Hardware architecture for list successive cancellation polar decoder

Abstract

This paper aims at designing an efficient hardware architecture for list successive cancellation (SC) polar decoder. Previous literatures have shown that, compared to conventional SC decoder, list SC decoder has the ability to approach the performance of maximum likelihood (ML) decoder. However, the efficient implementation of list SC decoder has not been proposed yet. To tackle this issue, first we propose a sub-optimal version of list SC decoding. Then different selections of list size L are evaluated. By introducing the pre-computation technique, the hardware architecture for a list SC decoder with L = 2 is proposed. Comparison results have shown that, for a rate-½ (1024, 512) polar code, the proposed decoder can achieve near-optimal decoding performance with less hardware cost and latency than the decoder with conventional design approach. We believe that the design approach presented in this paper will facilitate practical applications of list SC polar decoder.