5G Testing Training by TELCOMA Global

Copyright © TELCOMA. All Rights Reserved
Challenges in 5G
testing & evaluation
TELCOMAwww.telcomaglobal.com

Challenges posed by new
technologies to testing &
evaluation

New technologies to testing &
evaluation:
● The proposed new network architecture and emergence of
various types of transmission technology will pose new
challenges to 5G air interface technology standardization ,
program design and simulation.
● For physical layer transmission technology , 5G will introduce
new waveform & non orthogonal multiple access at the physical
layer to achieve the required traffic latency in air interface.

evaluation:
● To explore spatial freedom & improve the network throughput,
5G will introduce massive MIMO technology. In simulation
evaluation system , massive MIMO & MU-MIMO technology will
greatly increase computational interference complexity.
● The new channel propagation model will be introduced based
on high frequency band transmission technology, D2D
technology & massive MIMO technology.

evaluation:
● Need to design scheduling algorithm for heterogeneous
computing resources, accurately estimate the consumed time
of heterogeneous computing & interface data transmission and
meanwhile design the synchronized mechanism for computing
tasks to make full use of heterogeneous computing platform.

Get 5G Training with Certification
Visit our website
https://telcomaglobal.com/p/5g-testing-training-certification

Elements of Testing &
Evaluation

Four elements of testing & evaluation:
There are four elements of testing and evaluation requirements :
● Realness
● Comprehensiveness
● Rapidly
● flexibility

Realness

Realness :
● In the 5G testing & evaluation, the requirement for realness is
reflected in 5G wireless channel model, verification methods ,
user experience and other aspects.
● From the point of network topology, various link types in 5G will
co-exist in a same region, extending from traditional macrocell
and microcell to picocell , femtocell and support V2V, M2M
network architecture.

Realness :
● Two way mobility of D2D/V2V will introduce doppler model &
massively intensive scattering exists in both transmitter and
receiver & the stationary cycle is short , all of which need to be
considered in the channel model.
● Wide range propagation scenarios and diversified network
topologies emerge in 5G research , which pose challenges to
the channel model.

Challenges :
● Unique transmission characteristics of radio waves in higher
frequency & bandwidth.
● Plane wave propagation model assumption is no longer
applicable and scattered clusters non stationary feature is
reflected not only in the time axis , but it is also changing along
the array.

challenges :
● Software simulation assumes that the hardware design can
perfectly realize the software algorithms and thus we cannot
introduce the impact of hardware conditions on the
communication systems.
● In reality, algorithm design is often restricted by the hardware
conditions , and the software algorithms are often greatly
reduced , when realized in hardware.

Realness :
● The algorithms at the front of transmitter and receiver have to
consider power , the impact on other hardware components
and many other factors.
● Verification of the tested technology or prototype equipment in
real scenario through “algorithm realization “ ----> data
acquisition ----> system optimization ----> outfield verification ,
providing basis for standardization.

Comprehensiveness

Comprehensiveness :
Demand for testing & evaluating comprehensiveness mainly
involves two aspects :
● It lies in the comprehensive support for the evaluation of 5G
performance indicators.
● It lies in the comprehensive support for the diversification of
candidate technologies.

Comprehensiveness :
5G evaluation indicators system will include the objective indicators
like :
● Transmission
● Network
● User experience indicators

Comprehensiveness :
5G candidate technologies can be divided into two categories :
● Air interface technology
● Network technology

Rapidly

Rapidly :
● In order to meet the KPI’s in 5G, the computing performance of
the simulation system must grow by more than 1000 times
before meeting the requirements of the timely evaluation of
simulation task.

Rapidly :
● With such rapid growth in computational performance , a
systematic and brand new design & realization are needed in
simulation system’s hardware platform , software platform and
simulation application.
● For the simulation system, the key problem is how to complete
concurrent design & coding implementation on the new &
powerful hardware platform with powerful computational
capabilities.

Flexibility

Flexibility :
● Affected by flexible network architecture, network resource
virtualization management , parallel computing needs & other
factors , simulation & validation system needs to have enough
flexibility.
● Architecture design , module design and interface design have
the characteristics of coupling , modularization , interface
expansion , easy integration and so on.

Evolution of testing
technology

Evolution :
● Testing & evaluation are indispensable links of technology and
product inspection.
● Every new technology requires strict testing and evaluating
process.

Importance of testing technology :
● Testing is measurement, inspection & test.
● Testing is an important step for checking the entire product’s
quality and it proves important to improve product quality.
● Any new technology or product must be fully tested and
evaluated.
● Testing technology needs to run through the entire process ,
from pre-design stage to design stage to production
completion stage.

