a set of software simulation tools for mobile networks evaluation

Document technical information

Format pdf
Size 1.2 MB
First found Feb 12, 2016

Document content analysis

Category Also themed
not defined

Wireless wikipedia, lookup

Mobile phone wikipedia, lookup

Cellular network wikipedia, lookup

GSM wikipedia, lookup

WiMAX wikipedia, lookup

MIMO wikipedia, lookup

3GPP Long Term Evolution wikipedia, lookup

4G wikipedia, lookup




Overview of the MACHINE - a set of software
simulation tools for mobile networks evaluation
Krzysztof Bąkowski, Marcin Rodziewicz, Karolina Ratajczak
Abstract—MACHINE stands for Mobile networks evaluation
tools with Advanced CHannel and INterference modEling. In this
paper we present a set of software simulation tools entitled
MACHINE and intended for performance evaluation of mobile
wireless networks and radio access technologies. These tools were
developed at Poznań University of Technology in the Chair of
Wireless Communications and are extensively used in the study
of 4G and 5G wireless systems.
Index Terms— D2D, link level simulations, link to system
interface, LTE, MACHINE, MIMO, relaying, system level
simulations, two way relaying
IMULATION is a powerful tool that can be used in
evaluation of wireless networks and radio access
technologies. Indeed extensive simulations were performed by
various parties in the evaluation of 2G, 3G, and 4G standards
prior to their field tests. The level of details and scale of
modeled networks in these evaluations were increasing from
generation to generation of wireless cellular networks. It is
common to define complex models in a very detailed way to
enable comparison of results obtained by different researchers
or companies.
In this paper we present a set of software simulation tools
entitled MACHINE and intended for performance evaluation
of mobile wireless networks and radio access technologies.
These tools were developed at Poznań University of
Technology in the Chair of Wireless Communications and are
extensively used in the study of 4G and 5G wireless systems.
In section two we start with general overview of the
simulation tools and in the subsequent sections we present in a
more detailed way some of them. We conclude with
discussion of current development plans for the MACHINE.
Basic structure of the MACHINE is shown in Fig. 1. It
consists of two major parts: the simulated system model
(SSM) and radio environment model (REM). The REM is an
abstraction of the wireless networks and radio waves
propagation effects. Currently there are two major models
implemented. The first one is based on the IMT-Advanced
evaluation framework [1] and models cellular networks mostly
as hexagonal grids (there is one scenario that consists of a
single floor in a building). This tool was implemented during
work of the first author in the WINNER+ project. The second
REM, called MGM, is based on the METIS project’s test case
two (TC2) scenario [4, 18] and models a 3D map based city
structures. Both of these tools use the same tool for modeling
small scale fading effects - the SSE-MACHINE. The SSM
module of the MACHINE consists of three major parts. The
first one is a link level simulator (LLS-MACHINE) that is
based on the physical layer of the LTE system [12]. It
implements both uplink and downlink data and control
channels. LLS-MACHINE is able to simulate multiple links.
The abstraction model of the LTE link facilitates the second
major part of SSM - the link to system interface (L2SMACHINE). The L2S interface is utilized by the system level
simulator (SLS-MACHINE) which is the third major part of
SSM. The SLS models LTE-like cellular networks operating
in IMT-Advanced or METIS TC2-like scenarios.
Mobile networks evaluation tools with Advanced CHannel and INterference modEling
Simulatted Sy
Model - SSM
Link Level Simulator
Features: LTE-based, FDD/TDD, DL/UL Tr/
Link to System Interface
Features: MIESM, CQI,
LTE-link abstraction
System Level Simulator
Features: LTE based, Different
scenarios (cellular, cellular with
relays, cellular with D2D),
Heterogeneous Networks
adio Environment Model - REM
IMT-A Evaluation Framework
Features: Hexagonal network or
single floor deployment, Scenarios
– UMi, UMa, RMa, SMa, InH
Small Scale Effects
Features: WINNER/IMT-A based
