Report

Welcome to Simulation of communication systems (DT001A) [email protected] and [email protected] A project course about MATLAB with SIMULINK and Communications Blockset… MATLAB = Matrix Laboratory. Tool for numerical calculation and visualization. Commonly used for simulation of the communication system physical layer, signal and image processing research, etc. SIMULINK: Toolbox in Matlab that allows graphical data-flow oriented programming. …and about Network Simulation using tools such as Opnet, NS/2, etc Aim of the course To prepare the student for thesis project and work in the area of telecommunciations development and research. To give experience of performance analysis of communication systems and algorithms, at the physical layer and datalink layer. To give experience of simulation tools such as MATLAB, SIMULINK and/or Opnet. This may include modelling and simulation of traffic sources, channel models, modulation schemes, error coding schemes, equalizers, algorithms, protocols and network topologies. A real-world project is studied within an application area such as wireless sensor networks, cellular communications, modems for broadband access, wireless networks, shortrange communication, digital TV transmission, IP-TV or IPtelephony. Prerequisites Computer Networks A 7.5 ECTS credits or similar Computer Engineering B, Wireless Internet access (most important!) Computer Engineering AB-level, 30 ECTS credits TCP/IP networking Mathematical statistics Programming Other helpful courses: Transform theory, 7.5 ECTS credits. Electrical engineering A, Analog electronics or Circuit theory Electrical Engineering B, Telecommunications, 7.5 ECTS credits. Electrical engineering B, Signals and systems, 7.5 ECTS credits. Markov processes/Queueing theory Litterature Matlab, Simulink and Opnet documentation will be provided electronically. Please repeat physical layer issues and datalink layer issues in basic books in Computer Networks and Wireless Internet Access. Requirements All lectures and supervision lessons are mandatory. You should attend 80% of the mandatory lessons. You are expected to devote 20 hours/week to this course. Quzzes (multiple choice tests): At least 70% correct answers. Lab: About 20 hours of work. Homework problem. Oral presentations. Project Requirements on the project Review at least one research paper, and describe some standard and some existing simulation model. Simulate a communications standard, or check the simulations made in a research paper. At least modify an existing simulation model, for exampel a Simulink or Matlab demo, or build a model of your own (more difficult) Produce some plots for several parameter cases, showing for example BER, bit rate or delay as function of at least two different parameters, for example SNR, facing model, modulation scheme, etc. The simulation results should be stable (the plots smooth and not jerky), i.e simulate sufficiently long simulation time, or take the average of sufficiently large number of simulations. Draw some interesting conclusions from this. Grading is based on Keeping deadlines. Quzzes. Showing good understanding when andwering questions from teachers and other students about your presentations. Extent of own code. Research relevance. Own new results or conclusions. Assignment 1: Theory repetition The first assignment consists of old exam problems in Computer Networks A, Wireless Internet access B and Telecommunications B. Deadline: Friday course week 2. Be prepared to present your answers on the whiteboard. Assignment 2: Simulink lab exercize Takes about 10-15 hours to do. Deadline: Course week 3 Assignment 3: Present a standard and an existing simulation model Essentially chapter 2 (theory) and 3 (existing model that you start out from) of your report. Examples 1. 