Testing technology :
● Testing technology plays an important role in all stages of a
product or technology from prototype to standardization.
● The testing & measurement solutions at each stage are
important guarantees for successful application of new
technologies & products.

Development of testing instruments :
● Instruments are combination of various sciences & technologies
with many varieties and wide applications.
● To ensure more reliable performance & more powerful
integrated functions, the mature testing equipment & advanced
& reliable testing methods are needed. Powerful testing
instruments are indispensable to the stable & reliable running of
new wireless communication equipment.

Development of testing instruments :
The development of test equipments can be divided into three eras
:
● Instrument 1.0 era

Development of testing
instruments

Instrument 1.0 era :
● Microprocessors have improved performance & automation
degree of instruments , facilitating automatic range conversion ,
automatic zero adjustment , trigger level automatic adjustment,
automatic calibration , self- diagnosis and many other functions.

● With development of computers , test instruments began to
transform.
● Use of flexible instruments based on software design.
● In this, the users can redevelop the instruments according to
their own needs.

The combination of multiple techniques helped to create a number
of high performance testing systems that increased the flexibility of
testing and brought higher performance and lower cost.
● Bus technology
● Software defined radio technology
● Modular instrument technology
● Hybrid test system

Bus technology :
● Through the test bus, tha data communication between
different units and modules in the testing instruments are
realized.
● These standards allows testing engineers to assemble a variety
of automatic measurement systems with powerful functions by
very convenient means.
● Representative bus technologies includes VMEbus extensions
for instrumentation (VXI) , peripheral component interconnect
(PCI), peripheral component interconnect express (PCIe), LAN
extension for instrumentation (LXI).

Software defined radio technology :
● It is the most representative technology of instrument 2.0 era.
● The software radio is to take hardware as the basic platform of
wireless communications and implement the maximal functions
of wireless communications and personal communications with
software.
● The basic idea is to let all the tactile radios in use be based on
same hardware platform, install different software to form
different types of radio , complete functions of different natures
, and get the software programmable capability.

Software defined radio technology :
● The key idea of software radio is to construct a standardized,
modular universal hardware platform to realize various
functions by software, and to make the broadband A/D and
D/A convert to IF, near the RF side of the antenna , and strive to
carry out the digital processing from IF.
● Software radio requires a very high speed of hardware and
software processing.

Modular instrument technology :
● High speed communications test instruments featuring
modular, software and integration continue to spring up, which
complement and get closely integrated with traditional
methods expanding the applications continuously and forming
a technical highlight in the communication centred application
field.
● VXI was the earliest bus which introduced the concept of
modular instrument.

Modular instrument technology :
● Modular instruments integrated with computers and suitable for
PCI and PXI platforms are highly flexible plug-in computer
boards.
● Modular instruments are an important part of SDR technology.
● With the aid of modular instruments , engineers can choose
different kinds of modular instruments according to their
measurement needs and set up a test system.

Hybrid test system :
● The hybrid systems composed of different bus technologies
and instruments are gradually appearing in the testing field.
● In a hybrid test system, different components of multiple
automated test platforms are integrated in a system including
PCI, PXI, GPIB, VXI, USB, LAN and other different buses.

● It reflects the characteristics of openness, union and service.
● Cloud testing , software and hardware joint multi user system
simulation test is a series of applications confirming the
characteristics of instrument 3.0 technology.

Development trend of testing
technology

Co-existence of new & traditional test
equipment :
● Users can integrate test into the design process in a faster and
more flexible way, shorten the development time and improve
the test efficiency.

Leading trend
The leading trend of many-core / parallel technology & FPGA real
time technology :

Many core parallel technology :
● The number of cores in a server processor is increasing
drastically to more than 10 and the processor is developing
towards many core.
● The emergence and development of multi-core/many
processors is a choice of development of semiconductor
industry.
● The development of multi-core technology will promote the
development of test industry.

Real time FPGA technology :
● FPGA can realize the advantages of customized hardware tools
through the flexible software system.
● FPGA is suitable for uninterrupted filtering, modulation
/demodulation , encryption or other data processing.
● To ensure the inherent parallelism of data.

● It is defined by software first and then downloaded to a FPGA
chip for an actual execution.
● As design and testing requirements become higher and higher ,
the FPGA technology is introduced into the FPGA based
customized instruments.

