/CENELEC/ETSI strategic partnership has further developed the NIST Smart Grid
conceptual model, adapting it to the specific requirements of the European electricity grid. As a
result, a new domain related to DERs has been added, reflecting the importance and high
penetration of renewables generation in European power distribution networks [SGCG2012a].
As Figure 2-6 shows, the CEN/CENELEC/ETSI Smart Grid conceptual model considers that
DERs are electrically connected to the power distribution network and communicate with it, as
well as with the Markets, Operations and Service Provider domains. Nevertheless, it should be
noticed that the Customer domain encompasses both DG and energy storage at small scale (see
Figure 2-3).
Markets
Operations
Transmission
Bulk Generation
Service Provider
Distribution
Customer
Pan-European Energy
Exchange System
Domains
Electrical flows
DER
(e.g., wind, solar,
cogeneration)
Secure comms flows
Black: Original NIST Smart Grid model
Blue: Added to NIST Smart Grid model
Application Area of
Microgrid Architecture
Figure 2-6 –CEN/CENELEC/ETSI Smart Grid conceptual model
- 21 -
Chapter 2 – State of the art
The SG-CG has also developed the so-called SGAM (Smart Grid Architecture Model)
tailored to the requirements of the European electricity grid. As a result, the three-dimensional
architectural model comprising the domains, zones, and layers shown in Figure 2-7, has been
defined [SGCG2012a]. The SGAM allows a technologically neutral representation of all the
interoperability cases of the Smart Grid. The five defined layers represent – top to bottom - the
business objectives and processes, the functions, information exchange and data models,
communications technologies and protocols, and the physical and logical components. As a
token of the volume and importance of the Communication Layer, it is developed in a separate
document [SGCG2012b]. [SGCG2012b] defines the communications networks and their
deployment at the Component Layer and maps the identified communications technologies and
protocols onto the defined communications networks. This thesis is specially focused on the
Communications Layer, addressing also the Information Layer.
Figure 2-7 – CEN/CENELEC/ETSI SGAM [SGCG2012a]
In addition, another outstanding outcome of the SG-CG work is the elaboration and
classification of a first set of standards for the Smart Grid and the identification of
standardization gaps where standards are required [SGCG2012c].
It is also worthwhile to remark upon the work developed by ETSI on the standardization of
M2M (Machine-to-Machine) communications. This work has been recently transferred to the
partnership project OneM2M [OneM2M2014]. As a result of this work, a complete reference
architecture has been defined, including functional blocks and interfaces. Figure 2-8 shows the
main domains of the ETSI M2M reference architecture. The M2M Device Domain encompasses
the so-called capillary networks (in ETSI terminology), i.e., the SANs (Sensor and Actuator
Networks). The Network Domain represents the core of the M2M infrastructure and provides
bidirectional bulk data exchange over long distances. Finally, the Application Domain
encompasses the services which are delivered on top of the M2M infrastructure.
- 22 -
Chapter 2 – State of the art
Figure 2-8 – Main domains of the ETSI M2M reference architecture
Figure 2-9 provides a graphical overview of how these three domains can be mapped onto
the Smart Grid [ETSI2012].
Application Domain
3
API
Service
Aggregation
Deaggregation
API
…
API
M2M Network Domain
2
(including access network)
M2M Nwk capability
Control & Connectivity
M2M Device Domain
1
M2M Device/GW capability
Devices
(e.g., smart meters)
Energy
M2M GWs
Figure 2-9 – Mapping of ETSI M2M main domains onto the Smart Grid main layers [ETSI2012]
Reference [Lu2012] elaborates on how the ETSI M2M communications architecture can be
applied to the Smart Grid, as it is shown in Figure 2-10.
- 23 -
Chapter 2 – State of the art
Figure 2-10 –Application of ETSI M2M architecture to Smart Grid sub-systems [Lu2012]
2.3 Overview of related technologies
2.3.1 Smart meters
There is no single definition of smart metering. However, all smart-meter systems
comprise an electronic metering box and a communications link. At its most basic form, a smart
meter measures electronically how much energy is used and communicate this information to
another device which, in turn, allows the customers to view how much energy they are using
and how much it is costing to them [ESMA2012].
Smart-meter systems can be classified according to their communications capabilities. The
combination of the electronic meters with two-way communications technology is also
commonly referred to as AMI. Previous systems, which were limited to one-way
communications to collect meter data, are referred to as AMR systems. AMI has developed over
time, from its roots as a meter reading substitute to today’s two-way communications systems,
where smart meters are not just sensors any longer, but they become part of the core of the
power distribution network [EEI2011].
In general, AMI bring benefits both to the customers and to the operators or utilities. On
the one side, they provide customers with enriched information on their energy usage to aid
them in controlling the cost and the environmental impact. On the other side, they provide
utilities with very detail data which allow them to perform sophisticated tasks such as load
factor control, peak load management, or the development of pricing strategies based on
consumption information. In addition, they make the operation and maintenance of the power
distribution network easier, allowing enabling or disabling meters, or updating their firmware or
settings remotely [López2013c].
AMI are being deployed by utilities to increase operational efficiency (e.g., better pricing
information, more accurate bills, or faster outage detection and restoration), to improve energy
efficiency (e.g., increasing the user awareness of the energy consumption), and to meet a range
- 24 -
Chapter 2 – State of the art
of new customer requirements and market opportunities. However, AMI programs are not just a
utility operational issue, but they are also a core part of the energy policies of many
governmental authorities [Depuru2011].
As a matter of fact, currently AMI deployments are mainly driven by regulation. In the EU
(European Union), with the requirements of the article 13 of the so-called Energy Services
Directive (2006/32/ED) and the adoption of the Directive on the internal electricity market
(2009/72/EC), it became clear that the modernization of the European meter infrastructure and
the introduction of intelligent metering systems must become a reality [JRC2012a]. However,
this regulatory push meets an actual requirement, since increasing the awareness of the power
distribution network is crucial to enable more sophisticated mechanisms, such as DR (Demand
Response) or proper EV massive integration.
Nevertheless, the AMI market across Europe remains diverse in terms of maturity,
technology preferences, and market drivers. Some countries already have a very high degree of
smart meters penetration, e.g., Italy (with 94%) and the Nordic Countries (with 70%), but in
general the penetration rate is still medium. EU countries have an overall mandate to deploy
smart meters up to 80% of customers by 2020 (namely, directive 2009/73/EC). In some
countries the deployment plans are even more ambitious, e.g., the Spanish directive
IET/290/2012 forces utilities to renew 100% of their meters to smart meters with
telemanagement and ToU (Time of Use) capabilities by the end of 2018. As a result, it is
foreseen that 212 million of smart meters will be deployed in Europe between 2011 and 2020
[PR2012].
2.3.2. Smart appliances
The state-of-the-art domestic appliances are not only becoming more and more energy
efficient, but they are also offering new important features to achieve lower energy consumption
and costs. Smart appliances are household appliances or white goods with monitoring and
control capabilities, providing communication with other devices and interfaces.
The AHAM (Association of Home Appliance Manufactures) defines smart appliances as “a
modernization of the electricity usage system of a home appliance so that it monitors, protects
and automatically adjusts its operation to the needs of its owner” [AHAM2009].
Smart appliances could comprise typical white goods such as refrigerators, freezers,
dishwashers, oven and stoves, washing machines and tumble dryers, as well as air conditioners,
circulation pumps for heating systems, electric storage heating, and water heaters. The main
objective of such appliances is to use an intelligent power management strategy to optimize the
load of the power distribution grid responding to utility signals.
Such electric appliances can therefore be used in DR programs (with an override function
controlled by the user). This can include rescheduling the operation of washing or dishwashing
cycles, interruptions of the operation of appliances, or the use of refrigerators and freezers for
temporarily storing energy in order to avoid operation of the compressor during peak times
[Timpe2009]. Such DR programs can also detect and react against disturbances in the power
frequency of the grid by turning a group of appliances off or on for a few minutes in order to
allow the grid to stabilize [Momoh2012].
However, smart appliances are not widely available nor deployed yet, so most current
solutions to provide energy efficiency at house and building levels are based on dedicated
monitoring and control systems.
- 25 -
Chapter 2 – State of the art
2.3.3 Monitoring and control systems
Energy monitoring and control systems have been traditionally used to reduce industrial
electricity consumption, but they are being applied more and more to achieve the same goals in
corporate office buildings, commercial and service buildings, and even in residential buildings
and households.
For the energy consumption monitoring in households there are two main available options
with different objectives:

Whole-house power monitors.

Power outlet monitors.
Whole-house power monitors have sensor clamps around the incoming power conductors
to measure the whole-house consumption. The measured values are sent wirelessly and in real
time to a communications gateway, which forwards them either to a home display for direct
visualization or to a server so that data can be checked via a computer or smartphone. Thus,
whole-house power monitoring systems allow being aware of home energy consumption in realtime as well as double-checking that utility is charging correctly over a given billing period (as
long as they provide the required accuracy). In addition, these systems can provide advanced
notification of alarms, alerts and advice, using instant text message and email alerts, based on
user-defined parameters. Figure 2-11 shows an example of this kind of systems.
Gateway
Clamps
Display
Figure 2-11 –Example of whole-house power monitoring system [OWL2014]
The power outlet monitors is plugged into a power outlet and then the appliance is plugged
into it. Such devices can monitor the amount of energy consumed by the appliance and
associated costs (as long as the user introduces the prices and tariffs). The most common option
includes a display in the socket to show such information, as Figure 2-12 illustrates. Other
versions do not have a display in each socket, but they send the data wirelessly to a central
display that can receive data from several sockets, creating a kind of monitoring system for the
energy consumption of individual appliances.
- 26 -
Chapter 2 – State of the art
Figure 2-12 – Example of power outlet monitor [Efergy2014]
The most common devices for the control of energy consumption in households are the
remote controlled sockets, which consists of an electrical socket and a remote control. The
remote controlled socket is plugged into any normal electrical outlet and can be switched on and
off with the remote control. Such devices can bring to normal appliances part of the services
provided by smart appliances. Some devices can have a single remote control to several plugs,
as shown in Figure 2-13.
Figure 2-13 – Example of remote controlled sockets [Efergy2014]
The capabilities of power outlet monitors and remote controlled sockets can be
incorporated in a single device, the so-called smart plugs. Smart plugs represent the cornerstone
of the so-called HEMSs (Home Energy Management Systems). Such residential monitoring and
control systems for the electricity consumption can also include a whole-house power monitor
as well as sensors (e.g., temperature sensor, motion sensor, CO2 sensor) to measure ambient
variables that allow ensuring an agreed level of comfort. All these data are typically sent
wirelessly to a gateway that forwards them to a server for their visualization via a computer or
smartphone. This gateway also routes commands to the appropriate devices. Figure 2-14 shows
a typical configuration of this kind of systems.
- 27 -
Chapter 2 – State of the art
Figure 2-14 – Example of residential monitoring and control system for electricity consumption
[Cloogy2014]
Energy generation monitoring systems are also appearing recently in the market, reflecting
the increasing penetration of micro-generation at residential level. The energy generation
monitoring systems include devices to measure how much energy is being generated as well as
the performance of the micro-generation installation (the so-called inverters). The measured
values are sent typically wirelessly and in real-time to a gateway that forwards them either to a
display for direct visualization or to a server, so that they can be visualized via web applications
using any device with Internet connection. Figure 2-15 shows an example of this kind of
systems.
Figure 2-15 – Example of monitoring system for residential micro-generation [Enlighten2014]
- 28 -
Chapter 2 – State of the art
Finally, with the recent rise of self-consumption and NZEB (Nearly Zero-Energy
Buildings), there are also appearing solutions that combine residential electricity consumption
monitoring and control systems with residential energy micro-generation monitoring systems in
order to optimize and match consumption and generation locally. These systems can be seen as
an extension or a more sophisticated version of the aforementioned HEMS. Figure 2-16 shows
an example of this kind of systems.
Figure 2-16 – Example of residential monitoring and control system for energy consumption and
generation [SHM2014]
As a token of the importance that energy efficiency in the residential sector is winning,
much R&D has been carried out in this area during the last few years, which has led to an actual
market for such HEMS. Table 2-3 summarizes just a few relevant commercially available
solutions which have been developed in parallel to this thesis.
However, such HEMSs currently available in the market, initially tackle the problem of
energy efficiency and energy consumption-generation matching locally, without neither taking
into account what is going on nearby nor interacting with the energy provider or any other
interested third party which operates at broader scope (e.g., at neighborhood or district level),
which is the direction most of them are currently heading.
- 29 -
Chapter 2 – State of the art
Table 2-3 – Summary of some relevant commercially available HEMSs
Company
Communicatios
Description
technologies
Efergy
Wireless
Whole-house power monitoring system
Engage
[Engage2014]
Powerhouse
Wi-Fi
Whole-house power monitoring system
eMonitor
[eMonitor2014]
Dynamics
enphase
Zigbee
Residential PV (photovoltaic) microEnlighten
[Enlighten2014]
generation monitoring system
electronic
Z-Wave
Residential monitoring and control system
Housekeeper
[Hk2014]
housekeeper
for energy consumption
Intelligent
Zigbee
Initially, residential monitoring and control
Cloogy2
[Cloogy2014]
Sensing
system for energy consumption. Recently,
Anywhere
this solution has incorporated equipment for
monitoring of residential PV microgeneration and residential monitoring of gas
tanks and meters
greenWave
Zigbee, Z-Wave, Initially, residential monitoring and control
Energy
Reality
Jennet
system for energy consumption. Recently,
Management
[EM2014]
this solution has been upgraded to monitoring
and control system for consumption and
generation
fifthplay
Wi-Fi, RF
Modular framework which comprises
Smart Home
[SH2014]
equipment for residential monitoring and
control of energy consumption, as well as for
the integration of residential PV microgeneration and EV
OWL
Wireless
Family of products which allow monitor and
OWL Intuition
[OWL2014]
control of energy consumption as well as
energy generation monitoring at residential
level
SMA
Bluetooth
Solution for monitoring and control of
Sunny Home
consumption and generation at residential
Manager
[SHM2014]
level
GEO
Zigbee
Solution for monitoring and control of
Chorus+energynote
[Chorus2014]
consumption and generation at residential
level
Product
2.3.4 Communications architectures, technologies, and
protocols
ICT and M2M communications are crucial in all the aforementioned technologies and
represent the key enabler of the Smart Grid at the distribution and customer domains.
When it comes to the required communications infrastructure, the first question that needs
to be answered is which the most appropriate communications architecture is. However, there is
not a single answer to this question, since this decision depends on multiple factors (e.g., target
application, specific features and constraints of the underlying power infrastructure, regulation,
etc.).
Reference [Mao2011] analyses the advantages and drawbacks of two communications
architectures based on wireless 4G technologies for AMI applications. First, direct
communication between the smart meters and the so-called MDMS (Metering Data
Management System) is considered, focusing on the main issues that this approach presents
2
As a matter of fact, Cloogy takes as reference the residential monitoring and control system proposed in
this thesis and presented in chapter 3, which was jointly designed with Intelligent Sensing Anywhere.
- 30 -
Chapter 2 – State of the art
from the communications point of view. In this regard, the paper highlights the inefficiencies of
performing connection establishment and authentication procedures in a per-smart-meter-basis
and remarks upon some resultant problems that network operators would have to face in this
scenario, such as the required improvement of the RAN (Radio Access Network) to avoid
problems related to lack of bandwidth or coverage arising from the huge number of smart
meters. To solve these problems, the paper proposes a hierarchical architecture comprising two
network segments, where the intermediate device between the smart meters and the MDMS is
called AP (Aggregation Point). For the communications between the smart meters and the APs,
short range communications technologies (e.g., IEEE 802.15.4/Zigbee or Wi-Fi) can be used;
whereas wireless 4G technologies are still used for the communications between the APs and
the MDMS.
As a matter of fact, hierarchical heterogeneous communications architectures comprising
several network segments and combining different communications technologies present higher
flexibility, so they fit a wider range of Smart Grid applications and specific requirements and
constraints [Zaballos2011]. Figure 2-17 shows a proposal of this kind of communications
architecture for the specific case of the power distribution domain of the Smart Grid
[Fadlullah2011].
Figure 2-17 – Hierarchical and heterogeneous M2M communications architecture for the power
distribution domain of the Smart Grid [Fadlullah2011]
The communications technologies use in the different network segments of such
hierarchical heterogeneous M2M communications architectures may vary depending on the
country or region. Although this thesis is mainly focused on the EU, Table 2-4 illustrates such
different trends in NA (North America), the EU, and AU (Australia) [Lo2012], for the HAN and
NAN as defined in [IEEE2011].
- 31 -
Chapter 2 – State of the art
Table 2-4 – Summary of some of the communications technologies likely to be deployed in HAN and
NAN in different countries [Lo2012]
PLC
Sub-GHz
GPRS
WiMAX
X
Proprietary.
X
ISM
X
X
Other
HomePlug
X
15.4g
NAN
NA
EU
NA
EU
Asia
AU
Zigbee
Wi-Fi
HAN
X
X
X
X
X
X
X
X
X
X
However, the Smart Grid at the distribution and customer domains represents such a hot
and emerging market that Table 2-4 is far away from including all the available options. Table
2-5 aims to be more exhaustive, mapping the most relevant communications technologies onto
the most appropriate network segments defined in [IEEE2011], paying special attention to the
EU market [Güngör2011], [Fang2013], [Ancillotti2013], [Usman2013].
Backhaul
Wired
Wireless
Table 2-5 – Summary of communications technologies and their scope
Communications Technologies
HAN
NAN
(Occasionally,
Zigbee
(Initially based upon IEEE
X
using mesh
802.15.4)
topology)
X
Bluetooth
X
Z-Wave
EnOcean
(ISO/IEC 14543-3-10)
X
[EnOcean2014]
Wi-Fi
X
X
(IEEE 802.11)
WiMAX
X
(IEEE 802.16)
X
White Spaces
Satellite
Cellular
Insteon
X
[Insteon2014]
X
KNX
X
LonWorks
X
X
(e.g., BACnet,
(e.g., PRIME, G3PLC-based
HomePlug)
PLC, G.hnem)
Ethernet
X
X
X
X
X
(e.g., 1G/10G
Ethernet)
X
(e.g., Fast Ethernet)
X
DSL
Cable
(DOCSIS)
X
X
(e.g., FTTH)
Fiber
- 32 -
X
(e.g., SONET/SDH)
Chapter 2 – State of the art
Due to the huge number of devices that Smart Grids involve, cost and energy consumption
are two key constraints to be taken into consideration when deciding the most appropriate
technology for a given communications segment. On the one side, the cost of deploying and
managing the required monitoring and control infrastructure has to be lower than the cost of
building and maintaining new peak power plants and of increasing the T&D (Transmission and
Distribution) grid capacity. On the other side, the energy consumption of such monitoring and
control infrastructure has to be lower than the energy savings ensured by the system itself. In
addition, some other features, such as data rate or range, need to be considered depending on the
specific requirements of each communications segment.
Based on these considerations, there are some communications technologies that fit
specific communications segments better than others. In general, wireless communications
technologies are of special interest due to the well-known advantages that they present (e.g.,
flexibility or ease of deployment).
Standing out among the wireless communications technologies for HAN are
IEEE802.15.4/Zigbee and Bluetooth, since they are low-power communication technologies
that fit the range and data rate requirements of this network segment. As a token of that, several
of the commercial energy management platforms currently available in the market for dwellings
and SOHO (Small Office Home Office) presented in section 2.3.3 used Zigbee (e.g., Cloogy
[Cloogy2014], GreenWave Reality Energy Management [EM2014]) and Bluetooth (e.g., Sunny
Home Manager [SHM2014]).
IEEE802.15.4/Zigbee mesh networks can be also used as last mile solution
[Kulkarni2012]. This approach presents advantages from the point of view of the management
and maintenance of the infrastructure, although it may also present higher security risks. In
practice, this option presents fairly high adoption in areas such as California (US).
Standing out among the wired communications technologies for HAN and NAN are the
PLC (Power Line Communications)-based solutions, mainly due to the fact that they present
very low deployment costs, since in this case the power infrastructure represents also the
communications infrastructure. There are two different types of PLC communications, namely
broadband (B-PLC or BPL) and narrowband (NB-PLC). NB-PLC communications were
developed to mitigate some problems of BPL related to EMI (Electromagnetic Interference)
[Bartak2013] and to crossing from LV (Low Voltage) to MV (Medium Voltage) networks.
Currently, the second generation of NB-PLC technologies is already available in the market.
NB-PLC technologies represent the preferred solution of utilities for AMI applications
[Aidine2013]. The most relevant NB-PLC technologies for AMI are:

Meters & More, which is promoted by the Enel Group. Its lower layers are being
standardized by the IEC (International Electrotechnical Commission). Meters & More is
being widely deployed in those countries where the Enel Group presents high market share
(e.g., Italy).

OSGP (Open Smart Grid Protocol), which is promoted by Echelon. Its lower layers are
being also standardized by IEC. OSGP presents its higher penetration rates in the Nordic
countries.

G3, which is promoted by EDF and Maxim. Its PHY and MAC layers have already been
published as standard by ITU-T (International Telecommunication Union Telecommunication Standardization Sector) [ITU2012a]. G3 shows the highest penetration
rates in those countries where EDF presents high market share (e.g., France).
- 33 -
Chapter 2 – State of the art

PRIME (PoweRline Intelligent Metering Evolution), which is promoted by the PRIME
Alliance, led by Iberdrola. Its PHY and MAC layers have been also published as standard
by ITU-T [ITU2012b]. PRIME is being widely deployed in Spain, Portugal, UK (United
Kindom), Poland, Brazil, or Australia.
Table 2-6 summarizes the main features of some of these PLC-based communications
technologies [Güngör2011], [De Craemer2010], [Ekanayake2012], [Aidine2013].
Name
IEEE 1901
Table 2-6 –Summary of PLC communications technologies
Type
Details
BPL
High speed PLC. Up to 100 Mbps
BACnet
NB 1G
HomePlug
BPL
HomePlug
Green PHY
NB 1G
G3-PLC
NB 2G
PRIME
NB 2G
G.HNEM
NB 2G
Meters&More
NB 2G
OSGP
NB 2G
ASHRAE (American Society of Heating,
Refrigerating and Air Conditioning Engineers)
standard for building automation and control
networks
Non-standardized technology specified by the
HomePlug Powerline Alliance to interconnect
smart appliances using home electricity
network
Low-power, cost-optimized PLC networking
specification. Tested by American utilities such
as Duke Energy, Pacific Gas & Electric, and
Southern California Edison
Launched by ERDF and Maxim. Aim at
providing interoperability, cyber security and
robustness while reducing costs
Open, global standard for multi-vendor
compatibility
ITU-T standard which aims at harmonizing G3
and PRIME
NB-PLC technology promoted by the Enel
group as communications standard for AMI
NB-PLC technology promoted by Echelon as
communications standard for AMI
Scope/App
HAN/In-home
multimedia networks
HAN
HAN
HAN
NAN/AMI
NAN/AMI
NAN/AMI
NAN/AMI
NAN/AMI
Wi-Fi represents an interesting wireless option for the NAN especially for entities
potentially interested on providing energy services without owning the electrical infrastructure,
taking into account the fairly low cost and consumption and the high penetration of this
communications technology.
Regarding the backhaul, GPRS (General Packet Radio Service) seems to be the cellular
technology that better fits the specific requirements of this communications segment and is
widely used in the vertical solutions which are more and more being deployed nowadays.
Although it is not clear yet whether utilities will outsource the management of such
communications infrastructure or will operate it themselves as MVO (Mobile Virtual
Operators), utilities seem to prefer the first option. Wired broadband technologies, e.g., DSL
(Digital Subscriber Line), cable, or even fiber-based solution such as FTTH (Fiber To The
Home), represent appropriate technologies for approaches involving non-dedicated
communications infrastructures. In rural environments, WiMAX (Worldwide Interoperability
for Microwave Access), satellite communications, and White Spaces [Brew2011] represent
interesting technologies.
Reference [Robichon2013] proposes deploying an entirely new cellular network based on
CDMA-450 (Code Division Multiple Access at 450 MHz) exclusively devoted to the Smart
Grid applications provided by a given DSOr (Distribution System Operator). In this case, the
DSOr would be the owner of the communications infrastructure, but a telecom operator would
- 34 -
Chapter 2 – State of the art
be responsible for running it. The paper claims that this is a suitable solution from both
technical and economic perspectives for the specific boundary conditions of the Netherlands.
However, this would not be the case in other countries where the license to operate in this
frequency band is too expensive or where this frequency band is already allocated for any other
use.
The aforementioned communications technologies address either the lower
communications layers or the whole stack. However, at the application layer in particular, there
are also many available standards and protocols, most of them not being bounded to a single
network segment, but expanding across them. Table 2-7 summarizes the most relevant ones
[Güngör2011], [De Craemer2010], [Ekanayake2012].
Name
USNAP
M-Bus
DLMS/COSEM
(IEC 62056)
ANSI C12.19
OpenADR
DNP3
IEC 60870-5-101
IEC 60870-5-104
IEC 61970/61969
IEC 61850
MODBUS
IEC 60870-6
Table 2-7 - Summary of higher layers standards and protocols
Details
Standard to enable interoperability in HAN
European standard providing the requirements for remotely
reading all kind of utility meters. Wireless M-Bus has been
also specified recently
Application standard for data exchange of meter readings,
tariffs, and load control. Work over all the aforementioned 2nd
generation NB-PLC solutions
Flexible metering model for common data structures and
industry vocabulary for meter data communications
Originally developed by Lawrence Berkeley Labs. Open,
platform-independent, and transparent E2E standard for
demand response
Standard for communications between Control Centers, RTUs
(Remote Terminal Units), and IDEs (Intelligent Electronic
Devices). Widely used in the US and Canada
Standard for communications between Control Centers, RTUs,
and IDEs. Widely used in Europe
Include enhancements to 101 at every communications layer
Define a CIM which is necessary for exchanging information
between devices and networks. IEC 61970 works in the
transmission domain; whereas IEC 61969 works in the
distribution domain
Flexible open standard for communications between and
within substations
Messaging protocol in the application layer that provides
communication between devices connected over several buses
and networks. It can be implemented through Ethernet or using
asynchronous serial transmission over EIA 232, EIA 422, EIA
485
Protocol for data exchange between utility Control Centers.
Scope/App
HAN
AN/AMI
AMI
AMI
DR, DSM
SCADA
SCADA
SCADA
Energy
Management
Systems
Substation
Automation
Substation
Automation
Inter-Control
Centre
Communications
Figure 2-18 provides an overview of where some of the analyzed protocols and standards
are placed in the OSI (Open System Interconnection) model.
- 35 -
Chapter 2 – State of the art
Figure 2-18 – OSI layer placement of most of the analysed standards [Moura2013b], [De Craemer2010]
2.4 Main trends on standardization and research
From the ICT perspective, standardization represents a key issue to foster the effective
deployment of Smart Grids, since it allows equipment from different vendors to interoperate and
it encourages competition, creating a global market and thus reducing costs. However, in such
an emerging market, standardization has to be managed carefully, so that it leaves room for
research and innovation. As a result, an optimum balance between standardization and R&D is
critical to maximize Smart Grid potential [Yan2013]. Efforts for standardization and R&D need
to be undertaken at every layer, homogenization and interoperability being the key issues in this
context [Moura2013b].
As it has already been presented, there is a plethora of available specifications and
standards at the HAN and NAN level, which actually hampers potential deployments, instead of
promoting them, since they introduce uncertainty in the market, instead of reinforcing
investments.
The matter of the most appropriate communications architecture and technologies
eventually depends on the specific business case behind the target application, which in turn
depends on many factors, such as the special features of the target application itself or the
special features and the specific regulation of each country (which varies, e.g., even among EU
countries). As a result, effective methods to evaluate different options and select the most
appropriate ones before undertaking the important investments needed to deploy this kind of
infrastructures on a large, are required.
Simulations represent a powerful, flexible, and cost-effective solution to achieve this goal.
The research in this area can be classified depending on where it is focused. Thus, some
research efforts are mainly focused on evaluating the behavior and performance of the lower
layers of the communications protocols [Matanza2013], [Papadopoulos2013]. There are also
research works focused on evaluating some figures of merit of the communications
infrastructures themselves [Abdul Salam2012]. In this case, it is crucial to appropriately
characterize the traffic of the target application [Khan2013] in order to obtain meaningful
results. Finally, as the Smart Grid brings energy and ICT together, co-simulation of energy and
ICT infrastructures represents a very promising and challenging research area [Anderson2012a],
[Mets2011], [Lin2011], [Lévesque2012], [Celli2013], [Majumder2013].
Outstanding efforts are also being carried out by projects in Europe and the US to develop
a common language which enables interoperability at HAN level. Notably, two objectives are
set with regard to this within the scope of the EU eeBuilding Data Model collaboration (also
called as eeSemantics) space:
- 36 -
Chapter 2 – State of the art

The objective in the medium to long term is to develop a common protocol which allows
“Plug-and-Play" integration and interoperation of so-called EupP (Energy using and
producing Products) from different manufacturers. In order to achieve this goal, a common
data model needs to be developed first. In this sense, ontologies are becoming an
increasingly popular way of defining machine-readable data models within the Smart Grid
area [Grassi2011], [Santodomingo2012], [Wicaksono2012], in particular, and the M2M area
[Gyrard2013], in general.