small scale fading model,
antenna model
Madrid/Manhattan Grid Model
Features: 3D map; O2O, I2I, O2I
– propagation scenarios; Hybrid
Ray Traycing-Scenario path loss
model; Mobility model
Fig. 1. High level structure of the MACHINE - a set of simulation tools for
evaluation of mobile wireless networks.
From the programming point of view the core processing in
MACHINE is written in C++ with the help of IT++ library
[13]. Post processing of the obtained results is implemented
mainly in Python with the help of matplotlib [15] and numpy
[16] libraries. The tools can be developed and executed at least
on the MS Windows and Linux platforms. Typically the
simulation results are obtained by running in parallel multiple
simulations on the Chair of Wireless Communication’s
computation cluster capable of running simultaneously several
hundred tasks (each on a separate processor core).
This work was supported in part by "Działalność Statutowa - dotacja dla
Młodych Kadr" No. 08/81/147/DSMK/2014.
XVIII Poznańskie Warsztaty Telekomunikacyjne - Poznań, 12 grudnia 2014
The IMT-Advanced evaluation framework was specified in
[1] for evaluation of 4G technology proposals [4,8,9,10]. This
framework specifies a set of scenarios in which 4G compliant
technologies should operate fulfilling defined performance
indicators. The proposals should be evaluated through
simulation, as well as through analytical and inspection
procedures. The simulations are defined by providing network
layout, deployment, and channel model specification for each
scenario. There are four mandatory scenarios: indoor hotspot
(InH), urban microcell (UMi), urban macrocell (UMa), and
rural macrocell (RMa). The InH scenario assumes deployment
of two base stations in a single floor of an office building. The
remaining scenarios assume a hexagonal grid of base stations
with different inter site distances and maximum allowable
transmit powers. In each scenario users are deployed randomly
following uniform distribution. For each scenario a set of
parameters is specified for the generic IMT-Advanced channel
model. The ITU-R IMT-Advanced channel model is a
geometry-based stochastic model and is often called a doubledirectional channel model. In such a model, directions of the
rays departing from the transmit antennas and arriving at the
receive antennas are specified without the precise location of
the scatterers. Thanks to this approach, propagation
parameters and antenna parameters can be separated.
The coefficients of each channel are generated in a stepwise
procedure. First, the propagation scenario is selected, and the
environment parameters for this scenario are defined. After
that, the network layout is set up by generating base station
positions according to the network topology and by randomly
generating the positions of the user terminals and directions of
their motion. For a given scenario, the speed of all user
terminals is fixed. In the next step, large-scale parameters are
generated. First, line-of-sight (LOS) or non-line-of-sight
(NLOS) propagation conditions are set. After setting the
LOS/NLOS conditions, the appropriate path loss formula is
applied for each link. Subsequently, large-scale parameters are
generated for each link. These are delay spread (DS), angle
spread of arrival (ASA), angle spread of departure (ASD),
Ricean K factor (K), and shadow fading (SF). After largescale parameters are fixed for each link, the generation of
small-scale parameters takes place. For each propagation
scenario, the number of clusters is given in [1]. First, cluster
delays in a given link are generated based on the DS
parameter. After that, cluster powers are generated using a
particular cluster delay as a parameter. The last step in the
small-scale parameters generation is the calculation of
departure and arrival angles first for the clusters and then for
the rays of which the clusters are composed. When all largeand small-scale parameters are set, the final MIMO channel
impulse response coefficients are calculated.
The SSE-MACHINE implements generation of LSPs and
SSPs together with final calculation of channel impulse
response. The IMT-A-MACHINE implements the network
layout and deployment and together with SSE-MACHINE
constitute the IMT-Advanced evaluation framework for