802.11b PHY Simulink model and adaptive modulation and link control 2. 256 channel ADSL and bit loading. 3. Bluetooth Voice Transmission 4. Bluetooth Full Duplex Voice and Data Transmission – Also describe Bluetooth low energy mode. Can be extended to thesis project. 5. Digital Video Broadcasting Model (DVB-T). Also describe DVB-T2. 6. NFC (Simulink model by previous years students – see Mathworks file archive). 7. IEEE 802.11a WLAN Physical Layer. Also describe newer standards. 802.11n, 802.11ac, ad or v positioning. Stefan, Erik, Andreas Gabriel. 8. CDMA2000 Physical Layer. 9. WCDMA Coding and Multiplexing. 10. WCDMA Spreading and Modulation 11. WCDMA End-to-end Physical Layer. Hassan, Lamin. 12. Ultrawideband (UWB/wireless USB). See mathworks file archive. Fredrik, Markus, Seb 13. ZigBee Simulink or Prowler model and IEEE 802.15.4g (smart grid). See mathworks file central. Altahra. 14. ZigBee Prowler model and Multihop routing protocols (Prowler model) . You may demonstrate simulink model (see Matlab file central) or Prowler model. Perhaps you can add cooperative diversity. Yuxin Guo , Yu Tang, Suna Yin, Mengjun Qin. 15. Opnet Mac Protocol Mehrzad 16. Mobile Wimax 17. Long-term evolution (LTE) Phy Downlink with spatial modeling. Also describe LTE-A. cheng yang, shixian wen, wei liu 18. Long-term evolution (LTE) and eMBMS 19. Line codes. Comparison of RZ, NRZ, AMI, Manchester coding (used in 10 Mbps Ethernet), 4B5B (used in 100Base-TX Ethernet) and PAM5 (used in 1000Base-T Gigabit Ethernet): For a code demonstrating RZ, NRZ, AMI and Manchester, see http://apachepersonal.miun.se/~rogols/teaching/mks/lab2/LineEncoding.mdl This code also requires this MATLAB function: http://apachepersonal.miun.se/~rogols/teaching/mks/lab2/line_encoder.m . During the rest of the project you may further develope the code to deal with 4B5B and PAM5, and to measure the bit error rate. 20. 21. Acoustic modem (new model) Acoustic QR code (continue on project by previous year’s students) Assignment 3 (cont.) Oral presentation: Course week 3. Talk 5-10 minutes per person. Everyone should take notes, and everyone should ask questions and discuss the topic. Present: A standard (mention things like radio frequency, bandwidth, bit rate, modulation, error control method, multiplex method, multiple-access protocol, new/future versions) New versions of the standard or ongoing development Screen dumps – or demonstration of - an existing simulation Differences between simulation and full standard For higher grades: Also cite a related research paper or a textbook, for example a simulation method with results. See scholar.google.com or library. Within one week after that: Submit report chapter 2 (theory/previous research) and chapter 3 (existing model) Assignment 4: Quizzes Basic concepts, Matlab and Simulink concepts Requirement: At least 70% correct answers. You can do them over and over again until the deadline. Assignment 5: Opnet lab Zigbee and multihop simulation in Opnet. Takes about 4 hours to do. Assignment 6: Present project suggestion Oral presentation course week 6. Present Problem formulation (chapter 1) – what to parameters to evaluate Cite simulation done in a research paper (if you have not done so) Planned own modification or development of model (chapter 4) Submit or show report chapters 1 and 4 before christmas. Assignment 7: Final project presentation Demonstrate simulation code to teacher (and also in report appendice) Oral presentation in mid-January of Results (chapter 5): Plot performance for several cases. Conclusions (chapter 6). Discuss similarities and differences from result in a cited research paper. Provide a preliminary report when you give your oral presentation. MATLAB MATLAB = Matrix Laboratory. Tool for numerical calculation and visualization. Commonly used for simulation of the communication system physical layer, signal and image processing research, etc. 19 This is how MATLAB looks like Workspace Command history Command window 20 More MATLAB windows Figure window Array editor M-file editor 21 How to get help in MATLAB? help functionsname Shows unformatted text doc funktionsnamn Shows HTML documentation in a browser 22 SIMULINK SIMULINK: Toolbox in Matlab that allows graphical data-flow oriented programming. Repetition of some basic concepts Frequency spectrum Digitalisation, source coding Error coding Modulation Multiple-access methods Base-band model Distorsion, noise Signal-to-noise ratio Bit-error ratio Statistics Repetition of some basic concepts Digitalization PCM = Pulse Code Modulation = Digital transmission of analogue signals Number exemples from PSTN = the public telephone network 011011010001... 1 0 Anti aliasingfilter Sampler AD-converter with seerial output DAconverter Interpolation filter Loudspeaker Microphone 8000 300-3400Hz sampels band pass filter. Stops per sec everything over 4000Hz. 8 bit per sampel i.e. 64000 bps per phone call 28 = 256 voltage levels Aliasing Quantization noice Digital transmission Distorsion Effect of attenuation, distortion, and noise on transmitted signal. Point-to-point communication Layer 7 Layer 6 Mikrofon Högtalare Source coding Digitalizating compression 0110 Error management Layer 2 NACK ACK 0110010 Source decoding 0110 . Error control 0100010 0110010 Bitfel Layer 1 Flow control Flow control Modulation Demodulation Digital modulation methods Binary signal ASK = Amplitude Shift Keying (AM) FSK = Frequency Shift Keying (FM) PSK = Phase Shift Keying (PSK) 8QAM example: Below you find eight symbols used for a so called 8QAM modem (QAM=Quadrature Amplitude Modulation). The symbols in the first row represent the messages 000, 001, 011 and 010 respectively (from left to right). The second row representents 100, 101, 111 and 110. 000 001 011 010 2 2 2 2 0 0 0 0 -2 0 0.005 100 -2 0.01 0 0.005 101 -2 0.01 0 0.005 111 -2 0.01 0 2 2 2 2 0 0 0 0 -2 0 0.005 -2 0.01 0 0.005 -2 0.01 0 0.005 -2 0.01 0 0.005 110 0.01 0.005 0.01 Example 2 cont. a) The signal below is transmitted from the modulator. What bit sequency is transmitted? Spänning [Volt] Modulatorns utsignal 2 0 -2 0 0.005 0.01 0.015 0.02 0.025 Tid [sekunder] 0.03 0.035 0.04 b) The time axis is graded in seconds. What is the symbol rate in baud or symbols/s? c) What is the bit rate in bit/s? Bit rate vs baud rate Bit rate in bit/s: fb f S log 2 M Where M is the number of symbols and fs is the symbol rate in baud or symbols/s. Bit and baud rate comparison Units Bits /symbol Baud rate Bit Rate Bit 1 N N 4-PSK, 4-QAM Dibit 2 N 2N 8-PSK, 8-QAM Tribit 3 N 3N 16-QAM Quadbit 4 N 4N 32-QAM Pentabit 5 N 5N 64-QAM Hexabit 6 N 6N 128-QAM Septabit 7 N 7N 256-QAM Octabit 8 N 8N Modulation ASK, FSK, 2-PSK Figure 5.14 The 4-QAM and 8-QAM constellations Q (Quadrature phase) Q (Quadrature phase) I (Inphase) I (Inphase) Sine wave example Complex representation 5 Volt л/2 radians =I 90º Inphase and quadrature phase signal Sine wave as reference (inphase) signal: s(t ) I (t )sin(2 fct ) Q(t )cos(2 fct ). Cosine wave as reference (inphase) signal: u(t ) I (t )cos(2 fct ) Q(t )sin(2 f ct ). Complex baseband representation jQ C = I+jQ Amplitude: C I Q 2 |C| 2 C Arg C Phase: Q if I 0 arctan I , arg( I jQ) arctan Q , if I 0 I RF signal (physical bandpass signal, if a cosine is reference signal): s(t ) C cos(2 fct arg C). I Equivalent baseband signal s(t ) I (t )sin(2 fct ) jQ(t )cos(2 fct ). Figure 5.11 The 4-PSK characteristics Figure 5.12 The 8-PSK characteristics Figure 5.16 16-QAM constellations Spectrum of ASK, PSK and QAM signal Figure 3.9 Three harmonics Figure 3.10 Adding first three harmonics Example: Square Wave Square wave with frequency fo 4A 1 1 s(t ) {cos ot cos 3ot cos 5ot ...