● For 5G wireless communications , FPGA will be needed to
ensure that the response is real time and the inflow and outflow
speed of the data are high in real time system simulation and
high speed memory test applications.
● R-series data collection and FlexRIO products family provided
by NI have integrated the high performance FPGA into readily
available I/O broadcard for users to customize and repeat
configurations according to the applications.

Test ecosystem becomes the
trend

FPGA based IP to the PIN technology
● Adopting “system-level approach , integrating the concepts of
design and test, and expanding software architecture to FPGA “
is one of the effective means to balance the development and
improve the efficiency of communications testing .

FPGA based IP to the PIN technology
● The specific implementation of IP to the pin technology can be
expressed as a “V” graph.

Heterogeneous computing architecture
Heterogeneous computing architecture supporting parallel testing
and massive signal processing :
● The heterogeneous computing architecture is a system for
assigning data processing & program execution tasks among
different computing nodes, so that each node can handle the
most appropriate test & calculation task.

Heterogeneous computing architecture
● Along with the rapid development of high bandwidth and high
data rate of 5G mobile communications, the combination of
heterogeneous computing architecture and the multi-core
parallel programming technology will be indispensable main
technology with which 5G test is able to deal with massive data
processing and improve the parallel testing.

Challenges of Testing
technology

Challenges of multi-function & high
performance :
● The development of digital simulation technology and the
increase of the frequency band of RF test are urging the
development of testing technology.
● WLAN standards are applied in BW< 100 Mhz , so it doesn’t
need special methods of testing technology.

Challenges of multi-channel
● As per information theory, the more the number of antennas are
, the more obvious increase in the spectrum efficiency and
reliability.
● When the number of transmitting & receiving antennas is large ,
the MIMO channel capacity will have a close to linear increase
along with the minimum number of transmitting & receiving
antennas.

Challenges of multi-channel
● Requirements need a large number of OTA ( over the air ) tests
for support.
● Number of cooperative antenna at side of 5G BS should not be
less than 128.
● The number of antennas on this massive MIMO technology has
risen to hundreds, larger than traditional.

Challenges of high throughput
5G will promote service ability in three dimensions at the same time
:
● Improve the utilization rate of resources by more than 10 times
as compared to 4G.
● Increase the system throughput rate by about 25 times by
introducing UDN’s & deeper intelligence capabilities.
● Explore new frequency resources ( such as high frequency
band, millimeter wave & visible light etc).

Challenges of high throughput
● Massive data has put forward new requirements on the RF test
in both the method and the equipment.
● Keithley’s new generation MIMO test platform makes it simpler
and cheaper to increase new signal standards and MIMO
options.

Chapter 3

Channel Measurement and
modelling

Channel measurement & modelling
● The wireless channel model is numerical description of physical
propagation environments.
● It provides an effective and simple means to approximately
express the channel characteristics of the wireless transmission.

Requirements for 5G wireless channel
model :
● Applications and new technologies for 5G communications
poses many challenges in 5G wireless channel model.
● 5G channel model should support wide range propagation
scenarios, higher frequency and larger bandwidth

5G channel model
5G channel model should support :
● Wide range propagation scenarios and diverse network
topologies.
● Higher communications frequencies and larger bandwidth.
● Massive MIMO
● Spatial consistency and dual mobility
● High mobility

Wider range propagation scenarios
● 5G channel model should support mobile to mobile links and
networks.
● Network topology should not support only cellular networks but
also communications of D2D , M2M and V2V.

Higher communications frequency
● 5G may operate in frequency range from 350 Mhz to 100 Ghz.
● In 3GPP-HF channel model , if the BW of a channel is beyond
c/D , then such BW referred to as big BW.
● Propagation in HF band will undergo path loss due to shorter
wavelength.

Massive MIMO
● Planar wavefront assumption for conventional MIMO channels is
no longer applicable and should be replaced by spherical
wavefront assumption.
● Channel model must provide 3D angle information.

Spatial consistency & dual mobility
● 5G channel model should support spatial consistency.
● With co-existence and density increase of links , and with
application of D2D/ V2V , to support spatial consistency
become especially important for wireless channel models.

Channel modelling method

● In Deterministic channel model, there is prediction of
characteristics at every point in space.
● Stochastic channel modelling is a kind of method to obtain huge
measurement data set.
● GSCM is known as geometric based stochastic channel model.

● Channel parameters are divided into path loss , large scale
parameters such as shadowing , delay spread , angular speed
etc
● Small scale parameters such as delay , angle of arrival &
departure etc , which jointly reflect the channel fading
characteristics .
● Large scale parameters can be regarded as a statistical average
in a channel segment within which LSP or probability
distribution of LSP do not change significantly.

