Digital Communications 1: Source and Channel Coding

312 Pages·2015·11.691 MB·English

by  Didier Le Ruyet, Myl?ne Pischella

Table of Contents
Cover
Title
Copyright
Preface
List of Acronyms
Notations
Introduction
1: Introduction to Information Theory
1.1. Introduction
1.2. Review of probabilities
1.3. Entropy and mutual information
1.4. Lossless source coding theorems
1.5. Theorem for lossy source coding
1.6. Transmission channel models
1.7. Capacity of a transmission channel
1.8. Exercises
2: Source Coding
2.1. Introduction
2.2. Algorithms for lossless source coding
2.3. Sampling and quantization
2.4. Coding techniques for analog sources with memory
2.5. Application to the image and sound compression
2.6. Exercises
3: Linear Block Codes
3.1. Introduction
3.2. Finite fields
3.3. Linear block codes
3.4. Decoding of binary linear block codes
3.5. Performances of linear block codes
3.6. Cyclic codes
3.7. Applications
3.8. Exercises
4: Convolutional Codes
4.1. Introduction
4.2. Mathematical representations and hardware structures
4.3. Graphical representation of the convolutional codes
4.4. Free distance and transfer function of convolutional codes
4.5. Viterbi’s algorithm for the decoding of convolutional codes
4.6. Punctured convolutional codes
4.7. Applications
4.8. Exercises
5: Concatenated Codes and Iterative Decoding
5.1. Introduction
5.2. Soft input soft output decoding
5.3. LDPC codes
5.4. Parallel concatenated convolutional codes or turbo codes
5.5. Other classes of concatenated codes
5.6. Exercises
Appendix A: Proof of the Channel Capacity of the Additive White Gaussian Noise Channel
Appendix B: Calculation of the Weight Enumerator Function IRWEF of a Systematic
Recursive Convolutional Encoder
Bibliography
Index
End User License Agreement
List of Illustrations
Introduction
Figure I.1. Binary symmetric channel
Figure I.2. Image at the input and output of a binary symmetric channel
Figure I.3. Block diagram of a transmission chain
1: Introduction to Information Theory
Figure 1.1. Entropy of a binary source
Figure 1.2. Relations between entropies and average mutual information
Figure 1.3. Density probability and typical sequences set
Figure 1.4. Block diagram of the studied chain for source coding
Figure 1.5. Source coding
Figure 1.6. Tree associated with the source code of example 3
Figure 1.7. Kraft inequality
Figure 1.8. Entropy rate
Figure 1.9. Block diagram of the coder-decoder
Figure 1.10. Shannon distortion-rate for a Gaussian source of unitary variance
Figure 1.11. Uniform and non-uniform quantization L = 8 for a Gaussian source with
variance σ = 1
x
Figure 1.12. Binary symmetric channel
Figure 1.13. Conditional entropy H(X|Y) versus q and p
Figure 1.14. Discrete channel without memory
Figure 1.15. Erasure channel
Figure 1.16. Case C = H (X)
MAX
Figure 1.17. Case C = 0
Figure 1.18. Communication system with channel coding
Figure 1.19. Mutual information I(X;Y) versus q and p
Figure 1.20. Capacity of the binary symmetric channel versus p
Figure 1.21. Capacity of the additive white Gaussian noise channel
Figure 1.22. Spheres of noise illustration
Figure 1.23. Maximum spectral efficiency of an additive white Gaussian noise channel
Figure 1.24. Spectral efficiency versus E /N
b 0
2: Source Coding
Figure 2.1. Example of Huffman’s encoding
Figure 2.2. Probabilities of occurrence of the characters
Figure 2.3. Example of partitioning
Figure 2.4. Example of arithmetic coding
Figure 2.5. Tree associated with the strings memorized in the dictionary
Figure 2.6. Tree of the prototype strings
Figure 2.7. Sampling and reconstruction
Figure 2.8. Quantification uniform L = 8
Figure 2.9. Uniform quantization L = 8
Figure 2.10. Non-uniform quantization L = 8
Figure 2.11. Non-uniform quantization L = 8
Figure 2.12. Block diagram of the PCM coder
Figure 2.13. A law and μ law. For a color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 2.14. Example of realization of a source with memory function of the time
Figure 2.15. Example of realization of a source with memory projected on a plane
Figure 2.16. Example of scalar quantization. For a color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 2.17. Example of vector quantization.For a color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 2.18. Block diagram of the P order predictor
Figure 2.19. Delta modulator
Figure 2.20. Delta demodulator
Figure 2.21. Example of behavior of in a delta modulator delta
Figure 2.22. Ideal block diagram of a DPCM transmission chain
Figure 2.23. Block diagram of the DPCM coder
Figure 2.24. DPCM decoder
Figure 2.25. Block diagram of the transform coding