} 3 5 Component 1: Component 3: . . . Component 5: . . . s1 (t ) 4A cos o t 4A s3 (t ) cos 3ot 3 4A s5 (t ) cos 5o t 5 Figure 3.11 Frequency spectrum comparison Filtering the Signal Filtering is equivalent to cutting all the frequiencies outside the band of the filter • Types of filters – Low pass Low pass H(f) INPUT S1(f) OUTPUT S2(f)= H(f)*S1(f) H(f) f Band pass – Band pass H(f) INPUT S1(f) OUTPUT S2(f)= H(f)*S1(f) H(f) f – High pass High pass H(f) INPUT S1(f) H(f) f OUTPUT S2(f)= H(f)*S1(f) Figure 6.4 FDM (Frequency division multiplex) Figure 6.5 FDM demultiplexing example Figure 6.19 Time division multiplex (TDM) in the american telephone network Multi-path propagation Multiple access = channel access Several transmitters sharing the same physical medium, for example wireless network, bus network or bus network. Based on A physical layer multiplexing scheme A data link layer MAC protocol (medium access control) that avoids collisions, etc. Examples: TDMA (time division multiple-access) based on TDM FDMA (frequency division multiple-access) based on FDM CDMA based on spread spectrum multiplexing CSMA (carrier sense multiple-access) based on packet switching = statistical multiplexing OFDMA Cellular telephony generations 1G: (E.g. NMT 1981) Analog, FDMA circuit switched. 2G: (E.g. GSM 1991) Digital, FDMA+TDMA, 8 timeslots, circuit switched. 2.5G: (GPRS) Packet switched = statistical multiplexing. The old circuit switched infrastructure is kept. 3G: (e.g. WCDMA) FDMA + CDMA (= spread spectrum). 4G: (E.g. 3gpp LTE) All-IP. OFDM or similar. Spread spectrum DS-CDMA = Direct Sequence Code Division Multiple Access Chip sequencies Figure 13.15 Encoding rules Figure 13.16 CDMA multiplexer Figure 13.17 CDMA demultiplexer Figure 9.1 Discrete Multi Tone (DMT) Essentially the same thing as OFDM Used in ADSL modems Figure 9.2 ADSL Bandwidth division The 8PSK constellation 011 001 -sin 0.5 0 100 000 -0.5 101 -1 -1.5 -1.5 111 110 -1 -0.5 0 cos 0.5 1 A simple example: 4 sub-carriers 8 PSK 1.5 Subcarrier 1 Subcarrier 4 010 Sum signal 1 Subcarrier 3 1.5 Subcarrier 2 OFDM modulation 1 0 -1 1 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 0.2 0.4 0.6 0.8 1 1.2 Time [ms] 1.4 1.6 1.8 2 0 -1 1 0 -1 1 0 -1 5 0 -5 000 100 010 010 111 000 010 000 Technical data for DAB and DVB-T Adopted Coverage in parts of: Net bit rate R per frequency channel: Channel separation B: Link level spectrum efficiency R/B: Freq. range of today’s receivers: Maximum speed: Number of OFDM sub-carriers: Sub-carrier modulation: Inner Forward Error Correction Coding (FEC): Outer FEC: Time (outer) interleaving: DAB DVB-T 1995 Canada, Europe, Australia 576 - 1152 kbit/s 1997 Europe and Australia 4.98 - 31.67 Mbit/s 1.712 MHz 0.34 - 0.67 bit/s/Hz 8 MHz 0.62 - 4.0 bit/s/Hz 174 – 240 MHz , 1452 – 1492 MHz. About 200 - 600 km/h 1536, 384, 192 or 768. 