● Close-in reference short model
● ABG ( Alpha , beta , gamma) model

Measurement-based GSCM
● The GSCM modelling method separates antennas and
propagation channel.
● The clusters and rays in channel are parameterized by path loss,
shadowing & other parameters in large scale & small scale.

Stage 1
Preparation & measurement

Stage 2
Post processing of the measured data
● Different analysis methods are applied depending on the
required parameters.
● High resolution parameter estimation algorithms are required.

Stage 3
Generation of simulation model
● Clusters and ray parameters are generated according to their
PDF’s and statistical parameters.
● Measurement based GSCM modelling method has been widely
recognized & used.

Regular shape GSCM

Regular shape GSCM
● The characteristics of the propagation channel are described
according to the geometric relationships among the transmitter,
receiver & scatterers.

Regular shape GSCM
● The basic shapes of 2D distribution of scatterers mainly include :
one ring, eclipse and rectangular.
● The basic shapes of 3D distribution of scatterers include : one
sphere , two spheres etc.
● The geometric model in practical application can be
combination of these basic shapes and it is also possible to add
new geometric shapes.

CSCM

CSCM
● It describes the MIMO channel by the correlation matrix of
antennas.
● The existing CSCM model mainly includes :
● IID
● KBSM
● VCR
● Weichselberger channel model

Extended SV & Ray tracing
model

Extended SV model
● It can only describe the delay information of the channel and its
resolution is higher than 10 ns subjected to the sounding
equipments.

Ray tracing based model

Ray tracing based model
● The propagation prediction of EM wave can be solved by
analytical method based on electro-magnetic theory or by
electromagnetics.
● Ray tracing method is based on relatively simple geometrical
optics and uniform theory of diffraction.

Basic principle
● To use rays to simulate the electro-magnetic wave propagation
process , such as direct ray , diffraction , reflection & scattering.
● Calculation of the field strength of rays and path tracking of rays
are two main aspects of Ray tracing technique.
● RT treats the propagation of EM wave by means of rays
propagation.

Channel measurement

Channel measurement
● It is also known as channel sounding .
● It is a direct and most important mean to obtain channel
information & understand the characteristics of the channel.
● By exploiting the received signals , multiple propagation paths
can be identified.

Channel measurement methods
● Channel measurement methods can be divided into SISO,
MIMO, SIMO.
● MIMO method simultaneously measures multiple channels in
parallel.
● SISO only needs one RF chains at transmitter and receiver
respectively.
● SIMO is a compromise between SISO & MIMO.

Channel measurement activities
With the on going progress in 5G R&D, channel measurements are
divided into three categories :
● Measurement of massive MIMO
● Measurement of D2D/V2V/HSR
● Measurement of mmWave band

Massive MIMO measurement
● Massive MIMO is a promising technology for 5G communication.
● For measurement, BS was placed indoors and configured with a
polarized uniform cylindrical array with 128 ports.
● Measurements were carried out outdoor with a larger space
using two systems at in the same position and results were
compared.

Massive MIMO measurement
● Massive MIMO

Measurement of D2D, V2V &
HSR

Measurement of D2D, V2V & HSR
● D2D/V2V and high speed rail (HSR) communications are
important for 5G wireless communications.
● MIMO technology will be widely used in vehicle communication
systems because it can improve the system capacity &
guarantee reliable data transmission.
● Measurements were carried out during the day and night time in
order to observe the effects of pedestrians on the channel.

Three scenarios were designed as follows :
● 1st scenario was at crossroads , where the transmitter standed
statically near an intersection , while the receiver moved in
perpendicular direction of intersection with a speed of 30-40
km/h.
● 2nd scenario is at expressway with two lanes , where the
transmitting car was blocked in one lane & the receiving car
approached to the transmitter in another lane with speed of 70
kmph.

● 3rd scenario is two cars were derived at speed of 110 kmph in
one lane of a two lane highway where both cars were
obstructed by a tall van.

In the same sounding system was adopted to measure in more
scenarios including following situations :
● Road crossing
● General LOS obstruction
● Ramp merging
● Traffic congestion
● In tunnel
● On bridge

Measurement of mmWave band
● The measurement activities were designed and targeted to a
specific measurement environment, but the similar equipments
and procedures were used.
● Directional antennas was installed on a 3G rotating tripod.

Channel data processing
● In channel sounding, the measurement data , or the received
signals are superposition of altered versions of the transmit
signal passing through multiple paths with different delays,
gains and directions.

