Introduction To Information Theory Channel Capacity and Models
Introduction To Information Theory Channel Capacity and Models
theory
channel capacity and models
A.J. Han Vinck
University of Essen
May 2009
This lecture
Some models
Channel capacity
Shannon channel coding theorem
converse
some channel models
transition probabilities
memoryless:
- output at time i depends only on input at time i
- input and output alphabet finite
Example: binary symmetric channel (BSC)
1-p
Error Source
0 0
E
p
X Y X E
+ 1 1
Input Output
1-p
1-e
0 0 (light on) 0 0
e
X Yp
1-p
1 1 (light off) E
e
1
P(X=0) = P0 1-e 1
P(X=0) = P0
1-p-e
0 e 0
p
E
p e
1
1-p-e 1
burst error model (Gilbert-Elliot)
X Y
H(X) channel H(X|Y)
notes:
capacity depends on input probabilities
because the transition probabilites are fixed
Practical communication system design
Code book
Code receive
message word in
estimate
2k channel decoder
Code book
with errors
n
There are 2k code words of length n
k is the number of information bits transmitted in n channel uses
Channel capacity
Definition:
The rate R of a code is the ratio k/n, where
k is the number of information bits transmitted in n channel uses
2nh(p)
P( 1) (2k 1)
2n
2n(1h(p)R) 2n(CBSCR) 0
k
for R 1 h(p)
n
and n
channel capacity: the BSC
0.5 1.0
Bit error p
channel capacity: the Z-channel
1-e
0 0 I(X;Y) = H(X) – H(X|Y)
e
H(X) = h(P0 )
X Y
E H(X|Y) = e h(P0)
e
1
1-e 1
Thus Cerasure = 1 – e
P(X=0) = P0 (check!, draw and compare with BSC and Z)
Erasure with errors: calculate the capacity!
1-p-e
0 e 0
p
E
p e
1
1-p-e 1
0 0
1/3
example 1 1
1/3
2 2
Consider the following example
P1|1 y1
x1
P1|2
P2|1 Input alphabet X = {x1, x2, …, xn}
x2 y2
P2|2 Output alphabet Y = {y1, y2, …, ym}
: Pj|i = PY|X(yj|xi)
:
:
:
: In general:
:
xn calculating capacity needs more
Pm|n theory
ym
The statistical behavior of the channel is completely defined by
the channel transition probabilities Pj|i = PY|X(yj|xi)
* clue:
I(X;Y)
is convex in the input probabilities
Pe
k/n
C
Converse: For a discrete memory less channel
channel
Xi Yi
n n n n
I ( X ; Y ) H (Y ) H (Yi | X i ) H (Yi ) H (Yi | X i ) I ( X i ;Yi ) nC
n n n
i 1 i 1 i 1 i 1
k = H(M) = I(M;Yn)+H(M|Yn)
Xn is a function of M Fano
1 – C n/k - 1/k Pe
I(Xn;Yn) + 1 + k Pe
nC + 1 + k Pe
Pe 1 – C/R - 1/nR
Hence: for large n, and R > C,
the probability of error Pe > 0
We used the data processing theorem
Cascading of Channels
I(X;Z)
X Y Z
I(X;Y) I(Y;Z)
Assume:
binary sequence P(0) = 1 – P(1) = 1-p
t is the # of 1‘s in the sequence
Then n , > 0
Weak law of large numbers
Probability ( |t/n –p| > ) 0
Consequence:
1
h
0.9
Note:
0.8 h(p) = h(1-p)
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
p
Capacity for Additive White Gaussian Noise
Noise
Input X Output Y
2noise 2x
12 log 2 ( ) bits / trans.
2
noise
2noise S / 2 W
Cap W log 2 ( ) bits / sec .
2noise
1 z 2 / 2 2z
p( z ) e ; H( Z ) 21 log 2 (2 e 2z ) bits
2 2z
Middleton type of burst channel model
0 0
1 1
Transition
probability P(0)
channel 1
channel 2
Select channel k …
with probability channel k has
Q(k) transition
probability p(k)
Fritzman model:
1-p
G1 … Gn B
Error probability 0 Error probability h
Interleaving: from bursty to random
bursty
„random error“
0 0 0 1 0 column wise
1 0 0 1 1
1 1 0 0 1
De-Interleaving: block
read in column 1 0 1 e 1
read out
0 1 e e 0
wise 0 0 e 1 0
input sequence 0
input sequence 1 delay of b elements
input sequence m-1 delay of (m-1)b elements
in
Example: b = 5, m = 3
out