The objective in the short to medium term is to use middleware (e.g., the SmartLink
middleware platform developed under the scope of the EU FP6 project HYDRA and
enhanced under the scope of the EU FP7 project SEEMPubS [Osello2013]) which allows
managing devices from different vendors using different physical media and
communications protocols.
Examples of this trend towards interoperability in the HAN are found in Zigbee and
second-generation NB-PLC communications technologies. The Zigbee SEP (Smart Energy
Profile) 2.0 puts special emphasis on interoperability, working not only over IEEE 802.15.4 but
also over Wi-Fi and PLC, and incorporating 6LoWPAN (IPv6 over Low Power Wireless
Personal Area Network) in the Zigbee stack. In the case of NB-PLC technologies, the ITU-T is
making remarkable efforts to homogenize second-generation NB-PLC technologies, such as
PRIME or G3-PLC, under the standard G.HNEM [Oskman2011]. As a token of this, there are
already hardware equipped with both IEEE 802.15.4 and NB-PLC chipsets commercially
available [Greenvity2014].
Interoperability between heterogeneous systems (e.g., utilities and home/building energy
management systems), allowing holistic energy management, is also critical. As it has already
been pointed out, there are also many different protocols developed for specific purposes at the
application layer. Homogenization at the data structure level would allow such protocols to
interoperate (if required) and would encourage third parties to come up with a wide variety of
value-added services that would enrich and energize the Smart Grid market.
The US Green Button Data initiative represents an interesting example of this. Within the
Green Button Data approach, the utilities provide their customers with their energy-related
information in a standard machine-readable format. Since such information is in standard
format, third parties (e.g., ESCOs) will be able to offer added-value services by processing it
somehow. Then, the customers will allow those ESCOs whose services are of their interest, to
get access to their energy-related information in order to enjoy the services they offer, thus
partly solving privacy issues [GBD2014].
2.5 Overview of related projects
During the last few years, many R&D projects have been undertaken both at national and
European level to tackle the issues presented throughout this chapter. As a matter of fact, a
review recently published by the EU JRC (Joint Research Center) shows that, up to September
2012, a total of 281 Smart Grid projects and around 90 smart metering pilots and rollouts were
made across 30 countries in Europe, accounting for total investments of at least €5 billion and
€1.8 billion respectively [JRC2012b]. Based on this inventory of Smart Grid projects, the
European electricity union Euroelectric has recently launched an online platform on Smart Grid
activity in Europe, which includes an interactive, fully searchable map of Smart Grid projects
and the detailed project pages [SGprojects2014].
Next, some EU R&D projects of special relevance to this thesis are summarized:
- 37 -
Chapter 2 – State of the art

The REViSITE (Roadmap Enabling Vision and Strategy for ICT-enabled Energy
Efficiency) project [REViSITE2012] aims to contribute to the formation of a European
multidisciplinary “ICT for energy-efficiency” research community by bringing together the
ICT community and important and complementary application sectors, such as grids,
building/construction, manufacturing and lighting.

The IREEN (ICT Roadmap for Energy Efficient Neighborhoods) project [IREEN2013] is a
strategic project which examines the ways that ICT for energy efficiency can be extended
beyond individual homes and buildings to the wider context of neighbourhoods and
communities.

The Energy Warden (Renewable Energy Sourcing Decisions and Control in Buildings)
project [EW2012] aims at developing and marketing a simulator and modeling tool which
includes dynamic models for energy producing, storing and using units. The resulting tool
aims to provide decision aid in designing or retrofitting energy infrastructures at the
building domain.

The main objective of the HESMOS (ICT Platform for Holistic Energy Efficiency
Simulation and Lifecycle Management Of Public Use FacilitieS) project [HESMOS2013] is
to allow complex lifecycle simulations to be done during design, refurbishment and
retrofitting phases, where the largest energy saving potentials exist.

The SEEMPubS (Smart Energy Efficient Middleware for Public Spaces) [Osello2013]
project specifically addresses reduction in energy usage and CO2 footprint in existing public
buildings and spaces without significant construction works, by an intelligent ICT-based
service monitoring and managing of the energy consumption.

The Adapt4EE project [Adapt4EE2014] aims at providing a holistic approach to the design
and evaluation of the energy performance of construction products at an early stage and
prior to their realization. The Adapt4EE project is currently responsible for hosting and
managing the EC eeSemantics collaboration space.

The IntUBE (Intelligent use of buildings' energy information) project [IntUBE2011] aims to
develop tools for measuring and analyzing building energy profiles based on user comfort
needs.

The main objective of the AIM (A novel architecture for modelling, virtualizing and
managing the energy consumption of household appliances) project [AIM2010] is to foster
a harmonized technology for profiling and managing the energy consumption of appliances
at home.

The DEHEMS (Digital environment home energy management system) project
[DEHEMS2011] aims to bring together sensor data in areas such as household heat loss and
appliance performance as well as energy usage monitoring in order to give real-time
information on emissions and energy performance of appliances and services.

The PEBBLE (Positive-energy buildings thru better control decisions) project
[PEBBLE2012] aims to optimize loads to achieve generation-consumption matching.

The FIEMSER (Friendly Intelligent Energy Management System for Existing Residential
Buildings) project [FIEMSER2013] addresses the need of achieving energy positive
buildings through solutions based on a rational consumption of energy, local generation,
and an increase in the consciousness of the building owners towards their energy
consumption habits.
- 38 -
Chapter 2 – State of the art

The ENCOURAGE (Embedded iNtelligent COntrols for bUildings with Renewable
generAtion and storaGE) project [ENCOURAGE2014] aims to develop embedded
intelligence and integration technologies that will directly optimize energy use in buildings
and enable active participation in the future smart grid environment.

The BEyWatch (Building Energy Watcher) project [BeyWatch2011] aims to develop an
energy-aware and user-centric solution, able to provide intelligent energy
monitoring/control for white goods and power demand balancing at home/building and
neighbourhood level.

The objective of SmartCoDe (Smart Control of Demand for Consumption and Supply to
enable balanced, energy-positive buildings and neighbourhoods) project [SmartCoDe2012]
is to enable the application of advanced techniques for energy management in private and
small commercial buildings and neighbourhoods.

The ENERsip (ENERgy Saving Information Platform for Generation and Consumption
Networks) project [López2012b] aims to create an adaptive, customizable and serviceoriented energy monitoring and control system for near real-time generation and
consumption matching in residential, commercial buildings and neighborhoods by active
and proactively coordinating energy, communications, control, and computing.

The main target of the ADDRESS (Active Distribution network with full integration of
Demand and distributed energy RESourceS) [Belhomme2008] project is to enable the active
participation of small and commercial consumers in power system markets and provision of
services to the different power system participants.

The SmartHouse/SmartGrid project [SHSG2011] aims to validate and test how ICT-enabled
collaborative technical-commercial aggregations of smart houses provide an essential step
to achieve the needed radically higher levels of energy efficiency in Europe.

The NOBEL (Neighbourhood Oriented Brokerage Electricity and monitoring system)
project [NOBEL2012] aims to build an energy brokerage system that allows individual
energy consumers to communicate their energy needs directly to both large-scale and smallscale energy producers, thereby making energy use more efficient.

The MIRABEL (Micro-Request-Based Aggregation, Forecasting and Scheduling of Energy
Demand, Supply and Distribution) project [MIRABEL2013] is dedicated to develop an
approach on a conceptual and an infrastructural level that allows DSOs to balance the
available supply of renewable energy sources and the current demand in ad-hoc fashion.

The INTEGRIS (INTelligent Electrical Grid Sensor communications) project [Della
Giustina2011] proposes the development of a novel and flexible ICT infrastructure based on
a hybrid PLC-wireless integrated communications system able to completely and efficiently
fulfill the communications requirements foreseen for the Smart Grids of the future.

The PowerUp project [Caldevilla2013] aims to develop the V2G (Vehicle-to-Grid) interface
which ensures the seamless integration of EVs into the Smart Grid.
At national level, some Spanish R&D projects of special relevance to this thesis are
summarized next:

The main objectives of the GAD (Active Demand Side Management Project) project
[GAD2010] are to develop: 1) tools to optimize electricity consumption in households, thus
reducing electricity cost and environmental impact; 2) devices to show the price and origin
- 39 -
Chapter 2 – State of the art
of the energy to the customer; 3) technologies that improve power quality while aiding in
the integration of renewables.

The main objectives of the PRICE-GEN project [López2013c] are to: 1) design an optimal
and interoperable communications network architecture; 2) develop novel and smart sensing
and actuating equipment, which provide information on consumption and generation that
allows having a more accurate picture of the power distribution network in almost real-time;
3) validate this platform by means of a pilot scheme which involves the deployment of over
200,000 smart meters and their integration into the operational power distribution
infrastructure of a geographical area close to Madrid (Spain).

The DOMOCELL project [López2013b] aims to design and develop a smart charging
infrastructure to seamlessly integrate EV into the Smart Grid at residential level.
Table 2-8 summarizes, classifies and compares the considered R&D projects.
Table 2-8 – Summary and comparison of considered R&D projects
Projects
H
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
REViSITE
IREEN
Energy
Warden
HESMOS
SEEMPubS
Adapt4EE
IntUBE
AIM
DEHEMS
PEBBLE
FIEMSER
ENCOURAGE
BEyWatch
SmartCoDe
ENERsip
ADDRESS
Smart House/
Smart Grid
NOBEL
MIRABEL
INTEGRIS
PowerUp
GAD
PRICE-GEN
DOMOCELL
X
Scope
B N
X
X
X
X
X
X
X
X
ICT4EE
Guidelines
Sim
tools
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
DER
EV
DR
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
M2M
comms
X
X
X
X
X
X
AMI
DA&V
tools
X
X
X
X
X
X
H: Home; B: Building; N: Neighborhood; ICT4EE: ICT for Energy Efficiency; Sim: Simulation; DA&V:
Data Analysis and Visualization
As Table 2-8 shows, the considered projects approach the overall problem (i.e., energy
efficiency and integration of renewable micro-generation at the distribution and customer
domains of the Smart Grid) from different perspectives. There are some projects, such as
REViSITE and IREEN, whose main goal is to define guidelines and provide tools for effective
collaboration. These projects work at a high level of abstraction and present a wide scope, but
they do not get into the details of how to solve the specific problems. Regarding the projects that
do so, most of them deal with very specific pieces of the overall issue.
- 40 -
Chapter 2 – State of the art
The projects BEyWatch, SmartCoDe, ENERsip, and INTEGRIS, excel at presenting the
most comprehensive proposals. The ENERsip project differentiates itself from the BEyWatch
project and the SmartCoDe project because it explicitly deals with the core M2M
communications infrastructure required to tackle the aforementioned overall problem at
neighborhood level. The INTEGRIS project also does it. However, the INTEGRIS project
approaches the problem from the DSO perspective; whereas the projects BEyWatch,
SmartCoDe, and ENERsip are more focused on the customer, leaving the door open for any
interested party to run and operate the proposed platforms (e.g., the former project does not
consider HEMS nor BEMS; whereas the latter ones do so).
This thesis has been partly developed under the scope of the ENERsip project. As a matter
of fact, one of the main objectives of this thesis is to propose a M2M communications
architecture that supports energy efficiency and integration of renewable micro-generation
within the so-called energy-positive neighborhoods of the Smart Grid. In addition, the research
work carried out in this thesis has been taken as reference while being also influenced by the
Spanish projects DOMOCELL and PRICE-GEN.
2.7 Conclusions
The review of the state of the art presented throughout this chapter shows the efforts
carried out by the main standardization bodies and by the research community to tackle the
issues associated to the design and development of the Smart Grid from different perspectives,
such as the need for conceptual models and references architectures, the issue of selecting the
most suitable communications architectures and technologies, or the trend towards
homogenization and interoperability in ICT.
Regarding standards in particular, there are a myriad of them for Smart Grids, but this
chapter is focused on the most relevant ones to this thesis. To effectively look for any other
standard, the IEC has recently created a free Smart Grid Standards Mapping Tool that allows
easily identifying the standards that are needed for any part of the Smart Grid, including not
only IEC standards but also standards from other organizations [IEC2014].
This thesis has contributed to some of the aforementioned efforts in parallel with the
standardization and R&D works presented throughout this chapter. Figure 2-19 illustrates this
fact, showing a timeline of the most relevant milestones of this thesis along with the most
relevant standardization and research milestones3. Notably, the main contributions of this thesis
to the state of the art, as it will be described in detail in the next chapters, are:

We propose a novel M2M communications architecture to support energy efficiency and
energy consumption and generation optimization at neighborhood level, going beyond the
meters down to the consumption and generation monitoring and control networks.

We formally model the domain of knowledge of energy efficiency platforms for the socalled energy-positive neighborhoods by means of an ontology developed in OWL
(Ontology Web Language), with the aim that it becomes a reference data model for the
application of M2M communications to this context.

We characterize the traffic carried by the core of the designed M2M communications
architecture in realistic large-scale scenarios.
3
[Fadlullah2011] and [Zaballos2011] are included as remarkable proposals of M2M
communications architectures within the same scope of this thesis. [Khan2013] is included as a
comprehensive work on characterizing traffic requirements depending on the target application.
- 41 -
Chapter 2 – State of the art

Based on this traffic modeling, we evaluate the core of the proposed M2M communications
architecture from both economic and technical perspectives by means of theoretical analysis
and simulations.
- 42 -
1/2010
NIST Smart Grid
Conceptual Model
9/2011
IEEE 2030
7/2011
NIST CoS
4/2011
9/2011
[Fadlullah2011]
[Zaballos2011]
9/2012
ESOs
Conceptual model
SGAM
1st set of stds
2/2013
[Khan2013]
- 43 1/1/2010
12/1/2013
4/2011
Publication of proposed
M2M architecture
6/2012
Publication of traffic
modeling
1/2013
Publication of
ontology in
eeSemantics
Figure 2-19 – Timeline of the most relevant milestones of this thesis along with the most relevant standardization and research milestones
1/2014
Simulations results
Chapter 2 – State of the art
- 44 -
Chapter 3
Network Architecture
3.1 Introduction
The main drivers of the Smart Grid presented in chapter 1 (namely, energy efficiency and
the increasing penetration of DERs – Distributed Energy Resources – including renewable
generation and EVs – Electric Vehicles) have special impact on the power distribution domain
and on the customer domain (in particular, at residential level). The so-called energy-positive
neighborhoods, buildings, and households appear in this context as if such a “divide and
conquer” approach was applied to tackle the problem, in that it is addressed at these three levels.
This chapter is focused on addressing such Smart Grid drivers at neighborhood level. To be
more precise, the main goal of this chapter is to design a M2M (Machine-to-Machine)
communications architecture which supports energy efficiency within energy-positive
neighborhoods by enabling electricity consumption reduction at residential level and matching
the electricity consumption within a given neighborhood with the generation coming from the
local renewable sources distributed along the same neighborhood.
The remainder of the chapter is structured as follows. Section 3.2 presents the overall
system architecture, paying special attention to the M2M communications infrastructure.
Section 3.3 explains how our proposal is aligned with the standardization work reviewed in
chapter 2. Section 3.4 proposes the communications technologies to be used in each network
segment. Section 3.5 outlines the most relevant security features of the selected communications
Chapter 3 – Network Architecture
technologies and proposes the most appropriate solution in each case. Section 3.6 describes an
application-layer solution to provide E2E (End-to-End) addressability throughout the platform
and explains how it would work during deployment and operation. Finally, section 3.7
summarizes this chapter and draws conclusions.
3.2 System description
Figure 3-1 shows the overall system architecture of the proposed ICT (Information and
Communications Technologies) system to enable electricity consumption reduction and proper
integration of DER (Distributed Energy Resources) at neighborhood level. It can be seen that
the system is divided into four domains, which represents its main pillars from the ICT
perspective. The Building Domain comprises the physical infrastructures owned by the
commercial and residential customers of the power distribution network of the Smart Grid,
including consumption and generation equipment and the SANs (Sensor and Actuator
Networks) to monitor and control them. The User Domain encompasses everything related with
making the interaction of the users and the system as fruitful as possible. The Information
System Domain represents the “brain” of the system from the energy perspective, comprising
the logic that allows the optimal use of the available resources at neighborhood level at any
time. Finally, the Neighborhood domain encompasses the core communications infrastructure
that carried data and command back and forward, allowing that everything work correctly.
BUILDING DOMAIN
NILM
COMFORT
SENSORS
ADR
EP
NEIGHBORHOOD DOMAIN
PLUGS
M2M
Platform
INFRARED
BOX
I-BECI
INFORMATION SYSTEM DOMAIN
PS-BI
UAP
USER DOMAIN
UII
CNTR
DER
ENERGY
STORAGE
ADR
EP
External
Interface
802.15.4/ZB
802.11 (Wi-Fi)
EXTERNAL GRID
(DSO, TSO, ND)
SENSORS
(WS)
I-BEGI
GPRS/EDGE
Figure 3-1 – Overall system architecture
Thus, the Information System Domain and the User Domain are indeed related with IT
(Information Technology); whereas the Building Domain and the Neighborhood Domain are
tightly related with communications. As a matter of fact, the M2M communications architecture
proposed in this chapter is spread across these domains, as shown in Figure 3-1. This
architecture illustrates how the aforementioned “divide and conquer” approach is applied to
communications, in that it is a hierarchical architecture which comprises specific network
segments for households, buildings or group of households, and neighborhoods. This approach
makes the communications infrastructure more flexible and adaptive, and boosts scalability. The
proposed communications architecture is also heterogeneous in that it involves different
communications technologies in order to meet the specific communications requirements of
each network segment.
- 46 -
Chapter 3 – Network Architecture
3.2.1 Building domain
As it has already been mentioned, the Building Domain encompasses the physical
infrastructure (both electrical and ICT) owned by the commercial and residential customers of
the power distribution network of the Smart Grid. It is further divided into:

The so-called I-BECIs (In-Building Energy Consumption Infrastructures), which represent
the consumption infrastructures along with the SANs to smartly manage them;

And the so-called I-BEGIs (In-Building Energy Generation Infrastructures), which
represent the local generation facilities along with the SANs to smartly manage them.
I-BECIs and I-BEGIs can be combined or not, giving rise to different profiles of
customers:

Consumers: customers whose households or buildings are only composed of I-BECIs.

Producers: customers whose infrastructures comprise only I-BEGIs connected to the grid.

Prosumers: customers that own the so-called energy-positive households or buildings, also
known as ZEBs (Zero-Energy Buildings) or NZEBs (Net-Zero Energy Buildings), which
integrate both I-BECIs and I-BEGIs.
Every I-BECI or I-BEGI or combination of I-BECI and I-BEGI needs a device that
manages the communications inside the SAN or HEMS (Home Energy Management System)
and with the Information System where the overall optimization processes run. This device is
the so-called ADR EP (Automated Demand Response End Point). Apart from the already said
duties, the ADR EP also aims to allow managing in a uniform way the wide variety of devices
within the I-BECIs and I-BEGIs, hiding this complexity and heterogeneity to the Information
System. In order to achieve this goal, the ADR EP needs to be equipped with multiple hardware
interfaces and to support multiple communications protocols, including the potentially different
communications protocols within the I-BECIs/I-BEGIs and a common application protocol for
the communication with the Information System (e.g., an extension to IEC61850
[Apostolov2013]). Therefore, a star topology controlled by the ADR EP is proposed for the
HEMS (i.e., I-BECIs or I-BEGIs or both).
Regarding the SANs for the I-BECIs, taking into account that – as it was pointed out in
chapter 2 – smart appliances represent a solution in the long run, the so-called smart Plugs need
to be used both to monitor the consumption of the appliances and to control it accordingly to the
commands sent by the customer or automatically generated by the Information System. Thus,
such smart Plugs allow, e.g., cutting standby consumptions, which may account for up to the
7% of the total annual electricity consumption per household [De Almeida2011].
However, there is a group of devices that are of special interest for energy efficiency and
consumption-generation matching which may require a more sophisticated control via IR
(Infrared) communications: the HVAC (Heating Ventilating Air Conditioning) loads, which
present high consumption and high penetration at households. Therefore, a specific device, the
so-called IR Box or GW (Gateway), needs to be placed between this kind of equipment and the
ADR EP in order to allow managing them appropriately. The IR GW also allows seamless
integration of widely used devices such as TVs or DVDs into the HEMS.
The NILM (Non-Intrusive Load Monitoring) or NIALM (Non-Intrusive Appliance Load
Monitoring) [Zeifman2011] represents a very promising technology to enable backward
compatibility and to reduce the cost of the HEMS. The NILM is a technology that has been
- 47 -
Chapter 3 – Network Architecture
around for a long while [Warren1989]. It allows identifying (based on electrical signature) the
appliances which are running, even if they are not equipped with a smart plug with sensor and
communication capabilities. In order to achieve this goal, the theoretical NILM approach
determines when a specific appliance is turned ON based just on its electrical signature by
applying non-supervised DSP (Digital Signal Processing) methods to the overall electrical
signal. However, due to the complexity of such methods, a more straightforward approach
entails that the customer informs the NILM module about the appliance which has just been
turned ON in order to help it learn its electrical signature. As a result, this NILM module allows
disaggregating household energy consumption into individual appliances without the need of
using smart plugs for every single appliance; only those appliances susceptible to be controlled
(e.g., by participating in DR –Demand Response – programs) would need to be plugged in the
socket through smart plugs.
To complete the SANs of the I-BECIs, comfort is a keyword when talking about energy
efficiency, in that achieving energy efficiency may never compromise the basic comfort level
established by the customer. Therefore, the so-called comfort sensors are required to measure
different environmental variables, such as temperature, relative humidity or CO2 concentration,
which are taken into account when running the optimization algorithms in order to avoid
compromising the customers’ basic comfort levels. For instance, the data acquired by a CO2
concentration sensor can be used to control the AC (Air Conditioning) equipment in order to
reduce its electricity consumption as follows. When the CO2 concentration in a room is below a
maximum acceptable value, the AC can cool the air from inside the room, which requires much
less electricity than taking it from outside, since it is not so hot. Only when the CO2
concentration is above such a threshold, the AC will cool the air from outside, until CO 2
concentration goes below the maximum acceptable value again. A temperature sensor can be
also used to control the AC equipment efficiently and to integrate it into DR programs by
providing feedback on the actual temperature in a room, allowing actuating on the AC
consequently.
Regarding the I-BEGIs, photovoltaic panels and µwind turbines represent the most
extended technologies in buildings [Jellea2012], [Ayhana2012]. Both the photovoltaic panels
and the µwind turbines are equipped with Inverters, which play a double role: on the one side,
they adapt the electrical signal (i.e., converting DC into AC); on the other side, they work as
RTU (Remote Terminal Units), measuring some key parameters associated to the generation
equipment and sending them to the ADR EP. The photovoltaic panels may be also equipped
with panel temperature sensors, since this parameter influences their performance. In addition,
the so-called energy meters are needed to measure the energy that the installation is generating.
The SANs of the I-BEGIs need to be equipped with sensors to measure variables related to
weather conditions (e.g., temperature, humidity, wind direction and speed, solar irradiation) and
electrical variables (e.g., DC current or voltage, AC current or voltage) and to send them to the
Information System in order to allow accurate status monitoring and accurate generation
forecast, which are two key parameters when operating DR events. The set of sensors in charge
of monitoring weather conditions can be integrated all together into the same Weather Station
and the set of sensors in charge of measuring electrical variables are usually integrated into the
same Network Analyzer.
The I-BECI and I-BEGI presented in this section have been taken as reference in the EU
FP7 project ENERsip to develop the prototype networks shown in Figure 3-2 and Figure 3-3.
- 48 -
Chapter 3 – Network Architecture
Figure 3-2 – Prototype network for the monitoring and control of I-BECI [Carreiro2011]
Figure 3-3 – Prototype network for the monitoring and control of I-BEGIs [López2013a]
As it can be seen, the approach in this project was to develop independent ADR EPs for the
I-BECIs and for the I-BEGIs. In addition, the I-BECIs rely on wireless short distance
communications (IEEE 802.15.4/Zigbee [IEEE2006]); whereas the I-BEGIs rely on wired
communications which are currently widely used in these environments. However, the I-BEGI
shown in Figure 3-3 can be easily upgraded to support the same wireless protocol as in the IBECI - thus facilitating their integration – by using off-the-shelf RS-232-Zigbee modems (also
known as “Zigbee dongles”).
3.2.2 User domain
The User Domain encompasses the means through which the users and the system interact.
It is worthwhile mentioning at this point that energy efficiency is achieved both through
automated actions (e.g., DR event) and by influencing users’ behaviors. This is why is so
important the way in which the information is presented to the user, so that it is easily
understandable, as well as the tools that are provided to the users for them to make decisions,
which should be as human-friendly as possible.
This functionality is provided to users via the so-called UII (User Intuitive Interfaces). The
UII are applications that allow interacting with the platform in a human-friendly manner. Such
applications may run in smartphones or tablets or even in Smart TVs, since this technology is
- 49 -
Chapter 3 – Network Architecture
gaining significance and represent a great mean to reach people massively. It is also important
that such applications are somehow connected to social networks [Mankoff2010],
[Anderson2012b], in order to encourage the users to follow the recommendations proposed by
the system. In this way, social games could be presented to the users to try to motivate them to
be “greener”.
3.2.3 Information System domain
Gathering the consumption and generation data of the same district at a given moment of
time and processing them all together allow reaching global optimizations at neighborhood
level, instead of just local optimizations at household level, as it is the case in state-of-the-art
HEMS. In addition, since the customers are still allowed to configure a set of parameters and
thresholds and they are taken into account when running the optimization algorithms, local
optimizations can be also reached, whereas this is impossible the other way around.
However, the Information System is responsible not only for processing the data and
making the appropriate decisions, but also for making the appropriate information available to
the customers through the UIIs, as well as for dealing with the commands sent and
configurations made by the customers.
As a result, the Information System is further divided into two modules: the so-called PSBI (Power Saving Business Intelligent) and the so-called UAP (User Application Platform). The
PS-BI is the one responsible for collecting (through the M2M communication infrastructure) all
the data regarding both energy generation and consumption, processing them, and enabling an
efficient use of available resources anytime; whereas the UAP works as interface between the
PS-BI and the UII, enabling the provision of a whole set of value-added energy efficiency,
comfort monitoring and optimization services based on the user profile and the information
provided by the PS-BI.
In addition, the Information System may implement a standard interface to allow a system
running in a given neighborhood to interact with other relevant stakeholders of the electricity
market, such as aggregators, DSOrs (Distribution System Operators) or TSO (Transmission
System Operators), so that it may be managed and coordinated with other neighborhood grids,
thus providing interoperability and scalability.
3.2.4 Neighborhood domain
The Neighborhood Domain represents the core M2M communications infrastructure which
allows controlling, monitoring, and managing such a high volume of generation and
consumption devices spread over a wide area remotely. Therefore, it is responsible for carrying,
in a reliable manner, data coming from the Building Domain to the Information System Domain
and control and management commands going from the Information System Domain to the
Building Domain.
For the sake of scalability and flexibility, this core communications infrastructure
comprises two network segments. In the first hierarchical level, a group of ADR EPs are
managed by the so-called CNTR (Concentrator). In the second hierarchical level, the network of
CNTRs is managed by the so-called M2M Platform.
The CNTRs perform management tasks such as ID assignment, IP assignment, or
enabling/disabling ADR EPs. In addition, the CNTRs aggregate and forward the data coming
from the ADR EPs to the Information System through the M2M Platform and route the
commands coming from the Information System to the appropriate ADR EP. In this way, the
problems reported in [Mao2011], which were discussed in chapter 2, are avoided.
- 50 -
Chapter 3 – Network Architecture
The M2M Platform works both as OSS (Operation Support System) and as
communications gateway. On the one side, the M2M Platform is responsible for performing
typical OSS tasks, such as network inventory, network components configuration, fault
management, and service provisioning. On the other side, it forwards all the data coming from
the CNTRs to the Information System and routes the commands coming from the Information
System to the appropriate CNTR.
3.3 Relation to standardization activities
This section explains the relationships of the proposed system and M2M communications
architecture with the standardization work presented in chapter 2.
Firstly, taking into account the main goals and functionalities of the system explained in
section 3.2 and the summary of the domains defined in the NIST Smart Grid conceptual model
shown in the table 2-1 of chapter 2, the scope of the proposed system is bounded to the
Customer, Distribution, Service Provider, and Operations domains, as Figure 3-4 illustrates.
Figure 3-4 – Mapping of the scope of the target platform onto the NIST Smart Grid conceptual model
Secondly, the proposed M2M communications architecture can be mapped onto the CTIAP (Communications Technologies – Interoperability Architectural Perspective) of the overall
IEEE 2030 SGIRM (Smart Grid Interoperability Reference Model) [IEEE2011] as follows:

The I-BECIs and the I-BEGIs represent the HANs (Home Area Networks) in IEEE 2030
terminology. The ADR EP provides the functionality of ESI (Energy Service Interface).

The communications segment comprising the ADR EPs and the CNTRs represents the
NAN (Neighborhood Area Network) of the M2M communications infrastructure, as defined
in the CT-IAP of the IEEE 2030 SGIRM.

The communications segment composed by the CNTRs and the M2M Platform represents
the Backhaul of the M2M communications infrastructure, as defined in the CT-IAP of the
IEEE 2030 SGIRM.
- 51 -
Chapter 3 – Network Architecture
Figure 3-5 graphically identifies the HAN, NAN, and Backhaul network in the proposed
M2M communications architecture. It should be noted that the remainder of this dissertation
will refer to the IEEE 2030 terminology.
NAN
HAN
Backhaul
BUILDING DOMAIN
NILM
COMFORT
SENSORS
ADR
EP
NEIGHBORHOOD DOMAIN
PLUGS
M2M
Platform
INFRARED
BOX
I-BECI
INFORMATION SYSTEM DOMAIN
PS-BI
USER DOMAIN
UAP
UII
CNTR
DER
ENERGY
STORAGE
ADR
EP
External
Interface
802.15.4/ZB
EXTERNAL GRID
(DSO, TSO, ND)
802.11 (Wi-Fi)
SENSORS
(WS)
GPRS/EDGE
I-BEGI
H/NAN (Home/Neighborhood Area Network)
Figure 3-5 – Mapping of the proposed M2M communications architecture onto the IEEE 2030 SGIRM
Regarding the standardization work carried out by the ESOs (European Standardization
Organizations), the comparison is focused on the work developed by ETSI on M2M (recently
transferred to the partnership project OneM2M [OneM2M2014]), since the conceptual model
and the SGAM (Smart Grid Architectural Model) defined by the SG-CG (Smart Grid –
Coordination Group) take into account and present many similarities to the already considered
standardization work. Figure 3-6 graphically shows the relationship between the M2M domains
defined by ETSI and the domains presented in section 3.2.
M2M Device Domain
Network Domain
Application Domain
BUILDING DOMAIN
NILM
COMFORT
SENSORS
ADR
EP
NEIGHBORHOOD DOMAIN
PLUGS
M2M
Platform
INFRARED
BOX
I-BECI
INFORMATION SYSTEM DOMAIN
PS-BI
UAP
USER DOMAIN
UII
CNTR
DER
ENERGY
STORAGE
ADR
EP
External
Interface
802.15.4/ZB
802.11 (Wi-Fi)
EXTERNAL GRID
(DSO, TSO, ND)
SENSORS
(WS)
I-BEGI
GPRS/EDGE
Figure 3-6 – Mapping of the proposed M2M communications architecture onto the ETSI M2M domains
Our M2M communications architecture can be also mapped onto the ETSI M2M
architecture applied to the Smart Grid presented in [Lu2012]. The SANs within the I-BECI and
- 52 -
Chapter 3 – Network Architecture
I-BEGI can be seen as capillary networks at the Customer Domain. Regarding the remainder of
the M2M communications architecture, there are two possible options, as shown in blue in
Figure 3-7:

If the proposed platform were run by an Energy Service Provider, an additional nesting
level to represent the CNTR would be missing in the ETSI M2M architecture proposed in
[Lu2012].