Path parameters extraction
● There are mainly two categories of commonly used channel
parameters extraction algorithms i.e subspace - based methods
and maximum likelihood methods.
● Within first category are MUSIC ( Multiple signal classification)
algorithm , ESPRIT (estimation of signal parameter via rotational
invariance techniques) algorithm.

EM algorithm
● Basing on signal model, one can write the probability density
function (PDF) of the measurement data in terms of signal
parameters.
● EM algorithm can effectively estimate parameters by iteratively
carrying out the E step & M step

SAGE algorithm
● It is developed based on the classical EM algorithm.
● In M step , it divides all parameters of one path into several sets
and updates only one set of parameters at a time while keeping
other sets of parameters fixed.
● The process continues till all the parameters of one path are
updated.

FD-SAGE algorithm
● The SAGE algorithm mentioned is applicable for channel
sounding using time domain PN excitation signal.
● In some channel sounding , frequency excitation signal is
commonly used , such as broadband multi-carrier excitation or
VNA based single frequency stepping excitation.
● Frequency domain sage algorithm is developed for parameter
execution .

DMC and its estimation
● The Signal emitting from transmit antenna and arriving the
receiving antenna is composed of a LOS components, specular
components and dense multipath components.
● Each SC corresponds to a discrete and strong propagation path
formed by an independent scatterer reflecting EM waves.

Mobile channel estimation and tracking
● To model the dynamic channel accurately , it is required to
estimate and track the number of paths & multipath parameters.

Automatic clustering algorithm
● It has been found that the channel MPC tend to appear in
clusters i.e in groups of MPC’s with similar parameters.

Software simulation

Software simulation
● When compatibility and interoperability with other systems are
considered , its network structure and networking methods
would be more complicated.
● 5G candidate technologies are more abundant and its
application scenarios are more complicated.
● The software evaluation for the performance of 5G technical
schemes is facing unprecedented challenges.

Software simulation
● The wireless communication systems runs in a real environment
which is compatible & volatile.
● It is more feasible and effective to set up a software simulation
platform via computer & mathematical modelling , on which
software simulation programs can make performance
evaluation for the wireless communications modules or
systems.

Software simulation
● The new architecture designs the hardware & software jointly ,
decomposes the computational tasks through master - slave
node model and makes layered design for the whole software
based on modular design of application module , protocol stalk,
message service , connectivity management , mobility
management & system interface function.

Complete simulation system
● Requirement analysis : a networking scenario is needed to be
given, including network type , network element type, network
interface function , network configuration parameter, services
characteristics , performance indicators etc.
● System modelling : it includes abstracting & modeling the
network & system components & then outputting the
simulation model . these models describe the input output
relations of the network & system components.

Complete simulation system
● Evaluation criteria : through the analysis of original
requirements, the input information of simulation evaluation can
be obtained, which includes performance indicators, simulation
benchmark and evaluation criteria.
● Simulation realization : it includes three elements: simulation
system architecture , simulation methodology , simulation
process.

5G Software simulation
requirements

Simulation Requirements analysis
● Comprehensiveness is the functional requirement of the
software simulation system , which mainly refers to the
comprehensive support for 5G performance indicators and
candidate technologies .
● 5G performance indicators include KPI’s such as peak data rate ,
guaranteed min. user data rate, connection density, service
traffic volume density, wireless delay, end - to - end delay.

● Software simulation system must provide corresponding
modeling , statistic and evaluation of these new performance
indicators.
● 5G candidate technologies can be classified into two types :
namely air interface technology and network technology.

● New technologies in architecture level, such as SDN & SON,
have the greatest impact on the design of software simulation
system because it needs to model the new network
architecture, design new network element types, interfaces
between new n/w elements & new protocol stacks.

● Rapidly is the time efficiency requirements for software
simulation system.
● Accuracy is performance requirement for software simulation
systems. Common ways to improve the simulation efficiency
from the perspective of hardware configuration : high
configuration desktop , multi core servers , high performance
super computers.

Technological impact analysis

The main factors influencing the system simulation software design
and implementation can be divided into three categories:
● Architecture
● Function
● Performance
Performance shows the computational performance that can be
reached by system simulation software and its platform.

5G software link level
simulation

Link technology overview
● 5G transmission technologies will explore a series of new types
of multiple access & transmission mechanisms, to greatly
improve the spectrum efficiency & energy efficiency of wireless
systems.
● Massive MIMO has become one of the important 5G key
transmission technologies.