simulation based assessment of 4G candidate technologies.
SSE-MACHINE is also utilized by MGM-MACHINE.
One of the REM part of the MACHINE is the Madrid Grid
Model (MGM) which is an implementation of the
environment and channel models defined in the METIS
project [4,7,18].
The METIS project channel models are defined as so-called
propagation scenarios (PS). These propagation scenarios can
be divided into three sub-groups based on transmitter’s and
receiver’s locations. These are:
x outdoor–to-outdoor (O2O) propagation,
x outdoor–to-indoor (O2I) propagation
x indoor–to-indoor (I2I) propagation.
These three sub-groups of propagation scenarios can be
further divided depending on the type of the communication
link. Three types of links are recognized, i.e. macro base
station (MaBS) to user equipment (UE) links, micro base
station (MiBS) to UE links and UE to UE links. The MaBS to
UE link propagation scenarios correspond to a situation where
the BS is located over a building rooftop. The MiBS to UE
link propagation scenarios are defined for BSs located below
the buildings’ rooftops where the dominant part of the
propagation is due to the reflections between buildings. The
UE to UE link is defined to model the propagation for the
device-to-device (D2D) communication.
In the MGM, eight propagation scenarios are identified:
x PS#1 Urban Micro Outdoor-to-Outdoor
x PS#2 Urban Micro Outdoor-to-Indoor
x PS#3 Urban Macro Outdoor-to-Outdoor
x PS#4 Urban Macro Outdoor-to-Indoor
x PS#7 Indoor Office
x PS#9 D2D Urban Outdoor-to-Outdoor (also for
x PS#10 D2D Urban Outdoor-to-Indoor,
x PS#13 D2D Indoor Office.
The propagation scenarios were defined with two important
assumptions in mind. Firstly, a realistic and not synthetic
scenario is required. Secondly, a 3D propagation model should
be utilized in system evaluations. In the currently commonly
used propagation models, the LoS or NLoS conditions are
randomly selected. However, for scenarios that are realistic,
the sight conditions between transmitter and receiver should
be evaluated on a real-time basis. A more detailed description
of propagation scenarios can be found in [18]. Fig. 2 contains
examples of some of the defined PSs.
The PS could not exist without a suitable environment
model. For this purpose a Madrid Grid (MG) environment
model has been defined. The MG is a realistic urban
environment model. The realism is achieved by considering
different environments of buildings, roads, parks, bus stops,
metro entrances, sidewalks and crossing lanes. The model is
based on the structure of Madrid and captures typical
European city characteristics. Fig. 3. presents the streets and
buildings layout of the MG along with the BSs deployment
The MACHINE fully implements the channel, environment
and the mobility model defined in the D6.1 of the METIS
XVIII Poznańskie Warsztaty Telekomunikacyjne - Poznań, 12 grudnia 2014
project. The detailed description can be found in [18]. One of
the more interesting features of the implementation of the
MGM is the possibility of creating a custom city layout and
modeling the channel with METIS propagation scenarios. In
[7] we presented a results of simulations performed based on
Poznań Old Market.
Fig. 2. MGM propagation scenarios [7].
Fig. 3. MGM environment model [4].
The first tool developed within MACHINE simulation
framework was the Long Term Evolution (LTE) link level
simulator. It models a single or multiple communication links
that can be established between network nodes such as base
stations, mobile stations and relay stations. The LTE-LLS
implements a complete LTE transport and physical layer
processing procedures. In case of the data transmission the
DL-SCH (transport layer) and the PDSCH (physical layer)
channels are simulated. DL-SCH transport channel processing
is a combination of error detection, error correction, rate
matching, interleaving and transport channel mapping onto the
PDSCH physical channel. Forward error correction (FEC) is
provided by the turbo code, and error detection is based on
cyclic redundancy check (CRC) codes. If the block size at the
input of the turbo code encoder is greater than Z equal to 6144
bits, code block segmentation is performed. The rate matching
performs interleaving, as well as repetition or puncturing, in