However, if the proposed platform were run by a DSOr, the Customer Domain would be
embedded into the Distribution Domain, following the hierarchy of the electricity
infrastructure itself. Thus, the ADR EPs and CNTRs could be considered as ETSI M2M
GWs at Customer and Distribution Domains respectively, and our M2M Platform could be
seen as the ETSI M2M Core of the Distribution Domain.
M2M
Platform
M2M
Platform
CNTR
ADR EP
CNTR
Figure 3-7 – Mapping of the proposed M2M communications architecture onto the ETSI M2M
architecture applied to the Smart Grid [Lu2012]
3.4 Communications technologies
As it has already been mentioned, cost and power consumption are two key constraints to
be taken into account when deciding the most appropriate technologies for this kind of systems,
due to the huge number of devices they involve. Furthermore, some other features, such as data
rate or range, which depend on the specific requirements of each communication segment, are
also relevant and need to be considered.
Reflecting the outstanding importance of wireless communications in the distribution and
customer domains of the Smart Grid [NIST2011], this thesis is specially focused on them.
- 53 -
Chapter 3 – Network Architecture
3.4.1 Home Area Network
There is a plethora of communications standards for smart metering and sub-metering (i.e.,
HAN/HEMS) [Güngör2011]. In general, wireless solutions are preferred vis-à-vis wired ones,
since they reduce deployment and maintenance costs and allow higher flexibility. This is not the
case for PLC, since PLC does not require the deployment of additional infrastructure. However,
by the time the communications technology for this network segment was proposed, PLC was
not as technically mature as some of its market competitors, and thus it was not such a cost
competitive technology. In addition, PLC communications performance is very sensitive to
connecting/disconnecting electrical loads and to EMI (Electromagnetic Interference)
[Bartak2013].
The most relevant wireless technologies for this network segment are: Bluetooth (IEEE
802.15.1), Wi-Fi (IEEE 802.11), and IEEE 802.15.4 [Drake2010]. Bluetooth is a
communication technology designed to replace wires in the communication of multimedia
contents between devices. Although it is a quite inexpensive technology, it presents some
drawbacks, such as it provides limited range (very few meters) and too high data rates at the
expense of too high power consumption, which make it not appropriate for this kind of
applications1. IEEE 802.11 is a widely deployed and cheap technology. However, IEEE 802.11
is designed to transmit multimedia contents at high data rate. Therefore, its energy consumption
is optimized for transmitting and not for idling; whereas the monitoring sensors of the I-BECIs/I-BEGIs transmit few data from time to time and are in idle mode most of the time. As a result,
it is concluded that the most appropriate communication technology for this network segment is
IEEE 802.15.4 [IEEE2006], since it is defined to minimize power consumption and cost in
applications with low data rates and no latency constraints.
There are two main technologies that rely on PHY/MAC IEEE 802.15.4 standard:
6LoWPAN (IPv6 over Low power Wireless Personal Area Networks) and Zigbee. 6LoWPAN
is an open standard defined by the IETF (Internet Engineering Task Force) [IETF2007]. Its
main advantage is that it allows using IPv6 between wireless IEEE 802.15.4-compliant devices,
facilitating their integration into an IPv6-based Internet. However, in an IPv4- based Internet
scenario, such as the current one, this fact does not add much value to 6LoWPAN. In addition,
by the time this decision was made, 6LoWPAN was in the very early development phase.
Zigbee is an open industrial standard developed by the Zigbee Alliance [Zigbee2014]. It
had been around in the market for longer time than 6LoWPAN, so it was much more mature. As
a matter of fact, the most important hardware and appliance manufacturers, such as Freescale or
Texas Instruments, already commercialized Zigbee-capable chips and equipment. In addition,
the Zigbee Alliance had just launched the Zigbee SEP (Smart Energy Profile), which is
specially designed to meet energy efficiency scenarios’ requirements and it also incorporates the
IP into Zigbee (actually, it incorporates 6LoWPAN into the Zigbee stack). Therefore, the Zigbee
PRO feature set [Zigbee2007] is selected on top of 802.15.4.
3.4.2 Neighborhood Area Network
For the communication between the ADR EPs and the CNTRs, two technologies are
considered: IEEE 802.15.4/Zigbee (taking advantage of its mesh topology) and IEEE 802.11. In
this case, IEEE 802.11 is chosen over Zigbee since it clearly fits the requirements of this
communication segment better (i.e., IEEE 802.11 provides higher bandwidth and coverage). In
addition, IEEE 802.11 is considered as a potential cost-effective solution for the NAN due to its
massive use.
1
This analysis does not take Bluetooth Low Energy into account, since Bluetooth 4.0 was not released by
the time it was carried out.
- 54 -
Chapter 3 – Network Architecture
Since it is a quite static and star topology network, the most appropriate IEEE 802.11
working mode is the infrastructure mode, the CNTRs working as APs (Access Points) and the
ADR EPs being the clients.
3.4.3 Backhaul
For the communication between the CNTRs and the M2M Platform, the following
communication technologies were considered. ADSL (Asymmetric Digital Subscriber Line) is
considered not to be appropriate due to neighborhood’s network infrastructure installation
requirements. The LTE (Long Term Evolution), HSPA+ (High Speed Packet Access) and
WiMAX (Worldwide Interoperability for Microwave Access) technologies are not considered
appropriate mainly because they are not widely deployed yet and so they are not mature enough.
HSxPA (WCDMA - Wideband Code Division Multiple Access) presents some advantages, e.g.,
it offers increased peak data rates and reduced latency. However, its main problem is that it
requires some enhancements inside of the infrastructure that were not yet fully spread in Europe
by the time this analysis was performed. Therefore, GPRS/EDGE is considered the most
appropriate technology for this communication segment because it is widely deployed and
mature, so that cost is kept low and interoperability is boosted. In addition, GPRS/EDGE is
considered to be the available technology that fits the requirements of this kind of systems in
terms of data rate better.
The communication between the M2M Platform and the Information System will be based
on broadband wired technologies (e.g., fiber). Indeed, these pieces of equipment may be located
in the same site.
3.5 Security
Security and privacy are two key challenging issues for this kind of systems
[McDaniel2009], [Liu2012]. As a result, the main security features of the selected
communication technologies are outlined next, proposing also the most appropriate solution in
each case.
3.5.1 Home Area Network
As it has already been mentioned, Zigbee specification only addresses the network and
application layers. Although latest versions of Zigbee support several options at lower layers,
we assume IEEE 802.15.4 at PHY and MAC layers. Therefore, some IEEE 802.15.4 security
features and issues are listed first [Sastry2004]:

AES (Advance Encryption Standard) link layer security for authentication or encryption of
IEEE 802.15.4 frames is provided. AES is a block cipher operating on blocks of fixed
length (128 bits, in this case). To encrypt longer messages, several modes of operation may
be used. The earliest modes described, such as ECB, CBC, OFB and CFB, provide only
confidentiality, but they do not ensure message integrity. Other modes, such as the CCM*
mode, do ensure both confidentiality and message integrity.

Key management is not specified.

A sequential number (the so-called Sequential Freshness) is used to prevent from insertion
attacks.
- 55 -
Chapter 3 – Network Architecture

Integrity is checked by using a Hash code (the so-called MIC - Message Integrity Code).
Zigbee allows choosing whether to use it or not. It also allows to configure the length of
such a Hash code (32, 64 and 128 bits), being 64 bits the default MIC length.
Regarding Zigbee itself, Zigbee-2007 defines two different feature sets: Zigbee and Zigbee
PRO. In general, both Zigbee and Zigbee PRO provide:

Authentication and encryption at network and application layers by using an AES-128
symmetric key. To be more precise, authentication is based on symmetric keys and Trust
Center. The Trust Center is in charge of authenticating devices that want to join the
network, managing and distributing the symmetric keys and activating point-to-point
security between devices.

Network layer security for network command frames (route request, route reply, route error)
and application security for APS (Application Support Sub-layer) frames.

Freshness by using frame counters.

Message integrity.
Zigbee PRO defines two additional security modes [Gislason2008]:

Standard Security. This security mode is compatible with the basic Zigbee feature set, so
Zigbee and Zigbee PRO devices fully interoperate if this security mode is used. It is focused
on avoiding devices that do not belong to a given Zigbee network to access it. In order to do
so, every single message exchanges within the Zigbee network is encrypted using AES-128
symmetric keys which are only known by the devices that are authenticated in the network.
It uses two different sorts of keys: a Network Key, which is used for all network commands
from any device and for APS messages, and Link Keys, which are used for each pair of
communicating devices. The Network Key could be either provided (in clear) by the Trust
Center or programmed on the device. Figure 3-8 illustrates how Standard Security mode
works.
Factory or
Out-of-band installed
Unsecured
key-transport
NWK Key
Key-Transport Service
Basis of security between
two (or group of) devices
Unsecured key-transport
of NWK key
NWK Key is used
as basis of
security services
Authentication Service
Secure authentication that
a device shares a NWK key
Framework Security
Service
Secure all frames
(except key-transport)
Figure 3-8 – Zigbee PRO Standard Security Mode [Gislason2008]
- 56 -
Chapter 3 – Network Architecture

High Security. This security mode is only available for the Zigbee PRO feature set. It
prevents both from external and internal attacks. In order to do so, every single message
exchanges within the Zigbee network is encrypted using a key which is only known by the
sender and the receiver of the message. It uses the two keys explained in the Standard
Security Mode (i.e., Network Key and Link Keys) together with an additional Master Key.
The Master Key could be either provided (in clear) by the Trust Center or programmed on
the device. Figure 3-6 illustrates how High Security Mode works.
Unsecured
key-transport
Factory or
Out-of-band installed
Master Key
Basis for long-term security
between two devices
SKKE protocol
Secured key-transport
of Master key
Framework Security
Service
Link Key/NWK Key
Basis of security between two
(or group of) devices
Link key is used as basis of security services
Secure all frames
(except key-transport)
Secured key-transport
of ‘group’ link keys
Key-Transport Service
Authentication Service
Key-Transport Service
Secured key-transport of
‘master’ keys
Secure authentication that
a device shares a link key
Secured key-transport of
‘group’ link keys
Figure 3-9 – Zigbee PRO High Security Mode [Gislason2008]
Taking into account the relevance of the information managed by this kind of systems, the
Zigbee PRO High Security Mode is recommended, despite of the fact that it penalizes
backwards compatibility. In addition, following the same reasoning, it is strongly recommended
to use Master Keys factory installed on the devices, since this avoids the key distribution
mechanism and so reduces dramatically the vulnerability of the wireless network.
3.5.2 Neighborhood Area Network
IEEE 802.11 considers three different layer-2 encryption systems:

WEP (Wired Equivalent Privacy) is based on RC4 (Rivest Cipher 4). It allows using 64
bits-length keys or 128 bits-length keys, providing the latter higher level of security than the
former. Nowadays WEP encryption can be broken in just few minutes by software.
Therefore, it does not meet the security requirements of this kind of systems.

WPA (Wi-Fi Protected Access) introduces enhancements to WEP. Although data is still
ciphered using RC4 and 128 bits-length keys, WPA allows changing such keys dynamically
over the time (by means of the so-called TKIP - Temporal Key Integrity Protocol). WPA
with TKIP uses 48 bits IV (Initialization Vectors) to generate such keys, so the possibilities
of collecting sufficient number of 802.11 frames to crack the encryption are reduced. WPA
also allows using an Authentication Server (namely a RADIUS server) to distribute
different keys to each terminal in the network (using 802.1x/EAP – Extensible
Authentication Protocol). However, a less secure operation mode is also available, where all
the devices within the same Wi-Fi network share the same encryption key (PSK - PreShared Key). For the time being, some vulnerabilities have been already detected when the
shared key is too simple in PSK and in some messages exchange between the AP (Access
Point) and the terminals in TKIP.
- 57 -
Chapter 3 – Network Architecture

WPA2 (Wi-Fi Protected 2) fixes the vulnerabilities detected on WPA. It is fully compliance
with IEEE 802.11i standard and it represents the most secure encryption system currently
available for Wi-Fi networks.
Other available security solutions for Wi-Fi networks which are fully compatible with the
encryption systems described above are:

MAC filtering policies can be applied in Wi-Fi networks to allow only those devices whose
MAC addresses are well-known to access them. However, this mechanism is not very
effective since it can be easily cracked just by sniffing traffic for a while and then changing
the MAC address to such a well-known one (MAC spoofing).

In Wi-Fi networks working in infrastructure mode, AP can also be hidden in order to be
visible only for authorized devices.

Additional security at higher layers can be provided by setting VPN (Virtual Private
Networks) up [Khanvilkar2004].
In this case, IEEE 802.11 using WPA2 is recommended. MAC filtering or AP hiding can
be also applied to increase security. VPNs may be used to provide additional data protection at
network layer and above. However, the overhead as well as the increase of complexity (and its
potential impact on costs) introduced by this security mechanism have to be carefully analyzed.
3.5.3 Backhaul
Connectivity through cellular technologies requires complex and expensive infrastructures
that are going to be provided by a given NO (Network Operator)/SP (Service Provider).
Therefore, the operator of the platform will have to pay to such a NO/SP for using this service
(i.e., cellular connectivity) and the NO/SP will be responsible for providing robust security
mechanisms that ensure –at least - confidentiality, integrity and authentication. In fact, security
mechanisms in cellular technologies are standardized by the appropriate standardization bodies
(namely, 3GPP – 3rd Generation Partnership Project) and they are heavier and more robust than
the ones provided by the rest of the considered wireless technologies.
However, due to the highly sensitive data this kind of systems manages, they cannot rely
solely on the security provided by a 3rd party, but they must guarantee the privacy and
confidentiality of the sensitive information it carries. Therefore, additional encryption
mechanisms at higher layers (e.g., VPNs [Khanvilkar2004]) may be implemented. Nevertheless,
this decision has to be analyzed in detail, since encryption means redundancy and so more bits
to be transmitted, which in turn in this case means more money to be paid to the NO/SP.
Therefore, a trade-off between security and cost is needed.
3.6 Address management and end-to-end
addressability
Due to the facts that M2M systems involve a huge amount of devices, that the deployment
of this kind of systems is growing really fast, and that predictions point they will grow even
faster, efficient numbering and addressing represents definitely a major challenge in M2M
communications. The main trends to tackle this problem are: using E.164 numbers and using IP
addresses [ECC2010].
E.164 is an ITU-T (International Telecommunications Union - Telecommunication
Standardization Sector) recommendation which defines the international public communication
- 58 -
Chapter 3 – Network Architecture
numbering plan used in the PSTN (Public Switched Telephone Network) and other data
networks. E.164 numbers are needed in practice in order to deliver M2M services on top of
existing mobile infrastructures which are not capable to support IP-based mobile access. This
family of addressing solutions is based on the phone number associated to the GPRS SIM
(Subscriber Identity Module) card (i.e., the MSISDN - Mobile Station Integrated Services
Digital Network). The MSISDN, according to ITU-T Recommendation E.164, has 15 digits and
is composed of three fields:

CC: Country Code

NDC: National Destination Code

SN: Subscriber Number
There are different proposals on how to use such fields for M2M applications, each one
having its pros and cons [ECC2010].
However, as long as mobile networks are being upgraded to embrace IP as common
network protocol, IP addresses seem to be the most likely addressing solution for M2M
applications in the medium to long run. IPv4 addresses, which are 32-bits long, are the closest
solution in time, since IPv4 is the most used network protocol nowadays. IPv6 addresses, which
are 128-bits long, will be the solution of the future, since they better fit the requirements of
M2M applications in terms of number of devices, but IPv6 is not widely adopted nor used in the
target kind of applications yet [López2013c].
In this thesis, we assume IPv4 as network protocol down to the ADR EPs, which are
responsible for routing incoming packets to the appropriate consumption/generation device by
mapping the IPv4 address associated to their IEEE 802.11 interface onto the address space of
their HANs. This is a quite common situation until IPv6 is widely adopted, since there are not
enough IPv4 addresses available to be assigned to the huge number of devices that emerging
M2M applications entail.
As a result, IPv4 does not provide E2E addressability, so an additional mechanism is
needed for this purpose. In this section, an addressing solution at application layer is proposed
to allow identifying univocally and addressing every single communications device of the
designed M2M communications infrastructure taking advantage of its hierarchical structure.
Anyway, this addressing solution is independent of the underlying network protocol, so it would
work on top of either IPv4 or IPv6.
In such an application-layer addressing solution, each communications device is assigned a
16-bits ID. The scope of this ID is bounded to the communications element which is right above
in the hierarchical communications architecture. Thus, each M2M Platform has a pool of 216
IDs, so it can address univocally up to 216 CNTRs; every single CNTR, in turn, can manage up
to 216 ADR EPs; and, finally, every single ADR EP can manage up to 216
consumption/generation devices. Therefore, each consumption/generation device is globally
univocally identified by the 8-bytes ID which results from concatenating its own 2-bytes ID
with its associated ADR EP, CNTR and M2M Platform IDs. Figure 3-10 illustrates this idea.
- 59 -
Chapter 3 – Network Architecture
HAN
0x0002
ADR EPs
CNTRs
INFORMATION SYSTEM
(BACK OFFICE)
M2M PLATFORM
...
0x0001
...
0x097A0001
0x0003
...
0x097A
0x012E097A0001
0x0001012E097A0001
...
0x012E
0x0001
Washing Machine Brand B Model M – MyLovelyWashingMachine
Global address
0x0001012E097A0001
Local address (within HAN 0x097A)
0x0001
Figure 3-10 – E2E addressability in the proposed addressing solution
The ID distribution and management as well as the routing mechanisms of the proposed
addressing solution are described in the next sub-sections.
3.6.1 ID distribution and management
Although one M2M Platform is assumed per neighborhood (as a logical entity, since
physically it is advisable to have more than one for reliability and backup purposes), a potential
operator of this platform may manage more than one neighborhood. In such a case, each M2M
Platform is assigned a 2-bytes ID, which is stored in the Information System together with its
associated IMSI (International Mobile Subscriber Identity). It should be noted that if the
operator managed a single neighborhood, it would be optional to use a 2-bytes ID to univocally
identify the M2M Platform.
Each M2M Platform is equipped with a GPRS SIM card which contains the IMSI and a
128-bits authentication key. These two parameters represent the golden key that allow
authenticating and univocally identifying each M2M Platform in the very beginning, i.e., at
least, the first time it is switched on (1). If this procedure is successful (2), the Information
System assigns the M2M Platform its 2-bytes ID and an IP address (3). The 2-bytes ID is fixed
and it will never change. The IP address is assigned dynamically, so it may change. Figure 3-11
illustrates this mechanism.
INFORMATION SYSTEM
(BACK OFFICE)
M2M PLATFORM 1
(1) (IMSI1, K1)
M2M PLATFORMs
WHITELIST
(IMSI, K)
ID
(IMSI1, K1) 0x0001
...
(2)
(IMSIn, Kn) 0x000B
(3) (0x0001, 192.168.1.2)
(IMSI, K)
ID
@ IP
(IMSI1, K1) 0x0001 192.168.1.2
Figure 3-11 – M2M Platforms ID distribution and management
When deploying this kind of platform on a large scale, the CNTRs are provisioned in the
appropriate M2M Platform in advance, i.e., the system administrator provides the M2M
Platform with a table that contains the IMSI and 128-bits keys of all the CNTRs which are
- 60 -
Chapter 3 – Network Architecture
planned to be deployed as well as their associated 2-bytes IDs (1). It should be noted that this
table is also stored in the Information System. When a new CNTR is switched on, it
authenticates itself in the appropriate M2M Platform using its IMSI and the 128-bits key (2). If
this procedure is successful (i.e., it is not a fake CNTR) (3), the M2M platform assigns the
CNTR its 2-bytes ID and an IP address (4). Again, the 2-bytes ID is fixed and it will never
change, unless the CNTR is not valid anymore. However, the IP address is assigned
dynamically, so it may change from time to time or after rebooting the CNTR. Figure 3-12
illustrates such procedure.
CNTR1
INFORMATION SYSTEM
(BACK OFFICE)
M2M PLATFORM 1
CNTRs WHITELIST
(2) (IMSIm,Km)
(1)
(IMSI, K)
ID
(IMSIm, Km) 0x0001
...
(IMSIx, Kx) 0x012E
(4) (0x0001, 192.168.3.5)
ID
(3)
(IMSI, K)
ID
(IMSIm, Km) 0x0001
...
(IMSIx, Kx) 0x012E
@ IP
0x0001 192.168.3.5
Figure 3-12 – CNTRs ID distribution and management
If a single CNTR is to be installed (e.g., because a new geographical area needs to be
covered), the system administrator has to update the table of CNTRs in the appropriate M2M
Platform with this new entry. Otherwise, the new CNTR would try again and again without
success until this is done.
When deploying this kind of platform on a large scale, the ADR EPs are also preprovisioned in the appropriate CNTRs. Thus, the system administrator provides every single
CNTR with a table that contains the golden key to authenticate and identify univocally the ADR
EPs and their associated 2-bytes ID (1). The ADR EPs may be equipped with multiple
communication interfaces (e.g., IEEE 802.11, IEEE 802.15.4, IEEE 802.3, RS232, RS485 or
PLC). However, they use their IEEE 802.11 MAC address together with a key (which may be
pre-programmed on it) to authenticate themselves in the appropriate CNTR (2), because it is
mandatory for the ADR EPs to be equipped with an IEEE 802.11 for the communication with
the CNTR. If this procedure is successful (3), the CNTR assigns the ADR EP its fixed 2-bytes
ID and an IP address which may change dynamically (4). It should be noted that, since every
single ADR EP is associated to a given customer, in this case the Information System stores the
Customer ID together with all the communication nodes’ IDs which lead to its ADR EP (i.e.,
M2M Platform’s ID, CNTR’s ID, and ADR EP’s ID) (5). Figure 3-13 illustrates such a
procedure.
- 61 -
Chapter 3 – Network Architecture
ADR EP1
M2M PLATFORM 1 INFORMATION SYSTEM
(BACK OFFICE)
CNTR1
ADR EPs
WHITELIST
(1)
(2) (@MAC1,K1)
(@MAC, K)
(@MAC, K)
ID
(@MAC1, K1) 0x0001
...
(3)
(@MACy, Ky) 0x097A
ID
(@MAC1, K1) 0x0001
...
(@MACy, Ky) 0x097A
(5)
(4) (0x0001, 192.168.10.2)
ID
CUSTOMER #
@ IP
123456789
ADR EP
0x000100010001
0x0001 192.168.10.2
Figure 3-13 – ADR EPs ID distribution and management
Again, if a single ADR EP is to be installed (e.g., because a new customer is registered in
the system) the system administrator has to update the table of ADR EPs in the appropriate
CNTR with this new entry for the new ADR EP to get connected.
The procedure for registering new consumption/generation devices is sketched in Figure 314. First, the user has to register the new device in the platform (i.e., in the Information System)
through UIIs and UAP. The user will have to fill some information about the new device, such
as the type of device (e.g., washing machine, TV, air conditioning, photovoltaic panel), the
brand, the model, the location of the device within the house, and a human-friendly name, along
with a code which univocally identifies this device within the whole platform (1). This code will
be, in general, the physical/MAC address of the communication interface the device will use to
communicate with the ADR EP and it will be provided to the user attached to the device itself.
For additional security, this code may be the hash of the physical/MAC address of the device
and a random number.
Once the user has registered the new device in the Information System, the Information
System will provision this new device in the ADR EP associated to that user, i.e., the
Information System will send the new device’s code to the appropriate ADR EP (2). Then, the
new device can be turned on and start working without problems.
When the new device is turned on, first it authenticates itself in the ADR EP using its
associated code (3). Then, the ADR EP assigns a 2-bytes ID from its own pool of IDs to this
device (4) and it stores on its translation table not only the pair [code, 2-bytes ID] but also the
appropriate interface (5). Finally, the ADR EP advertises the pair [code, 2-bytes ID] to the
Information System (6).
- 62 -
Chapter 3 – Network Architecture
CONSUMPTION
END-DEVICE
CNTR1
ADR EP1
M2M
PLATFORM1
INFORMATION SYSTEM
(BACK OFFICE)
(1)
(2) CodeN
(2) CodeN
(2) CodeN
CUSTOMER # = 123456789
(MyWashingMachine, Miele, W864, CodeN)
(3) CodeN
(4) ID = 0x0001
(5)
Code
ID
Intrfc
CodeN
0x0001
1
(6) (CodeN, 0x0001)
(6)
(6)
(6) (CodeN, 0x0001)
CUSTOMER # = 123456789
(MyWashingMachine, Miele, W864, CodeN,
0x0001)
Figure 3-14 – Device ID distribution and management
3.6.2 Routing
When the data flow from the Building Domain to the Information System Domain/User
Domain (e.g., sensor reporting a periodic measurement), the procedure is straightforward since
every single device in the communication chain has only one option as NH (Next Hop). The
device sends the data to its associated ADR EP. The ADR EP concatenates its 2-byte ID to the
device’s ID as the source of the data and forwards them to its associated CNTR. The CNTR
concatenates its 2-bytes ID to the ADR EP’s and the device’s IDs, as the new source of the data,
and forwards them to the appropriate M2M Platform. Finally, the M2M Platform forwards the
packet to the Information System (namely, to the Back Office Server) [López2011c]. Figure 315 shows such a data flow.
ADR EPs
CNTRs
0x0004
M2M PLATFORM
...
...
0x097A
0x097A0004
@IP NH
...
...
0x012E
0x012E097A0004
@IP NH
INFORMATION SYSTEM
(BACK OFFICE)
0x0001012E097A0004
@IP NH
0x0001
Figure 3-15 – Forwarding of data in the uplink
When the data flow from the User Domain/Information System Domain to the Building
Domain, the procedure is a bit more complex. If, for instance, a given user wants to send a
request to a specific appliance, first, the Information System will traverse the tree associated to
this Customer ID, it will fetch the appropriate sequence of 2-bytes IDs which represent this
appliance univocally, and it will send the request to the appropriate M2M Platform, as it is the
NH. The M2M Platform will check its table that maps CNTR 2-bytes IDs with IP addresses, it
will fill the destination IP address properly and it will route the packet to the appropriate CNTR.
The CNTR will perform exactly the same procedure to route the packet to the appropriate ADR
EP. Finally, the ADR EP will check its table that maps device 2-bytes IDs with HAN-valid
addresses and it will route the packet to the appropriate appliance [López2011c]. Figure 3-16
illustrates this procedure.
- 63 -
Chapter 3 – Network Architecture
0x0001
ADR EPs
CNTRs
M2M PLATFORM
@M
AC
...
...
0x00010001
x
...
192.168.10.2
0x0001
@Native
ID
@MACx 0x0001
0x000100010001
0x0001000100010001
192.168.3.5
192.168.1.2
...
0x0001
Intrfc
ID
INFORMATION SYSTEM
(BACK OFFICE)
@ IP
1
0x0001 192.168.10.2
0x0001
ID
@ IP
0x0001 192.168.3.5
CUSTOMER # = 123456789
END-DEVICE = 0x0001000100010001
ID
@ IP
0x0001 192.168.1.2
Figure 3-16 – Routing of commands in the downlink
It should be noted that within this approach, the use of IPv4 public addresses (which
represent a scarce resource) is not required, but everything can be managed by means of IPv4
private addresses.
3.7 Conclusions
In this chapter we propose a M2M communications architecture to support energy
efficiency and integration of renewable micro-generation within the so-called energy-positive
neighborhoods of the Smart Grid, which represents one of the main contributions of this thesis.
The proposed M2M communications architecture comprises three network segments, for
the sake of flexibility and scalability. It is also a heterogeneous communications infrastructure
in that different communications technologies are combined to meet the specific
communications requirements of each network segment. In addition, it is compared with the
most relevant standardization work presented in chapter 2, identifying how they are related.
Reflecting the outstanding importance of wireless communications in the distribution and
customer domains of the Smart Grid [NIST2011], the proposed M2M communications
architecture is fully based on wireless communications technologies, such as
IEEE802.15.4/Zigbee, 802.11, and GPRS. Throughout the chapter, the most relevant security
features of such communications technologies are outlined and the most appropriate solution is
proposed in each case.
As an end to the chapter, we also propose an application-layer solution to provide E2E
addressability throughout the platform.
The performance of the proposed communications technologies in large scale scenarios is
evaluated by means of simulations in chapter 6. The security concerns brought in section 3.5.2
and 3.5.3 related to the impact of using VPN on the performance and operational costs of the
platform are also addressed in chapter 6.
- 64 -
Chapter 4
Formal Modeling
4.1 Introduction
As it has been shown in Chapter 3, the so-called smart energy-positive neighborhoods
represent complex infrastructures which involve huge number of devices of different nature and
with different functionalities. In addition, their communications capabilities enable delivering a
wide variety of services which bring into play a wide variety of interested parties or
stakeholders. The IT-IAP (Information Technology - Interoperability Architectural Perspective)
of the SGIRM (Smart Grid Interoperability Reference Model) defined in IEEE 2030
[IEEE2011] remarks upon the importance of well-defined data models in such complex
scenarios in order to:

Avoid data inconsistency and duplication.

