Simulation of Communication Systems

1
Simulation of Communication Systems
Professor Z. Ghassemlooy Optical Communications
Research Group http//soe.unn.ac.uk/ocr/ School
of Computing, Engineering and Information
Sciences University of Northumbria at Newcastle,
UK
Eng. of S/W Pro., India 2009
2
Outline of Presentation

• Communications Systems
• Simulation software types
• Case Studies based on Matlab
• Concluding Remarks

3
Northumbria University at Newcastle, UK
4
Telecommunications Research Areas
Eng. of S/W Pro., India 2009
5
Photonics - Applications

• Photonics in communications expanding and
  scaling

Metropolitan
Home access
Board -gt Inter-Chip -gt Intra-Chip

• Photonics diffusing into other application
  sectors

Health(bio-photonics)
Environment sensing
Security imaging
6
School of Computing, Engineering and Information
Sciences Research
Optical Communications
Optical Fibre Communications
Photonic Switching

• Pulse Modulations
• Equalisation
• Error control coding
• Artificial neural network
• Wavelet based receivers

• Fast switches
• All optical routers

• Chromatic dispersion
• compensation using
• optical signal processing
• Pulse Modulations
• Optical buffers
• Optical CDMA

• Subcarrier modulation
• Spatial diversity
• Artificial neural
• network/Wavelet
• based receivers

6
Eng. of S/W Pro., India 2009
7
OCRG People

• Staff
• Prof. Z Ghassemlooy
• J Allen
• Dr R Binns
• Dr K Busawon
• Dr W. P. Ng
• Visiting Academics
• Prof. V Ahmadi, Univ. Of Tarbiate Modaress ,
  Tehran, Iran
• Dr M. H. Aly, 2Arab Academy for Scie. and Tech.
  and Maritime Transport, Egypt
• Prof. J.P. Barbot, France
• Prof. I. Darwazeh, Univ. College London
• Prof. H. Döring, Hochschule Mittweida Univ.
  of Applied Scie. (Germany)
• Prof. E. Leitgeb, Graz Univ. of Techn.
  (Austria)
• PhD Students
• M. Amiri, A. Chaman-Motlagh, M. F. Chiang, M. A.
  Jarajreh, R. Kharel, S. Y Lebbe, W.
• Loedhammacakra, Q. Lu, V. Nwanafio, E. K. Ogah,
  W. O. Popoola, S. Rajbhandari, A.
• Shalaby, X. Tang

7
Eng. of S/W Pro., India 2009
8
Simulation Introduction

• In recent years there has been a rapid growth in
  application of computer simulation in
  communication engineering.
• Hardware becoming more complex and costly
• A way forward to many researcher and teachers is
  to implements ideas in the software environment.
• This allows testing of the system using idealised
  processing elements, which may take a significant
  time to design and realise in hardware.

9
Simulation Introduction

• Can support the hardware design by giving
  optimised component values, for the critical
  parts, and an early indication of the performance
  of the system
• Allowing users to study or try things that would
  be difficult or impossible in real life
• Simulations are particularly useful when a
  real-life process
• is too dangerous,
• takes too long,
• is too quick to study,
• is too expensive to create.

10
Simulation Tools - Some Features

• Reliability - Depend on the validity of the
  simulation model, therefore verification and
  validation are very important
• Reproducibility of results
• User friendly, simple and flexible (allowing
  user defined functions)
• Extensive details of theory adopted
• High speed, precession and accuracy
• Hidden source code Up to date library
• Debugging capabilities and Scalability
• Can readily be upgraded and updated
• Cost effective and time saving

11
Simulation Tools - Disadvantages

• Poor modelling or poor data collection can lead
  to
• inaccuracy or
• completely misleading results
• Obsession - can lead to superficial understanding
  and no experimental verification
• However, simulation tools have become integral
  part of todays research and teaching activities
• Mainly for cost reasons

12
Simulation Software Application in Engineering
Eng. of S/W Pro., India 2009
13
Simulation Software Key Features

• Numerical Integration procedures
• E.g. Matlab has a number of procedures
• Rung-Kutta 45 Most advanced and ideal for
  analogue systems
• Rung-Kutta 45
• Stiff Adam with a fixed step integration Used
  for discrete systems
• Euler The most basic and used for slow varying
  discrete systems
• Ability to plot and display graphs
• 2D, 3D visualisation
• Simplicity for programming
• Compatibility with other software

14
Simulation Tools Types

• Matlab/Simulink
• Orcad/Pspice
• VPI
• Mathcad
• OptSim 4.0 simulation and design of advanced
  fiber optic communication systems
• OptiSystem large scale system software
• OptiFDTD

15
Matlab/Simulink

• A high-performance language for technical
  computing
• Integrates computation, visualization, and
  programming in an easy-to-use environment
• Typical uses include
• Math and computation
• Algorithm development
• Data acquisition
• Modelling, simulation, and prototyping
• Data analysis, exploration, and visualization
• Scientific and engineering graphics
• Application development, including graphical user
  interface building
• Compatible with excel, uses Maple and is
  compatible with other software packages such as
  C, C, VPI, etc.