order to match the size of a block generated at the output of
the transport channel to the number of bits allocated for the
transmission of the physical channel. The number of bits
carried by the PDSCH physical channel is determined by the
modulation scheme, the number of resource blocks allocated
for transmission, and the transmission mode. PDSCH
processing consists of scrambling, modulation, layer mapping,
precoding, mapping onto resource grid and conversion to
complex-valued OFDM baseband signals for each antenna
port. These antenna ports do not correspond to physical
antennas, but rather are logical entities distinguished by their
reference signal sequences. Multiple antenna port signals can
be transmitted on a single transmit antenna. Correspondingly,
a single antenna port can be spread across multiple transmit
antennas. Both layer mapping and precoding are associated
with multiple antenna transmission and reception (MIMO).
These two steps map up to two incoming code words to up to
four transmit antennas (LTE Rel. 8). At the receiver side,
complementary operations are performed after channel
estimation and data equalization.
The LTE-LLS consists of several parts. Fig. 4 presents the
top-level architecture of the LTE link level simulation package
implementation [2]. The core processing is written in C++
programming language in an object oriented way. This main
part utilizes IT++ library [13], which is a C++ library of
mathematical, signal processing, and communication classes
and functions. The validation of the core part source code is
done by unit testing. Unit tests of the main project modules are
placed in a separate C++ project. Unit testing gives the ability
to verify that implemented methods works as expected for any
given set of inputs by returning the proper values or handling
failures (in case of invalid input). The testing project uses one
of the most commonly used C++ unit testing framework - the
Google Test (GTest) [14
The performance of the LTE link level simulator was
aligned with the results obtained by 3GPP partners during the
design of the LTE system. We compared our results with [19]
where ideal simulation results for PDSCH in AWGN channel
are presented. After that we configured our simulation tool
according to [20] where UE demodulation simulation
assumptions were agreed by 3GPP partners. The obtained
results were compared with those presented in [21].
The more detailed description of the developed LTE-based
link level simulator can be found in [2].
files and
comand line
Parsing and
Output files
with results
XVIII Poznańskie Warsztaty Telekomunikacyjne - Poznań, 12 grudnia 2014
Core processing
blocks written in
C++ and supported
by IT++, BLAS,
Validation and calibration
based on Google C++ testing framework
Fig. 4. LTE link level simulation package structure [2].
Due to excessive processing involved in determining the
result of a transmission of a data block through the mobile
radio channel it is practically not realistic to perform it for
each link in the radio network. Therefore proper abstraction of
a single link transmission is needed in order to limit the
processing requirements. Typically from the current channel
impulse responses of the desired and interfering links a single
metric is constructed that enables to map this conditions onto
block error probability. Several methods were developed for
this purpose. L2S-MACHINE utilized the mutual information
effective SINR mapping (MIESM) approach.
Even though SLS-MACHINE is based on the LTE
technology it is able to simulate features beyond the current
specification of the LTE standards. Apart from typical
simulation setups with cellular network consisting of base
stations (BSs) and mobile stations (MSs), it is capable of
simulating two-way relaying and D2D scenarios.
In principle SLS-MACHINE was designed to simulate
orthogonal frequency division multiple access (OFDMA)
based systems. The tool models algorithms and procedures of
such a system utilizing interfaces provided by one of the
REMs described before and L2S interface. There are four
primary operations that are realized by SLS. First, based on
the feedback received from various elements of the radio
network, scheduling is performed. Scheduling distributes time,
