Outline
Channel Capacity and Coding
Burst error
Theorem Definition Example
Chapter 8: Digital Communication Systems
Vahid Meghdadi
University of Limoges
[email protected]
Reference : Digital communications by John Proakis;
Wireless communication by Andreas Goldsmith
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Channel Capacity and Coding
Burst error
Channel Capacity and Coding
Capacity of infinite bandwidth channel
Capacity of finite bandwidth channel
Burst error
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
Channel capacity and coding
I A BPSK transmission over AWGN gives the following error
probability !
r
2Eb
pb = Q
N0
I The capacity of an AWGN is C = B log2 (1 + P/N0 B).
I Capacity of infinite bandwidth channel
P 1 P
C ∞ = lim B log2 (1 + )= bits/sec
B→∞ P/N0 B ln 2 N0
Pτ
I Sending k information bits during τ seconds: Eb = k
I On the other hand, the information rate is Dt = k/τ .
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
Capacity of infinite bandwidth channel
I We can write Eb = P/Dt , therefore:
C∞ 1 Eb
=
Dt ln 2 N0
I If Dt < C ∞ , an error free transmission is possible.
C∞ Eb
>1⇒ > ln 2 = 0.69 = −1.6dB
Dt N0
I It means that for a signal to noise ration greater than -1.6 dB,
we can find a coding scheme for error free transmission.
I Note that for a hard detection, a 2 dB degradation is
Eb
achieved: N 0
> π2 ln 2 = 1.069 = 0.4dB.
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
Capacity versus BER
BER
10-1
Limit de Shannon (soft)
Uncoded BPSK
10-2
10-3 Hard
10-4
10-5
Eb/N0 (dB)
-2 2 4 6 8 10
Coding gain
The coding gain (for pe = 10−5 ) is about 11.2 dB.
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
Capacity of finite bandwidth channel
I C = B log2 (1 + P/N0 B).
I With BW=B the maximum symbol rate is rs = 2B.
I This symbol rate is often called Baud rate wich is not the
same as bit rate with is the rate of information bits.
information Coded Transmitted
bits bits symbols
Binary Channel Modulator
source k coding n (mapping) N
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
transmission rate
I k information bits are sent during τ seconds.
k
Dt =
τ
I It means, during the same time, n coded bits and N symbols
are transmitted.
N Dt rs
rs = ⇒ =
τ k N
I Defining the rate R = k/N, which is “code rate” times
“mapping rate”. The amount of the information bit per
second:
Dt = Rrs = 2BR
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
Example
Using a block code of rate (7,4) and a 16-QAM modulation and a
channel with BW=10 kHz:
4 4 16
rs = 20ksym/sec, R = ∗ =
7 1 7
16
Dt = 20000 ∗ = 45714bits/sec
7
Using a convolutional code CC(7,5) with QPSK modulation:
R = 1 ⇒ Dt = 20000 ∗ 1 = 20 kbits/sec
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Capacity of infinite bandwidth channel
Channel Capacity and Coding
Capacity of finite bandwidth channel
Burst error
Minimum SNR
Given a rate, what is the minimum signal to noise ratio required in
order to obtain an error free communication?
The transmission rate must be less than the channel capacity:
P
Dt < C ⇒ 2BR < B log2 1 +
N0 B
!
Eb Dt
= B log2 1 + Dt
N0 2R
2REb
= B log2 1 +
N0
This leads to:
Eb 22R − 1
>
N0 2R
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Channel Capacity and Coding
Burst error
Burst error
information Coded Transmitted
I Burst error occurs
bits in some communication
bits links. symbols
Binary Channel Modulator
I Decoders
source performance
k degrades ifnerrors(mapping)
coding occur in burst.
N
I An interleaver is used to break down a burst error to
individual errors distributed in time.
Channel Channel
Interleaver Channel De-interleaver
coding decoder
Vahid Meghdadi Chapter 8: Digital Communication Systems
Outline
Channel Capacity and Coding
Burst error
information Coded Transmitted
bits bits symbols
Concatenated code
Binary Channel Modulator
source k coding n (mapping) N
I Some times two level of coding is used to enhance the
correcting performance.
I A block code and a convolutional code give always burst
errors.
Channel
Interleaver De-interleaver
Channe
Channel
coding decode
I Therefore, the inner code produces burst error.
I Interleaver is required between inner and outer codes.
Outer Inner
Interleaver
coder Coder
Channel
Outer Inner
De-interleaver
decoder decoder
Vahid Meghdadi Chapter 8: Digital Communication Systems