Link technology overview
● Massive MIMO can bring the huge array gain and the
interference suppression gain through large-scale antenna
arrays, thus greatly improving the system spectrum efficiency
and the edge user spectrum efficiency.
● In 5G network deployment , will have distinctive heterogeneous
characteristics

Link simulation realization

Simulation key factors
● The link level simulation is mostly used to evaluate the physical
layer transmission performance of the wireless communication
systems.
● Through modular simulation design , the link level simulation
can realize the performance comparison of different
transmission schemes with a variety of transmitter structures
and receiver algorithms.

Simulation key factors
● 5G transmission will pose higher requirements for the
computing abilities and the simulation speed of the link level
simulation.
● The factors that need most consideration in the link level
simulation of the main candidate technologies include the
following aspects :

Simulation of Massive MIMO channel
● In terms of 5G link level simulation, it is necessary to carefully
consider various characteristics of massive MIMO channel such
as spatial correlation , the coupling, the near field effect etc.
● The empirical channel model can also be constructed through
analysis, comparison and fitting of the measured channel data in
combination with theoretical analysis.

Simulation for the neighboring
interference
● In 5G scenarios with ultra dense nodes, the co-channel
interference will limit the network capacity.
● The interference signals can also be theoretically modelled
through theoretical derivation in order to simplify the simulation
complexity.

Simulation for the novel multiple access
technology
● It needs a comprehensive evaluation on the impact on the
spectral efficiency and the impact on the robustness of the
non-ideal factors such as frequency offset and channel
estimation errors.

Simulations with high-performance
multi-core parallelization
● 5G system will be configured with massive number of antennas.
● The multi-thread parallel simulation, the high performance
multi-core server , the software and hardware co-simulation
and other advanced simulation methods will be conducive to
the fast and accurate link level simulation evaluation.

Simulation process overview

● The signal processing modules at the transmitter mainly include
channel coding , constellation mapping , multiple antenna
precoding , multiple access , reference signal generation ,
framing etc.
● The signal processing modules at the receiver mainly include
cell search and synchronization , de-framing , channel
estimation , multiple antenna detection , demodulation , channel
decoding etc.

● In the link level simulation, programming is required to realize
each of the above mentioned function modules.
● The wireless link level simulation system usually includes two
sets of simulation systems for DL & UL.

●

● It may include system level parameters such as the frame
structure , the system bandwidth , cell information , resource
allocation methods etc.
● Realization of each function module in UL & DL and the
interface definition between modules are the most basic parts
of link level simulation.

Introduction to simulation
cases

Massive MIMO performance simulation
● Turbo ⅓ code rate & quadrature phase shift keying (QPSK)
modulation are used.
● Adaptive modulation & coding (AMC) & HARQ mechanisms are
not initiated.
● In the massive MIMO simulation, the number of BS antenna is
set to 128, and the number of users took K= 20,30,40,50
respectively.

● Reasonable parallel simulation can greatly reduce the
simulation time.
● To accelerate the process, the parallel optimization is carried
on.
● The link simulation consists of two layers of cycles, SNR cycle
and frame cycle.
● Data in the SNR cycle and frame cycle are independent.
● Parallelization of the SNR and frame cycles is simple and
effective way to improve simulation timeliness.

● The theoretical parallel speedup result of the new parallel
simulation method can be faster in orders of the number of SNR
times by the number of frames.
● Parallelization of the SNR and frame cycles is simple and
effective way to improve the simulation timeliness.

Heterogeneous network
energy efficiency simulation

Heterogeneous n/w energy efficiency
simulation
● The wireless heterogeneous network is an effective means to
alleviate the contradiction between data growth and energy
consumption.
● The indicators have considered the impact of the new BS layout
on the energy consumption increase at the network side and
the energy consumption changes at the terminal side at the
same time.

Simulation parameters
Macro cell parameters
Carrier frequency 2.0 GHz
Bandwidth 20 MHz
Path loss model COST-231 - Walfish - Ikegami
model
Cell radius 5 Km
Minimum SNR requirement 10 db
Noise -160 dbm/Hz

Low power consumption cell parameters
Carrier frequency 2.0 GHz
Bandwidth 20 MHz
Pathloss model COST 231- Walfish - ikegami
model
Minimum SNR requirements 10 db
Noise -150 dbm/hz

Micro-cell user parameters
Active user density 10^-5 user / m^2
Working state power consumption 1.2 W
Resting state power consumption 0.6 W
Packet size 100 bit

Tail-Biting convolution code
decoder simulation

Tail - biting simulation
● In the short packet transmission, short codes are usually
needed for channel coding such as tail-biting convolution
codes.
● tail - biting convolution codes usually use circular viterbi
algorithm (CVA) for decoding.
● Detection of the circular trap can be used to help control the
CVA decoding process, so as to get the fast convergent iterative
decoding algorithm.