Make exchange of data and updates of legacy software easier.

Enhance the ROI (Return On Investment) by enabling more applications to use the data and
improve their value through novel analytics.
Therefore, the main goal of this chapter is to formally define the vocabulary and taxonomy
and capture the engineering and business semantics of the domain of knowledge of the energy
efficiency platforms for energy-positive neighborhoods. To be more concrete, this chapter aims
Chapter 4 – Formal Modeling
to formally represent the main architectural entities and interfaces of energy efficiency
platforms for energy-positive neighborhoods, as well as their potential services and
stakeholders, and the relationships between them.
The ISO/IEC/IEEE 42010 [ISO2011] is an international standard for architecture
description of systems and software engineering, so it fits the initial aforementioned scope of
our target model. UML (Unified Modeling Language) is one of the ADL (Architecture
Description Language) considered by ISO/IEC/IEEE 42010. UML is an object modeling and
specification language widely used in software engineering. It is used to visually express use
cases, hierarchy and composition relationships or sequences of events, and it allows some
automatic programming code generation.
Ontologies and OWL (Ontology Web Language) represent another option to define our
target model. OWL is an ontology language based on RDF (Resource Description Framework)
[Manola2004] that allows the expression of classes and sub-classes, properties with their
domains and ranges, and other features such as symmetry, disjointedness, or transitivity (e.g., if
a is b and b is c, then a is c). Hence, OWL also allows formally representing real-world systems,
the architectural entities composing those systems, and the relationships between them, so it fits
the initial scope of our target model as well.
However, OWL presents some advantages compared to UML, namely [Gómez2007],
[Witte2007], [Pan2012]:

The expressiveness of OWL is higher, allowing robust specification of complex
relationships among structural entities.

OWL allows knowledge reusability.

OWL allows automated reasoning (i.e., inferring information from the existing knowledge
without it having to be explicitly expressed).

The so-called ontology queries allow software applications to load the ontology from the
OWL file dynamically while running. Therefore, if the ontology changes, there is no need to
recompile the application.
As a result, although both ISO/IEC/IEEE 42010 along with UML and OWL are equally
valid for the purpose of our target model, in this thesis we have chosen OWL due to its greater
potential, which in turn increases the impact of the thesis itself. As a matter of fact, OWL-based
ontologies are becoming an increasingly popular way of defining machine-readable data models
within the Smart Grid area [Grassi2011], [Santodomingo2012], [Wicaksono2012], in particular,
and the M2M area [Gyrard2013], in general.
The remainder of the chapter is structured as follows. Section 4.2 presents the ontology
proposed in this thesis. Section 4.3 illustrates our ontology through a specific use case. Finally,
section 4.4 compares the ontology proposed hereby with some related works, discusses how
future work can make the most out of our ontology, and draws conclusions.
4.2 Ontology description
Figure 4-1 shows the taxonomy of energy efficiency platforms for energy-positive
neighborhoods which results from our ontology. This taxonomy includes the hierarchical levels
of the system architecture presented in Chapter 3, from the domains down to the functional
blocks. However, in this case the functional blocks are particularized in devices, which are
classified depending on their nature. In addition, beside the main architectural concepts,
- 66 -
Chapter 4 – Formal Modeling
stakeholders, and services, a concept to represent different sites is considered with the aim of
making the ontology closer to implementation and real-world scenarios.
As it has been mentioned is section 4.1, the ontology is developed using OWL as modeling
language. From the three variants of OWL (on increasing order of expressiveness: OWL Lite,
OWL DL – Description Logic -, and OWL Full), OWL DL is used and FaCT++ [Tsarkov2006]
is used as reasoner. The free and open source OWL editor Protégé [Protégé2013] is used as
development environment. The Protégé Onto-Graph plug-in allows visualizing graphically and
in a hierarchical way the defined classes and properties. In addition, the Protégé editor allows
attaching descriptions of the defined classes and properties, so that all the information presented
hereby is available at the editor and can be consulted easily and in a human-friendly manner.
4.2.1 Classes
The classes that are defined in our model and that will be described throughout this section
are (see Figure 4-1):

Domain

Subsystem

Device

Site

Stakeholder

Service
- 67 -
Ontology
Subsystem
Domain
Building
Neighborhood
IS
Device
User
I-BECI
Site
House
I-BEGI
CNTR
Network
M2M
Platform
PS-BI
UAP
-68CNTR
GW
RCB
Campus
µGeneration
Instalation
UII
Communication
ADR EP
Flat
Stakeholder
BO Server Appliance
RC
EMV&R
RP
Sensing
& Control
Electrical
ESS
µGen
Source
IR GW
Relay
Actuator
Inverter
Sensor
Plug
Comfort
NILM
Service
WS
Figure 4-1 – Taxonomy of an energy efficiency platform for energy-positive neighborhoods
BP
LEP
Aggr DSOr
RA&CA
LM MEM DSO
Chapter 4 – Formal Modeling
4.2.1.1 Domain
This class represents the highest level of abstraction in the proposed hierarchical
architecture. As it was described in chapter 3, in this thesis we divide a typical energy efficiency
platform for energy-positive neighborhoods into the following domains:

Building domain;

Neighborhood domain;

Information System domain;

User domain.
The Building domain encompasses the consumption and generation infrastructures, along
with the SANs (Sensor and Actuator Networks) to monitor and control them, typically
associated to smart homes and buildings. The Neighborhood domain encompasses the core of
the M2M communications infrastructure, responsible for enabling the required bulk data
exchange. The Information System domain represents the intelligence of the platform from the
energy perspective. Finally, the User domain encompasses everything related to the interaction
of the stakeholders with the platform (e.g., presentation of data, sending of commands).
4.2.1.2 Subsystem
Going deeper into the proposed hierarchical architecture, the Subsystems are found. The
Subsystems encompass a group of functional blocks whose duties are tightly related and share
some common features. The I-BECI (In-Building Energy Consumption Infrastructure) and the IBEGI (In-Building Energy Generation Infrastructure) comprise a set of consumption and
generation devices respectively, along with their associated SAN and communications gateway
(so-called ADR EP – Automated Demand Response End Point). The CNTR (Concentrator)
Network encompasses a group of CNTRs that manage the access to the core of the M2M
communications infrastructure. The M2M Platform represents the “brain” of the entire M2M
communications network, working both as OSS (Operations Support System) and as gateway to
the PS-BI (Power-Saving – Business Intelligence). The PS-BI represents the “brain” of the
system from the energy perspective and includes extensible set of software services dealing with
energy use optimization. The UAP (User Application Platform) encompasses the set of pieces of
software that work as interface between the PS-BI and the UIIs (User Intuitive Interfaces),
adapting contents appropriately. Finally, the UIIs include the set of front-end applications that
interact directly with the stakeholders of the platform.
4.2.1.3 Device
This class represents the lowest level of abstraction in the proposed hierarchical
architecture. Thus, if Domain is the class closer to the conceptual model, then Device is the
class closer to the development and implementation issues (i.e., to the software and hardware).
According to their nature, the Devices are classified into the following categories:

Communication Device;

Electrical Device;

Sensing and Control Device.
- 69 -
Chapter 4 – Formal Modeling
Communication Device represents those devices whose main objective is to enable the
bidirectional communication between the Building Domain and the Information System and
User Domains. Two attributes are defined to all the Communications Devices: IP address and
16-bits ID (Identifier) [López2011a]. Note that the Protégé editor allows specifying the type of
every attribute, e.g., Boolean, Integer, Double, or String, which is the case for these two
attributes. The Communication Devices are:

ADR EP;

CNTR;

GW (Gateway);

BO (Back-Office) Server.
The functionality of ADR EPs and CNTRs has already been explained in Chapter 3. The
GW represents the implementation of the M2M Platform as a physical piece of equipment.
Therefore, the GW performs the typical tasks of OSS (Operation Support System), e.g., network
inventory, network components configuration, fault management, or service provisioning, and is
also responsible for enabling the bidirectional communication between the CNTRs and the BO
Server.
The BO Server represents the piece of software responsible for the communications within
the PS-BI. The BO Server is responsible for sending the commands generated by the business
and energy-related processes of the PS-BI to the GW and for receiving the data coming from the
GW and making them available to such processes (e.g., by storing them in a database).
Electrical Device represents the pure energy consumption and generation devices
considered as the most representative equipment in this kind of platforms. The Electrical
Devices are:

Appliance;

ESS (Energy Storage System);

µGeneration Source.
Some attributes that are defined as common to all the Appliances are: Average
Consumption, Standby Consumption, and Power (all them are Double). In addition, there are
also attributes associated with specific Appliances. For instance, ON/OFF (Boolean) and
Temperature Set Point (Double) are attributes of Air Conditioning and Heating. Note that these
two attributes are important for DR (Demand Response) events, allowing a coarse (ON/OFF)
and a fine (Temperature Set Point) adjustment. ON/OFF/Standby (Integer) is a specific attribute
of TV and DVD. In this case, Standby consumption is considered due to the aforementioned
importance of such a working mode in ICT (Information and Communications Technologies)
equipment when it comes to energy efficiency. ON/OFF (Boolean), Starting Hour and Deadline
(Integers) are attributes of Dishwasher and Washing Machine. In this case, Starting Hour and
Deadline represent two key parameters to perform load shifting by means of DR events without
compromising the agreed level of comfort.
The ESSs are key elements to handle efficiently the energy generated by µGeneration
Sources as well as to match generation with consumption in near-future scenarios. Some
attributes that are defined as common to all the ESSs are: Is Charging (Boolean), Is Releasing
Energy (Boolean), Storage Energy (Double), and Connection Power (Double). The most
relevant ESSs in near future scenarios are Battery and EV (Electric Vehicle) [Pang2012], the
- 70 -
Chapter 4 – Formal Modeling
latter having specific attributes such as Charging Rate (Double), Starting Hour, and Deadline
(Integers). Charging Rate deals with the charging mode of the EV (e.g., fast and slow); whereas
Starting Hour and Deadline are again key parameters for DR and V2G [Ma2012].
Some attributes that are defined as common to all the µGeneration Sources are: Is
Generating (Boolean), Rated DC Voltage and Current (Double), Rated AC Voltage and Current
(Double), and Power Factor (Double). As it was already pointed out in Chapter 3, Photovoltaic
Panels and µWind Engines are considered as the most extended µGeneration Sources in the
building sector [Jellea2012], [Ayhana2012].
Sensing and Control Device represents the Actuators and Sensors considered as the most
relevant ones for this kind of systems. Since the Sensing and Control Devices send data and
receive commands, they are also assigned an IP address and an ID as attributes. The considered
Actuators are: the Relay and the IR (Infrared) GW. The considered Sensors are: the WS
(Weather Station), the Comfort Sensor, and the NILM (Non-Intrusive Load Monitoring) Sensor
(also known as NIALM - Non-Intrusive Appliance Load Monitoring - in this specific domain
[Zeifman2011]). In addition, the Plug and the Inverter works both as an Actuator and as a
Sensor.
The Inverters receive commands to control their associated µGeneration Source. In
addition, they measure the energy generated by such a µGeneration Source and transmit it either
periodically or on-demand. The Plugs act on the power supply of the Appliances by cutting it
OFF or ON. In addition, the Plugs measure the electricity consumption of those appliances and
send it to the PS-BI, allowing accurate monitoring and abnormal behavior identification.
The NILM Sensor allows identifying (based on electrical signature) the appliances which
are running, even if they are not equipped with a Plug with sensor and communication
capabilities. The Comfort Sensors measure different environmental variables, such as
temperature, relative humidity or CO2 concentration, which are taken into account when
achieving energy savings without compromising the agreed comfort levels. The WS measures
variables related to weather conditions in order to provide the PS-BI with relevant parameters
for accurate energy generation forecast.
The IR GWs enable managing the IR-controlled appliances, such as HVAC, TVs or DVDs,
remotely. Finally, the Relays receive commands targeting their associated ESSs and acts on
them consequently, allowing controlling them remotely.
4.2.1.4 Site
This class aims at modeling the most relevant profiles of consumption and generation
infrastructures present in energy-positive neighborhoods. Flat represents a flat belonging to a
block of flats, so it only has a consumption infrastructure. House represents a house belonging
to a housing development, so beside the typical consumption infrastructure, it may have a
µgeneration infrastructure (e.g., small photovoltaic installation in the roof). RCB (Residential
Commercial Building) represents, for instance, a building of offices, the headquarters of a
company, or a shopping center. This kind of buildings may comprise many rooms, several
floors, a basement, a roof, and a facade. The RCB has a consumption infrastructure and are
assumed to always have a µgeneration infrastructure (e.g., BIPV – Building-integrated
Photovoltaics). µGeneration Installation represents an installation composed by Photovoltaic
Panels and/or µWind Engines. Finally, Campus represents a set of buildings, which are
managed or belong to the same entity (e.g., a university campus comprising several buildings).
- 71 -
Chapter 4 – Formal Modeling
4.2.1.5 Stakeholder
This class represents the user segments that may play a role in the energy-positive
neighborhoods.
The RC (Residential Consumer) is connected to the low voltage grid and only has energy
consuming devices. Their goals include increasing energy efficiency (which means reduction of
energy costs) while maintaining a certain level of comfort. They are also willing to participate in
a dynamic market, e.g., through DR programs, if properly motivated and rewarded.
The RP (Residential Prosumer) not only has energy consuming devices but also energy
producing equipment. Besides the goals mentioned previously for the RC, the RP are also
interested in taking maximum economic advantage of their production capabilities.
The BP’s (Building Prosumer) main objectives are: to manage their building with certain
contractually agreed service level at minimal cost; to contribute to a reduction of the
environmental impact of the building operation; and to integrate their BEMS (Building Energy
Management System) already deployed – if any – with the energy efficiency platform targeting
the whole energy-positive neighborhood.
The LEP (Local Energy Producer) manages a local microgrid, which may include
generation and storage equipment. Their goals include the selling of the produced energy, as
well as the operation of their energy production devices at minimal cost.
The DSOr (Distribution System Operator) task is to optimize the operation of the power
distribution network, which means to anticipate and prevent any abnormal situation possibly
resulting in electricity blackouts. The DSOr needs to control the power quality in the network by
balancing electricity generation and consumption at any time and is willing to reward Prosumers
who actively contribute to this goal. Although the balancing of electricity supply and demand
can be managed not only by DSOr but also by other entities, within the scope of our ontology
the DSOr is kept as the only user in charge of such operations.
The Aggr (Aggregator) is an intermediary that ensures services to the DSOr (e.g.,
integration of DER, DR), grouping contracts with individual consumers and managing them.
Aggr boosts scalability and increase the system impact by allowing a larger level of energy
consumption and production that can be regulated by the DSOr.
4.2.1.6 Service
This class models the services this kind of platforms will potentially offer through the
Information System (mainly based on PS-BI).
RA&CA (Remote Access and Control of Appliances) includes a group of services that will
allow creating an initial configuration of the network of energy consuming devices, turning a
selected device ON or OFF, or changing its properties (e.g., the HVAC system operating
temperature).
EMV&R (Energy Monitoring Visualization and Reporting) encompasses a range of
services that will provide users with near real-time/periodic/on-demand information about their
energy consumption or generation and the associated economic and environmental impact.
The LM (Load Management) category of service will allow users to specify individual
devices or groups of devices to be included in the automated load management program.
Consequently, those devices will be directly controlled by the platform, ensuring the
minimization of costs to the end users, by shifting loads to periods with lower tariffs.
- 72 -
Chapter 4 – Formal Modeling
The MEM (Microgrid Energy Management) category of services will provide the
appropriate Stakeholders with near real-time monitoring data from the consumption and
generation sites, at the local communities' and neighbourhoods' level, improving the balance
between generation and consumption in the microgrids.
The DSO (Distribution System Operation) category of services will provide DSOr with
near real-time information about distributed generation and consumption of electricity in a given
location, highlighting the deviations from the expected energy consumption behavior and
providing accurate short-term forecasts of the energy generation. This group of services will
also provide the DSOr with the required conditions to operate DR programs.
4.2.2 Properties
Table 4-1 summarizes the features of the defined properties, which are described next. It
should be noticed that sub-classes of a given class (i.e., parent class) inherit the properties and
attributes from it. Likewise, implicit logical rules hold by the concatenation of explicit
relationships (e.g., a Flat cannot have a Photovoltaic Panel, since Photovoltaic Panel is a sort of
µGeneration Source which is part of the I-BEGI and Flat is only equipped with I-BECI).
Property
Is Composed Of
Composes
Interacts With
Supports
Is Equipped
With
Is Owned By
Owns
Is Addressed To
Is Located In
Is Part Of
Communicates
With
Table 4-1 - Summary of the properties defined in our ontology
Domain
Range
Features
Expression
Domain
Subsystem
1:N
Domain → Subsystem
Subsystem
Domain
Inverse of Is Subsystem → Domain
Compose Of
{Subsys|Stkhld}
{Subsys|Stkhld}
Symmetric
{Subsys|Stkhld} ↔
{Subsys|Stkhld}
PS-BI
Service
1:N
PS-BI → Service
Site
I-BE{C|G}I
Site →
I-BE{C|G}I
Site
Stakeholder
Site → Stakeholder
Stakeholder
Site
Inverse of Is
Stakeholder → Site
Owned By
Service
Stakeholder
Service →
Stakeholder
{Device|Site}
Site
{Device|Site} → Site
Device
{Device|Subsystem}
Transitive
Device →
{Device|Subsystem}
Device
Device
Symmetric
Device ↔ Device
4.2.2.1 Is Composed Of
The domain of this property is the class Domain and its range is Subsystem. I.e., one
Domain is composed of one or more Subsystems. Thus, the Building Domain is composed of
the I-BECI and the I-BEGI; the Neighborhood Domain is composed of the CNTR Network and
the M2M Platform; the Information System Domain is composed of the PS-BI and the UAP;
and the User Domain is composed of the UII.
4.2.2.2 Composes
This is the inverse property of Is Composed Of, so its domain is the class Subsystem and
its range is Domain.
- 73 -
Chapter 4 – Formal Modeling
4.2.2.3 Interacts With
This is a symmetric property that models the interaction between Subsystems themselves
as well as Subsystems and Stakeholders. As for being symmetric, e.g., if the property
“Subsystem A interacts with Subsystem B” is defined, then the property “Subsystem B interacts
with Subsystem A” is also implicitly defined. In most of the cases, this property is defined
between Subsystems themselves. However, the Stakeholders can also interact with the UIIs.
4.2.2.4 Supports
This property models the fact that the considered Services rely on the PS-BI. As a result, its
domain is the PS-BI Subsystem and its range is the class Service.
4.2.2.5 Is Equipped With
The domain of this property is the class Site and its range covers the I-BECI and I-BEGI
Subsystems. It is used to specify if a given Site is equipped with one or more I-BECIs or with
one or more I-BEGIs, or both, which eventually will determine the potential owners of this Site.
Thus, it is considered that: a Flat is equipped with exactly 1 I-BECI and exactly 0 I-BEGI, since
a flat belongs to a block of flats and photovoltaic windows are not considered (although they
may be relevant in the mid-term once they reach a competitive price); a House is equipped with
exactly 1 I-BECI and with min 0 I-BEGI; a RCB is equipped with exactly 1 I-BECI and exactly
1 I-BEGI; a µGeneration Installation is equipped with exactly 0 I-BECI and with exactly 1 IBEGI; and a Campus is equipped with min 1 I-BECI and 1 I-BEGI, since it is assumed to be
composed of at least one building (i.e., RCB).
4.2.2.6 Is Owned By
This is a unidirectional property to state what Sites can be owned by which Stakeholders,
based on the energy consumption/generation infrastructures of the given Site.
4.2.2.7 Owns
This is the inverse property of Is Owned By. The RC may own a Flat or a House (if not
equipped with I-BEGI). The RP owns a House. The BP may own a RCB or a Campus. The LEP
owns a µGeneration Installation. The DSO cannot own any Site, but it actually manages and
controls the energy consumption and generation within a district composed of hundreds or
thousands of Sites. Finally, the Aggr does not own any Site either, but they manage the
consumption/production data associated to a group of Sites.
4.2.2.8 Is Addressed To
This is a unidirectional property to map the Services onto the Stakeholders. Table 4.2
shows such a mapping.
- 74 -
Chapter 4 – Formal Modeling
Table 4-2 - Mapping of the Services onto the Stakeholders they are addressed to
RC
RP
BP
LEP
Aggr
X
X
X
X
RA&CA
X
X
X
X
X
EMV&R
X
X
X
LM
X
X
X
X
MEM
DSO
DSOr
X
X
X
4.2.2.9 Is Located In
This is a unidirectional property to formally state that the Devices are located physically in
the Sites. In addition, some Sites may be located in other Sites (e.g., Flat may be located in
RCB, if it is a Residential Building, RCB may be located in Campus). As a result, the domain of
this property is Device or Site and its range is Site.
4.2.2.10 Is Part Of
This is a transitive property to formally bind the Devices to the Subsystems. Thus, the
Appliances, the Comfort Sensors, the NILM Sensor, and the IR GW are part of the I-BECI; the
µGeneration Sources, the ESS, and the WS are part of the I-BEGI; the ADR EP is part of both
the I-BECI and the I-BEGI; the CNTR is part of the CNTR Network; the GW is part of the
M2M Platform; and the BO Server is part of the PS-BI.
However, there are some special Devices, such as the Inverter, the Relay, and the Plug,
which are part of other devices, namely the µGeneration Source, the ESS, and the Appliance,
respectively. As for being transitive, if the Inverter is part of the Appliance and the Appliance is
part of the I-BECI, the Plug is part of the I-BECI.
4.2.2.11 Communicates With
This is a symmetric property between Devices with communication capabilities (i.e.,
Communication Devices and Sensor and Actuator Devices). Thus, the Sensors and Actuators
communicate with the ADR EP; the ADR EP communicates with the CNTR; the CNTR
communicates with the GW; and the GW communicates with the BO Server. As for being
symmetric, the communication is bidirectional.
4.3 Use Case
One of the additional advantages of using OWL and Protégé as editor is that they allow
defining not only the architecture in a form of a schema, but also creating instances of the
classes defined in this schema. This in turn allows for better understanding of the system in the
early design stages as well as for clear description of the use cases.
Thus, the main objective of this section is to illustrate the dynamics of this kind of
platforms by means of a use case derived from the presented ontology. The use case involves a
RP (John Residential Prosumer) who subscribes one of the services offered by the platform
from the category RA&CA.
Figure 4-2 shows the overall picture of the use case. First, it can be checked that the
developed ontology captures not only engineering issues but also business and humaninteraction aspects. Furthermore, the main Domains and Subsystems, as well as the relationships
between them, are clearly identified. It can be seen that John Residential Prosumer owns a
House (House_UC1). As his name suggests, John Residential Prosumer is a prosumer, so the
House_UC1 is equipped with a consumption infrastructure (I-BECI_UC1) and a µgeneration
- 75 -
Chapter 4 – Formal Modeling
installation (I-BEGI_UC1). The I-BECI_UC1 and the I-BEGI_UC1 are specific Subsystem
instances that compose the conceptual Building Domain of the use case. The I-BECI_UC1 and
the I-BEGI_UC1 interact with the Concentrator_Network_UC1, which in turn interacts with the
M2M_Platf_UC1. Again, the Concentrator_Network_UC1 and the M2M_Platf_UC1 are
specific Subsystem instances that compose the conceptual Neighborhood Domain of the use
case. The M2M_Platf_UC1 interacts with the PS-BI_UC1, which in turn interacts with the
UAP_UC1, both Subsystem instances composing the conceptual Information System Domain of
the use case. Finally, the UAP_UC1 interacts with the UII_UC1 (that compose the conceptual
User Domain of the use case), which interacts with John in order to allow him to monitor and
control his in-house devices.
Figure 4-2 – Overall picture of the use case
Note that if the use case considered many Stakeholder instances (e.g., hundreds of RPs),
the conceptual Building Domain would be composed of many I-BECI and I-BEGI instances and
the conceptual User Domain would be composed of many UII instances; whereas the conceptual
Neighborhood and Information System Domains would be still composed of one instance of the
appropriate Subsystems. This is because the Subsystems that compose the conceptual
Neighborhood and Information System Domains represent a common infrastructure
(communications infrastructure and IT infrastructure respectively) shared within a given district
running on this platform. Thus, in a use case that considered two different districts running on a
platform like this (either managed by the same DSOr or not), there would be two different
instances of the Neighborhood and Information System Domains, and so the Subsystems that
compose them would be also duplicated.
Figure 4-3 shows the elements that compose the I-BEGI_UC1 and the I-BECI_UC1 as well
as how they communicate with each other. The I-BEGI_UC1 represents a small photovoltaic
installation comprising the Solar_Panel_UC1 and the ADR_End_Point_G_UC1. The
Solar_Panel_UC1 incorporates the Inverter_UC1, which allows monitoring and controlling it
and is responsible for communicating with the ADR_End_Point_G_UC1. The I-BECI_UC1 is
composed of the Infrared_GW_UC1, the Air_Conditioning_UC1, the Washing_Machine_UC1,
and the ADR_EP_C_UC1. The Infrared_GW_UC1 communicates with both the
ADR_End_Point_C_UC1 (e.g., to receive commands) and the Air_Conditioning_UC1 (e.g., to
forward the received commands to the appliance). The Washing_Machine_UC1 incorporates the
Plug_UC1, which allows monitoring its consumption and controlling it and is responsible for
- 76 -
Chapter 4 – Formal Modeling
communicating with the ADR_End_Point_C_UC1. Both the ADR_End_Point_C_UC1 and the
ADR_End_Point_G_UC1 communicate with the Concentrator_UC1, which is part of the
Concentrator_Network_UC1.
Figure 4-3 – Instances of the I-BECI and I-BEGI in the use case
Finally, Figure 4-4 shows how the different elements of the architecture interact (at
different level of abstraction) in order to enable the service. Let us assume that John Residential
Prosumer wants to know whether he forgot to turn the air conditioning off or not. In order to do
so, John will send a query to the Air_Conditioning_UC1 and it will receive an answer. At a high
level of abstraction, John will interact with his UII_UC1, which in turn will interact with the
UAP_UC1, and so on and so forth until the query reaches the I-BECI_UC1. The interaction will
be the other way around when the answer comes back to John. At a low level of abstraction, the
BO_Server_UC1 will send the query to the Gateway_UC1, which in turn will route the query to
the Concentrator_UC1, which in turn will route the query to the ADR_End_Point_C_UC1,
which will route the query to the Infrared_GW_UC1, which will finally forward it to the
AC_UC1 (see Figure 4-4). The answer will follow the opposite direction until the
BO_Server_UC1, although in the uplink the Communication Devices will always forward it
instead of routing it, since the next hop is well-known. The BO_Server_UC1 will deliver the
answer to the PS-BI_UC1. The PS-BI_UC1 and the UAP_UC1 will process it and the result will
be shown to John through his UII_UC1. If he checks that the Air_Conditioning_UC1 is on,
John will be able to send a command to turn it off and he will receive an acknowledgement,
following the same aforementioned procedure.
- 77 -
Chapter 4 – Formal Modeling
Figure 4-4 – Use case interaction at different levels of abstraction
Many other use cases can be derived from current version of the ontology, illustrating, e.g.,
the dynamics of different services or the end-to-end addressing solution presented in Chapter 3
in large-scale scenarios [López2012a]. The defined ontology will ensure that the defined rules
hold in such use cases, avoiding incongruities.
4.4 Conclusions
There are some recent research works that use ontologies to improve energy efficiency in
buildings and households, encouraged by the fact that energy efficiency has become a
mandatory requirement in buildings as well as by the fact that NZEB (Nearly Zero-Energy
Buildings) and self-consumption seem to be the future in this sector. The works reported in
[Wicaksono2012] and [Grassi2011] are the most tightly related to our ontology.
Reference [Wicaksono2012] combines a Building Automation System with an OWL
ontology to improve energy efficiency in buildings. However, µgeneration facilities are not
considered. Thus, the services offered by the Building Automation System are mainly related to
RA&CA and EMV&R. The ontology is generated in two stages. The base ontology representing
terminological knowledge in building automation is created manually by experts. This fits the
approach of our ontology, which is focused on rendering the shared vocabulary and taxonomy
of energy efficiency platforms for energy-positive neighborhoods and is also generated
manually. Such a base ontology can be automatically extended through the interpretation of 2DAutoCAD-drawings of the buildings and data mining approaches. Novel knowledge-driven
energy analysis based on the ontology is used to understand energy usage patterns and notify
users about any energy inefficiency.
Reference [Grassi2011] proposes the definition of an OWL-based ontology framework for
modeling several aspects such as operative scenario, context, QoS (Quality of Service), user
preferences, and energy production and consumption in a unique global knowledge base used to
support the implementation of efficient control logics. Five ontologies are defined, namely the
Context Ontology, the Service Ontology, the User Ontology, the Device Ontology, and the
Energy Ontology, all them being covered by our ontology. However, [Grassi2011] is focused
just on households, although extending the scope of the ontology to wider application scenarios,
- 78 -
Chapter 4 – Formal Modeling
such as the ones considered in this thesis (i.e., energy-positive buildings and districts), is pointed
out as future work due to their higher potential.
Therefore, the ontologies presented in [Wicaksono2012] and [Grassi2011] are mainly
focused on buildings and [Wicaksono2012] does not consider micro-generation; whereas our
ontology represents a holistic approach to energy efficiency in buildings, formally defining the
vocabulary and taxonomy and capturing the engineering and business semantics from the
energy-positive neighborhood perspective.
The ontology developed in this thesis was applied in the EU FP7 project ENERsip
[López2012b], bringing many benefits to effectively manage all the phases of the project life
cycle. First, it allowed defining the common terminology of the project. This is especially
important in Smart Grid related projects, since they involve engineers coming from different
fields with different background (e.g., ICT and energy), and it becomes even more important if
such projects involve medium to large development teams from different countries. Secondly,
this ontology allowed wrapping the platform specifications up in a single model which can be
graphically visualized. As a result, it also served as a valuable reference during the development
and validation phases of the project. In addition, the obtained formal model facilitates sharing
information and knowledge with projects and standardization bodies working on the same or
closely related topics.
Our ontology has been also made public through the EC (European Commission)
eeBuildings Data Model community (also called as eeSemantics), so that other researches can
re-use it and further improve it. Our ontology can be extended and improved and can be useful
to other researches in many different ways. An inexperience researcher in this area can use it to
get started and get a clear idea of its main elements and the relationships among them at a
glance. It can be also taken as basis for developing software to simulate and evaluate either the
performance of the platform as a whole or the performance of some specific sub-system or
domain (e.g., the Building Domain in self-consumption scenarios) [Stecher2008], [Anjum2012].
It can bring context to generic services or applications, so that they can run seamlessly over the
modeled domain (or a part of it) once they are connected to the ontology and understand it. In
addition, services and applications developed based on our ontology (or connected to it) can
also take advantage of typical software maintenance tasks, such as architectural evolution,
which are supported by ontologies through ontology queries and DL reasoning [Gómez2007],
[Witte2007].
As a matter of fact, our ontology has been already proposed, with other related works, as
starting point for the study on “Available semantics assets for the interoperability of SMART
APPLIANCES. Mapping into a common ontology as a M2M application layer semantics”,
launched as invitation to tender by the EC in June 2013 [EC2013].
Finally, within the scope of the standard IFC (Industry Foundation Classes) data model for
data sharing in the construction and facility management industries [ISO2013], it is currently
being considered to convert IFC architecture to OWL. If this is finally the case, our ontology (or
part of it) could be considered to be reused and included in such a standard.
- 79 -
Chapter 4 – Formal Modeling
- 80 -
Chapter 5
Practical Modeling
5.1 Introduction
Choosing the most suitable communications technologies for a given combination of
communications architecture and application represents a challenging task in Smart Grid
scenarios, due to the wide range of available options and to the specific requirements from both
technical and economic perspectives that need to be met [Güngör2013], [Yan2013], [Liu2012].
As a result, effective methods to evaluate and compare how such available communications
technologies meet these specific requirements are crucial in order to select the most appropriate
ones before undertaking the important investments needed to deploy this kind of infrastructures
on a large scale.
This evaluation may be carried out by deploying real pilot schemes involving high volume
of devices. Although this approach may yield the most accurate results, it implies high costs and
lacks of flexibility, in that only the deployed technologies can be evaluated. However, such
evaluation can be also approached by means of simulations, which represent a powerful,
flexible and cost-effective solution to achieve the same goal.
A crucial consideration when it comes to simulations is that they are based on a model of
the actual issue under study. Therefore, the better that model fits the actual issue, the more
relevant and meaningful the results obtained from the simulations will be.
Chapter 5 – Practical Modeling
This chapter is focused on modeling real world scenarios of the energy-positive
neighborhoods of the Smart Grid which allow obtaining meaningful results from potential
works based on them. The model considers near real-time bidirectional communications in
realistic scenarios. In addition, in order to maximize the impact on the Smart Grid area, it not
only considers current or short-term scenarios, but also foreseen medium to long-term ones.
Thus, the conclusions from potential works based on this model will be valid for a longer period
of time and they will allow making appropriate decisions in advance [López2012a].
The model presented in this chapter is mainly based on data from actual power distribution
infrastructures and it is customized for the EU FP7 project ENERsip, since it was taken as
reference in part of the comprehensive evaluation plan of this project [López2011b]. However,
the characterized scenarios are valid for any energy-positive neighborhood and can be easily
adapted just by suitably tuning the identified parameters.
The remainder of the chapter is structured as follows. Section 5.2 outlines the methodology
we propose to properly model any communications overlay which works on top of an
infrastructure devoted to any purpose. Section 5.3 presents the context and scope of the model
developed in this thesis along with the characterized scenarios. Finally, section 5.4 draws
conclusions highlighting for what purposes this model can be used, in general, as well as how it
is used in this thesis.
5.2 Methodology
The methodology followed throughout this chapter can be applied to characterize in
practice not only the communications infrastructure required to meet the requirements and
achieve the goals of the so-called energy-positive neighborhoods, but also any communications
overlay deployed on top of an infrastructure devoted to any purpose. The main steps of such a
methodology are:
1.
Map the communications overlay onto the underlying infrastructure. This step allows
extrapolating well-known information of the underlying infrastructure (e.g., how it is
physically or geographically deployed and organized) to the communications
infrastructure.
2.
Set the scope of the target model (i.e., which parts of the communications infrastructure are
going to be considered and which parts are not going to be considered).
3.
Identify parameters that can be relevant to properly characterize how the communications
infrastructure works.
4.
Cluster such parameters in relevant scenarios and quantify them given certain boundary
conditions. In our case, for instance, the model is bounded to the Portuguese power
distribution network (since it is based on data provided by EDP – Energias de Portugal)
and the ENERsip specifications and implementation.
As main outcome of applying this methodology, a set of more sophisticated scenarios
resulting from combining the basic ones is obtained. Each of those more sophisticated scenarios
models the communications infrastructure in realistic situations which are characterized by the
specific values the considered parameters take.
- 82 -
Chapter 5 – Practical Modeling
5.3 Communications infrastructure modeling
5.3.1 Context and scope
This section covers the steps 1 and 2 of the already presented methodology. First, the
proposed M2M communications architecture is going to be mapped onto the typical power
distribution infrastructure, which is shown in Figure 5-1.
Secondary Power
Distribution Network
Consumer
Substation
TP
Primary Power
Distribution
Network
TP
Prosumer
Producer
HV (High Voltage) lines (> 70 kV)
MV (Medium Voltage) lines (2kV – 70 kV)
LV (Low Voltage) lines (220V)
Figure 5-1 – Typical electricity distribution infrastructure
Such typical power distribution network is mainly composed of:

Customers, which in the case of energy-positive neighborhoods can be either Consumers or
Producers or Prosumers;

TPs (Transformer Points)1, which are dedicated to transforming the voltage supplied by the
medium voltage distribution grid into voltage values suitable for supplying low voltage
Customers (e.g., residential customers);

Substation, which is dedicated to transforming the voltage supplied by the high voltage
distribution grid, used to carry electricity throughout long distances, into voltage values
suitable for supplying medium voltage lines.
The TPs are responsible for supplying low voltage to clusters of Customers; whereas the
Substation is in charge of many TPs, and thus, of a high number of Customers. Thus, the scope
1
They are also known as Transformation Centers or Feeders.
- 83 -
Chapter 5 – Practical Modeling
of the TPs is bounded to a group of Customers; whereas the scope of the Substation is bounded
to big neighborhoods or small cities.
There is a clear correspondence between the designed M2M communications architecture
and the power distribution grid, as Figure 5-2 shows. Thus, the ADR EPs (Automated Demand
Response End Points) are associated to the Customers and the CNTRs (Concentrators) are
associated to the TPs. The M2M GW (Gateway) is logically associated to the Substation that
manages the target neighborhood. However, using a backhaul network allows the M2M GW to
be physically located at the Substation or wherever else the data centers of the entity operating
the system (e.g., DSOr – Distribution System Operator -, retail electric provider, aggregator,
ESCO – Energy Service COmpany) are.
As it is highlighted in continuous green line in Figure 5-2, our model will be focused on the
core of the designed M2M communications architecture, which comprises the wireless
communications segments from the ADR EPs to the M2M GW, i.e., the NANs (Neighborhood
Area Networks) and the Backhaul network as defined in the CT-IAP (Communications
Technologies – Interoperability Architectural Perspective) of the IEEE 2030 SGIRM (Smart
Grid Interoperability Reference Model) [IEEE2011]. The communications technologies
considered for these communications segments are IEEE 802.11 and GPRS (General Packet
Radio Service), respectively.
NAN (Neighborhood
Area Networks)
HAN (Home Area Networks)
Backhaul
Customers
NILM
PLUGS
INFRARED
BOX
Information
System
TPs
I-BECI
COMFORT
SENSORS
802.11
GPRS
RENEWABLE
GENERATION
NETWORK
ANALYZER
I-BEGI
WEATHER
STATION
CNTR
M2M GW
ADR EPs
Figure 5-2 – Mapping of the designed M2M communications architecture onto the power distribution
infrastructure
5.3.2 Characterized scenarios
This section covers steps 3 and 4 of the methodology presented in section 5.2. In order to
actually identify relevant parameters, first the traffic patterns of the M2M communications
infrastructure (both for the uplink and for the downlink) need to be characterized as statistical
distributions. Regarding the downlink, the users’ behaviors and the outputs of the decision
maker module of the Information System (i.e., the PS-BI – Power Saving Business Intelligence
- module) are considered to be random and memoryless, so they can be characterized as
exponential distributions. Consequently, the aggregated downlink traffic can be characterized as
Poisson distributions. Therefore, the following parameters need to be estimated in order to
properly model the downlink traffic:

µIS: Average periodicity of sending requests either by the users or automatically generated
by the PS-BI module.

SIS: Size of the application messages sent either by the users or automatically generated by
the PS-BI module (the size of the acknowledgement can be assumed to be the same).
- 84 -
Chapter 5 – Practical Modeling
Regarding the uplink, the consumption and generation data are sent periodically. Thus,
every uplink stream can be characterized as a uniform distribution during the first sending
period and, from then on, as a deterministic distribution with a given sending periodicity.
Therefore, the following parameters need to be estimated in order to properly model the uplink
traffic:

AC and AG: number of ADR EP-C (Consumption) and ADR EP-G (Generation) per CNTR,
respectively.

SC and SG: size of the consumption and generation data at the application layer,
respectively.

T: periodicity which ADR EPs send data with.

C: number of CNTRs per M2M GW.
It is worthwhile to remark upon the fact that downlink traffic may influence uplink traffic
by means of the acknowledgements. However, they can be decoupled in practice by using a
specific module at the ADR EPs – independent of the module responsible for generating uplink
traffic - responsible for echoing the received messages.
Several scenarios are considered in order to assign values to such parameters appropriately.
First, it is distinguished between Urban (U) and Rural (R) scenarios, since in the power
distribution grid the number of Customers/TP (Cust/TP), the number of TPs/Substation
(TP/Subs) and the maximum acceptable distance between Customers and TPs (Dmax) vary
remarkably between both. Based on the mapping presented in section 5.3.1, Cust/TP is related
to the number of ADR EPs/CNTR and TP/Subs is directly the number of CNTRs/M2M GW
(C). Dmax is relevant in order to figure out whether IEEE 802.11 coverage is enough or not.
Finally, for the sake of comprehensiveness, the minimum density of Customers per TP (dmin),
computed as shown in (1), may be also of interest.
⁄
[
⁄
]
Table 5-1 summarizes the values of such parameters for each of these scenarios.
Table 5-1 - Main parameters for Urban and Rural scenarios
Parameter
Urban
360 (1)
Cust/TP
500 (1)
Dmax (m)
2
458.36
dmin (Cust/km )
150 (1)
C
Rural
100 (1)
700 (1)
64.96
220 (1)
(1) Data provided by EDP
As it was already motivated in section 5.1, current or short-term (ST) scenarios and
medium to long-term (LT) scenarios are also considered. The main parameters that vary from
one of these scenarios to another are:

T: Periodicity of sending consumption and generation data of the ADR EPs.

Type of appliances.

Plugs/I-BECI: Number of appliances per I-BECI. For the purpose of the model, the I-BECI
architecture presented in Chapter 3 is simplified by assuming that every appliance is
monitored and controlled by a Plug.
- 85 -
Chapter 5 – Practical Modeling

SC and SG: size of the consumption and generation data at the application layer,
respectively.

Penetration of micro-generation ((I-BECI|I-BEGI)/Cust). This parameter varies also from
Urban to Rural scenarios. It will be always higher in the Rural scenarios than in the Urban
ones due to the type of dwellings (e.g., houses where photovoltaic panels can be installed in
the roof are more common in the former; whereas buildings of flats are more common in the
latter). It will be also higher in the long-term than in the short-term, since the penetration of
micro-generation and self-consumption is foreseen to increase in the forthcoming years.

AC and AG: number of ADR EP-C and ADR EP-G per CNTR, respectively. In this chapter,
it is assumed that there are independent communications gateways for the I-BECI (ADR
EP-C) and for the I-BEGI (ADR EP-G). Thus, AC is equal to Cust/TP, since it is assumed
that every Customer is equipped with an I-BECI; whereas AG is computed by multiplying
AC by the estimation of the penetration of micro-generation (assumed always < 1).
T is estimated as 15 minutes, for the short-term scenario, and 5 minutes, for the long-term
scenario. In addition, in order to evaluate somehow the impact of DR (Demand Response)
events on the communications infrastructure performance, it might be considered that a given
percentage of the overall ADR EP-C send information every minute (T = 1 minute) during a
given period of time (prior to the DR event). The latter assumption holds both for short-term
and long-term scenarios.
Regarding the number of appliances per I-BECI, in principle one appliance per type of
appliance is considered. The number of appliances per I-BECI as well as the type of the
appliances is determined by coupling those appliances with higher impacts on energy efficiency
and DR with those appliances which present higher ownership rate. Based on the data collected
in the project REMODECE (Residential Monitoring to Decrease Energy Use and Carbon
Emissions in Europe) [De Almeida2011], 5 appliances per I-BECI are considered for the shortterm scenario, such appliances being refrigerator, washing machine, dishwasher, air
conditioning, and water heater; whereas 10 appliances per I-BECI are considered for the longterm scenario, such appliances being refrigerator, freezer, washing machine, clothes dryer,
dishwasher, router, TV, DVD player, air conditioning, and water heater.
SG and SC are taken from the actual implementation of ENERsip. In the case of SG, the
following information is provided:

An Inverter sends 520 Bytes periodically.

A Counter (used to measure the energy generated by a given µGeneration Source) sends
150 Bytes periodically.

A WS (Weather Station) sends 360 Bytes periodically.
As it has already been explained throughout this dissertation, both the Inverter and the
Counter are associated to a µGeneration Source. The most relevant µGeneration Sources are
photovoltaic panels and µwind turbines. Since the penetration of photovoltaic panels is foreseen
to be much higher than the penetration of µwind turbines, the configuration of the I-BEGI
considered for the short-term scenario comprises 1 photovoltaic panel installation (i.e., 1
Inverter and 1 Counter) and 1 WS. However, the configuration of the I-BEGI considered for the
long-term scenario comprises 1 photovoltaic panel installation, 1 µwind turbine installation (i.e.,
2 Inverters and 2 Counters) and 1 WS. As a result, assuming that the ADR EP-G aggregates the
outgoing traffic, the following values are obtained:

SG|ST = 1030 Bytes
- 86 -
Chapter 5 – Practical Modeling

SG|LT = 1700 Bytes
In the case of the I-BECI, it is assumed that the sampling periodicity in the Plugs is the
same as the sending periodicity in the ADR EP-C (i.e., one sample per plug is sent by the ADR
EP-C at a time) and that the ADR EP-C aggregates the outgoing traffic. The number of bytes
varies depending on the number of samples aggregated in the message, being approached by the
straight line formula shown in (2).
As a result, the following values are obtained:

SC|ST = 540 Bytes

SC|LT = 895 Bytes
The estimation of the penetration of micro-generation is based on the know-how of the
ENERsip consortium. Table 5-2 summarizes the main parameters for both short and long-term
scenarios:
Table 5-2 - Main parameters for Short-term and Long-term scenarios
Parameter
Short-term
Long-term
15
5
T (min)
5
10
Plugs/I-BECI
540
895
SC (Bytes)
1030
1700
SG (Bytes)
1
1
I-BECI/Cust
U: 0.1
U: 0.4
I-BEGI/Cust
R: 0.4
R: 0.8
U: 360
U: 360
AC
R: 100
R: 100
U: 36
U: 144
AG
R: 40
R: 80
The intensity of usage of the platform is tuned in order to model the downlink traffic,
giving rise to three scenarios more:

Low usage

Medium usage

High usage
Such scenarios model, e.g., peaks and dips on the intensity of usage of the platform. The
main parameters estimated for these scenarios are summarized in Table 5-3.
Table 5-3 - Main parameters in Low usage, Medium usage and High usage scenarios
Feature
Low usage
Medium usage
High usage
1 day
2 hours
5 minutes
µIS
256
256
256
SIS (Bytes)
As a result, at a first glance, the model presented throughout this chapter comprises 12
different scenarios, as Table 5-4 shows, which still can be tuned to evaluate different figures-ofmerit of the proposed M2M communications architecture.
- 87 -
Chapter 5 – Practical Modeling
Urban
Rural
Table 5-4 - Scenarios considered in the model presented in this chapter
Short-term
Long-term
1) Low usage
4) Low usage
2) Medium usage
5) Medium usage
3) High usage
6) High usage
7) Low usage
10) Low usage
8) Medium usage
11) Medium usage
9) High usage
12) High usage
5.4 Conclusions
This chapter presents a methodology that can be applied to characterize in practice not only
the communications infrastructure required to meet the requirements and achieve the goals of
the so-called energy-positive neighborhoods, but also any communications overlay deployed on
top of an infrastructure devoted to any purpose. As a matter of fact, this methodology is also
applied in the on-going Spanish R&D (Research and Development) project PRICE to obtain
guidelines for proper design and deployment of AMI (Advanced Metering Infrastructures)
[López2013c].
As a result of applying this methodology, the proposed M2M communications architecture
is modeled in practice, as it is explained throughout the chapter. This model considers near realtime bidirectional communications both in short-term and in long-term scenarios. Although this
model is customized for the Portuguese power distribution infrastructures and the EU FP7
project ENERsip, it can be easily adapted to any other situation just by suitably tuning the
appropriate parameters.
Power distribution networks are quite similar throughout Europe. Therefore, the typical
values of Cust/TP, TP/subs, and Dmax of the Portuguese power distribution networks are
representative for the rest of Europe. However, they are not representative in North America.
First, European transformers are larger and there are more Cust/TP and TP/Subs. Hence, AC
and C would be lower in North America than the values considered in this chapter. Second,
North American secondary power distribution networks are single-phase and are standardized
on 120/240V; whereas European secondary power distribution networks are three-phase and are
standardized on 220, 230, or 240 V, which represent twice the North American standard. With
twice the voltage, a circuit feeding the same load can reach four times the distance.
Furthermore, taking into account that three-phase secondary can reach over twice the length of a
single-phase secondary, a European secondary can reach up to 8 times the length of a North
American secondary for a given load and voltage drop [Short2005]. Therefore, Dmax could be up
to 8 times lower in North America than the value considered in this chapter, so the situation in
North America is even more advantageous for using IEEE 802.11 in this network segment.
Regarding micro-generation and self-consumption, the situation in term of total installed
capacity is not the same in all the countries of the EU. Regarding residential PV in particular
[EPIA2013], the top 5 European markets in term of overall installed capacity are Italy,
Germany, Belgium, UK, and Denmark. However, our model considers the penetration rate of
these technologies as a percentage of the overall number of households/buildings with the aim
that the estimated values are as representative as possible. Nevertheless, countries like Belgium,
Denmark or the Netherlands still stand out when talking about penetration rates of residential
PV. In the US, the differences are also remarkable, standing out states like California.
Since this model has been carefully developed to fit real-world scenarios, potential works
based on it will provide meaningful results to any entity interested on operating this kind of
platforms, avoiding one of the main problems identified in the state-of-the-art when it comes to
- 88 -
Chapter 5 – Practical Modeling
simulating or testing communications infrastructures for the Smart Grid [Abdul Salam2012],
[Shrestha2012].
Thus, in general potential simulations based on this model may yield results along the lines
of, e.g.:

Evaluate how the communications infrastructure performs (e.g., in terms of percentage of
available resources consumed) under the characterized scenarios.

Evaluate the maximum number of users that can be handled in each of such scenarios
assuming full availability of communications resources (i.e., the communications
infrastructure is exclusively devoted to carry traffic from a given energy efficiency
platform).

Identify possible bottlenecks in the communications architecture.

Evaluate whether a public communications network, or a private communications network,
or a hybrid solution, is the best approach for this kind of systems.

Evaluate the performance of the system under different design decisions or different
network conditions (e.g., with and without data aggregation, during a DR event).

Evaluate the performance of the selected communications technologies in the
aforementioned scenarios

Evaluate the performance of other alternative communications technologies, such as NBPLC (Narrow-Band Power Line Communications) solutions, WiMAX (Worldwide
Interoperability for Microwave Access), or UMTS (Universal Mobile Telecommunications
System), in such scenarios.
This model is taken as reference in chapter 6 to assess the operational costs of using
different security solutions in the backhaul network as well as to evaluate the performance of
the selected communications technologies based on different metrics.
- 89 -
Chapter 5 – Practical Modeling
- 90 -
Chapter 6
Evaluation
6.1 Introduction
As it has been already highlighted throughout this dissertation, communications for the
Smart Grid need to meet specific requirements from both the technical and economic
perspectives, such as throughput, reliability, scalability, security, or low deployment and
operational costs [Güngör2013], [Yan2013], [Liu2012]. Hence, it is crucial to evaluate how
different communications technologies meet such requirements before undertaking the
important investments needed to deploy this kind of communications infrastructures on a large
scale.
This chapter aims to shed some light on this issue. In particular, the main goal of this
chapter is to evaluate, from both the technical and economic perspectives, the core of the M2M
(Machine-to-Machine) communications architecture proposed in chapter 3 taking as reference
the model presented in chapter 5. Such a core communications infrastructure is fully based on
widely deployed wireless communications technologies, such as IEEE 802.11 and GPRS
(General Packet Radio Service).
The remainder of the chapter is structured as follows. Section 6.2 outlines the general
considerations and assumptions made in this chapter regarding the model presented in chapter 5.
Section 6.3 analyzes and compares IPSec (Internet Protocol Security) and TLS/SSL (Transport
Layer Security / Secure Socket Layer) both from the technical point of view and in terms of the
Chapter 6 – Evaluation
potential impact on operational costs of using them as VPN (Virtual Private Network)
technologies. Section 6.4 evaluates by means of simulations the performance of IEEE802.11b,
in terms of goodput (i.e., throughput at the application layer), and the performance of GPRS, in
terms of transmission time (which is in turn related to bandwidth).
6.2 General considerations and assumptions
The model presented in chapter 5 is focused on the core of the M2M communications
architecture proposed in chapter 3, which encompasses from the ADR EPs (Automated Demand
Response End Points) up to the M2M GW (Gateway), and considers near real-time bidirectional
communications in realistic large-scale scenarios.
However, decoupling the uplink and the downlink reduces the complexity and increases the
granularity of potential assessments, allowing addressing them separately as well as putting
them eventually together to double-check the results obtained previously. In particular, this
chapter is focused on the uplink (i.e., consumption and generation data flowing from the ADR
EPs to the M2M GW), because the uplink traffic may be so high as to challenge the
communications infrastructure itself, thus representing the major concern for the entities
interested on operating this kind of platforms (e.g., DSOr – Distribution System Operator -,
retail electric provider, aggregator, ESCO – Energy Service COmpany) in the short to medium
term.
In addition, we assume that data reach the application layer at the CNTR (Concentrator),
which implies some advantages:

Relevant functionalities, such as data aggregation, can be evaluated at the CNTR.

The NAN (i.e., ADR EPs – CNTR) and the Backhaul network (CNTRs – M2M GW) can be
addressed separately.
Thus, in this chapter we distinguish between Urban (U) and Rural (R) scenarios, since based on the data from real power distribution infrastructures provided by EDP (Energias de
Portugal) - the number of customers/TP (Cust/TP), the number of TPs/Substation (TP/Subs),
and the maximum acceptable distance between customers and TPs (Dmax), vary remarkably
between both.
In addition, we consider not only current or short-term (ST) scenarios but also medium to
long-term (LT) scenarios, so that the obtained conclusions are valid for a longer period of time
and are used to take the appropriate decisions in advance. The main differences between shortterm and long-term scenarios have to do with:

The periodicity which ADR EPs send data with (T) and the size of such data (S). T will be
lower in the long-term, thus being closer to real-time. S will be higher in the long-term,
since more devices with communications capabilities are assumed both in the I-BECIs (InBuilding Energy Consumption Infrastructures) and in the I-BEGIs (In-Building Energy
Generation Infrastructures). In addition, S is also different for I-BECIs and I-BEGIs (SC and
SG, respectively), since the sensors and actuators networks within them are composed of
different devices.

The penetration of micro-generation. This parameter will be always higher in the rural
scenarios than in the urban ones due to the type of dwellings (e.g., houses where
photovoltaic panels can be installed in the roof are more common in the former; whereas
buildings of flats are more common in the latter). Furthermore, it will be also higher in the
long-term than in the short-term scenarios, since the penetration of micro-generation and
- 92 -
Chapter 6 – Evaluation
self-consumption is foreseen to increase in the forthcoming years. In this chapter we assume
independent communications gateways for the I-BECI (ADR EP-C) and for the I-BEGI
(ADR EP-G). Thus, the number of ADR EP-C (AC) is equal to the number of Cust/TP;
whereas the number of ADR EP-G (AG) is computed by multiplying AC by the estimation of
the micro-generation penetration (always lower than 1).
Table 6-1 summarizes the values of the aforementioned parameters in the four scenarios
considered in this chapter, where C refers to the number of CNTRs/M2M GW (i.e., TP/Subs)
and D refers to the maximum acceptable distance between customers and TPs (Dmax).
Table 6-1 - Summary of the parameters relevant to the scenarios considered in Chapter 6
Scenarios
Short-term (ST)
Long-term (LT)
AC/AG = 360/36
SC/SG = 540B/1030B
T/D/C = 15’/500m/150
AC/AG = 100/ 40
SC/SG = 540B/ 1030B
T/D/C = 15’/700m/220
Urban (U)
Rural (R)
AC/AG = 360/144
SC/SG = 895B/1700B
T/D/C = 5’/500m/150
AC/AG = 100/80
SC/SG = 895B/1700 B
T/D/C = 5’/700m/220
Figure 6-1 illustrates the model considered in this chapter and highlights its scope in
continuous green line.
NAN (Neighborhood
Area Networks)
HAN (Home Area Networks)
Backhaul
Customers
NILM
PLUGS
INFRARED
BOX
Information
System
TPs
I-BECI
COMFORT
SENSORS
802.11
GPRS
RENEWABLE
GENERATION
NETWORK
ANALYZER
I-BEGI
WEATHER
STATION
CNTR
M2M GW
ADR EPs
App (XML)
App (XML)
IP
LLC
MAC
802.11b
IP
LLC
MAC
802.11b
App (XML)
GPRS
UDP
FTP
TCP
IP
SNDCP
RLC/MAC
GSM RF
GPRS
UDP
App (XML)
FTP
TCP
IP
SNDCP
RLC/MAC
GSM RF
Figure 6-1 – Details of the model considered in chapter 6
6.3 Evaluation of end-to-end security protocols
The main goal of this section is to assess the impact of using security protocols which
support VPNs on the operational costs of an energy-efficiency platform for energy-positive
neighborhoods which relies on the M2M communications architecture proposed in chapter 3.
Thus, secure communications channels are to be established between pairs of entities of this
communications architecture. Therefore, bearing in mind that this chapter is focused on the core
of the already mentioned communications architecture, such secure channels can be established
from the ADR EPs directly to the M2M GW or from the CNTRs to the M2M GW, as Figure 6-2
(a) and (b) illustrates.
- 93 -
Chapter 6 – Evaluation
If the secure tunnels were established from the ADR EPs straight to the M2M GW, the
CNTRs would not be able to aggregate data, which would affect negatively the scalability and
operational costs of the platform. Thus, this case is actually divided into establishing secure
tunnels from the ADR EPs to the CNTR and from the CNTRs to the M2M GW, which implies
the highest numbers of tunnels and so the most complex scenario to manage, as Figure 6-2 (c)
shows.
Information
System
.
.
.
GPRS
M2M GW
CNTR: M
ADR EPs: N
(a)
Information
System
.
.
.
GPRS
M2M GW
CNTR: M
ADR EPs: N
(b)
Information
System
.
.
.
GPRS
M2M GW
CNTR: M
ADR EPs: N
(c)
Figure 6-2 – (a) NxM direct secure tunnels from the ADR EPs to the M2M GW; (b) M secure tunnels
from the CNTRs to the M2M GW; (c) NxM secure tunnels from the ADR EPs to the CNTRs + M secure
tunnels from the CNTRs to the M2M GW
Regarding the secure tunnels from the ADR EPs to the CNTRs, it might be interesting to
evaluate the impact of the overhead introduced by the security protocol on the performance of
the wireless link. This overhead will not increase the operational costs though, since in principle
it is assumed that the operator of the platform will be responsible for this network segment.
Hence, the operator itself will be also responsible for configuring the basic security mechanisms
within this network segment (i.e., WPA2 – Wi-Fi Protected Access 2).
- 94 -
Chapter 6 – Evaluation
Regarding the secure tunnels from the CNTRs to the M2M GW, the overhead introduced
by the security protocol does have an impact on the operational costs, since the backhaul
connectivity will be a service offered by a 3rd party (e.g., a telecom operator). In addition, in this
case the operator of the platform must use such security mechanisms at higher layers, since the
basic security mechanisms are out of its scope and it cannot rely solely on the security provided
by such a 3rd party.
This section is focused on the case of establishing VPNs from the CNTRs to the M2M
GW. Two scenarios are in turn considered within this specific case:

Fast Forwarding (FF): the CNTRs forward the packets coming from the ADR EPs to the
M2M GW on a per-packet basis, using a TCP (Transport Control Protocol) connection for
this purpose.

