You are on page 1of 56

# USING THE MATLAB

COMMUNICATIONS TOOLBOX
TO LOOK AT CYCLIC CODING
Wm. Hugh Blanton
East Tennessee State University
http://faculty.etsu.edu/blanton

The Theory
General Concepts

Channel Coding

Designed to improve communications
performance by enabling the
transmitted signals to better withstand
the effects of various channel
impairments, such as noise,

Hamming Distance
The number of bit positions in which two binary words
differ.

110

010

011

This figure has a minimum
distance between code words of
1.

111

000

100

001
101

Hamming Distance of 2
A code word that differs by
a Hamming distance of 2
will detect an erroneous
code word.

0011

0001

0111

r=1

0000

0110

If the minimum distance is 3. 1 1 The error is corrected by assigning the received code word to the nearest (in Hamming distance) valid code word. 1 . the code words are separated by at least three edges. Note that if a single error occurs. the erroneous code word will be closer to one of the correct code words.

it is impossible to detect an error. but not corrected.The important concepts associated with Hamming distance are: 1. the error can be detected. When the Hamming distance is 3. When the Hamming distance is 1. . 2. a single error can be detected and corrected. 3. much less correct the error. When the Hamming distance is 2.

More Theory Mathematical Concepts .

.LINEAR BLOCK CODES  Definition – A systematic (n.k) linear block code is a mapping from a k-dimensional message vector to an n-dimensional codeword in such a way that part of the sequence generated coincides with the k message digits.

where n = 7. k-dimensional information vector. The linear block code that everyone uses is the (7.n-dimensional information vector.4) code. where k = 4. Information Code 0000 0000000 0001 1010001 0010 1110010 0011 0100011 0100 0110100 0101 1100101 0110 1000110 0111 0010111 1000 1101000 1001 0111001 1010 0011010 1011 1001011 1100 1011100 1101 0001101 1110 0101110 1111 1111111 .

...k.   0 0 . 0 . . the difference  p11 p   21    p n1 p12 p 21 .4) code? – Of course with a Generator Matrix! G  P I k  Note that m = n ... 1 ... 0 0 1 .....Generator Matrix  How do you generate a (7...... p1m . p nm 1 0 . p 2 m pn2 .

4) code is generated by the matrix 1 0 G 1  1 1 0 1 0 0 0  1 1 0 1 0 0 1 1 0 0 1 0  0 1 0 0 0 1 . A (7.

v  u[G ]  The (7. 2. 3. u is a [1 × k] vector representing the information word v is a [1 × n] vector representing the code word [G] is a [k × n] generating matrix .4) block code is generated using the matrix equation v  u[G] – where 1.

the code word is computed as follows: 1 0 v  u[G ]  0 0 0 1  1  Transformation 1 from 4-bit information word to 7-bit code word 1 0 1 0 0 0 1 1 0 1 0 0 1 1 0 0 1 0  0 1 0 0 0 1 v  1 0 1 0 0 0 1 Note that the additions are modulo-2 additions . For the information word (0 0 0 1).

– Thus. this code can detect double bit errors and correct single bit errors . The Hamming distance for the (7.4) code is 3.

.Parity Check Matrix  Another important matrix associated with block codes is the parity check matrix. [H].

p n 2   ..  .. 0   .. The parity check matrix is formed by starting with the identity matrix and appending the transpose of the nonidentity portion of [G]:  H  I k PT  1 0 .. p nm  .. ...... 1 p11 p 21 p12 p 22 p1m p2m p n1  . 0 0 1 .  0 0 .......

 For the (7.4) block code. 1 0 0 1 0 1 1    H  0 1 0 1 1 1 0  0 0 1 0 1 1 1 .

any errorless. or syndrome. – If the received code word contains an error.The Syndrome  The parity check matrix has the property v[H] T = 0  That is. . the resulting vector will match the corresponding bit that caused the error. yields a zero vector. received code word multiplied by the transpose of the parity check matrix. [H].

 Consider one of the valid code words of the (7. vH  T 1 0  0   1 1 0 0 1 0 1 1 0  1 1  0 0  1 0 0 1  1 0 1 1  1 1 0 1 .4) parity check matrix.4) block code. v = 1100101 and the transpose of the (7.

