Complexity Reduction of Reliability Based Decoding of Linear Block Codes

Abstract

The Reliability Based Algorithm for linear block codes is investigated. The decoding complexity is increased when the length of information bits of linear block codes is increased. A Simplified statistical approach to evaluate the error performance bound of Reliability Based Algorithm of Linear Block Codes is investigated. First, one novel statistic is proposed which depicts the number of error contain in the ordered received noisy codewords. Also, another new statistic is proposed which compare the hamming distance between the permuted received word and the reprocessing codeword only on the parity check section. Then, the Probability Density Function (PDF) and the Cumulative Distribution Function (CDF) for each statistics are derived. Also, the new Probabilistic Threshold Test in combination with basic order-I reprocessing is proposed which compare the decision statistic evaluated only on the parity check section for all reprocessing codeword corresponding to TEP of any permissible order. Finally, with proposed approach, reduction of the number of Test Error Patterns (TEPs) is obtained to calculate the required codeword which reduces the decoding complexity.