Performance of Self-Corrected APP-Based Decoding Algorithm for Decoders with Quantized Input
Aleksei Kharin, Igor Volkov, Aleksei Ovinnikov, Evgeny A. Likhobabin, Vladimir V. Vityazev · 2018
In this paper we explore the impact of self-correction (SC) modification on performance of a posteriori probability (APP) based decoding algorithm for LDPC codes. Influence of self-correction technique was investigated for decoders with 3, 2 and 1 bit input quantization. It was shown that performance of decoders with quantized input can be significantly improved in comparison with regular APP-based decoding algorithm, and in some cases with Min-Sum algorithm. Convergence of algorithm is almost the same as for Min-Sum decoding. Performance and convergence results are presented for two CCSDS (8176, 7156) and IEEE 802.3an (2048, 1723) LDPC codes for APP-based, SC APP-based and Min-Sum decoding algorithms.