frequency, power, and space (MIMO streams) resources
between radio transmission capable nodes. Based on these
assignments, appropriate links are updated by calculating
channel impulse responses preceded by terminals position
update if required. After that calculation of post-reception
signal to interference and noise power ratios can be realized.
In OFDMA-based systems interference pattern changes in
each transmission time interval (TTIS) due to the dynamic
scheduling. This pattern is specific to used transmission
protocols, devices, and their operation modes (eg. D2D, twoway relaying). Finally, feedback metrics can be calculated.
This step can involve, calculation of effective SINR, channel
quality indicators CQIs, precoding matrices indices (PMIs),
determination of positive or negative acknowledgements
(ACK/NACKs), and rank indicators (RIs) to name a few.
During the simulation various statistics are calculated and
Development of comprehensive simulation platform is a
challenging task. It requires vast knowledge of various
elements of complex radio systems. The development of
MACHINE is a result of the shared expertise of its authors in
various aspects of wireless communication technology,
ranging from radio propagation and environment modeling
(both stochastic and ray-tracing methods), through physical
layer processing algorithms (including MIMO processing) and
scheduling algorithms, up to the standards knowledge. These
team effort continues along the research on 5G systems.
Current work is mainly concentrated on the system level
The described software IMT-Advanced channel simulator
was initially developed within the EUREKA/CELTIC project
WINNER+ (No. CELTIC CP5-026). The authors thank the
members of the WINNER+ IMT-Advanced Evaluation Group
for many helpful hints and remarks during the software
development and its calibration. Also, part of the research
leading to the presented results has received funding from the
EU Seventh Framework program FP7-ICT-2012 under grant
agreement No. 317669, also referred to as METIS. Work on
LLS-MACHINE was supported in part by a scholarship held
by the first author within the project “Scholarship support for
PH.D. students specializing in majors strategic for
Wielkopolska’s development”, Sub-measure 8.2.2 Human
Capital Operational Programme, co-financed by European
Union under the European Social Fund. Finally, the authors
would like to express their appreciation of work of other
coworkers, supervisors and students that supported the
development of MACHINE. This list involves but is not
limited to Paweł Sroka, Prof. Krzysztof Wesołowski,
Zbigniew Długaszewski, Adrian Langowski, and Emilia
“Guidelines for evaluation of radio interface technologies for IMTAdvanced”, Report ITU-R M.2135-1, Geneva 2009
[2] K. Bąkowski, K. Ratajczak, "Overview of the LTE link level simulator
developed at Poznań University of Technology", XVII Poznańskie
Warsztaty Telekomunikacyjne PWT’2013, Poznań, 13 grudnia 2013
[3] K. Ratajczak, K. Bąkowski, "Analiza wydajności techniki MU-MIMO w
systemie LTE-Advanced dla modelu kanału radiowego IMT-Advanced",
Przegląd Telekomunikacyjny i Wiadomości Telekomunikacyjne, nr
6/2013, str. 258 -261
[4] K. Bąkowski, K. Wesołowski, M. Rodziwicz, "Simulation Tools for the
Evaluation of Radio Interface Technologies for IMT-Advanced and
Beyond", Chapter 13 in "Simulation Technologies in Networking and
Communications: Selecting the Best Tool for the Test", CRC Press,
[5] K. Ratajczak, K. Bąkowski, K. Wesołowski, "Two-way Relaying for 5G
Systems - Comparison of Network Coding and MIMO Techniques",
IEEE Wireless Communications and Networking Conference, IEEE
WCNC 2014, 6-9 April 2014, Istanbul, Turkey
[6] K. Bąkowski, K. Ratajczak, "Porównanie kodowania sieciowego z
techniką MU-MIMO w transmisji dwukierunkowej przez przekaźnik dla
modelu kanału radiowego IMT-Advanced", KKRRiT 2014
[7] M. Rodziewicz, K. Bąkowski, "Modelowanie tłumienia propagacyjnego
w środowisku trójwymiarowym na przykładzie mapy Starego Rynku w
Poznaniu", KKRRiT 2014
[8] K. Bąkowski, "Modelowanie kanału radiowego za pomocą linii
opóźniającej z klastrami dla systemu IMT-Advanced", Przegląd
Telekomunikacyjny i Wiadomości Telekomunikacyjne, nr 8-9/2012, str.
769-776, ISSN 1230-3496 (XXVIII Krajowego Sympozjum