1  0 0  1 1 0 0 1 0 1 1 0  1  1  0 0  1 1 0 0 1 0 1 1 0  1  1 0 0  1 0 0 1  1 0  1  0  0  0  0  0  1  0 1 1  Modulo-2 Addition 1 1 000  0 10 1 0 1  0 11 1 0 11 0 0 1  1 0  0  1  0  0  1  0  0  0 1 1  1 1  0 1 .

1  0 0  1 1 0 0 1 0 1 1 0  1  1 vH  T 0 0  1 0 0 1  1 0  0  0  0  0  1  0  1  0 1 1  1 1  0 1 1  0 0   1 1 0 0 1 0 1 1 0  1  1 0 0  Syndrome 1 0 0 1  1 0  0 0 0 1 1  The null vector 1 1 indicates no error.  0 1 .

v = 1100101 gets corrupted to 1000101. 1  0 0  T vH   1 0 0 0 1 0 1 1 0  1  1 0 0  1 0 0 1  1 0 1 1  1 1  0 1 . Suppose the valid code words.

1  0 0  1 0 0 0 1 0 1 1 0  1  1  0 0  1 0 0 0 1 0 1 1 0  1  1 0 0  1 0 0 1  1 0  1  0  0  0  0  0  1  0 1 1  Modulo-2 Addition 1 1 000  0 10 1 0 1  0 11 1 0 11 0 0 1  1 0  0  0  0  0  1  0  0  1 1 1  1 1  0 1 .

1  0 0  1 0 0 0 1 0 1 1 0  1  The second row 1 corresponds to the second bit from left. vH  T 0 0  1 0 0 1  1 0  0  0  0  0  1  0  1  0 1 1  1 1  0 1 1  0 0   1 0 0 0 1 0 1 1 0  1  1 0 0  Syndrome 1 0 0 1  1 0  0 1 0 1 1 Match syndrome with  1 1 corresponding code in  0 1 parity check matrix .

the corresponding corrupted bit can be detected and and subsequently corrected. – By matching the non-zero syndrome with the contents contained in the rows of the transposed parity matrix. Each row in the transposed parity check matrix corresponds to a bit in the code word. .

. The process will convert the erroneous code 1000101 to the correct code 1100101.

Cyclic Codes .

the circuitry can become quite complex for very long blocks of code.CYCLIC CODES  The implementation of linear block codes requires circuits capable of performing matrix multiplication and of comparing the result of various binary numbers. . – Although integrated circuits have been developed to implement the most common codes.

the cyclic code can be electronically implemented relatively easily. . A special case of the block code.

A (7.4) cyclic code generating matrix 1 0 G 0  0 1 0 1 0 0 0  1 1 0 1 0 0 0 1 1 0 1 0  0 0 1 1 0 1 .

4) cyclic code generating matrix 1 0 G 0  0 1 0 1 0 0 0  1 1 0 1 0 0 0 1 1 0 1 0  0 0 1 1 0 1 Non-systematic—No Identity matrix.A (7. .

4) cyclic code generating matrix is made systematic by performing row. G sys 1 0  1  1 1 0 1 0 0 0  1 1 0 1 0 0 1 1 0 0 1 0  0 1 0 0 0 1 .A nonsystematic (7. column operations until you obtain the Ik partition.

MATLAB COMMUNICATIONS TOOLBOX DERIVATION .

MATLAB COMMUNICATIONS TOOLBOX DERIVATION  Although row. . many engineering technology students are not familiar with the process.column matrix manipulation is a straightforward process for students that have taken a linear algebra course.

 Although. the discussion of linear block codes is not mathematically complicated. it is mathematically tortuous with many mathematical turns. .

4) cyclic code is relatively benign with a manageable number of calculations. the discussion for the (7. Finally. .

 For larger codes. . the calculations become overwhelming. especially for the engineering technology student.

.  the MATLAB Communications Toolbox provides a set of cyclic code functions. In order to – better manage larger codes. – relieve the mathematical stress on engineering technology students. and – focus on the application of block codes rather than the mathematics.

1  x  x3  x 4  x5  x 7  x8 that corresponds to the code word 1 1 0 1 1 1 0 1 1 1 x1 x2 x3 x4 x5 x6 x7 x8 .7) cyclic code with the following generator polynomial.Consider the problem with a (15.

is: 1 0  0  G  0 0  0 0  1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 0  0 0 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0  0 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 . G.The nonsystematic cyclic code generating matrix.

