Improving performance of multithreshold decoder for self-orthogonal codes

Abstract

The article discusses self-orthogonal error-correcting codes (SOC) for the decoding of which multithreshold algorithms (MTD) are usually used. To decode SOC the algorithms used for low-density parity-check (LDPC) codes can also be applied. The article shows that using a min-sum decoder for SOC over a channel with additive white Gaussian noise (AWGN) in case of binary phase shift keying allows to receive additional coding gain about 1..1,5 dB in comparison with MTD usage. At the same time computing complexity in a min-sum algorithm turns out to be 6...7 times higher than MTD. For SOC decoding the work offers a combined decoder including the elements of MTD and min-sum algorithms. The first several decoding iterations require the usage of min-sum decoder while later MTD is added. The results of offered decoder simulation show about 1 dB increase of coding gain in comparison with MTD for SOC over a channel with AWGN with binary phase shift keying in case of twofold increase of computing complexity. The gain received depends on the SOC used, the number of min-sum decoding iterations and MTD.