A low-complexity successive cancellation list decoding for polar codes with variable thresholds
摘要
The successive cancellation list (SCL) decoding algorithm has attracted significant interest in polar code due to its superior performance by retaining multiple decoding paths with increasing the complexity of decoding. To reduce the decoding complexity, we propose two low-complexity SCL decoding algorithms, including the codewords grouping-based threshold modified sequential probability ratio test (CG-TMSPRT) algorithm and the bit-based threshold modified sequential probability ratio test (B-TMSPRT) algorithm. Both algorithms trim some decoding paths by the proposed splitting rules, which consider adaptive double thresholds for the SCL decoding procedure. Based on the SCL decoding procedure, the CG-TMSPRT algorithm considers double thresholds, which vary with the group of code length, while the B-TMSPRT algorithm changes double thresholds with decoding bit according to Gaussian approximation. The simulation results show that both proposed algorithms significantly reduce the decoding complexity while keeping good decoding performance. Compared with the CG-TMSPRT algorithm, the B-TMSPRT algorithm has better decoding performance and higher computational complexity.