470 - 862 MHz DQPSK Convolutional coding with code rates 1/4, 3/8 or 1/2. None Convolutional interleaving of depth 384 ms. 36 - 163 km/h The 2K mode: 1705 The 8K mode: 6817 QAM, 16QAM or 64QAM Convolutional coding with code rates 1/2, 2/3, 3/4, 5/6 or 7/8. RS(204,188,t=8) Convolutional interleaving of depth 0.6 - 3.5 ms. Orthogonal Frequency Division Multiplex (OFDM) Summary of advantages Can easily adapt to severe channel conditions without complex equalization Robust against narrow-band co-channel interference Robust against Intersymbol interference (ISI) and fading caused by multipath propagation High spectral efficiency Efficient implementation using FFT Low sensitivity to time synchronization errors Tuned sub-channel receiver filters are not required (unlike conventional FDM) Facilitates Single Frequency Networks, i.e. transmitter macrodiversity. Summary of disadvantages Sensitive to Doppler shift. Sensitive to frequency synchronization problems. Inefficient transmitter power consumption, due to linear power amplifier requirement. Bit error rate (BER) = Bit error probability = Pb Packet error rate (PER) = Packet error probability for packet length N bits: Pp = 1 – (1-Pb)N Error-correcting codes (ECC), also known as Forward-error correcting codes (FCC) A block code converts a fixed length of K data bits to a fixed length N codeword, where N > K. A convolutions code inserts redundant bits into the bit-stream. Code rate ¾ means that for every 3 information bit, totally 4 are transferred, i.e. every forth of the transferred bits is redundant. Bit rates Gross bit rate = Transmission rate. Symbol rate = Baud rate ≤ Gross bit rate In spread spectrum: Chip rate ≥ Bit rate ≥ Symbol rate. In FEC: Net bit rate = Information rate = Useful bit rate ≤ Code rate * Gross bit rate Maximum throughput ≤ Net bit rate Goodput ≤ Throughput Nyquist formula Gives the gross bit rate,without taking noise into consideration: Symbol rate < Bandwidth*2 Bit rate < Bandwidth * 2log M The above can be reached for line coding (base band transmission) and so called singlesideband modulation. Howeverm in practice most digital modulation methods give: Symbol rate = Bandwidth Signal to noise ratios S/N= SNR = Signal-to-noise ratio. Often same thing as C/N=CNR = Carrier-to-noise ratio SNR in dB = 10 log10 (S/N) S/I = SIR = Signal-to-interference ratio. Often the same thing as C/I=CIR = Carrier-to-interference ratio. I is the cross-talk power. CINR = C/(I+N) = Carrier-to-noise and interference ratio Eb/N0 = Bit-energy (Power in watt divided by bitrate) divided by Noise density (in Watt per Hertz) Es/N0 = Symbol-energy (Power in Watt divided by bitrate) divided by Noise density (in Watt per Hertz) Shannon-Heartly formula Gives the channel capacity, i.e. the maximum information rate (useful bit rate) excluding bit error rate. I=B log2 (1+C/N) Where C/N is carrier-to-noise ratio (sometimes called S/N) Some statistical distributions Gaussian noise Voltage Time Gaussian = Normal distribution Probability density funciton Additive White Gaussian Noise (AWGN) channel White noise = wideband (unfiltered) noise with constant noise density in Watt/Hertz Pink noise = lowpass-filtered noise. Additive = linear mixing. Signal Noisy signal + Noise source Bernoulli distribution Channel Noise 0 Tx 0 1 0 1BSC 1010010 Bernoulli Binary Error Rate Rx Calculation Error Rate Calculation Bernoulli Binary Generator Binary Symmetric Channel Info Random sequence of independent 0:s and 1:s. Display Exponential distribution Commonly used for time between phone calls and length of phone calls. Simple model for calculation and simulation, but does not reflect data traffic bursty nature. Multi-path propagation Rayleigh distribution Model of rayleigh fading, i.e. amplitude gain caused by multi-path propagation with no line-of-sight More commons distributions Ricean distribution (fading with line-ofsight) Poisson distribution (number of phone calls during a phone call) Self-similar process (bursty data traffic) Rectangular distribution Discrete distributions, for example the distribution of a dice