And the systematic cyclic code generating matrix is G sys 1 1  1   0 0  1 1  1 0 1 1 1 0 1 1 0 0 0 0 0 0  0 1 1 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0  1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0  1 0 0 1 1 0 1 0 0 0 0 0 1 0 0 1 1 1 0 1 1 0 0 0 0 0 0 1 .

7. MATLAB yields a set of all generating polynomials >> p=cyclpoly(15. By using the function cyclpoly.'all') p= 1 0 0 0 1 1 1 1 0 1 1 1 0 1 1 0 0 1 1 0 0 1 0 1 1  x  x3  x 4  x5  x 7  x8 1 1 1 .

the nonsystematic generating matrix.genpoly. is derived as follows: >> genpoly=[1 1 0 1 1 1 0 1 1]. genmat. >>[parmat.genmat]=cyclgen(15.By using the function cyclgen.'nonsys') .

genmat = Columns 1 1 0 1 0 0 0 0 0 0 0 0 0 0 1 through 13 0 1 1 1 0 1 1 1 0 0 1 1 0 0 1 0 0 0 0 0 0 Columns 14 through 15 0 0 0 0 0 0 0 0 0 0 1 0 1 1 1 0 1 1 0 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 0 1 1 0 1 1 1 0 0 1 1 0 1 1 0 0 0 1 1 0 1 0 0 0 0 1 1 0 .

.which is 1  0 0  genmat  0 0  0  0 1 0 1 1 1 0 1 1 0 0 0 0 0 0  1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 0  0 0 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0  0 0 0 0 1 1 0 1 1 1 0 1 1 0  0 0 0 0 0 1 1 0 1 1 1 0 1 1 in matrix form.

1  0 0  genmat  0 0  0  0 1  0 0  G  0 0  0  0 1 0 1 1 1 0 1 1 0 0 0 0 0 0  1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 0  0 0 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0  0 0 0 0 1 1 0 1 1 1 0 1 1 0  0 0 0 0 0 1 1 0 1 1 1 0 1 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0  1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0 0 0  0 0 1 1 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 1 1 1 0 1 1 0 0  0 0 0 0 1 1 0 1 1 1 0 1 1 0  0 0 0 0 0 1 1 0 1 1 1 0 1 1 .

the systematic generating matrix. can be computed using cyclgen without the conditional parameter (nonsys): >> genpoly=[1 1 0 1 1 1 0 1 1].genmat]=cyclgen(15. genmatsys. >>[parmat.Likewise.genpoly) .

genmatsys = Columns 1 through 12 1 1 0 1 1 1 0 1 1 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 1 1 0 0 1 1 0 1 1 1 Columns 13 through 15 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 1 1 0 1 0 0 1 0 0 1 0 1 0 0 1 1 1 0 0 1 1 1 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 .

which is 1  1 1  genmatsys 0 0  1  1 1 0 1 1 1 0 1 1 0 0 0 0 0 0  0 1 1 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0  1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 0 0  1 0 0 1 1 0 1 0 0 0 0 0 1 0  0 1 1 1 0 1 1 0 0 0 0 0 0 1 in matrix form. .

768 possibilities is a frightening task. picking the right 128 valid codes out of 32. .768 (215) possible code words of which 128 (27) are valid code words.7) code produces 32. – Needless to say. The (15.

 The Communications Toolbox provides an encode function that alleviates much of this mathematical anxiety. >> data=(0:127). >> c=de2bi(data.'cyclic'. >> valid_code=encode(c.7.15.'left-msb').genpoly) .

valid_code = 0 0 1 0 1 1 0 1 0 0 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 0 0 1 1 0 0 1 0 1 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 0 information code 0 1 0 1 0 .

Conclusion .

– The underlying result is that the electronic engineering technology student can apply more effort in understanding the theory rather than the mathematics. especially when considering more complex problems. The MATLAB Communications Toolbox derives the same generating matrices that are derived by mathematical manipulation. .

 By reducing the mathematical anxiety. . and the instructor can move from theoretical foundation of cyclic codes to the circuit implementation of cyclic codes. students can concentrate more attentively on the actual nuances of block coding.