Figure 2.26. Example of scalar quantization after a Karhunen–Loève transform. For a
color version of the figure, see www.iste.co.uk/leruyet/communications1.zip
Figure 2.27. Subband coding
Figure 2.28. Zigzag serialization
Figure 2.29. Example of speech signal (“this is”) in the time domain
Figure 2.30. Simplified block diagram of the LPC coder
Figure 2.31. Block diagram of the CELP coder
Figure 2.32. Another version of the CELP coder
Figure 2.33. Absolute threshold curve
Figure 2.34. Masking level in frequency. Fora color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 2.35. Block diagram of an MPEG audio coder
Figure 2.36. Impulse response of the prototype filter
Figure 2.37. Frequency response of the filterbank in the bandwidth [0; 22 Khz]
Figure 2.38. Probability density
Figure 2.39. Density probability
Figure 2.40. Block diagram of the predictive coding system
Figure 2.41. Block diagram of the modified coder
3: Linear Block Codes
Figure 3.1. Parity check code C (3, 2)
2
Figure 3.2. Hamming sphere
Figure 3.3. Bounds on the minimum distance for linear block codes with q = 2
Figure 3.4. Bounds on the minimum distance for linear block codes with q = 256
Figure 3.5. Error exponent function versus rate
Figure 3.6. Poltyrev bound rate versus length N
Figure 3.7. Sphere packing bound ratio E /N versus N
b 0
Figure 3.8. Two versions of the Tanner graph for the Hamming code (7, 4)
Figure 3.9. Branch of a trellis section
Figure 3.10. Trellis diagram obtained from the parity check matrix of the code C
3
Figure 3.11. Trellis diagram of Hamming code (7,4)
Figure 3.12. Block diagram of a transmission chain with an additive white Gaussian
noise channel
Figure 3.13. Branch metric calculation
Figure 3.14. Cumulated metric calculation after the reception of the 1st bit of the word
r. For a color version of the figure, see www.iste.co.uk/leruyet/communications1.zip
Figure 3.15. Cumulated metric calculation after the reception of the 4th bit of the word
r. For a color version of the figure, see www.iste.co.uk/leruyet/communications1.zip
Figure 3.16. Cumulated metric calculation
Figure 3.17. Determination of the estimated sequence. Fora color version of the figure,
see www.iste.co.uk/leruyet/communications1.zip
Figure 3.18. Branch metric calculation
Figure 3.19. Cumulated metric calculation. For a color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 3.20. Cumulated metric calculation. For a color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 3.21. Determination of the estimated sequence. For a color version of the figure,
see www.iste.co.uk/leruyet/communications1.zip
Figure 3.22. Iterative decoding on the Tanner graph
Figure 3.23. Decision regions associated with the codewords
Figure 3.24. Example
Figure 3.25. Coding gain of different codes
Figure 3.26. Performance comparison of hard and soft input decoding
Figure 3.27. a) Stop-and-wait protocol b) go-back-N protocol with N = 4
f
Figure 3.28. Hardware structure of the division by 1 +p + p2
Figure 3.29. Sequencing of the division by 1+p + p2
Figure 3.30. Hardware structure of the Hamming encoder with g(p) = 1+p + p3
Figure 3.31. Hardware structure of the Hamming encoder with premultiplication for
g(p) = 1 + p + p3
Figure 3.32. Complete hardware structure of the Hamming encoder for g (p) = 1 +p +
p3
Figure 3.33. Trellis diagram of the Hamming code defined by the polynomial g(p) = 1 +
p + p3
Figure 3.34. Hardware structure of the decoder for the Hamming code with g(p) = 1 +
p + p3
Figure 3.35. Hardware structure of the decoder for the Hamming code (7,4)
Figure 3.36. SER = f(SER ) for the (255,239,17) Reed-Solomon code
O I
Figure 3.37. Trellis diagram of the (8,4) Reed-Muller code
Figure 3.38. Trellis diagram of the (7, 4) code
Figure 3.39. Trellis diagram of the (3,2) code
Figure 3.40. Trellis diagram
4: Convolutional Codes
Figure 4.1. Convolutional encoder
Figure 4.2. Non-recursive convolutional coder of rate 1/2 M =2 = 7, = 5
g1 g2
Figure 4.3. Non-recursive convolutional encoder of rate 1/2 M = 6 = 133, = 171
g1 g2
Figure 4.4. Systematic recursive convolutional encoder 1/2 M = 2
Figure 4.5. Generic non-recursive convolutional encoder of rate 1/n
Figure 4.6. State transition diagram for the non-recursive convolutional code = 7,
g1 g2
= 5
Figure 4.7. Elementary trellis of a non-recursive convolutional encoder = 7, = 5
g1 g2
Figure 4.8. Trellis diagram for the non-recursive convolutional coder = 7, = 5
g1 g2
Figure 4.9. TWL graph of a systematic convolutional coder of rate 1/2
Figure 4.10. Trellis diagram of the non-recursive
Figure 4.11. Modified state diagram of a non-recursive
Figure 4.12. Viterbi decoding for a non-recursive convolutional encoder = 7, = 5