Compressive sensing
simulation

Compressive sensing simulation
● It is an important way of data processing, which can recover the
raw data from extremely few sample values.
● The computational complexity of compression sampling is
much lower at the transmission than that of receiver.

Compressive sensing simulation
● CS technology has a great application prospect in the sensor
network data aggregation application.
● Treelet based compressive data aggregation (T-CDA) is also a
data collection method.

User-oriented link adaption in
D2D network coding
multicast

User-oriented link adaption
● D2D is a candidate technology in 5G.
● In user - oriented link adaptive method , two multicast channels
whose link quality corresponds to the maximum value of the
modulation type and the minimum value of coding type are
chosen.
● In order to verify the algorithm’s BER and spectrum efficiency,
link simulations are carried out

5G software system level
simulation

Test evaluation methods

Test evaluation methods
● A good evaluation system requires basic features of
completeness , simplicity, better usability etc.
● For the comprehensive , accurate and efficient evaluation of 5G
network technologies, the study should at least include
following aspects :
● Design of new performance indicators
● Redesign of traditional indicators in 5G network
● Network function virtualization

Design of new performance indicators
New KPI’s introduced by 5G networks are :
● Connection density
● Traffic volume density
● Minimum guaranteed rate

Redesign of traditional indicators
● Virtual resources model is introduced after introduction of the
virtual technologies.
● There are big breakthroughs and innovations in some 5G new
technologies in terms of design of network resources model
and new concepts and designs are introduced , which are not
compatible with designs in the past.

Network function virtualization
● There can be three layers : infrastructure layer , control layer
and application layer.
● Infrastructure layer : it is supported by various kinds of network
equipment nodes such as macro eNB and micro eNB.
● Control layer : it is composed of a series of distributed
management node.
● Application layer : it is made up of many different services and
applications.

Network function virtualization
● Infrastructure layer is supported by various kinds of network
equipment nodes, such as macro eNB and micro eNB.
● The services are scheduled and allocated by management
nodes.

Key simulation technologies

Key simulation technologies
Key technologies of 5G system simulation design are :
● Dynamic simulation modelling technology
● Management technology of virtualized computational
resources.
● Multi-core parallel simulation technology
● Hardware acceleration simulation technology
● Real time transmission technology

Dynamic simulation modelling
It is embodied in following aspects :
● Networking scenario is complicated & changeable
● Network model is evolving

Dynamic simulation modelling process
It is composed of following steps :
● The simulation model is decomposed. It is composed of 5
layers.
● Simulation parameter library is generated according to the
model and the requirements.
● The corresponding function libraries are mapped by the model.

Dynamic simulation modelling process
It is composed of following steps :
● According to the simulation requirements, the mapped function
library and parameter library are organically organised to
become a complete simulation process.
● By dynamically configuring the parameter library, function
library and the simulation process , we can get the specific
simulation tasks, which can get the specific simulation tasks ,
which directly faces the users and need to provide friendly
configuration management interface.

Virtual computational resources
management technology
The virtual computational resources management technology is
broken into three parts :
● The simulation requirements are mapped into computational
tasks that can be deployed independently.
● Various kinds of hardware resources can be virtualized into
three kinds of virtual resources : computational, storage &
communication resources.
● Binding the virtual resource dynamically to computational tasks.

Multiple -core parallel simulation
technology
● Simulation platform design based on multi-core parallel
computation covers hardware, operating systems, parallel
technology, simulation software, models and algorithm designs
etc.
● From the hardware level, parallel server scheme or high
performance host system can be chosen to support high
computing power.

technology
● Operating system allocates process, storage and other
hardware resources for parallel tasks, realizing the inter-process
communications.
● Parallel technologies generally include messaging, shared
storage and data parallel.

technology
● Parallelization of simulation software is the key work of
multi-core parallel design of simulation platform, which need to
consider the following design requirements :
● Simulation software is decomposed in parallel from the aspects
of function, algorithm and operands.
● The reasonable division design of simulation function modules
can reduce the communications data between parallel sub
tasks.

technology
Multiple CPU-GPU heterogeneous platform has multiple layers of
parallel execution ability in task level and data level.
The model’s mapping process has the following three levels :
● Mapping from simulation model instance to logical process.
● Mapping from logical process to thread.
● Mapping from thread to processor cores.

