PERFORMANCE COMPARISON OF NON-INTERLEAVED BCH CODES AND INTERLEAVED BCH CODES

This project covers the research about the BCH error correcting codes and the
performance of interleaved and non-interleaved BCH codes. Both long and short
BCH codes for multimedia communication are examined in an A WGN channel.
Algorithm for simulating the BCH codes was also being investigated, which includes
generating the parity check matrix, generating the message code in Galois array
matrix, encoding the message blocks, modulation and decoding the message blocks.
Algorithm for interleaving that includes interleaving message, including burst errors
and deinterleaving message is combined with the BCH codes algorithm for
simulating the interleaved BCH codes. The performance and feasibility of the coding
structure are tested. The performance comparison between interleaved and noninterleaved
BCH codes is studied in terms of error performance, channel performance
and effect of data rates on the bit error rate (BER). The Berlekamp-Massey Algorithm
decoding scheme was implemented. Random integers are generated and encoded with
BCH encoder. Burst errors are added before the message is interleaved, then enter
modulation and channel simulation. Interleaved message is then compared with noninterleaved
message and the error statistics are compared. Initially, certain amount of
burst errors is used. "ft is found that the graph does not agree with the theoretical bit
error rate (BER) versus signal-to-noise ratio (SNR). When compared between each
BCH codeword (i.e. n = 31, n = 63 and n = 127), n = 31 shows the highest BER while
n = 127 shows the lowest BER. This happened because of the occurrence of error
bursts and also due to error frequency. A reduced size or errors from previous is used
in the algorithm. A graph similar to the theoretical BER vs SNR is obtained for both
interleaved and non-interleaved BCH codes. It is found that BER of non-interleaved
is higher than interleaved BCH codes as SNR increases. These observations show that
size of errors influence the effect of interleaving. Simulation time is also studied in
terms of block length. It is found that interleaved BCH codes consume longer
simulation time compared to non-interleaved BCH codes due to additional algorithm
for the interleaved BCH codes.