Telekomunikacji i Teleinformatyki, KSTiT’12, 12-14 września 2012,
Warszawa – Miedzeszyn)
[9] K. Bąkowski, K. Wesołowski, "A Software Channel Simulator with
Time Evolution of Parameters for the IMT-Advanced System",
European Wireless EW'2011, 27-29 April 2011, Vienna, Austria
[10] K. Bąkowski, K. Wesołowski, "Change the Channel: A New Model for
the IMT-Advanced System", IEEE Vehicular Technology Magazine,
June 2011, Volume 6, Number 2, p. 82-91
[11] K. Ratajczak, Z. Długaszewski, K. Bąkowski, "Odbiornik MIMO dla
systemu 3GPP LTE", Przegląd Telekomunikacyjny i Wiadomości
Telekomunikacyjne, nr 6/2011, str. 381-384, ISSN 1230-3496 (Krajowa
XVIII Poznańskie Warsztaty Telekomunikacyjne - Poznań, 12 grudnia 2014
Konferencja Radiokomunikacji, Radiofonii i Telewizji KKRRiT'2011,
8-10 czerwca 2011, Poznań)
K. Bąkowski, "Modelowanie warstwy fizycznej systemu LTE w łączu w
dół", Przegląd Telekomunikacyjny i Wiadomości Telekomunikacyjne,
nr 6/2010, str. 376-379, ISSN 1230-3496 (Krajowa Konferencja
Radiokomunikacji, Radiofonii i Telewizji KKRRiT'2010, 16-18 czerwca
2010, Kraków)
Open Source library, GPL license (2013, Nov. 15). IT++ Available:
List of unit testing frameworks:
Open Source library, PSF license (2013, Nov. 15). matplotlib Available:
Open Source library, (2013, Nov. 15).
Numpy Available:
IST-WINNER II Deliverable 1.1.2 v.1.2. WINNER II Channel Models,
Technical Report 2007, (http://www.istwinner.org/deliverables.html)
P. Agyapong, V. Braun, M. Fallgren et al., Simulation guidelines, ICT
METIS, Deliverable D6.1, October 30, 2013
R4-071640, Nokia, Ideal simulation results for PDSCH in AWGN. TSG
RAN WG4 Meeting #44bis
R4-071800, Motorola, Agreed UE demodulation simulation
assumptions. TSG RAN WG4 Meeting #44bis
R4-072218, Ericsson, Collection of PDSCH results. TSG RAN WG4
Meeting #45
communications with particular focus on underlay Device-toDevice communication.
Karolina Ratajczak (M.Sc.) received
her Bachelor's and Master’s degrees in
Electronics and Telecommunications, and
the Bachelor's degree in Computer
Science at Poznan University of
Technology, Poland, in 2011, 2012 and
2013, respectively. Since October 2012,
she has been a senior researcher at the
Chair of Wireless Communications in the
Faculty of Electronics and Telecommunications at Poznan
University of Technology. Her research interests cover the
wide spectrum of wireless communications, with the main
focus on the multi-hop transmission, multi-user MIMO
precoding, interference alignment and network coding. She is
a team member of the European research project METIS.
Krzysztof Bąkowski (M.Sc.) received
his M.Sc. degree in Electronics and
University of Technology (PUT) in 2009.
Since that time he has been employed at
the Chair of Wireless Communications at
PUT. In 2009-2011 he worked in the EU
scientific project CELTIC/EUREKA
WINNER+ "World Wireless Initiative New Radio" where he
was a member of the IMT-Advanced evaluation group. In
2011-2012 he participated in a large scientific national project
entitled "Future Internet Engineering". Currently he is
participating in project entitled "Mobile and wireless
communications Enablers for Twenty-twenty (2020)
Information Society" (METIS) conducted within the EU
Seventh Framework Programme (FP7). Since 2012 he has also
been involved in several research projects commissioned by
Nokia Solutions and Networks. He is an author and co-author
of several simulation tools used for research on 3G, 4G, and
5G systems.
Marcin Rodziewicz (M.Sc.) received his
M.Sc. degree in Electronics and
University of Technology (PUT) in 2009.
Since October 2009 he has been a PhD
student and an employee in the Chair of
Wireless Communications at PUT. In
2009-2011 he was working in the
CELTIC/EUREKA "World Wireless
Initiative New Radio" (WINNER+) project. In 2011-2012 he
participated in a significant national project entitled "Future
Internet Engineering". Currently he is participating in the FP7
"Mobile and wireless communications Enablers for Twentytwenty (2020) Information Society" (METIS) where he is
involved in research on Device-to-Device communication. His
research interests cover the wide spectrum of wireless
XVIII Poznańskie Warsztaty Telekomunikacyjne - Poznań, 12 grudnia 2014

Similar documents


Report this document