Reed Solmen Code1

Chapter 7
Reed-
Reed -Solomon Code
Spring 2009
Ammar Abu-Hudrouss Islamic

University Gaza ١
Introduction
 A Reed Solomon code is a special case of a BCH code in which

the length of the code is one less than the size of the field over
which the symbols are defined.
 It consists of sequences of length q — 1 whose roots include 2t

consecutive powers of the primitive element of GF(q).
 The Fourier transform over GF(q) will contain 2t consecutive

zeros.
 Note that because both the roots and the symbols are specified
in GF(q), the generator polynomial will have only the specified
roots; (NO Conjugates)
Channel Coding Theory

Slide ٢
١
Introduction
 To construct the generator for a Reed Solomon code, and we

decide that the roots will be from i to i+2t-1, the generator
polynomial will be
    
g x   X   i X   i 1  X   i 2t 2 X   i  2t 1 
 Example:
Suppose we wish to construct a double-error correcting, length 7
RS code;
we first construct GF(8) using the primitive polynomial X3 + X + 1 as
shown. We decide to choose i = 0, placing the roots from 0 to 3.
The generator polynomial is
  
g x   X   0 X   1 X   2 X   3  
g x   X 4   2 X 3   5 X 2   5 X   6

Slide ٣
Time Domain Encoding
 The encoding of a Reed Solomon code can be done by a long

division method similar to that of Chapter 4
 Example
 For the Reed Solomon code above, encode the data sequence
111001111.
 The data maps to the symbols 5 0 5.
 Four zeros are appended, corresponding to the four parity

checks to be generated, and the divisor is the generator
sequence.

Slide ٤
٢
 The remainder is 6  5 0 1 so that

 the codeword is 5 0 5 6 5 0 1 . Expressed as a binary
sequence this is 11101111101111001010.

Slide ٥
The encoder circuit for a Reed Solomon code is shown
g1 g0
gn-k-1 g2
+ + + +
U(X)

Slide ٦
٣
For our example the encoder will be
2 5 5 6
+ + + +
U(X)

Slide ٧
Decoding Reed Solomon Code
 The frequency domain algebraic decoding method explained in

Chapter 6.
 Error value calculation can be done using the Forney algorithm.

 Where we found that
Az E z   f z z 2t  z  0  k  2t  1
 Where (z) is the connection polynomial, E (z) is syndrome in

frequency domain and (z) is t - 1 order evaluator polynomial

Slide ٨
٤
 We calculate the error evaluator polynomial and also '(z), the

formal derivative of the connection polynomial. This is found to
be
even t   t 1 z t 2   t 3 z t 4    1
odd t   t z t 1   t 2 z t 3    1
 In other words, get rid of the zero coefficient of A and then

set all the odd terms in the resulting series to zero.
 The error value in position m is now
em  z  / z 1i '  z 
 evaluated at z = m. The parameter i is the starting location of

the roots of the generator polynomial.
Slide ٩
 Example Consider the codeword 5 0 5 6 5 0 1 previously

generated for the double-error correcting (7, 3) RS code. We
create errors in positions 5 and 3, assuming that we receive
5 4 5 3 5 0 1 . The frequency domain syndrome of this
sequence is
S0   5   4   5   3   5   0   1   0
S1   5 . 6   4 . 5   5 . 4   3 . 3   5 . 2   0 .   1   1
S 2   5 . 5   4 . 3   5 . 1   3 . 6   5 . 4   0 . 2   1   0
S3   5 . 4   4 . 1   5 . 5   3 . 2   5 . 6   0 . 3   1  0

Slide ١٠
٥
 Now we form the key equation for which the solution is
 0  2   1 1   0  0
 1 2   0 1  0
 Which leads to
1   2 2  1
 And the connection polynomial is
 z    1 z 2   2 z  1  0
 The roots for the connection polynomial is 4 and 2

Slide ١١
 Having found two roots for a connection polynomial of degree 2

indicates successful error correction with errors located at
positions -4 and -2, i.e. positions 3 and 5. We now calculate the
error evaluator polynomial, taking the powers from 0 to t — 1 of
S(z)A(z).
z   S01  S1 0 z  S0  0
 z    4 z   0
 Therefore
 z   4z  0
em  
z' z  z  m  2z z   m
 Evaluating at m = 3 and 5

Slide ١٢
٦
 This leads to
 4 3   0
e3  2 3
4

 4 5   0
e5  5
 2 5
 The received symbol 3 at position 3 is therefore corrected to

6. The received symbol 4 at position 5 is corrected to 0. This
successfully completes the decoding.

Slide ١٣
Frequency domain encoding of RS Code
 As the Fourier transform of a Reed Solomon code word contains

n — k consecutive zeros, it is possible to encode by considering
the information to be a frequency domain vector, appending the
appropriate zeros and inverse transforming.
 Example
 We will choose again a (7, 3) double-error correcting RS code
over GF(23). Let the information be 2, 5, 0
 The frequency domain code word is 2500000, and the inverse
is 5056501
 We now create a two-symbol error, say 5 in position 5 and 4 in
position 3, as in our previous example. The received sequence is
5453501

Slide ١٤
٧
 The decoding proceed with finding the Fourier transform of the

sequence which gives 5230010 with syndrome being
0010
We now form the key equation with t = 2 and A0 = 1 :

 0  2   1 1   0  0
 1 2   0 1  0  0
The solution is
1   2 2  1

Slide ١٥
 We use the shift register shown to generate the code sequence

in the frequency domain
 In our example 0 = 1, 1 = 2, and 2 = . E3 = 0, E2 = .