g1 g2
Figure 4.13. Trellis diagram of a punctured convolutional encoder
5: Concatenated Codes and Iterative Decoding
Figure 5.1. Factor graph of the parity check code (3,2)
Figure 5.2. Factor graph of the repetition code (3,1)
Figure 5.3. Message flows for the calculation of μ (c) and μ (c)
c→T T→c
Figure 5.4. Level curves of μ (c) as a function of and a) sum-
T→c
product algorithm and b) minimum-sum algorithm
Figure 5.5. Tanner graph of the considered code (7,4)
Figure 5.6. Sum-product algorithm applied to the factor graph of the code (7, 4)
Figure 5.7. Minimum-sum algorithm for the code (7, 4)
Figure 5.8. Quantities α(m), γ(m′,m) and β (m) on the trellis of a convolutional code
t t t+1
Figure 5.9. Trellis of the coder
Figure 5.10. Calculation of α
Figure 5.11. Calculation of β
Figure 5.12. Calculation of ln α
Figure 5.13. Calculation of ln β
Figure 5.14. Communication system
Figure 5.15. Example of a factor graph for a recursive convolutional encoder of rate
1/2 used for the forward-backward algorithm
Figure 5.16. Message passing when performing the forward-backward algorithm
Figure 5.17. Tanner graph of a (N, d , d ) regular LDPC code The rate of an
c T
LDPC code is:
Figure 5.18. Tanner graph of a (12, 3, 4) regular LDPC code
Figure 5.19. Bipartite graph with girth of 4
Figure 5.20. Parity check matrix of a (638,324) QC-LDPC code
Figure 5.21. Parity matrix H with a lower triangular form
Figure 5.22. Graphical illustration of the evolution of the erasure probability as a
function of the iterations for a regular LDPC code with d = 6, d = 3 . For a color
c T
version of the figure, see www.iste.co.uk/leruyet/communications1.zip
Figure 5.23. Graphical illustration of the evolution of the erasure probability as a
function of the iterations for a irregular LDPC code. For a color version of the figure,
see www.iste.co.uk/leruyet/communications1.zip
Figure 5.24. Factor graph of the (N,d , d ) regular LDPC code
c T
Figure 5.25. Example of local tree for a (d , d ) LDPC code
c T
Figure 5.26. Message exchange for the calculation of and
Figure 5.27. Bit error rate versus E /N for the (4320, 3557) regular LDPC code with
b 0
Di = 17 and d = 3
T
Figure 5.28. Comparison of the mean and variance calculated using the Gaussian
approximation and using the density probability evolution. For a color version of the
figure, see
Figure 5.29. Parallel concatenated convolutional encoder
Figure 5.30. TWL graph of a turbo encoder
Figure 5.31. Parity check matrix of a turbo coder composed of two RSC coders (7,5)
with RT = 1/3
Figure 5.32. Comparison of the average performance of average turbo codes with rate
R = 1/3 composed of RSC (7,5) encoders (dashed lines) and RSC (15,17) encoders
t
(continuous line) and a uniform interleaver of size K = 100 and K = 1000
Figure 5.33. Contribution of the input sequences of low weight on the bit error rate of a
turbo coder of rate R = 1/3 composes of RSC (15,17) encoders and a uniform
t
interleaver of size K = 100. For a color version of the figure, see
www.iste.co.uk/leruyet/communications1.zip
Figure 5.34. Contribution of the input sequences of low weight on the bit error rate of a
turbo coder of rate R = 1/3 composes of RSC (15,17) encoders and a uniform
t
interleaver of size K = 1000
Figure 5.35. Structure of the iterative decoder
Figure 5.36. Example of factor graph for a turbo code of rate 1/3
Figure 5.37. BER = f(E /N ) for a turbo code of rate 1/2 composed of two RSC
B 0
encoders and an S-random interleaver of size K = 1024
Figure 5.38. Curves BER = f(E /N ) for a turbo code of rate 1/2 composed of two
B 0
RSC encoders and a pseudo random interleaver of size K=65536
Figure 5.39. EXIT charts for an RSC encoder (13,15) of rate R = 1/2
Figure 5.40. Evolution of the average mutual information as a function of the iterations
for a turbo code
Figure 5.41. Example of primary cycle
Figure 5.42. Interleavers a) L-random K = 320 and b) QPP K = 280
Figure 5.43. Code PCBC
Figure 5.44. TWL graph of the PCBC codes
Figure 5.45. Serial concatenated convolutional encoder
Figure 5.46. RA coder R = 1/2
T
Figure 5.47. Product code
List of Tables
1: Introduction to Information Theory
Table 1.1. Variable length code of example 1
Table 1.2. Variable length code of example 2
Table 1.3. Variable length code of example 3
Table 1.4. Probabilities P (X, Y), P (Y) and P (X|Y) for the binary symmetric channel
r r r
Table 1.5. Probabilities P (X,Y), P (Y) and P (X|Y) for the erasure channel
r r r
2: Source Coding
Table 2.1. Huffman’s encoding table
Table 2.2. Probabilities of occurrence of characters
Table 2.3. Dictionary of strings