16
Orcad/Pspice

• To model circuits with mixed analogue and digital
  devices
• Software-based circuit breadboard for test and
  refinement
• Can perform
• AC, DC, and transient analyses
• Parametric, Monte Carlo, and sensitivity/worst-cas
  e analyses i.e. circuit behaviour in a changing
  environment
• Digital worst-case timing analysis to resolve
  timing problems occurring with only certain
  combinations of slow and fast signal
  transmissions, etc.
• Not compatible with excel

17
Mathcad

• A desktop software for performing and documenting
  engineering and scientific calculations
• Equations and expressions are displayed
  graphically (WYSIWYG)
• Capabilities
• Solving differential equations - several possible
  numerical methods
• Graphing functions in two or three dimensions
• Symbolic calculations including solving systems
  of equations
• Vector and matrix operations including eigenvalues
   and eigenvectors
• Curve fitting
• Finding roots of polynomials and functions
• Statistical functions and probability
  distributions
• Calculations in which units are bound to
  quantities
• One cant use symbolic parameters only numerical
  parameters

18
OptiSystem

• Is used for
• designing, testing and optimization of virtually
  any type of optical links in the physical layers
• based on a large collection of realistic models
  for components and sub-systems
• OptiFDTD (finite-difference time-domain)
• propagation of optical fields through nano- to
  micro-scaled devices by directly solving
  Maxwells equations numerically

19
OptiSystem contd.

• OptiBPM
• Based on the beam propagation method (BPM)
• a semi-analytical technique that solves an
  approximation of the wave equation
• Waveguide other similar optical devices
• Light propagation predominantly in one direction
  over large distances

20
Virtual Photonics Inc.

• Used in optical networks and optical devices
  modelling
• Support C and Matlab
• Will talk about this in my second lecturer!

21
Case Studies - MATLAB
A typical communication system block diagram
Channel
22
Case Study 1 - AM/FM communication system s

• Aim To simulate a communication system link
• Tasks
• Channel modeling
• Comparing received and transmitted signals
• System performance evaluation
• System optimization
• Final system design

23
AM/FM Simulation - System Parameters

• Know parameters
• Carrier frequency, and power
• Signal bandwidth
• Modulation index
• Channel bandwidth and loss
• Link length
• Transmitter/receiver antenna type and gain
• Performance parameters
• Output signal-to-noise vs carrier to noise ratio
• System linearity
• Harmonic distortions

24
FM Simulation Block Diagram
FM modulator
Amplifier
Transmitter
Message
Channel
Receiver
Amplifier
FM demodulator
Low pass filter
Recovered Message
25
FM Simulation - Matlab-Simulink

• Provided that the mathematics underlying each
  block is fully appreciated, one could use any
  programming languages including high level
  computer languages C, C, Java or scientific
  programming languages Matlab, MathCAD ,
  Mathematica, Octave to name a few
• Matlab/Simulink
• One of the most popular simulation tool
  available
• Simulink is more user friendly for beginners as
  there are many drag and drop block functions.
• However Simulink also sometimes limits
  flexibility to users.

26
FM Simulation - Results
27
FM Simulation - Performance Evaluation

• The easiest way to evaluate the performance is by
  visual inspections
• For example, one can hardly differentiate
  between the transited message and recover message
  in the previous example
• Message signal at different SNRs is shown below-
  observe the improvement in the performance with
  increasing SNRs

28
FM Simulation - Performance Evaluation

• Visual inspection is the simplest and in many
  cases gives an insight to the system, BUT it is
  very error prone
• Alternative method of analysis should be used
• Considered error signal defined as error (m
  - mr)2
• The error signal at SNRs of 15, 20 and 40 is
  shown below
• The performance difference between the SNRs of
  15 and 20 is apparent

29
FM Simulation - Performance Evaluation

• Simulation software may provide many interesting
  results, but the expertise and experience of the
  user play's a major role
• In previous plot - very little difference
  between 20 dB and 40 dB
• An experienced user may choose the log-scale to
  plot error to gain more information, shown below
• Compared to the pervious plot, difference in
  performance for 20 db and 40 dB is clear from
  this plot

30
Case study 2- Digital Communications

• Depending upon the channel, receiver may
  incorporated other signal processing tools like
  equalizing filter, low pass filter and so on
• The output bits are compared to the transmitted
  to bit to calculated the error
• The bit error rate (BER) is the metric used in
  all digital communication system to compare and
  evaluate the system performance
• BER depends on the SNR (valid only for particular
  signalling format)