Aggregation (Aggr): the CNTRs store all the packets received from the ADR EPs
throughout a given period and send them all together using a FTP (File Transfer Protocol)
connection.
Regarding the security protocols themselves, there are many mechanisms to provide E2E
(End-to-End) security at the different layers of the protocol stack [Khanvilkar2004],
[Berger2006]. VPN can be implemented at the link layer using L2TP (Layer 2 Tunneling
Protocol). IPSec represents the most widely deployed solution to do so at the network layer.
TLS/SSL is the most widely used protocol for this purpose at the transport layer. And SSH
(Secure Shell) is widely used at the application layer for secure remote access. In this chapter,
we analyze and compare IPSec and TLS/SSL.
6.3.1 Technical comparison
Table 6-2 summarizes and compares some relevant technical features of IPSec and
TLS/SSL. It can be seen that both IPSec and TLS/SSL provides the basic security features
required by our target application (i.e., authentication, trustworthiness, and confidentiality). The
main drawbacks of IPSec are the complexity of configuration and the NAT (Network Address
Translation) incompatibility; whereas one of the main drawbacks of TLS/SSL is the complexity
of using PKI (Public Key Infrastructure). Regarding the fact that TLS/SSL only supports some
TCP applications, there is no problem in our case, since FTP is one of the TCP applications
supported by TLS/SSL. As result, it is concluded that, from a technical point of view, there is no
compelling reason to rule one of these protocols out.
Table 6-2 - Summary of IPSec and TLS/SSL technical comparison
Feature
IPsec
TLS/SSL
Yes
Yes
Authentication
Yes
Yes (More robust, since the
Trustworthiness
HMAC is longer)
Yes (if ESP)
Yes
Confidentiality
Complex
Straightforward
Configuration
Yes
No
Interoperability problems
(NAT)
All
Some
TCP apps support
Yes
Only DTLS
UDP support
No
Yes
PKI
Yes
Only OpenSSL
Compression
Yes
No
Client-specific software
Some times
Yes
Multi-environment support
No
Yes (VPN support to specific
Apps filter
apps)
- 95 -
Chapter 6 – Evaluation
6.3.2 Economic comparison
This section analyses the impact of using IPSec or TLS/SSL on the operational costs of a
potential energy efficiency platform which relies on the proposed M2M communications
architecture.
In order to do so, first the MSS (Maximum Segment Size) of TCP needs to be determined,
since this influences the number of packets sent through the GPRS link and the ratio of data vs.
control headers (i.e., overhead). There are quite a few papers on the use of TCP over GPRS.
Initially, the trend was to use low MSS (e.g., 512 B [Meyer1999] and 413 B [Rendón2001]).
However, although low MSS may be appropriate for interactive applications, [Benko2004]
proves that using high MSS (1400-1600 B) maximizes the goodput in applications of massive
data exchange, as it is our case.
Taking this range of TCP MSS as reference, we compute the MSS used in this chapter by
subtracting from the 1482 B pointed out in [Aschenbruck2004] as optimum MTU (Maximum
Transmission Unit) of the SNDCP (Sub Network Dependent Convergence Protocol) layer of
GPRS, the size of the headers up to the transport layer. Table 6-3 summarizes the overhead
introduced by IPSec and TLS/SSL [Alshamsi2005]. We consider always the worst case for the
analysis carried out in this section, i.e., 44 B for IPSec (plus the 20 B of the additional header
introduced in the tunnel mode) and 25 B for TLS/SSL.
Table 6-3 - Overhead introduced by IPSec and TLS/SSL [Alshamsi2005]
Protocol
Mode
Size (Bytes)
ESP
32
IPSec tunnel mode
ESP and AH
44
ESP
36
IPSec transport mode
ESP and AH
48
HMAC-MD5
21
TLS/SSL
HMAC-SHA-1
25
Figure 6-3 shows the protocol stack that implements both the CNTRs and the M2M GW in
each of the considered cases, specifying the size of the headers in bytes. The layers considered
in this section to compute the number of bytes carried by GPRS are marked in Figure 6-3 with
forward slash.
ENERsip App
(XML)
FTP (6)
TCP (20)
IP (20)
ENERsip App
(XML)
SNDCP (3)
RLC/MAC (7/8)
GSM RF
FTP (6)
TLS/SSL (25)
TCP (20)
IP (20)
SNDCP (3)
RLC/MAC (7/8)
GSM RF
(a)
(b)
ENERsip App
(XML)
TCP (20)
IP (20)
ENERsip App
(XML)
TLS/SSL (25)
IPSec (44)
IP (20)
IPSec (44)
SNDCP (3)
RLC/MAC (7/8)
GSM RF
TCP (20)
IP (20)
SNDCP (3)
RLC/MAC (7/8)
GSM RF
(c)
(d)
IP (20)
Figure 6-3 – Protocol stack at CNTR and M2M GW for: (a) IPSec &Aggr. (b) TLS/SSL & Aggr. (c)
IPSec & FF. (d) TLS/SSL & FF
- 96 -
Chapter 6 – Evaluation
In order to translate the volume of traffic onto cost, two commercial Spanish M2M tariffs
are used: 1) 100 MB/10 €/month, and 2) 20 MB/3 €/month.
Next, we evaluate the impact of IPSec and TLS/SSL on the operational cost following
these steps: 1) the volume of bytes carried by GPRS is computed per a single CNTR and per
month for each scenario; 2) the obtained bytes are translated onto cost using the aforementioned
M2M tariffs; 3) the cost per neighborhood and per year is computed by multiplying the cost per
CNTR and per month by C (cf. Table 6-1) and by 12 [El achab2013].
Table 6-4 details the results of our analysis for each of the considered scenarios. VNS
represents the volume of traffic (in MB) carried by the GPRS network in one month without
using any security protocol. VS represents the volume of traffic (in MB) carried by the GPRS
network in one month using the corresponding security protocol. RNS represents the ratio
between the application-layer data and VNS (in %). RS represents the ratio between the
application-layer data and VS (in %). OS is computed as the difference between VS and VNS, so it
represents the overhead introduced by the security protocol (in %). CNS represents the monthly
cost of carrying VNS (in €). CS represents the monthly cost of carrying VS (in €). Finally, DC is
computed as the difference between CS and CNS, so it represents the cost of using the
corresponding security solution in a given scenario.
Table 6-4 - Summary of the results of the analysis of the impact on the operational costs of using IPSec or
TLS/SSL
Urban (U)
Aggr
FF
Rural (R)
Aggr
FF
Short-term (SL)
IPSec
SSL/TLS
VNS =656,25
VNS = 656,25
VS =686,74
VS = 667,96
RNS =96.88 %
RNS =96.88 %
RS =92.58 %
RS =95.18 %
Os = 4.3 %
Os =1.7%
CNS = 69
CNS = 69
CS = 70
CS = 70
DC = 1
DC = 1
VNS =679,28
VNS =679,28
VS =748,89
VS =706,48
RNS =93.6 %
RNS =93.6 %
RS =84.89 %
RS =90 %
Os =8.71 %
Os =3.6%
CNS = 70
CNS = 70
CS =79
CS =73
DC = 9
DC = 3
VNS =269,94
VNS =269,94
VS =282,62
VS =274,74
RNS =96,86%
RNS =96,86%
RS =92.5 %
RS =95,17 %
Os =4.36 %
Os =1.69 %
CNS = 30
CNS = 30
CS = 30
CS = 30
DC = 0
DC = 0
VNS =276,86
VNS =276,86
VS =301,46
VS =286,47
RNS =94.4 %
RNS =94.4 %
RS =86.7 %
RS =91.27 %
Os =7.7 %
Os =3.6 %
CNS = 30
CNS = 30
CS = 33
CS = 30
DC = 3
DC = 0
- 97 -
Long-term (LT)
IPSec
SSL/TLS
VNS =4821,65
VNS = 4821,65
VS =5047,17
VS =4907,11
RNS =96.895 %
RNS = 96.895 %
RS =92.56 %
RS = 95.207%
Os =4.335 %
Os = 1.688 %
CNS = 486
CNS = 486
CS =509
CS =493
DC = 23
DC = 7
VNS =4885,51
VNS =4885,51
VS =5227,23
VS =5018,99
RNS =95.628 %
RNS = 95.628 %
RS =89.38 %
RS =93.085 %
Os =6.248 %
Os =2.543 %
CNS = 490
CNS = 490
CS =526
CS =500
DC = 36
DC = 10
VNS =1917,95
VNS =1917,95
VS =2007,61
VS =1951,67
RNS =96.877 %
RNS =96.877 %
RS =92.55 %
RS =95.2 %
Os =4.327 %
Os =1.67 %
CNS = 193
CNS = 193
CS = 203
CS = 199
DC = 10
DC = 6
VNS =1943,76
VNS =1943,76
VS =2080,87
VS =2021,04
RNS =95.59 %
RNS =95.59 %
RS =89.29%
RS =91.9 %
Os =6.3 %
Os =3.69 %
CNS = 199
CNS = 199
CS = 210
CS = 206
DC = 11
DC = 7
Chapter 6 – Evaluation
Table 6-5 shows the difference between the annual cost of using Fast Forwarding and the
annual cost of using Aggregation (CS|FF – CS|Aggr) in each scenario for a single CNTR. Table 6-5
also shows this difference in each scenario for the whole district/neighborhood1.
Table 6-5 - Difference in terms of cost (in €) per CNTR and per district during one year between using
Fast Forwarding and using Aggregation in each scenario
Short-term (ST)
Long-term (LT)
IPSec
TLS/SSL
IPSec
TLS/SSL
9*12=108
3*12=36
17*12=204
7*12=84
Urban (U)
108*150=
36*150=
204*150=
84*150=
16200
5400
30600
12600
3*12=36
0
7*12 = 84
7*12=84
Rural (R)
36*220=
84*220=
84*220=
7920
18480
18480
In order to facilitate the understanding of the impact of using Fast Forwarding or
Aggregation on the operational costs of the platform, Figure 6-4 graphically shows the
difference between the annual cost of using Fast Forwarding and the annual cost of using
Aggregation in each scenario for a whole district. It can be seen that the difference of cost –
although almost negligible for a single CNTR - can be appreciable at neighborhood level,
notably in urban and long-term scenarios. In addition, it can be also checked that the difference
is always higher when using IPSec, since it introduces higher overhead. In conclusion, Figure 64 illustrates the savings that can be achieved by using Aggregation. Nevertheless, it is
worthwhile to remark that the results obtained in this analysis represent a lower bound of the
savings that Aggregation could bring, since we just aggregate data during one period.
Figure 6-4 – Annual savings per district between implementing Fast Forwarding or
Aggregation at CNTRs for each security protocol in each scenario
Table 6-6 shows the difference between the annual cost of using IPSec and the annual cost
of using TLS/SSL (CS|IPSec – CS|TLS/SSL) in each scenario for both a single CNTR and the whole
district/neighborhood.
1
It should be noted that District is used to refer to the whole power infrastructure managed by a given
Substations, where the consumption-generation optimization algorithms run.
- 98 -
Chapter 6 – Evaluation
Table 6-6 - Difference in terms of cost (in €) per CNTR and per district during one year between using
IPSec and TLS/SSL in each scenario
Short-term (ST)
Long-term (LT)
Aggr
FF
Aggr
FF
6*12 = 72
16*12 = 192
26*12 = 312
Urban (U)
0
72*150 = 10800
192*150= 28800
312*150 = 46800
3* 12 = 36
4*12 = 48
4*12 = 48
Rural (R)
0
36*220 = 7920
48*220 = 10560
48*220 = 10560
Again, to aid understanding the impact of using IPSec or TLS/SSL on the operational
costs, Figure 6-5 graphically shows this difference in each scenario for a whole district. It can be
checked that the difference of costs between using IPSec or TLS/SSL is always higher when
using Fast Forwarding, since data sending is very inefficient in this situation, so the difference
between the overhead introduced by IPSec and by TLS/SSL is even higher. It can be also
checked that, in the case of using Aggregation, the potential savings of using TLS/SSL instead
of IPSec are especially relevant in long-term scenarios.
Figure 6-5 –Annual savings per district between implementing IPSec and TLS/SSL at CNTRs in each
scenario depending on whether Fast Forwarding or Aggregation is used
Therefore, it can be concluded that Aggregation and TLS/SSL as VPN technology is the
best combination in order to minimize the operational costs of the platform. Hence, such a
combination will be assumed in the next section.
Nevertheless, it should be noticed that the overhead – and so the costs - can be lower by
implementing compression mechanisms (in the case of TLS/SSL, only OpenSSL supports it)
and that the volume of data sent through the GPRS network – and so the costs – can be lower if
only the data that change compared to the previous period are sent, which can be implemented,
e.g., using JSON (JavaScript Object Notation).
- 99 -
Chapter 6 – Evaluation
6.4 Evaluation of communications infrastructure
performance
This section evaluates the performance of IEEE 802.11 and GPRS using goodput and
transmission time as metrics, respectively.
6.4.1 Simulation setup
The simulations were run using OMNeT++ 4.2.2 as network simulation framework
[Varga2008]. Notably, the models and network components available in the INET framework 4
were used and modified when needed.
The simulations were run in a laptop equipped with a microprocessor Intel Core 2 Duo
T7250 at 2 GHz, 1 Gb of RAM (Random Access Memory), and 2 Mb of cache memory, using
Ubuntu 10.04 as OS (Operating System). The simulation time was set to 1 hour after checking
that simulations converge after one period (i.e., 15 minutes in short-term scenarios and 5
minutes in long-term scenarios). Each scenario was simulated 100 times for both the NAN and
the backhaul network in order to get statistically meaningful results. Therefore, the obtained
data were considered to fit a Gaussian distribution based on the central limit theorem and hence
(1) can be used to compute 95% confidence intervals, µ being the mean, σ being the standard
deviation, and n being the length of the data (i.e., n=100). Octave 3.2 was used to process the
data.
(   1.96 

n
,   1.96 

n
)
(1)
Figure 6-6 shows the network used to simulate the NAN.
Figure 6-6 – Network used to simulate the NAN
The functionalities of the modules shown in Figure 6-6 are outlined next:
- 100 -
Chapter 6 – Evaluation

channelcontrol: this module manages the PHY (PHYsical) and MAC (Media Access
Control) layer of IEEE 802.11.

flatNetworkConfigurator: this module assigns IP addresses within the same network
segment.

MobileHost: this module is taken as reference to implement the functionalities of ADR EPCs, ADR EP-Gs, and CNTR. Figure 6-7 shows the internal structure of the MobileHost. It
can be seen that this module implements from an IEEE 802.11 interface in ad-hoc mode up
to a set of UDP (User Datagram Protocol) applications. From such a set of UDP
applications, UDPBasicApp is used to implement the ADR EPs, since UDPBasicApp allows
fixing the sending periodicity and length of the messages; whereas a passive UDP socket
(UDPSink) is used to implement the CNTR. In addition, a DatarateChannel was included
between the UDP layer and the application layer in the upper direction in order to measure
the aggregate throughput at the application layer in the CNTR (i.e., the so-called goodput).
Figure 6-7 – Internal structure of the module MobileHost
The wireless router Linksys WRT160NL [WRT160NL] from Cisco is taken as reference
for configuring the PHY and MAC parameters of the 802.11 link, since this device was actually
used to implement the prototype of the ADR EP in the ENERsip project. The family of
standards 802.11 encompasses many protocols [Hiertz2010]. The Linksys WRT160NL, in
particular, supports 802.11b/g/n. In this work, we focus on 802.11b and g. Table 6-7
summarizes the most relevant parameters of IEEE 802.11b and g in the Linksys WRT160NL.
EIRP
Sensitivity
Table 6-7 - Summary of Linksys WRT160NL datasheet
802.11b
802.11g
1 Mb
11 Mb
6 Mb
54 Mb
19 ± 1.5 dBm
15 ± 1.5 dBm
Min = 56.2 mW
Min = 22.387 mW
Max = 112.2 mW
Max = 44.67 mW
- 92 dBm
- 86 dBm
-84 dBm
-74 dBm
- 101 -
Chapter 6 – Evaluation
In order to check whether all these protocols provide the range required by our target
application (D) or not, the well-known Friis equation (2) with α=2 is applied, since Free-Space
is used as propagation model in OMNeT++ [Khosroshahy2007].
receivedPo wer 
powerSend * 2
16 *  2 * dist 
(2)
Table 6-8 summarizes the minimum transmission power (powerSend) required to reach D
with a received power equals to the sensitivity. If powerSend is higher than the maximum EIRP
(Effective Isotropic Radiated Power), the protocol under question cannot be used in our target
application, as it is the case of 802.11g at 54 Mb.
Table 6-8 - Summary of minimum transmission powers required for our target application
Scenarios
Distance
powerSend (mW)
(m)
802.11b 1 Mb 802.11b 11 Mb 802.11g 6 Mb
802.11g 54 Mb
500
Urban (U)
1.59 < EIRP
6.35 < EIRP
10.056 < EIRP
100.56 > EIRP
700
Rural (R)
3.125 < EIRP
12.44 < EIRP
19.71 < EIRP
197.1 > EIRP
Finally, IEEE 802.11b at 11 Mb was selected for our simulations. Table 6-9 shows the
configuration of the PHY and MAC parameters of IEEE 802.11b used in such simulations.
Regarding the Backhaul network, when developing the simulations, we realized that in
practice there cannot be interferences in the GPRS links, since GPRS uses dedicated channels,
instead of shared medium. As a result, the parameter C from Table 6-1 is not needed in these
simulations. This reduces the complexity of them, since instead of having to simulate C GPRS
connections; it is enough to simulate just one. Figure 6-8 shows the network used to simulate
the backhaul communications.
Figure 6-8 – Network used to simulate the Backhaul
From the modules shown in Figure 6-8, only the functionalities of the
WirelessHostSimplified need to be described, since the functionalities of the channelcontrol and
the flatNetworkConfigurator have already been presented. WirelessHostSimplified is taken as
reference to implement the functionalities of the CNTRs and the M2M GW. Figure 6-9 shows
the internal structure of this module. As it can be seen, the WirelessHostSimplified implements
three different PHY/MAC layers, namely wlan (i.e., IEEE 802.11 in infrastructure mode), eth
- 102 -
Chapter 6 – Evaluation
(i.e., Ethernet), and ppp (i.e., Point-to-Point). The ppp module is selected to model the GPRS
link. At the application layer, the WirelessHostSimplified implements a set of TCP applications.
TCPBasicClientApp and TCPGenericSrvApp are the most suitable applications to implement an
FTP application. TCPBasicClientApp allows specifying the time gap between requests (i.e.,
sending periodicity) and the size of the reply (i.e., size of data). Thus, once a
TCPGenericSrvApp receives a request, it just replies a message with the size specified in that
request.
Figure 6-9 – Internal structure of the module WirelessHostSimplified
As a result, despite the fact that in practice the M2M GW works as FTP server and the
CNTRs work as FTP clients, in our simulations the TCPBasicClientApp was used to implement
the M2M GW and the TCPGenericSrvApp was used to implement the CNTR. Thus, the time
that the CNTR needs to send the data to the M2M GW can be computed as the difference
between the time when they are received at the M2M GW and the time when the request is
received at the CNTR (since the parameter ReplyDelay of TCPGenericSrvApp is set to 0). A
DatarateChannel was included between the TCP layer and the application layer in the upper
direction indeed to measure the bytes received at the application layer in the M2M GW and in
the CNTR.
The GPRS link itself is modeled as a DatarateChannel with the following features, which
are also summarized in Table 6-9:

Delay = 1 µs.

Data rate = 26.8 kbps. This parameter is set based on the uplink data rate of an actual GPRS
network reported in [Shrestha2012]. A dedicated infrastructure is assumed, i.e., the GPRS
link only carries data associated to our application.

Probability of error = 0.001, based on the theoretical availability of the channel
[NIST2010b].
- 103 -
PHY
Table 6-9 - Summary of the most important parameters for each communications technology
Parameter
Value
Delay
1 µs
Uplink data rate
26.8 Kbps
Probability of error
0.001
Carrier frequency
2.4 GHz
Transmitter power
79.43 mW
Path loss (α)
2
Sensitivity
-86 dBm
Bit rate
11 Mbps
Retry limit
7
Contention window
32
MAC
802.11b
GPRS
Chapter 6 – Evaluation
6.4.2 Performance evaluation
Table 6-10 summarizes the goodput measured at the CNTR in the four simulated scenarios
and compares it with the theoretical approximation computed using (3).
Theoretica lGoodput 
Ac * Sc * 8  Ag * Sg * 8
T
(3)
Table 6-10 - Summary of NAN results (Goodput in bps)
Urban (U)
Rural (R)
Theoretical
Simulated
Theoretical
Simulated
Short-term (ST)
Long-term (LT)
2057.6
(2043.94 , 2048.05)
846.22
(843.54 , 845.79)
15120
(15070.09 , 15077.91)
6013.3
(6011.81 , 6012.32)
It can be checked that the obtained results fit the expected results in every scenario2. Figure
6-10 shows the goodput for just one simulation run in every scenario. It can be observed that the
simulated goodput converges to a value which is close to the theoretical one in each scenario.
As a result, it can be concluded that the protocol IEEE802.11b at 11 Mb meets the requirements
(in terms of goodput) of the NAN in all the considered scenarios. Due to the fact that the
maximum goodput is in the order of tens of Kb, IEEE 802.11b at 1 Mb or IEEE802.11g at 6 Mb
could also be used.
2
In order to avoid simulation problems in the worst case scenario (i.e., ULT), explicit memory
deallocation at MAC layer is needed.
- 104 -
Chapter 6 – Evaluation
Figure 6-10 – NAN
results for each scenario for a simulation time of 3600 s
Table 6-11 summarizes the time that the CNTR spends to transfer the data to the M2M GW
(TCNTR) in the four simulated scenarios and compares them with the time computed theoretically
using the fragmentation criteria reported in [Aschenbruck2004] and the bytes of the layers
marked in Figure 6-3 with back slash.
Table 6-11 - Summary of Backhaul network results (Transmission Time in s)
Urban (U)
Rural (R)
Theoretical
Simulated
Theoretical
Simulated
Short-term (ST)
Long-term (LT)
TCNTR = 73.477 < 900
(77.41 , 78.93)
TCNTR = 30.224 < 900
(31.67 , 32.37)
TCNTR = 179.93 < 300
(190.18 , 191.28)
TCNTR = 71.564 < 300
(75.47 , 76.12)
It can be checked that the simulated TCNTR is always slightly higher than the theoretical
value, mainly due to the probability of error and to slightly different fragmentation policies. In
addition, TCNTR is always lower than the sending period, which means that GPRS meets the
requirements (in terms of bandwidth) of the Backhaul network in all the considered scenarios.
6.5 Conclusions
This chapter evaluates some features of such widely used wireless communications
technologies as IEEE 802.11 and GPRS in a Smart Grid application which consists of gathering
consumption and generation data periodically in order to reduce electricity consumption and to
maximize consumption-generation matching at neighborhood level. This assessment is carried
out analytically and by means of simulations considering realistic large-scale scenarios, so the
obtained results are valid to be taken into account as guidelines for potential deployments.
The chapter analyses the impact of using IPSec and TLS/SSL as VPN technologies on the
operational costs of the platform. The main conclusions of such analysis are: 1) that TLS/SSL
- 105 -
Chapter 6 – Evaluation
minimizes the operational costs, the difference being especially remarkable in Long-term
scenarios; and 2) that using Aggregation at the CNTR also allows reducing the operational costs
of the platform at neighborhood level considerably.
In addition, it is worthwhile to remark that the overhead – and so the costs - can be lower
by implementing compression mechanisms (in the case of TLS/SSL, OpenSSL should be used)
and that the volume of data sent through the GPRS network – and so the costs – can be lower if
only the data that change compared to the previous period are sent, which can be implemented,
e.g., using JSON (JavaScript Object Notation).
The main conclusion of the simulations with regards to the NAN is that IEEE 802.11b
meets the requirements in terms of goodput of this network segment in all the considered
scenarios. This conclusion is of special interest to the Smart Grid community taking into
account the low cost and wide adoption of IEEE 802.11b. Furthermore, the data rate demanded
by the application is much lower than the data rate provided by IEEE 802.11, which leaves
room to use such a IEEE 802.11 infrastructure either to offer additional services (e.g.,
multimedia connectivity using IEEE 802.11e) or to operate more challenging applications such
as DR. However, the performance of IEEE 802.11 in these scenarios still needs to be checked
on a large scale in the field, since the simulations do not take into account the effects of
interferences, which are usually high in the ISM (Industrial, Scientific and Medical) band at 2.4
GHz, nor complex propagation models.
In addition, in order to avoid potential problems related to the coverage of IEEE 802.11,
sub-GHz Wi-Fi (IEEE 802.11AH) [Aust2012] may be also considered in the future for the NAN
in rural scenarios, where distances are higher, but data rates are lower. In such rural scenarios, it
is also worth considering White Spaces [Brew2011].
Regarding the Backhaul network, it was proved that GPRS meets the requirements in terms
of bandwidth of this network segment in all the considered scenarios. Therefore, GPRS
represents a very attractive technology considering that it is the most mature and widely
deployed cellular technology in Europe.
However, the relevance of the obtained results can be improved by using a more accurate
model for the GPRS link or by testing the considered scenarios in a real GPRS network.
- 106 -
Chapter 7
Conclusions
7.1 Introduction
This chapter is structured as follows. Section 7.2 summarizes the main contributions and
conclusions of this thesis. Section 7.3 highlights the most outstanding contributions of this
thesis to the state of the art and the synergies with R&D (Research and Development) projects,
EC (European Commission) initiatives, standardization activities, and industry. Finally, section
7.4 outlines potential research lines that can be undertaken using the work developed in this
thesis as baseline.
7.2 Conclusions
Broadly speaking, this thesis represents a remarkable contribution to the area of tailored
M2M (Machine-to-Machine) architectures that meet the specific requirements of the power
distribution and customer domains of the Smart Grid, supporting energy efficiency and proper
integration of DG (Distributed Generation) and EVs (Electric Vehicles) by enabling
sophisticated mechanisms such as DR (Demand Response). The specific contributions of this
thesis are summarized as follows:
Chapter 7 – Conclusions

In chapter 2 we carry out a survey on the most relevant standardization activities developed
in parallel to this thesis and on the most outstanding R&D trends within the Smart Grid
area, identifying gaps and challenges [Moura2013b].

In chapter 3 we propose a novel M2M communications architecture to support energy
efficiency and optimum coordination of DER (Distributed Energy Resources) within the socalled energy-positive neighborhoods (i.e., neighborhoods which ensure a substantial part of
their consumption by local generation) [López2011a], and we map it onto the
standardization work presented in chapter 2. The proposed M2M communications
architecture aims to meet Smart Grid specific requirements such as scalability,
interoperability, or low deployment and operational costs. Reflecting the outstanding
importance of wireless communications in the Smart Grid, such M2M communications
architecture is fully based on well-known and widely adopted wireless communications
technologies, such as IEEE802.15.4/Zigbee, IEEE 802.11, and GPRS (General Packet
Radio Service). This network architecture is taken as reference to design M2M
communications infrastructures for related applications such as smart EV charging
[López2013b].

In chapter 4 we formally model the domain of knowledge of energy efficiency platforms for
energy-positive neighborhoods by means of an ontology developed in OWL (Ontology Web
Language). This ontology not only represents the main architectural entities and interfaces
of such energy efficiency platforms for energy-positive neighborhoods but also their
potential services and stakeholders, and the relationships between them. The developed
ontology has been made public through the EC eeBuildings Data Model community (also
called eeSemantics) so that the research community can make the most out of it.

In chapter 5 we characterize the core of the M2M communications architecture presented in
chapter 3 in realistic large-scale scenarios. The scenarios modeled in this chapter are
customized for the Portuguese power distribution infrastructure and for the EU FP7 project
ENERsip [López2012a]. However, they can be adapted to any other electricity distribution
infrastructure and energy efficiency platform just by suitably changing the values of the
identified parameters. As a matter of fact, the methodology followed in this chapter can be
applied to characterize in practice any communications overlay deployed on top of an
infrastructure devoted to any purpose and it ensures obtaining results that actually mean and
bring value to the interested parties.

Finally, in chapter 6 we evaluate the core of the M2M communications architecture
proposed in chapter 3 taking as reference the model presented in chapter 5. This evaluation
is twofold. On the one side, we analyze and compare IPSec (Internet Protocol Security) and
TLS/SSL (Transport Layer Security/Secure Socket Layer) from both technical and
economic perspectives [El achab2013]. The main outcome of this analysis is that using
TLS/SSL as VPN (Virtual Private Network) technology along with data aggregation at the
CNTR (Concentrator) is the best option to minimize operational costs at neighborhood
level. On the other side, we evaluate by means of simulations the performance of the
selected communications technologies using as metrics the goodput (i.e., throughput at the
application layer), in the case of IEEE 802.11b, and the transmission time (which is in turn
related to bandwidth), in the case of GPRS (General Packet Radio Service). Standing out as
main outcome of these simulations is that IEEE 802.11b meets the requirements in terms of
goodput of the NAN (Neighborhood Area Networks), which is of special interest to the
Smart Grid community taking into account the low cost and wide adoption of this
technology, and that GPRS meets the requirements in terms of bandwidth of the backhaul
network, thus confirming that it represents a very attractive technology considering that it is
the most mature and widely deployed cellular technology in Europe.
- 108 -
Chapter 7 – Conclusions
As a result, this thesis adds value to the emerging Smart Grid universe, notably as regards
communications architectures and technologies and common data models. Much R&D activities
are still on-going with the aim of determining the most appropriate communications
architectures and technologies, involving both simulations and actual deployments. In truth and
in fact this will eventually depend on the specific business case behind the target application,
which in turn may depend on many factors, such as the special features of the target application
itself or the special features and the specific regulation of each country (which varies even
among EU countries) [Moura2012].
Beside the underlying communications technologies, application protocols are also crucial
to push the Smart Grid forward in the distribution and customer domains. In particular, common
data models which can be further developed down to standard application protocols would
enable moving beyond current vertical solutions and systems and reaching a level of
interoperability that allows heterogeneous devices from different vendors relying on different
communications technologies but targeting the same application to interact seamlessly. This
would give rise to a more stable and wider market that would foster investments and economies
of scale, which in the end imply more reliable products at lower costs.
This brings us to the key piece of the Smart Grid puzzle: the customers. Customer
engagement is definitely crucial for the Smart Grid to become a reality at distribution level and
this is far away from being just a matter of technology. As it has been made clear throughout
this dissertation, customer involvement is mandatory to properly operate the Smart Grid and the
Smart Grid brings benefits to customers. However, in this respect the power distribution
industry needs to undertake a revolution like the one the telecom industry underwent with
cellular communications. Nowadays most of us cannot live without them because they also
bring many benefits to our day-to-day. But did not we live without being always connected a
couple of decades ago? We did it without problems. Such a need was artificially created in us.
In the case of the telecom industry, the telecom operators were responsible for this. In the case
of the power distribution industry, who will lead the change that converts the Smart Grid into a
mass phenomenon?
7.3 Thesis impact
The main results from the research work presented throughout this dissertation have been
disseminated by means of the following publications:

Top-tier journals:
o G. López, V. Custodio, F. J. Herrera, J. I. Moreno, “Machine-to-Machine
Communications Infrastructure for Smart Electric Vehicle Charging in Private Parking
Lots”, International Journal of Communication System, Wiley, November 2013. DOI:
10.1002/dac.2705. Impact Factor: 0.712. JCR (48/78), category: Telecommunications.
o P. Moura, G. López, J. I. Moreno, A. de Almeida, “The role of Smart Grids to foster
energy efficiency”, Energy Efficiency, Volume 6, Issue 4, Pages 621-639, November
2013. Impact Factor: 1.150. JCR (47/81), category: Energy & Fuels.