 Cyclic the shift register generates 133, which are the
frequency domain error at locations 4,5 and 6.

Slide ١٦
٨
 The next two values generates 01 which are the syndrome
components S0 and S1. so the code will repeat
 The complete error sequence in the frequency domain is

3310010 .
 Adding this to the Fourier transform of the received sequence

which is 5230010 .
 The result is 2500000, which is the original codeword without

errors with four zeros syndrome

Slide ١٧
 Example 2
 For the previous example consider three errors have occurred,
and the received sequence is 6 4 5 3 5 0 1.
 The Fourier transform is
R z    3 z 6   5 z 5   6 z 4   3 z 3   2 z 2   3 z   3
 The key equations are found to be
 3  2   3 1   2  0
 3 2   21   3  0
 Or 0 = 1, 1 = 0, and 2 = 2.

Slide ١٨
٩
 The shift register become
 After initializing the shift register with 3 2 . The output is 2

6 5 followed by 5  which is not S0. this means that there is
an error which we cannot correct.

Slide ١٩
 Example 3
 For the previous example consider single errors have occurred,
and the received sequence is 5 4 5 6 5 0 1.
R z    6 z 6   3 z 5   5 z 4   6 z 3   1 z 2   3 z   5
 The key equations are found to be
 5  2   3 1   1  0
 3 2   11   6  0
 Or 0 = 1, 1 = 5, and 2 = 0.

Slide ٢٠
١٠
 The shift register become
 After initializing the shift register with 6 . The output is 4 2

0 followed by 5 which is the value of s0.
 The complete error sequence is 420 6 5 0 1 .
 Adding this to 5 4 5 6 5 0 1.The result is 2500000,

Slide ٢١
Erasure Decoding
 Reed-Solomon has the ability to recover event known as erasure
 Erasure is a symbol which is more likely to be in error
 Example
 Consider the same previous example but with one error a
position 5 and two erasures in positions 6 and 1
 The received sequence will be taken as 0 4 5 6 5 0 1.
R z    6 z 6   6 z 5   3 z 4   1 z 3   6 z 2   5 z   0

Slide ٢٢
١١
Erasure Decoding
 The lower terms of R forms the syndrome

 The erasure polynomial is
  
 z    6 z  1  1 z  1  z 2   5 z  1
 The error locator polynomial is
 z   1 z  1
 and the product is

   
 z z   1 z 3   0   5 1 z 2   5  1 z  1
 By convolution with the syndrome polynomial

 0 1   5   3 1   4   6 1   1  0

Slide ٢٣
Erasure Decoding
 Which means that 1 = 5. which leads to

 z  z    5 z 3   1 z 2  1
 The following circuit can be used for recursive extension

Slide ٢٤
١٢
Erasure Decoding
 Loading with values 5 6 1 and shifting gives the sequence 1 1

0 and then regenerating the syndrome 0 5 6. The decoding is
successful.
 The terms 0 1 1 are added to component 6 6 3 from the

Fourier transform of the received sequence yo give the
recovered information 2 5 0
 If we choose Forney algorithm to correct errors in time domain
z   S z z  z  mod z 4   5 z 2   5 z   0
 We need also the formal derivative of (z)(z) which is 5 z2

Slide ٢٥
Erasure Decoding
 Therefore at position i the error value is

 5z2   5z   0
ei  z  i
 5 z3
 Evaluating this at -6 and -1 which are the roots of (z) and -5
which is root of (z) and found by Chien search. The results are
e1   0 , e5   5 , e6   5
 The received sequence then is 5056501, corresponding to

the original codeword.