31
Modelling Approach

• A discrete model based on mathematical analysis
  is generated and model using the simulation
  software
• Discrete-time equivalent system of digital
  communication system is defined as
• ri Ebni if bi1
• ri ni if bi0
• ri is the sampled output
• Eb is the energy per bit and ni is the
  additive white Gaussian noise
• Performance evaluation
• bit error rate
• eye-diagram

32
Digital Systems Matlab Simulink
33
Digital Simulation - Performance Evaluation

• BER of different modulation techniques for
  indoor optical wireless system

34
Digital Simulation - Notes

• To properly model the system, it is necessary to
  understand mathematics involved in each and every
  module
• Code are written to approximate the mathematical
  equations. The code are grouped together and put
  as a block for simple user interface
• Example Matlab codes for noise signal

35
Digital Simulation Matlab Codes
Fixed and variable parameters clear clc close
all fs 6.0e6 sampling frequency 6 MHz ts
1/fs Sampling time fc clock signal
frequency ac clock signal peak amplitude n
2(6fs/fc) Maximum number of points w.r.t
the 6 cycles of clock signal fc nc
6 Number cycls of clock signal to be
shown tmax nctc Maximum number of point in 6
cycles of fc fmax (2nfc/fs) Maximum
frequency range final ts(n-1) maximum time
t 0tstmax time vector for sketching
waveform in time domain
36
Digital Simulation Matlab Codes
Data signal generated from the Clock Signal L
length (sq) All the values of clock signal is
assigned to a new variable l da sq Set
initial values out1 temp1 for i1L-1 if
sq(i) -2.5 sq(i1) 2.5 Reverse
output voltage polarity temp out -1
outtemp end Change value of out to
/-1 if outgt0 out1 else out
-1 end da(i)out data signal at half the
clock frequency end Set value of final element
of da da(L)out Plot data signal
37
Optical Wireless Communication
What does It Offer ?

• Abundance of unregulated bandwidth - 200 THz in
  the 700-1500
• 
  nm range

No multipath fading - Intensity modulation and
direct
detection
High data rate In particular line of sight (in
and out doors)
Improved wavelength reuse capability
Flexibility in installation
Secure transmission
Flexibility - Deployment in a wide variety of
network architectures. Installation on roof to
roof, window to window, window to roof or
wall to wall.
38
Access Network Bottleneck
(Source NTT)
Eng. of S/W Pro., India 2009
39
Free Space Optics

• Cloud
• Rain
• Smoke
• Gases
• Temperature variations
• Fog and aerosol

The transmission of optical radiation through the
atmosphere obeys the Beer-Lambertss law
Preceive Ptransmit exp(-aL)
a Attenuation coefficient
This equation fundamentally ties FSO to the
atmospheric weather conditions
39
Eng. of S/W Pro., India 2009
40
Case Study 3 Optical Wireless Systems
DC bias
m(t)
m(t)bo
d(t)
Subcarrier modulator
Summing circuit
Optical transmitter
Serial/parallel converter
Data in
Atmospheric channel
ir
d(t)
Photo- detector array
Spatial diversity combiner
Subcarrier demodulator
Parallel/serial converter
. .
Data out
40
Eng. of S/W Pro., India 2009
41
Subcarrier Modulation - Transmitter
Eng. of S/W Pro., India 2009
42
Subcarrier Modulation - Receiver
42
Eng. of S/W Pro., India 2009
43
Error Performance Bit Error Performance
BPSK BER against SNR for M-ary-PSK for log
intensity variance 0.52
BPSK based subcarrier modulation is the most
power efficient
Eng. of S/W Pro., India 2009
43
44
Receiver Models
Eng. of S/W Pro., India 2009
45
Wavelet-AI Receiver - Advantages and Disadvantages

• Complexity
• - many parameters computation power
• High sampling rates
• - technology limited
• Speed
• - long simulation times on average machines
• Similar performance to other techniques
• Data rate independent
• - data rate changes do not affect structure (just
  re-train)
• Relatively easy to implement with other pulse
  modulation techniques

46
Wavelet-AI Receiver
SNR Vs. the RMS delay spread/bit duration
47
Final Remarks

• Simulation software provide scientist and
  engineers with additional tools to implement,
  assess and modify ideas with a press of a button
• Detailed mathematical understanding is essential
• High speed and parallel processing is the way
  forward
• Should never be a substitute to real practical
  systems

48

• Thank you for your attention !
• Any questions?

49
Acknowledgements

• To R Kharel, S Rajbhandari, W Popoola, and other
  PhD students,
• Northumbria University and CEIS School for
  Research Grants