Top-tier conferences:
o G. López, P. Moura, V. Custodio, J. I. Moreno, “Modeling the Neighborhood Area
Networks of the Smart Grid”, IEEE ICC 2012, Ottawa, Canada, June 2012. Microsoft
Academic Search (5/247), category: Networks & Communications.
- 109 -
Chapter 7 – Conclusions
o G. López, P. Moura, J. I. Moreno, A. de Almeida, “ENERsip: M2M-based platform to
enable energy efficiency within energy-positive neighborhoods”, IEEE INFOCOM
2011 Workshop on M2M Communications and Networking, Shanghai, China, April
2011. Microsoft Academic Search (1/247), category: Networks & Communications.

Conferences in the area of Smart Grids or M2M communications:
o E. El achab, G. López. J. I. Moreno, “Evaluación de mecanismos de seguridad en
entornos de Smart Grid”, JITEL 2013: XI Jornadas de Ingeniería Telemática, Granada,
Spain, October 2013.
o P. Moura, G. López, J. I. Moreno, A. de Almeida, “Impact of Residential Demand
Response on the Integration of Intermittent Renewable Generation into the Smart Grid”,
EEDAL2013: 7th International Conference on Energy Efficiency in Domestic
Appliances and Lighting, Coimbra, Portugal, September 2013.
o G. López, J. Moreno, P. Moura, A. de Almeida, M. Perez, L. Blanco, “Monitoring
System for the Local Distributed Generation Infrastructures of the Smart Grid”, CIRED
2013: 22nd European Conference and Exhibition on Electricity Distribution,
Stockholm, Sweden, June 2013.
o P. Moura, G. López, A. Carreiro, J. I. Moreno, A. de Almeida, “Evaluation
Methodologies and Regulatory Issues in Smart Grid Projects with Local GenerationConsumption Matching”, EEMSW2012: International Workshop on Energy Efficiency
for a More Sustainable World, São Miguel, Portugal, September 2012.
o A. Carreiro, G. López, P. Moura, J. I. Moreno, A. de Almeida, J. Malaquias, “In-house
monitoring and control network for the Smart Grid of the future”, ISGT Europe 2011:
2nd IEEE PES International Conference and Exhibition on Innovative Smart Grid
Technologies, Manchester, UK, December 2011.
o G. López, P. Moura, M. Sikora, J. I. Moreno, “Comprehensive validation of an ICT
platform to support energy efficiency in future smart grid scenarios”, SMFG2011:IEEE
International Conference on Smart Measurements for Future Grids, Bologna, Italy,
November 2011.
Additionally, the research conducted throughout this thesis takes as baseline previous
research on the application of M2M communications and WSNs (Wireless Sensor Networks) to
e-health. Such previous work was published in the following top-tier journals and conferences:

V. Custodio, F. J. Herrera, G. López, J. I. Moreno, “A Review on Architectures and
Communications Technologies for Wearable Health-Monitoring Systems” Sensors, Pages
13907-13946, October 2012. Impact Factor: 1.953. JCR (8/57), category: Instruments &
Instrumentation

G. López, V. Custodio, J. I. Moreno, “LOBIN: E-Textile and Wireless-Sensor-NetworkBased Platform for Healthcare Monitoring in Future Hospital Environments”, IEEE
Transactions on Information Technology in Biomedicine, Volume 14, Issue 6, Pages 14461458, November 2010. Impact Factor: 1.978. JCR (21/132), category: Computer Science,
Information Systems.

G. López, V. Custodio, J. I. Moreno, “Location-Aware System for Wearable Physiological
Monitoring Within Hospital Facilities”, IEEE PIMRC 2010: 21st Annual IEEE International
Symposium on Personal, Indoor and Mobile Radio Communications, Istanbul, September
2010. Microsoft Academic Search (18/247), category: Networks & Communications.
- 110 -
Chapter 7 – Conclusions
This thesis has been carried out under the scope and influence of the R&D projects
summarized in Table 7-1.
Table 7-1 - R&D projects related to this thesis
Funding scheme
Topic
Duration
Consortium
INNPACTO
AMI
2011-2014
Iberdrola, Unión Fenosa
(IPT-2011-1507Distribución, ZIV, Arteche,
920000)
Current, CIRCE, UC3M
FP7
EE
2010-2012
Tecnalia,
Israel Electric Corp.,
ENERsip
(Grant agreement n°
DG
Motorola Solutions Israel Ltd.,
247624)
Honeywell Prague Laboratory,
AMPLIA, Intelligent Sensing
Anywhere, ISASTUR, VITO,
ISR-UC, UC3M
AVANZA2
EV
2009-2011
Unión Fenosa Distribución, Red
DOMOCELL
(TSI-020100-2009Eléctrica Española, Amplía, Nlaza,
849)
Neoris, CITEAN, UPV, UC3M
EE: Energy Efficiency
Name
PRICE-GEN
During the development of the thesis, I have also carried out two internships at ISR-UC
(Institute of System and Robotics – University of Coimbra), in particular at the "Intelligent
Energy Systems " research group led by Dr. Aníbal de Almeida, working close together with
Dr. Pedro Moura. As a result of such internships, our expertise on ICT has been complemented
with their expertise on energy and the link between the Telematics Engineering Department of
the UC3M and the ISR-UC has been strengthened. In addition, the research conducted during
these internships has led to the publication of several papers and the internships themselves
allow this thesis to be awarded with the International Doctorate certification.
Finally, the main synergies between this thesis and EC initiatives, standardization activities
and industry, are summarized next:

Participation in EC initiatives. As it has already been mentioned in section 7.2, the ontology
developed under the scope of this thesis has been inputted to the eeBuildings Data Model
community (also called as, eeSemantics). As a first impact of this, our ontology was
proposed with other related works as starting point for the study on “Available semantics
assets for the interoperability of SMART APPLIANCES. Mapping into a common ontology
as a M2M application layer semantics”, launched as invitation to tender by the EC in June
2013 [EC2013].

Participation in standardization activities. Part of the research work carried out under the
scope of this thesis has been presented in the following workshops organized by ETSI:
o G. López, J. I. Moreno, “PRICE Project: M2M communications architecture for largescale AMI deployment”, 4th ETSI M2M Workshop, Mandelieu-la-Napoule, France,
November 2013.
o G. López, J. I. Moreno, “Smart Energy-positive Neighbourhoods for the Smart Grid.
Architecture, Communications Technologies and Address Management for the
ENERsip platform”, 2011 ETSI Workshop on “Standards: An Architecture for the
Smart Grid”, Sophia Antipolis, France, April 2011.
In addition, the knowledge acquired during this thesis has enabled me to be promoted as
representative of the COIT (Spanish Official Professional Association of
Telecommunications) in AENOR (Spanish Association for Standardization and
Certification), participating on a regular basis in the AEN/CTN 207/SC 13 “Aparatos de
medida de la energía eléctrica y del control de cargas”. As a remarkable impact of my
- 111 -
Chapter 7 – Conclusions
participation in AENOR as representative of the COIT so far, we have recently inputted our
proposal of priority ranking for the new standardization gaps identified in the second phase
of the standardization mandate M/490, which will be addressed by the SG-CG (Smart Grid
– Coordination Group) during 2014.

Knowledge transferred to industry. The I-BECI (In-Building Energy Consumption
Infrastructure) presented in chapter 3 of this dissertation, which was designed jointly with
ISA (Intelligent Sensing Anywhere) [Carrerio2011], has been further developed by this
company, becoming a commercial product which is already available in the market
[Cloogy2014].
7.4 Future work
Next, we outline some future research lines that result from this thesis:

In this thesis, the evaluation of the performance of the proposed M2M communications
architecture is focused on the uplink, since the importance of the uplink is higher in the
short term and the uplink traffic pattern is well-known and may be so high as to challenge
the communications infrastructure itself. Therefore, this work can be further developed by
including the downlink, putting special emphasis on evaluating the impact of DR events on
the performance of the M2M communications infrastructure.

This thesis is especially focused on wireless communications technologies, reflecting their
outstanding importance within the Smart Grid area. Notably, IEEE 802.11 and GPRS are
considered and evaluated, since they represent two of the most widely adopted wireless
communications technologies nowadays. As result, this work can be extended by evaluating
other communications technologies (e.g., PRIME, UMTS, LTE or WiMAX) either for the
application considered in this thesis or for any other Smart Grid application (by suitably
tuning the parameters identified in the practical model developed as part of this thesis).