Slide ٢٦
١٣
Welch-Berlekamp Algorithm
 Consider the previous double error –correction examples we

transmit a codeword from a (7,3) RS code with roots 3,2 , 1 ,
and 0.
 To start with, we need to calculate some values needed as input

for the algorithm
n  k 1
gX 
 gx i i     
 X  0 X 1 X  2  X 3  5 X 2  6 X  3
j 0 X   nk 1
 Therefore g3 = 3 , g2= 3 , g1= 3, and g0 = 3
 For error value calculation, we compute
   
C   0 1  nk 1  0   1  0   2   0   nk 1 
Slide ٢٧
 For our example C = 6. we now need for each data location to
find the value hi = C /g(i) . The values are found to be
h6   6 /  4   2
h5   6 /  1   5
h6   6 /  0   6
 Assume now we receive 5453501,. The first step is to compute

the syndrome by long division.
 This found to be s3 = 0 , s2= 2 , s1= 0, and s0 = 6
 The input to WB algorithm is the set of point (Sj, j ) where

 Sj = sj/gj
Slide ٢٨
١٤
 The input points are therefore (0, 3),(4, 2),(1, 1),(3, 0).
 We need to find two polynomials Q(X) and N(X) for which
   
Q  j Sj  N  j for 0  j  n  k  1
 And the length L[Q(X) ,N(X)], defined as the maximum of

deg[Q(X)] and deg[N(X)]+1 has the minimum possible value

Slide ٢٩
 The steps of the algorithms are now

1. Set Q0(X ) = 1, N0(X ) = 0, W0(X ) = X, V0(X ) = 1 and d = 0.
2. Evaluate D1 = Qd(d )Sd + Nd(d )
3. If D1 = 0, set Wd+1 = Wd(X+ d ), Vd+1 = Vd(X+ d ) and go to step
6. otherwise, set D2 = Wd(d )Sd + Vd(d )
4. Set Qd+1 = Qd(X+ d ), Nd+1 = Nd(X+ d ) , Wd+1 = Wd + QdD2/D1,
Vd+1 = Vd + NdD2/D1,
5. Check whether L[Wd , Vd] was less than or equal to L[Qd , Nd];
if it was then swap Qd+1 , Nd+1 with Wd+1 , Vd+1
6. increment d
7. If d < n – k, return to step 1; otherwise, Q(X) = Qd(X), N(X)=
Nd(X).

Slide ٣٠
١٥
 For our example
 The error locator polynomial is X2 + 2 X+ 1 which has roots 3

and 5 . The error values in the data positions are
ek  hk
 
N k
 
Q'  k

Slide ٣١
 Where Q’(X) is the formal derivative of Q(X). In this case Q’(X)

= 2 .
 
e5   5  5   3 /  2   5
 The received information therefore is 5 0 5.

Slide ٣٢
١٦

Reed Solmen Code1

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Reed Solmen Code1

Uploaded by

Copyright:

Available Formats

Chapter 7

Ammar Abu-Hudrouss Islamic

 A Reed Solomon code is a special case of a BCH code in which

 It consists of sequences of length q — 1 whose roots include 2t

 The Fourier transform over GF(q) will contain 2t consecutive

Channel Coding Theory

 To construct the generator for a Reed Solomon code, and we

Channel Coding Theory

Time Domain Encoding

 The encoding of a Reed Solomon code can be done by a long

 The data maps to the symbols 5 0 5.

 Four zeros are appended, corresponding to the four parity

Channel Coding Theory

 The remainder is 6  5 0 1 so that

Channel Coding Theory

Time Domain Encoding

The encoder circuit for a Reed Solomon code is shown

Channel Coding Theory

For our example the encoder will be

Channel Coding Theory

Decoding Reed Solomon Code

 The frequency domain algebraic decoding method explained in

 Error value calculation can be done using the Forney algorithm.

Az E z   f z z 2t  z  0  k  2t  1

 Where (z) is the connection polynomial, E (z) is syndrome in

Channel Coding Theory

 We calculate the error evaluator polynomial and also '(z), the

 In other words, get rid of the zero coefficient of A and then

em  z  / z 1i '  z 

 evaluated at z = m. The parameter i is the starting location of

Decoding Reed Solomon Code

 Example Consider the codeword 5 0 5 6 5 0 1 previously

Channel Coding Theory

 Now we form the key equation for which the solution is

 The roots for the connection polynomial is 4 and 2

Channel Coding Theory

Decoding Reed Solomon Code

 Having found two roots for a connection polynomial of degree 2

Channel Coding Theory

 The received symbol 3 at position 3 is therefore corrected to

Channel Coding Theory

Frequency domain encoding of RS Code

 As the Fourier transform of a Reed Solomon code word contains

Channel Coding Theory

 The decoding proceed with finding the Fourier transform of the

We now form the key equation with t = 2 and A0 = 1 :

Channel Coding Theory

Frequency domain encoding of RS Code

 We use the shift register shown to generate the code sequence

 In our example 0 = 1, 1 = 2, and 2 = . E3 = 0, E2 = .

Channel Coding Theory

 The complete error sequence in the frequency domain is

 Adding this to the Fourier transform of the received sequence

 The result is 2500000, which is the original codeword without

Channel Coding Theory

Frequency domain encoding of RS Code

 The key equations are found to be

 Or 0 = 1, 1 = 0, and 2 = 2.

Channel Coding Theory

 The shift register become

 After initializing the shift register with 3 2 . The output is 2

Channel Coding Theory

Frequency domain encoding of RS Code