Hardware acceleration simulation
technology
● It uses hardware modules rather than the software modules to
make full use of the inherent fast hardware features.
● Hardware uses high performance FPGA board which has strong
computing power and logical processing ability.
● FPGA board has stronger floating point computing power than
CPU server & stronger task management, resource scheduling
and other logic handling abilities than GPU server.

Hardware acceleration simulation
technology
The implementation process of hardware acceleration simulation is:
● Key technology research
● FPGA based hardware accelerator card system is designed.
● Configuration data is loaded.

Real time transmission technology
● Network architecture can extend hardware processing ability of
the simulation platform through the way of extending server
nodes.
● After the high strength computational tasks are parallelized, the
computational time of each independent parallel subtask
becomes shorter, which is usually within hundreds of
microseconds.

Massive MIMO system level simulation
Simulation parameter description
● It uses MU-MIMO model to simulate LTE DL system
performance.
● Channel matrix formed with 128 BS tx antennas, 1 rx antenna , 15
users scheduled in a single cell at the same time.
● CPU memory : 256GB
● Windows server
● MATLAB

Radio resource optimization Of UDN
● Site number required to be coordinated will also increase in
resources allocation, making resource allocation more difficult.
● The site deployment in hotspot areas shows a trend of high
density and no programming.

Statistical modelling simulation
This simulation is of UL interference
● In OFDMA based frequency multiplexing network, inter-cell
interference has become one of the key factors that restrict the
improvement of system performance.
● In the past network deployment, it is often difficult to choose
appropriate control parameters and only conservative settings
can be made according to limited experience.

Local mobile cloud assisted computation
offloading
● Computation offloading technology is an important application
of mobile cloud computing.
● It offloads the user’s local computational tasks to the cloud with
rich resources and extends the computing power of the mobile
terminal with limited resources, so as to use the new
computation intensive applications.

Software visualization of 5G
network simulation

Architecture summary
● Simulation platform for 5G based on the universal software and
hardware platform. This platform adopts the distributed
master-slave parallel processing architecture.
● The master computational node is the management centre of
the simulation platform.
● The slave nodes are managed by master nodes, take on
computational tasks of the master node, and report the
simulation results.

Client 1
Client 2
Client n
Distributed computing service
Master
computing
node
Slave
computing
node 1
Slave
computing
node 2
Slave
computing
node n

● Sim

Evaluation test of hardware
and software co-simulation

Overview
● The concept of hardware and software co-simulation was
proposed as early as deployment of HDL.
● It reflects the idea of authenticity and rapidness , which is an
effective testing and evaluation method to deal with rapid
development of 5G.

Requirements
● Authenticity
● Rapidness

Composition
● In this, the physical layer of the simulation platform is partially
substituted by the real physical layer and the transmission
network , which can increase the reality and instantaneity of the
system.
● Mapping relation from the system level simulation software to
the real physical layer should be customized in order with the
specific requirements of the simulation evaluation.

Hardware test evaluation
platform

5G hardware test evaluation platform
● The constitution of typical hardware test platform of 5G
evaluation includes parallel channel sounder platform of
channel measurement and modelling .
● A MIMO OTA platform of designated channel model, a platform
of software and hardware of open source community and
terminal and base station system based on the general purpose
processor.

Key technical challenges
● Synchronization across multiple channels.
● Real - time storage of massive raw measurement data
● Parallel channel calibration
● High speed continuous storage of raw data

OTA test platform
● The method to evaluate radiation performance of traditional
SISO is mature.
● Its evaluation mainly aims for two indicators , making OTA test
for total radiated power (TRP) and total radiated sensitivity (TRS).
● Many performance parameters of passive MIMO antenna, such
as efficiency, gain are not different from traditional SISO
antennas.

OTA test platform
● MIMO antennas contain many antenna elements, passive
parameters are introduced to describe the relationships among
several antenna elements.
● The MIMO OTA test scheme is essentially a different simulation
method of multipath in space propagation environment ,
producing wireless communication environment close to the
reality to deal with the key challenge in massive antenna MIMO
OTA testing technique.

Field Trail network

Requirements and technical challenges
● Various application scenarios
● Technical challenges in field trail
● Development status of 5G testbed in foreign countries
● Development & evolution of HetNet convergence

Wireless network data intelligence
analysis
● Background and necessity
● Key technologies
● Technical roadmap
● Uniqueness
● Status
● Application scenarios
● Innovation points

Working mode of intelligent network
optimization analysis system

Thanks ...

5G Testing Training by TELCOMA Global

More Related Content

5G Testing Training by TELCOMA Global