Taking the network simulation framework used in this thesis (i.e., OMNeT++) as reference,
it can be investigated how it could be coupled with other power network simulator (e.g.,
PSLF) to enable co-simulations (i.e., simulate power networks and their associated
communications overlay at the same time), which definitely represents a very hot topic
within the Smart Grid area currently and in the forthcoming years.
- 112 -
Table of Acronyms
AC
AC
ADDRESS
ADL
ADR
ADR EP
ADR EP-C
ADR EP-G
AENOR
Aggr
AHAM
AIM
AMI
AMR
ANSI
ASHRAE
BEMS
BEyWatch
BIPV
BO
BP
BPL
CDMA
CEN
CENELEC
CNTR
COIT
COSEM
CSWG
CT-IAP
DC
DER
DEWG
DG
DEHEMS
DL
Air Conditioning
Alternating Current
Active Distribution network with full integration of Demand and distributed
energy RESourceS
Architecture Description Language
Automated Demand Response
Automated Demand Response End Point
Automated Demand Response End Point – Consumption
Automatic Demand Response End Point – Generation
Spanish Association for Standardization and Certification
Aggregator
Association of Home Appliance Manufactures
A novel architecture for modelling, virtualizing and managing the energy
consumption of household appliances
Advanced Metering Infrastructure
Automated Meter Reading
American National Standard Institute
American Society of Heating, Refrigerating and Air Conditioning Engineers
Building Energy Management System
Building Energy Watcher
Building-integrated Photovoltaics
Back-Office
Building Prosumer
Broadband Power Line Communications
Code Division Multiple Access
European Committee for Standardization
European Committee for Electrotechnical Standardization
Concentrator
Spanish Official Professional Association of Telecommunications
Companion Specification for Energy Metering
Cyber Security Working Group
Communications Technologies – Interoperability Architectural Perspective
Direct Current
Distributed Energy Resource
Domain Expert Working Group
Distributed Generation
Digital environment home energy management system
Description Logic
Table of Acronyms, Table of Symbols, and References
DLMS
DNP
DOCSIS
DR
DSL
DSM
DSO
DSOr
DSP
EAP
EC
ECC
EIRP
EISA
EMC
EMS
EMV&R
ENCOURAGE
Energy Warden
ENERsip
EPIA
EPRI
ESCO
ESO
ESS
ETSI
EU
EupP
EV
E2E
FIEMSER
FP7
FTP
FTTH
GAD
GHG
GPRS
GW
HAN
HEMS
HESMOS
HVAC
I-BECI
I-BEGI
ICT
ID
IEA
IEC
IED
IEEE
IFC
IMSI
INTEGRIS
IntUBE
IP
IPSec
Device Language Message Specification
Distributed Network Protocol
Data Over Cable Service Interface Specification
Demand Response
Digital Subscriber Line
Demand-Side Management
Distribution System Operation
Distribution System Operator
Digital Signal Processing
Extensible Authentication Protocol
European Commission
Electronic Communications Committee
Effective Isotropic Radiated Power
Energy Independence and Security Act of 2007
Electromagnetic Compatibility
Energy Management System
Energy Monitoring Visualization and Reporting
Embedded iNtelligent COntrols for bUildings with Renewable generAtion and
storaGE
Renewable Energy Sourcing Decisions and Control in Buildings
ENERgy Saving Information Platform for Generation and Consumption
Networks
European Photovoltaic Industry Association
Electric Power Research Institute
Energy Service Company
European Standardization Organization
Energy Storage System
European Telecommunications Standards Institute
European Union
Energy using and producing Products
Electric Vehicle
End-to-end
Friendly Intelligent Energy Management System for Existing Residential
Buildings
Seventh Framework Programme
File Transfer Protocol
Fiber To The Home
Active Demand Side Management Project
Greenhouse Gas
General Packet Radio Service
Gateway
Home Area Network
Home Energy Management System
ICT Platform for Holistic Energy Efficiency Simulation and Lifecycle
Management Of Public Use FacilitieS
Heating Ventilation and Air Conditioning
In-Building Energy Consumption Infrastructure
In-Building Energy Generation Infrastructure
Information and Communications Technologies
Identifier
International Energy Agency
International Electrotechnical Commission
Intelligent Electronic Device
Institute of Electrical and Electronics Engineering
Industry Foundation Classes
International Mobile Subscriber Identity
INTelligent Electrical Grid Sensor communications
Intelligent use of buildings' energy information
Internet Protocol
Internet Protocol Security
- 114 -
Table of Acronyms, Table of Symbols, and References
IR
IREEN
ISO
IT-IAP
ITU-T
IV
JSON
LDV
LEP
LM
LTE
LV
L2TP
MAC
MDMS
MEM
MIRABEL
MSISDN
MSS
MTU
MV
MVO
M2M
NAN
NAT
NB-PLC
ND
NH
NILM
NIALM
NIST
NZEB
OS
OSGP
OSI
OSS
OWL
PAP
PEBBLE
PHEV
PHY
PKI
PLC
PRICE
PRIME
PS-BI
PS-IAP
PSK
PSTN
QoS
RAM
RAN
RA&CA
RC
RCB
RC4
RDF
Infrared
ICT Roadmap for Energy Efficient Neighborhoods
International Organization for Standardization
Information Technology – Interoperability Architectural Perspective
International Telecommunication Union - Telecommunication Standardization
Sector
Initialization Vector
JavaScript Object Notation
Light Duty Vehicle
Local Energy Producer
Load Management
Long Term Evolution
Low Voltage
Layer 2 Tunneling Protocol
Medium Access Control
Metering Data Management System
Microgrid Energy Management
Micro-Request-Based Aggregation, Forecasting and Scheduling of Energy
Demand, Supply and Distribution
Mobile Station Integrated Services Digital Network
Maximum Segment Size
Maximum Transfer Unit
Medium Voltage
Mobile Virtual Operator
Machine-to-Machine
Neighborhood Area Network
Network Address Translation
Narrow Band - Power Line Communications
National Dispatcher
Next Hop
Non-Intrusive Load Monitoring
Non-Intrusive Appliance Load Monitoring
National Institute of Standards and Technologies
Nearly Zero-Energy Buildings
Operating System
Open Smart Grid Protocol
Open Systems Interconnection
Operation Support System
Ontology Web Language
Priority Action Plan
Positive-energy buildings thru better control decisions
Plug-in Hybrid Electric Vehicle
PHYsical
Public Key Infrastructure
Power Line Communications
Joint Project of Intelligent Networks in the Henares Corridor
PoweRline Intelligent Metering Evolution
Power Saving – Business Intelligence
Power Systems – Interoperability Architectural Perspective
Pre-Shared Key
Public Switched Telephone Network
Quality of Service
Random Access Memory
Radio Access Network
Remote Access and Control of Appliances
Residential Consumer
Residential Commercial Building
Rivest Cipher 4
Resource Description Framework
- 115 -
Table of Acronyms, Table of Symbols, and References
REMODECE
REViSITE
ROI
RP
RTU
R&D
SAN
SDH
SEEMPubS
SEP
SGAC
SGAM
SGIP
SGIRM
SGTCC
SIM
SmartCoDe
SNCDP
SOHO
SONET
SSH
SSL
TC
TCP
TKIP
TLS
ToU
TP
TSO
T&D
UAP
UDP
UII
UK
UML
UMTS
UNFCCC
US
USNAP
V2G
VPN
VPP
WEP
WG
Wi-Fi
WiMAX
WLAN
WPA
WPA2
WS
WSAN
WSN
ZEB
3GPP
6LoWPAN
Residential Monitoring to Decrease Energy Use and Carbon Emissions in
Europe
Roadmap Enabling Vision and Strategy for ICT-enabled Energy Efficiency
Return On Investment
Residential Prosumer
Remote Terminal Unit
Research and Development
Sensor and Actuator Network
Synchronous Digital Hierarchy
Smart Energy Efficient Middleware for Public Spaces
Smart Energy Profile
Smart Grid Architecture Committee
Smart Grid Architecture Model
Smart Grid Interoperability Panel
Smart Grid Interoperability Reference Model
Smart Grid Testing and Certification Committee
Subscriber Identity Module
Smart Control of Demand for Consumption and Supply to enable balanced,
energy-positive buildings and neighbourhoods
Sub Network Dependent Convergence Protocol
Small Office Home Office
Synchronous Optical Networking
Secure Shell
Secure Socket Layer
Technical Committee
Transmission Control Protocol
Temporal Key Integrity Protocol
Transport Layer Security
Time of Use
Transformation Point
Transmission System Operator
Transmission & Distribution
User Application Platform
User Datagram Protocol
User Intuitive Interface
United Kingdom
Unified Modeling Language
Universal Mobile Telecommunications System
United Nations Framework Convention on Climate Change
United States
Universal Smart Network Access Port
Vehicle-to-Grid
Virtual Private Network
Virtual Power Plan
Wired Equivalent Privacy
Working Group
Wireless Fidelity
Worldwide Interoperability for Microwave Access
Wireless Local Area Network
Wi-Fi Protected Access
Wi-Fi Protected Access 2
Weather Station
Wireless Sensor and Actuator Network
Wireless Sensor Network
Zero Energy Building
3rd Generation Partnership Project
IPv6 over Low Power Wireless Personal Area Network
- 116 -
Table of Acronyms, Table of Symbols, and References
Table of Symbols
AC
AG
Aggr
B
C
CNS
CS
CS|Aggr
CS|FF
CS|IPsec
CS|TLS/SSL
Cust/TP
DC
Dmax
dmin
FF
(I-BECI|I-BEGI)/Cust
Kb
LT
Mb
MB
OS
Plugs/I-BECI
R
RNS
RS
SC
SC|LT
SC|ST
SG
SG|LT
SG|ST
SIS
ST
T
TCNTR
TP/Sub
U
Number of ADR EP-C (Consumption) per CNTR
Number of ADR EP-G (Generation) per CNTR
Aggregation
Byte
Number of CNTRs per M2M GW
Cost of carrying VNS through the GPRS network (in €)
Cost of carrying VS through the GPRS network (in €)
Cost of carrying VS through the GPRS network (in €) using Aggregation
Cost of carrying VS through the GPRS network (in €) using Fast Forwarding
Cost of carrying VS through the GPRS network (in €) using IPSec
Cost of carrying VS through the GPRS network (in €) using TLS/SSL
Number of Customers per TP
Cost of using the corresponding security solution
Maximum acceptable distance between Customers and TPs
Minimum density of Customers per TP
Fast Forwarding
Penetration of micro-generation
Kilobits per second
Long-term scenario
Megabits per second
MegaByte
Overhead introduced by the security protocol (in %)
Number of appliances per I-BECI
Rural scenario
Ratio between the application-layer data and VNS (in %)
Ratio between the application-layer data and VS (in %)
Size of the consumption data at the application layer
Size of the consumption data at the application layer in long-term scenarios
Size of the consumption data at the application layer in short-term scenarios
Size of the generation data at the application layer
Size of the generation data at the application layer in long-term scenarios
Size of the generation data at the application layer in short-term scenarios
Size of the application messages sent either by the users or automatically
generated by the PS-BI module
Short-term scenario
Periodicity which ADR EPs send data with
Time that the CNTR spends to transfer the data to the M2M GW
Number of TPs/Substation
Urban scenario
- 117 -
Table of Acronyms, Table of Symbols, and References
VNS
VS
µ
µIS
σ
Volume of traffic (in MB) carried by the GPRS network in one month without
using any security protocol
Volume of traffic (in MB) carried by the GPRS network in one month using
the corresponding security protocol
Mean
Average periodicity of sending requests either by the users or automatically
generated by the PS-BI module
Standard deviation
- 118 -
Table of Acronyms, Table of Symbols, and References
References
[Abdul Salam2012]
S. Abdul Salam, S. Mahmud, G. Khan, H. Al-Raweshidy, “M2M
communication in Smart Grids: Implementation scenarios and performance
analysis”, IEEE Wireless Communications and Networking Conference
Workshops, Paris, France, 2012
[Adapt4EE2014]
Adapt4EE project Official Web Site: http://www.adapt4ee.eu/adapt4ee/
[Accessed 2014/1/5]
[AHAM2012]
“Smart Grid White Paper: The Home Appliance Industry’s Principles &
Requirements for Achieving a Widely Accepted Smart Grid”, AHAM, 2009
[Aidine2013]
A. Aidine, A. Tabone, J. Muller, “Deployment of Power Line Communication
by European Utilities in Advanced Metering Infrastructure”, IEEE ISPLC
2013, Johannesburg, South Africa, 2013
[AIM2010]
AIM project Official Web Site: http://www.ict-aim.eu/ [Accessed 2014/1/5]
[Ancillotti2013]
E. Ancillotti, R. Bruno, M. Conti, “The role of communication systems in
smart grids: Architectures, technical solutions and research challenges”,
Computer Communications, Vol. 36, Issues 17-18, Pages 1665-1697, 2013
[Anderson2012a]
D. Anderson, Z. Chuanlin, C. Hauser, V. Venkatasubramanian, D. Bakken, A.
Bose, “Real-Time Simulation for Smart Grid Control and Communications
Design”, IEEE Power&Energy Magazine, 10 (1), Pages 49-57, January 2012
[Anderson2012b]
K. Anderson, S. Lee, C. Menassa, “Effect of social network type on building
occupant energy use”, BuildSys 2012: Fourth ACM Workshop on Embedded
Sensing Systems for Energy-Efficiency in Buildings, Pages 17-24, 2012
[Anjum2012]
N. A. Anjum, J. A. Harding, R. I. Young, K. Case, “Mediation of foundation
ontology based knowledge sources”, Computers in Industry, Volume 63, Pages
433-442, 2012
[Apostolov2013]
A. Apostolov, “Integration of Distributed Energy Resources in Smart Grids”,
CIRED 2013: 22nd European Conference and Exhibition on Electricity
Distribution, Stockholm, Sweden, June 2013
- 119 -
Table of Acronyms, Table of Symbols, and References
[Aravinthan2011]
V. Aravinthan, B. Karimi, V. Namboodiri, W. Jewell, “Wireless
Communication for Smart Grid Applications at Distribution Level – Feasibility
and Requirements”, IEEE PES GM: Power and Energy Society General
Meeting, Detroit, MI USA, July 2011
[Aschenbruck2004]
N. Aschenbruck, M. Frank, W. Hansmann, P. Martini, C. Scholz, J. Tölle,
“Integration of 3G Protocols into the Linux Kernel to Enable the Use of
Generic Bearers”, High Speed Networks and Multimedia Communications,
Pages 533-544. Springer Berlin Heidelberg, 2004
[Aust2012]
S. Aust, R. Prasad, I. Niemegeers, “IEEE 802.11ah: Advantages in standards
and further challenges for sub 1 GHz Wi-Fi”, IEEE International Conference
on Communications, Ottawa, Canada, 2012, Pages 6885-6889
[Ayhana2012]
D. Ayhana, S. Sağlam, “A technical review of building-mounted wind power
systems and a sample simulation model”, Renewable and Sustainable Energy
Reviews, Volume 16, Issue 1, Pages 1040–1049, 2012
[Bartak2013]
G. F. Bartak, A. Abart, “EMI of emissions in the frequency range 2 kHz - 150
kHz”, CIRED 2013: 22nd European Conference and Exhibition on Electricity
Distribution, Stockholm, Sweden, June 2013
[Belhomme2008]
R. Belhomme, R. Cerero, G. Valtorta, A. Paice, F. Bouffard, R. Rooth,
”ADDRESS – Active Demand for the Smart Grid of the Future”, CIRED
seminar 2008: Smart Grids for Distribution, Frankfurt, Germany, June 2008
[Benko2004]
P. Benko, G. Malicsko, A. Veres, “A Large-scale, Passive Analysis of End-toEnd TCP Performance over GPRS”, IEEE INFOCOM 2004, Hong-Kong, 2004
[Berger2006]
T. Berger, “Analysis of current VPN technologies”, IEEE ARES 2006: 1st
International Conference on Availability, Reliability and Security, 2006
[BEyWatch2011]
BEYWatch
project
summary:
http://www.innovationseeds.eu/Virtual_Library/Results/BEYWATCH.kl#.Us8
-ubS0aQY [Accessed 2014/1/5]
[Brew2011]
Brew M., Darbari F., Crockett L.H., Waddell M.B., Fitch M., Weiss S., Stewart
R.W., “UHF White Space Network for Rural Smart Grid Communications”,
IEEE SmartGridComm: International Conference on Smart Grid
Communications, October 2011
[Caldevilla2013]
A. Caldevilla, “EU FP7 project PowerUp and Beyond”, 4th ETSI M2M
Workshop, Mandelieu-la-Napoule, France, November 2013
[Carreiro2011]
A. Carreiro, G. López, P. Moura, J. I. Moreno, A. de Almeida, J. Malaquias,
“In-house monitoring and control network for the Smart Grid of the future”,
ISGT Europe 2011: 2nd IEEE PES International Conference and Exhibition on
Innovative Smart Grid Technologies, Manchester, UK, December 2011
[Carvallo2011]
A. Carvallo, J. Cooper, “The Advanced Smart Grid Edge Power Driving
Sustainability”, Artech House, 2011. ISBN-13: 978-1608071272
[Celli2013]
G. Celli, E. Ghiani, F. Pilo, G. G. Soma, “Reliability assessment in smart
distribution networks”, Electric Power Systems Research, Vol. 104, Pages 164175, 2013
[Chorus2014]
Chorus
Official
Web
Site:
http://store.greenenergyoptions.co.uk/dp/B00CFNAXVG [Accessed 2014/1/5]
- 120 -
Table of Acronyms, Table of Symbols, and References
[Claudy2011]
M. C. Claudy, C. Michelsen, A. O’Driscoll, “The diffusion of microgeneration
technologies – assessing the influence of perceived product characteristics on
home owners' willingness to pay”, Energy Policy, Volume 39, Issue 3, Pages
1459-1469, March 2011
[Clement2011]
K. Clement-Nyns, E. Haesen, J. Driesen, “The impact of vehicle-to-grid on the
distribution grid”, Electric Power Systems Research, Volume 81, Issue 1,
Pages 185-192, 2011
[Cloogy2014]
Cloogy Official Web Site: http://www.cloogy.com/en/ [Accessed 2014/1/5]
[De Craemer2010]
K. De Craemer, G. Deconinck, “Analysis of State-of-the-art Smart Metering
Communications Standards”, Proceedings of the 5th Young Researchers
Symposium, 2010
[DEHEMS2011]
DEHEMS project Official Web Site: http://www.dehems.eu/ [Accessed
2014/5/1]
[Della Giustina2011]
D. Della Giustina, S. Zanini, S. Pugliese, L. Cremaschini, “The role of
INTEGRIS in the A2A Reti Elettriche's roadmap toward the smart grid ”,
SMFG2011:IEEE International Conference on Smart Measurements for Future
Grids, Bologna, Italy, November 2011
[Depuru2011]
S. S. S. R. Depuru, L. Wang, V. Devabhaktuni, “Smart meters for power grid:
Challenges, issues, advantages and status”, Renewable and Sustainable Energy
Review, Volume 15, Issue 6, Pages 2736-2742
[Drake2010]
J. Drake, D. Najewicz, W. Watts, “Energy Efficiency Comparisons of Wireless
Communication Technology Options for Smart Grid Enabled Devices”,
General Electric, December 2010
[EC2009]
Standardization mandate M/441, “Standardisation mandate to CEN,
CENELEC and ETSI in the field of measuring instruments for the development
of an open architecture for utility meters involving communications protocols
enabling
interoperability”,
EC,
March
2009.
On-line:
http://www.etsi.org/images/files/ECMandates/m441%20EN.pdf
[Accessed
2014/1/5]
[EC2010]
Standardization mandate M/468, “Standardisation mandate to CEN,
CENELEC and ETSI concerning the charging of electric vehicle, EC, June
2010.
On-line:
http://ec.europa.eu/energy/gas_electricity/smartgrids/doc/2010_06_04_mandat
e_m468_en.pdf [Accessed 2014/1/5]
[EC2011a]
“Energy Roadmap 2050”, Communication from the Commission to the
European Parliament, the Council, the European Economic and Social
Committee and the Committee of the Regions, 2011
[EC2011b]
Standardization mandate M/490, “Standardization Mandate to European
Standardisation Organisations (ESOs) to support European Smart Grid
deployment”,
EC,
March
2011.
On-line:
http://ec.europa.eu/energy/gas_electricity/smartgrids/doc/2011_03_01_mandat
e_m490_en.pdf [Accessed 2014/1/5]
[EC2013]
“Study on the available semantics assets for the interoperability of Smart
Appliances. Mapping into a common ontology as a M2M application layer
semantics” EC Tender Specifications, SMART 2013/0077. On-line:
http://ec.europa.eu/digital-agenda/en/news/invitation-tender-study-availablesemantics-assets-interoperability-smart-appliances-mapping
[Accessed
2014/1/5]
- 121 -
Table of Acronyms, Table of Symbols, and References
[ECC2010]
“Numbering and Addressing in M2M Communications”, Draft ECC Report
153, Vilnius, June 2010
[EEI2011]
“Smart Meters and Smart Meter Systems: A Metering Industry Perspective";
An EEI-AEIC-UTC White Paper Smart Grid white paper; March 2011
[Efergy2014]
Efergy Official Web Site: http://efergy.com/eu/ [Accessed 2014/1/5]
[Ekanayake2012]
Ekanayake J., Liyanage K., Wu J., Yokoyama A., Jenkins N., “Smart Grid:
Technology and Applications”, IEEE Press, 2012. ISBN-13: 978-0470974094
[El achab2013]
E. El achab, G. López. J. I. Moreno, “Evaluación de mecanismos de seguridad
en entornos de Smart Grid”, JITEL 2013: XI Jornadas de Ingeniería
Telemática, Granada, Spain, October 2013
[EM2014]
GreenWave Reality Energy Management solution Official Web Site:
http://www.greenwavereality.com/solutions/energymgmt/ [Accessed 2014/1/5]
[eMonitor2014]
eMonitor Official Website: http://www.powerhousedynamics.com/residentialenergy-efficiency/ [Accessed 2014/1/5]
[ENCOURAGE2014]
ENCOURAGE Project Official Web Site: http://www.encourage-project.eu/
[Accessed 2014/1/5]
[Engage2014]
Engage Official Website: http://engage.efergy.com/ [Accessed 2014/1/5]
[Enlighten2014]
Enlighten Official Website: http://enphase.com/enlighten/ [Accessed 2014/1/5]
[EnOcean2014]
EnOcean Official Website: http://www.enocean.com/en/home/ [Accessed
2014/1/5]
[EPIA2012]
“Connecting the Sun: Solar Photovoltaics on the road to large-scale grid
integration”, European Photovoltaic Industry Association, September 2012
[EPIA2013]
“Global Market Outlook for Photovoltaics 2013-2017”, European Photovoltaic
Industry Association, September 2012, May 2013
[EPRI2011]
“Estimating the Costs and Benefits of the Smart Grid – A Preliminary Estimate
of the Investment Requirements for a Fully Functioning Smart Grid”, Electric
Power Research Institute, 2011
[ESMA2012]
ESMA, “What is Smart Metering?”, European Smart Metering Alliance, 2012
[ETSI2012]
“Machine-to-Machine communications (M2M); Applicability of M2M
architecture to Smart Grid Networks; Impact of Smart Grids on M2M
platform”, ETSI Technical Report 102 935, September 2012.
[EUCR2008]
“20 20 by 2020: Europe’s climate change opportunity”, EU Communication
Report, 2008
[EURELECTRIC2011]
“10 Steps to Smart Grids”, Union of the Electricity Industry –
EURELECTRIC, 2011
[Eurostat2013]
Eurostat
Energy
http://epp.eurostat.ec.europa.eu/portal/page/portal/energy
2013/11/12]
[EW2012]
Energy
Warden
project
Official
Web
http://buildingwarden.com/energywrdn/ [Accessed 2014/1/5]
- 122 -
Statistics:
[Accessed
Site:
Table of Acronyms, Table of Symbols, and References
[De Almeida2011]
A. De Almeida, P. Fonseca, B. Schlomann, N. Feilberg, “Characterization of
the Household Electricity Consumption in the EU, Potential Energy Savings
and Specific Policy Recommendations”, Energy and Buildings, Volume 43,
Issue 8, Pages 1884–1894, August 2011
[Fadlullah2011]
Z. Fadlullah, M. Fouda, N. Kato, A. Takeuchi, N. Iwasaki, Y. Nozaki,
“Towards Intelligent Machine-to-Machine Communications in Smart Grid”,
IEEE Communications Magazine, Volume 49, Issue 4, Pages 60-65, April
2011
[Farhangi2010]
H. Farhangi, “The Path of the Smart Grid”, IEEE Power and Energy Magazine,
Volume 8, Issue 1, Pages 18 – 28, January-February 2010. ISSN: 1540-7977
[Faria2012]
R. Faria, P. Moura, J. Delgado, A. de Almeida, “A sustainability assessment of
electric vehicles as a personal mobility system”, Energy Conversion and
Management, Volume 61, Issue 1, Pages 19 – 30, 2012. ISSN: 0196-8904
[FIEMSER2013]
FIEMSER project summary: http://www.ice-wish.eu/uk/ec-cofunded-projectsrelated-to-ict-and-building/fiemser.asp [2014/1/5]
[Firth2008]
S. Firth, K. Lomas, A. Wright, R. Wall, “Identifying trends in the use of
domestic appliances from household electricity consumption measurements”,
Energy and Buildings, Volume 40, Issue 5, Pages 926–936, 2008
[GAD2010]
GAD project Official Web Site: http://gad.ite.es/index_en.html [Accessed
2014/1/5]
[Gellings2009]
C. Gellings, “The Smart Grid: enabling energy efficiency and demand
response”, CRC, New York, 2009. ISBN-13: 978-1439815748
[Gislason2008]
D. Gislason, “Zigbee Technical Overview”, Wireless Japan, Tokyo, Japan,
2008
[Goebel2013]
C. Goebel, D. Callaway, “Using ICT-controlled plugin electric vehicles to
supply grid regulation in California at different renewable integration levels”,
IEEE Transactions on Smart Grid, Volume 4, Issue 2, Pages 729-740, June
2013. ISSN: 1949-3053
[Gómez2007]
A. Gómez-Pérez, M. Fernández-López, O. Corcho, “Ontological Engineering:
With Examples from the Areas of Knowledge Management, E-Commerce and
the Semantic Web”, Advanced Information and Knowledge Processing,
Springer-Verlag, 2007. ISBN 978-1-85233-551-9
[Grassi2011]
M. Grassi, M. Nucci, F. Piazza, “Towards an ontology framework for
intelligent smart home management and energy saving”, IEEE International
Symposium on Industrial Electronics, Pages 1753–1758, 2011
[GBD2014]
Green Button Data
[Accessed 2014/1/5]
[Greenvity2014]
Greenvity Official Website: http://www.greenvity.com/products-gv7011.htm
[Accessed 2014/1/5]
[Güngör2011]
V. C. Güngör, D. Sahin, T. Kocak, S. Ergüt, C. Buccella, C. Cecati, G. P.
Hancke, “Smart Grid Technologies: Communication Technologies and
Standards”, IEEE Transactions on Industrial Informatics, Volume 7, Issue 4,
Pages 529-539, November 2011
Official
- 123 -
Website:
http://www.greenbuttondata.org/
Table of Acronyms, Table of Symbols, and References
[Güngör2013]
V. C. Güngör, D. Sahin, T. Kocak, S. Ergüt, C. Buccella, C. Cecati, G. P.
Hancke, “A Survey on Smart Grid Potential Applications and Communication
Requirements”, IEEE Transactions on Industrial Informatics, Volume 9, Issue
1, Pages 28-42, February 2013
[Gyrard2013]
A. Gyrard, C. Bonnet, K. Boudaoud, “A Machine-to-Machine Architecture to
Merge Semantic Sensor Measurements”, WWW2013, Rio de Janeiro, Brazil,
2013.
[Hernández2013]
L. Hernández, C. Baladron, J. M. Aguiar, B. Carro, A. Sánchez-Esguevillas, J.
Lloret, D. Chinarro, J. J. Gómez-Sanz, D. Cook, “A Multi-Agent System
Architecture for Smart Grid Management and Forecasting of Energy Demand
in Virtual Power Plants”, IEEE Communications Magazine, Volume 51, Issue
1, Pages 106-113, January 2013. ISSN: 0163-6804
[HESMOS2013]
HESMOS project Official Website: http://hesmos.eu/ [2014/1/5]
[Hiertz2010]
G. R. Hiertz, D. Denteneer, L. Stibor, Y. Zang, X. Pérez-Costa, B. Walke, “The
802.11 Universe”, IEEE Communications Magazine, Volume 48, Issue 1,
January 2010
[Hk2014]
electronic
Housekeeper
Official
Web
http://www.electronichousekeeper.com/ [Accessed 2014/1/5]
[IBM2011]
“The Value of Smarter Energy: Orchestrating the Network End-to-End”, IBM
Center for Applied Insights
[IEA2011]
“Technology Roadmaps: Smart Grids”, International Energy Agency, 2011
[IEA2013]
“Transition to Sustainable Buildings: Strategies and Opportunities to 2050”,
International Energy Agency, 2013
[IEC2014]
IEC Smart Grid Standards Mapping Tool Official
http://smartgridstandardsmap.com/ [Accessed 2014/1/5]
[IEEE2006]
“Wireless Medium Access Control and Physical Layer Specifications for LowRate Wireless Personal Area Networks,” IEEE Std 802.15.4-2006
[IEEE2011]
“Guide for Smart Grid Interoperability of Energy Technology and Information
Technology Operation with the Electric Power System (EPS), and the End-Use
Applications and Loads”, IEEE standard 2030, September 2011
[IEEE2014]
Approved
&
Proposed
IEEE
Smart
Grid
standards:
http://smartgrid.ieee.org/standards/ieee-approved-proposed-standards-relatedto-smart-grid [Accessed 2014/1/5]
[IETF2007]
G. Montenegro, K. Kushalnagar, J. Hui, D. Culler, “Transmissions of IPv6
packets over IEEE 802.15.4 Networks”, RFC 4944, September 2007. Online:
http://www.rfc-editor.org/rfc/rfc4944.txt [Accessed 2014/1/10]
[Insteon2014]
Insteon Official Website: http://www.insteon.com/ [Accessed 2014/1/5]
[IntUBE2011]
INTUBE project summary: http://cordis.europa.eu/projects/rcn/86722_en.html
[Accessed 2014/1/5]
[IREEN2013]
IREEN project Official Web Site: http://www.ireenproject.eu/ [Accessed
2014/1/5]
[ISO2011]
ISO/IEC/IEEE 42010 Official Website: http://www.iso-architecture.org/42010/
[Accessed 2013/11/10]
- 124 -
site:
Web
Site:
Table of Acronyms, Table of Symbols, and References
[ISO2013]
ISO/PRC 16739. Industry Foundation Classes (IFC) for data sharing in the
construction
and
facility
management
industries.
On-line:
http://www.iso.org/iso/home/store/catalogue_ics/catalogue_detail_ics.htm?csn
umber=51622 [Accessed 2013/11/18]
[ITU2012a]
ITU-T standard G.9903, “Narrowband orthogonal frequency division
multiplexing power line communication transceivers for G3-PLC networks”,
2012
[ITU2012b]
ITU-T standard G.9904, “Narrowband orthogonal frequency division
multiplexing power line communication transceivers for PRIME networks”,
2012
[Jellea2012]
B. Jellea, C. Breivikb, H. Røkenes, “Building integrated photovoltaic products:
A state-of-the-art review and future research opportunities”, Solar Energy
Materials and Solar Cells, Volume 100, Pages 69–96, May 2012
[JRC2012a]
“Guidelines for Cost Benefit Analysis of Smart Metering Deployment”, JRC
Scientific and Technical Research, 2012
[JRC2012b]
“Smart Grid projects in Europe: Lessons learned and current developments”,
JRC
Scientific
and
Policy
Report,
2012.
On-line:
http://ses.jrc.ec.europa.eu/sites/ses.jrc.ec.europa.eu/files/documents/ld-na25815-en-n_final_online_version_april_15_smart_grid_projects_in_europe__lessons_learned_and_current_developments_-2012_update.pdf
[Accessed
2014/1/5]
[Kezunovic2011]
M. Kezunovic, “Translational Knowledge: From Collecting Data to Making
Decisions in a Smart Grid” Proceedings of the IEEE, Volume 99, Issue 6,
Pages 977-997, 2011
[Khan2013]
R. H. Khan, J. Y. Khan, “A comprehensive review of the application
characteristics and traffic requirements of a smart grid communications
network”, Computer Networks, Volume 57, Issue 3, Pages 825–845, 2013
[Khanvilkar2004]
S. Khanvilkar, A. Khokhar, “Virtual Private Networks: An Overview with
Performance Evaluation”, IEEE Communications Magazine, Volume 42, Issue
10, Pages 146-154, 2004
[Khosroshahy2007]
M. Khosroshahy, T. Turletti, K. Obraczka, “Snapshot of MAC, PHY and
Propagation Models for IEEE 802.11 in Open-Source Network Simulators”,
INRIA Research Report, September 2007. Online: http://hal.inria.fr/inria00175359/en/ [2014/1/12]
[Kulkarni2012]
P. Kulkarni, S. Gormus, F. Zhong, F. Ramos, “AMI Mesh Networks—A
Practical Solution and Its Performance Evaluation”, IEEE Transactions on
Smart Grid, Volume 3. Issue 3, Pages 1469-1481, 2012
[Lévesque2012]
M. Lévesque, D. Q. Xu, G. Joós, M. Maier, “Communications and power
distribution network co-simulation for multidisciplinary smart grid
experimentations”, ANSS 2012, San Diego, USA, 2012
[Lin2011]
H. Lin, S. Sambamoorthy, S. Shukla, J. Thorp, L. Mili, “Power system and
communication network co-simulation for smart grid applications”, in IEEE
PES Innovative Smart Grid Technologies, 2011, Hilton Anaheim, CA, Pages 16
[Liu2012]
J. Liu, Y. Xiao, S. Li, W. Liang, C. L. Chen, “Cyber Security and Privacy
Issues in Smart Grids”, IEEE Communications Surveys and Tutorials, Volume
14, Issue 4, Pages 981-997, 2012
- 125 -
Table of Acronyms, Table of Symbols, and References
[Lo2012]
C.-H. Lo, N. Ansari, “The Progrssive Smart Grid System from both Power and
Communications Aspects”, IEEE Communications Surveys & Tutorials,
Volume 14, Issue 3, Pages 799-821, 2012
[López2011a]
G. López, P. Moura, J. I. Moreno, A. de Almeida, “ENERsip: M2M-based
platform to enable energy efficiency within energy-positive neighborhoods”,
IEEE INFOCOM 2011 Workshop on M2M Communications and Networking,
Shanghai, China, April 2011
[López2011b]
G. López, P. Moura, M. Sikora, J. I. Moreno, “Comprehensive validation of an
ICT platform to support energy efficiency in future smart grid scenarios”,
SMFG2011:IEEE International Conference on Smart Measurements for Future
Grids, Bologna, Italy, November 2011
[López2011c]
G. López, J. I. Moreno, “Smart Energy-positive Neighbourhoods for the Smart
Grid. Architecture, Communications Technologies and Address Management
for the ENERsip platform”, 2011 ETSI Workshop on “Standards: An
Architecture for the Smart Grid”. Sophia Antipolis, France, April 2011.
[López2012a]
G. López, P. Moura, V. Custodio, J. I. Moreno, “Modeling the Neighborhood
Area Networks of the Smart Grid”, IEEE ICC 2012, Ottawa, Canada, June
2012
[López2012b]
G. López, P. Moura, B. Kantsepolsky, M. Sikora, J. I. Moreno, A. de Almeida,
“European FP7 Project ENERsip: Bringing ICT and Energy Together”, Global
Communications Newsletter, November 2012
[López2013a]
G. López, J. Moreno, P. Moura, A. de Almeida, M. Perez, L. Blanco,
“Monitoring System for the Local Distributed Generation Infrastructures of the
Smart Grid”, CIRED 2013: 22nd European Conference and Exhibition on
Electricity Distribution, Stockholm, Sweden, June 2013
[López2013b]
G. López, V. Custodio, F. J. Herrera, J. I. Moreno, “Machine-to-Machine
Communications Infrastructure for Smart Electric Vehicle Charging in Private
Parking Lots”, International Journal of Communication System, Wiley, 2013.
DOI: 10.1002/dac.2705
[López2013c]
G. López, J. I. Moreno, “PRICE Project: M2M communications architecture
for large-scale AMI deployment”, 4th ETSI M2M Workshop, Mandelieu-laNapoule, France, November 2013
[Ma2012]
Y. Ma, T. Houghton, A. Cruden, D. Infield, “Modeling the Benefits of Vehicleto-Grid Technology to a Power System”, IEEE Transactions on Power
Systems, Volume 27, Issue 2, Pages 1012-1020, 2012
[Majumder2013]
R. Majumder, G. Bag, G. Velotto, A. Marinopoulos, “Closed loop simulation
of communication and power network in a zone based system”, Electric Power
Systems Research, Volume 95, Pages 247-256, 2013
[Mankoff2010]
J. Mankoff et al, “StepGreen.org: Increasing energy saving behaviors via social
networks”, Proceedings of the 4th International AAAI Conference on Weblogs
and Social Media, May 2010
[Manola2004]
F. Manola, E. Miller, B. McBride. RDF Primer. W3C Recommendation 2004.
On-line: http://www.w3.org/TR/2004/REC-rdf-primer-20040210/ [Accessed
2013/6/10]
[Mao2011]
R. Mao, V. Julka, “Wireless Broadband Architecture Supporting Advanced
Metering Infrastructure”, IEEE VTC Spring, Budapest, Hungary, 2011
- 126 -
Table of Acronyms, Table of Symbols, and References
[Matanza2013]
J. Matanza, S. Alexandres, C. Rodríguez-Morcillo, “Performance evaluation of
two narrowband PLC systems: PRIME and G3”, Computer Standards &
Interfaces, Volume 36, Issue 1, Pages 198-208, 2013
[McDaniel2009]
P. McDaniel, S. McLaughlin, “Security and Privacy Challenges in the Smart
Grid”, IEEE Security and Privacy, Volume 7, Issue 3, Pages 75-77, May/June
2009
[Mets2011]
K. Mets, T. Verschueren, C. Develder, T. Vandoorn, L. Vandevelde,
“Integrated simulation of power and communication networks for smart grid
applications”, in IEEE CAMAD 2011 International Workshop on Computer
Aided Modeling and Design of Communication Links and Networks, 2011,
Kyoto, Japan, Pages 61-65
[Mets2012]
K. Mets, R. D’hulst, C. Develder, “Comparison of intelligent charging
algorithms for electric vehicles to reduce peak load and demand variability in a
distribution grid”, Journal of Communications and Networks, Volume 14, Issue
6, Pages 672-681, 2012
[Meyer1999]
M. Meyer, “TCP Performance over GPRS”, IEEE WCNC 1999, New Orleans,
USA, 1999
[Miller2013]
H. G. Miller, P. Mork, “From Data to Decisions: A Value Chain for Big Data”,
IT Professional, Volume 15, Issue 1, Pages 57-59, 2013
[MIRABEL2013]
MIRABEL project
[Accessed 2014/1/5]
[Momoh2012]
J. Momoh, “Smart Grid: Fundamentals of Design and Analysis”, IEEE Press,
2012. ISBN-13: 978-0470889398
[Motamedi2012]
A. Motamedi, H. Zareipour, W. D. Rosehart, “Electricity Price and Demand
Forecasting in Smart Grids” IEEE Transactions on Smart Grid, Volume 3,
Issue 2, Pages 664-674, 2012
[Moura2010a]
P. Moura, A. de Almeida, “Multi-Objective Optimization of a Mixed
Renewable System with Demand-Side Management”, Renewable &
Sustainable Energy Reviews, Volume 14, Issue 5, Pages 1461-1468, 2010
[Moura2010b]
P. Moura, A. de Almeida, “The Role of Demand-Side Management in the Grid
Integration of Wind Power”, Applied Energy, Volume 87, Issue 8, Pages 2581
- 2588, 2010
[Moura2012]
P. Moura, G. López, A. Carreiro, J. I. Moreno, A. de Almeida, “Evaluation
Methodologies and Regulatory Issues in Smart Grid Projects with Local
Generation-Consumption Matching”, EEMSW2012: International Workshop
on Energy Efficiency for a More Sustainable World, São Miguel, Portugal,
September 2012
[Moura2013a]
P. Moura, G. López, J. I. Moreno, A. de Almeida, “Impact of Residential
Demand Response on the Integration of Intermittent Renewable Generation
into the Smart Grid”, EEDAL2013: 7th International Conference on Energy
Efficiency in Domestic Appliances and Lighting, Coimbra, Portugal,
September 2013
[Moura2013b]
P. Moura, G. López, J. I. Moreno, A. de Almeida, “The role of Smart Grids to
foster energy efficiency”, Energy Efficiency, Volume 6, Issue 4, Pages 621639, November 2013
Official
- 127 -
Web
Site:
http://www.mirabel-project.eu/
Table of Acronyms, Table of Symbols, and References
[NIST2010a]
“NIST Framework and Roadmap for Smart Grid Interoperability Standards,
Release 1.0”, NIST Special Publication 1108, January 2010
[NIST2010b]
“Wireless Functionality and Characteristics Matrix for the Identification of
Smart Grid Domain Application”, NIST PAP2, January 2010
[NIST2011]
“Guidelines for Assessing Wireless Standards for Smart Grid Applications”,
NIST PAP2, February 2011
[NIST2012]
“NIST Framework and Roadmap for Smart Grid Interoperability Standards,
Release 2.0”, NIST Special Publication 1108R2, February 2012
[NIST2014]
NIST Smart Grid Collaboration Wiki for Smart Grid Interoperability
Standards:
http://collaborate.nist.gov/twikisggrid/bin/view/SmartGrid/SGIPCatalogOfStandards [Accessed 2014/1/5]
[OneM2M2014]
OneM2M Official Website: http://www.onem2m.org/ [Accessed 2014/1/5]
[Osello2013]
A. Osello et al, “Energy saving in existing buildings by an intelligent use of
interoperable ICTs”, Energy Efficiency, Volume 6, Issue 4, Pages 621-639,
November 2013
[Oksman2011]
V. Oksman, J. Zhang, “G.HNEM: the new ITU-T standard on narrowband PLC
technology”, IEEE Communications Magazine, Volume 49, Issue 12, Pages
36-44, 2011
[OWL2014]
OWL Official Web Site: http://www.theowl.com/ [Accessed 2014/1/5]
[Pan2012]
J. Z. Pan, S. Staab, U. Assman, J. Ebert, Y. Zhao, “Ontology-Driven Software
Development”, Springer, 2012, ISBN: 978-3-642-31226-7
[Pang2012]
C. Pang, P. Dutta, M. Kezunovic, “BEVs/PHEVs as Dispersed Energy Storage
for V2B Uses in the Smart Grid”, IEEE Transactions on Smart Grid, Volume 3,
Issue 1, Pages 473-482, 2012
[Papadopoulos2013]
T. A. Papadopoulos, A. I. Chrysochos, G. K. Papagiannis, “Narrowband power
line communication: Medium voltage cable modeling and laboratory
experimental results”, Electric Power Systems Research, Volume 102, Pages
50-60, 2013
[PEBBLE2012]
PEBBLE project Official Web Site: http://www.pebble-fp7.eu/ [Accessed
2014/1/5]
[PR2012]
“Smart Meters in Europe - Advanced Metering Infrastructure for Electric
Utilities in Europe: Business and Technology Issues, Country Profiles, Key
Industry Players, and Market Forecasts”, Pike Research, 2012.
[Protégé2013]
Protégé Official Website: http://protege.stanford.edu/ [Accessed 2013/11/18]
[Rendón2001]
J. Rendón, F. Casadevall, D. Serarols, “Snoop TCP Performance over GPRS”,
IEEE VTC Spring 2001, Rhodes, Greece, 2001
REViSITE project Official Web Site: http://www.revisite.eu/ [Accessed
2014/1/5]
[REViSITE2012]
[Robichon2013]
G. Robichon, “CDMA-450: a new approach for DSO’s in Europe”, 4th ETSI
M2M Workshop, Mandelieu-la-Napoule, France, 2013
[SGprojects2014]
EU
Smart
Grids
Projects
Portal:
https://portal.smartgridsprojects.eu/Pages/default.aspx [Accessed 2014/1/5]
- 128 -
Table of Acronyms, Table of Symbols, and References
[SmartCoDe2012]
SmartCoDe project Official Web Site: https://www.fp7-smartcode.eu/overview
[Accessed 2014/1/5]
[Sakr2011]
Z. Sakr, A. Liu, D. M. Batista, M. Alomari, “A Survey of Large Scale Data
Management Approaches in Cloud Environments”, IEEE Communications
Surveys and Tutorials, Volume 13, Issue 3, Pages 311-336, 2011
[Santodomingo2012]
R. Santodomingo, J. A. Rodríguez-Mondéjar, M. A. Sanz-Bobi, “Using
Semantic Web Resources to Translate Existing Files Between CIM and IEC
61850”, IEEE Transactions on Power Systems, Volume 27, Number 4, 2012
[Sastry2004]
N. Sastry, D. Wagner, “Security Considerations for IEEE 802.15.4 Networks”,
WiSE’04: ACM Workshop on Wireless Security 2004
[SGCG2012a]
“Smart Grid Reference Architecture”, SG-CG, November 2012
[SGCG2012b]
“Smart Grid Reference Architecture – Annex K”, SG-CG, November 2012
[SGCG2012c]
“First Set of Standards”, SG-CG, November 2012
[SH2014]
Fifthplay
Smart
Home
solution
Official
Web
Site:
http://www.fifthplay.com/en/smart-homes/the-solution [Accessed 2014/1/5]
[SHM2014]
Sunny
Home
Manager
Official
Web
http://www.sma.de/en/produkte/monitoring-systems/sunny-homemanager.html [Accessed 2014/1/5]
[Short2005]
T. A. Short, “Electric Power Distribution Equipment and Systems”, CRC
Press, November 2005. ISBN-13: 978-0849395765
[Shrestha2012]
G. M. Shrestha, J. Jasperneite, “Performance Evaluation of Cellular
Communication Systems for M2M Communication in Smart Grid
Applications”, Computer Networks, Pages 352-359. Springer Berlin
Heidelberg, 2012
Smart House/Smart Grid project Official Web Site: http://www.smarthousesmartgrid.eu/ [2014/1/5]
[SHSG2011]
Site:
[Sioshansi2012]
F. Sioshansi, “Smart Grid: Integrating Renewable, Distributed and Efficient
Energy”, Elsevier, 2012. ISBN-13: 978-0123864529
[Stecher2008]
R. Stecher, C. Niederée, Nejdl, P. Bouquet, “Adaptive ontology re-use: finding
and re-using sub-ontologies”, International Journal of Web Information
Systems Volume 4, Issue 2, Pages 198-214, 2008
C. Timpe, “Smart Domestic Appliances Supporting the System Integration of
Renewable Energy, SMART-A project, 2009
[Timpe2009]
[Toffler1980]
A. Toffler, “The Third Wave”, Bantam, 1980. ISBN-13: 978-0553246988
[Tsarkov2006]
D. Tsarkov, I. Horrocks, “FaCT++ Description Logic Reasoner: System
Description”, Automated Reasoning, Lecture Notes in Computer Science,
4130, Pages 292-297, Springer, 2006
[Usman2013]
A. Usman, S. H. Shami, “Evolution of Communication Technologies for Smart
Grid applications”, Renewable and Sustainable Energy Reviews, Volume 19,
Pages 191-199, March 2013
[UNKP1997]
“Kyoto Protocol”, United Nations Framework Convention on Climate Change,
2009
- 129 -
Table of Acronyms, Table of Symbols, and References
[Varga2008]
A. Varga, R. Hornig, “An overview of the OMNeT++ simulation
environment”, Simutools08 1st International Conference on Simulation tools
and Techniques for Communications, Networks and Systems & Workshops,
2008,
Brussels,
Belgium,
Pages
1-10.
On-line:
http://www.omnetpp.org/doc/workshop2008/omnetpp40-paper.pdf. [Accessed
2014-1-12]
[Warren1989]
S. D. Warren, L. D. Brandeis, “Residential Energy Monitoring and
Computerized Surveillance via Utility Power Flows”, IEEE Technology and
Society Magazine, June 1989
[Wicaksono2012]
H. Wicaksono, K. Aleksandrov, S. Rogalski, “An Intelligent System for
Improving Energy Efficiency in Building Using Ontology and Building
Automation Systems. Automation”, InTech, 2012. ISBN: 978-953-51-0685-2
[Witte2007]
R. Witte, Y. Zhang, J. Rilling, “Empowering Software Maintainers with
Semantic Web Technologies”, ESWC'07: 4th European Conference on The
Semantic Web: Research and Applications, Pages 37-52, 2007
[WRT160NL2009]
Cisco Linsys WRT160NL Datasheet. 2009
[Yan2013]
Y. Yan, Y. Quian, H. Sharif, D. Tipper, “A Survey on Smart Grid
Communication Infrastructures: Motivations, Requirements and Challenges”,
IEEE Communications Surveys and Tutorials, Volume 15, Issue 1, Pages 5-20,
2013
[Zaballos2011]
A. Zaballos, A. Vallejo, J. M. Selga, “Heterogeneous communication
architecture for the smart grid”, IEEE Network, Volume 25, Issue 5, Pages 3037, 2011
[Zeifman2011]
M. Zeifman, K. Roth, “Nonintrusive Appliance Load Monitoring: Review and
Outlook”, IEEE Transactions on Consumer Electronics, Volume 57, Issue 1,
Pages 76-84, 2011
[Zigbee2007]
“Zigbee 2007/PRO specification”, Zigbee Alliance, October 2007
[Zigbee2014]
Zigbee Alliance Official Web Site: http://www.zigbee.org/ [Accessed
2014/1/5]
[ZWave2014]
Z-Wave Official Web Site: http://www.z-wave.com/ [Accessed 2014/1/5]
- 130 -

Similar documents

×

Report this document