Performance of Different Polar Code Decoding Over NOMA Systems

In this paper, we investigate the performance of different polar decoding algorithms over an uplink non-orthogonal multiple access (NOMA) system. The polar code which has low complexity can be decoded in two ways: sequential decoding like successive cancellation (SC) and parallel decoding like belief propagation (BP). We further introduce their list versions: SC list (SCL), CRC-aided SC list (CA-SCL) and BP list (BPL) which have better error performance. Simulation results show that BP outperforms SC by 0.2 dB over Gaussian channel, and by 0.5 dB over Rayleigh fading channel. Meanwhile, SCL achieves performance gain of 0.2 dB over BPL in both channels. Moreover, latency analysis shows that BP can save 21.6% clock cycles (CCs) to SC and BPL can further achieve 47.7% CCs reduction compared to SCL and CA-SCL.