Performance of Modified Iterative Decoding Algorithm for Product codes

Abstract

In this paper, a Modified Iterative Decoding Algorithm (MIDA) is investigated for decoding of Product codes. Product codes provide a good code rate as well as a good BER performance. Moreover, they have a brute power to correct burst errors, which are very harmful in digital transmission. It is because product codes are correct row-wise as well as column-wise. Conventional iterative decoding algorithm initially proposed by Al-Askary based upon Maximum Likelihood (ML) scheme, has a huge complexity which is unaffordable for long codes used in real time transmissions. MIDA being a sub-optimum algorithm reduces its complexity in an efficient way by utilizing the basic idea of Syndrome Decoding of Linear Block Codes. So decoding becomes more efficient also the complexity reduction factor increases with each iteration. The worst case complexity of MIDA is closed to that of conventional algorithm but in average case there is huge reduction in complexity, with negligible performance degradation.