An Improved Simplified Soft Cancellation Decoding Algorithm for Polar Codes Based on Frozen Bit Check
Mingyu Cai, Shufeng Li, Zhicheng Liu · 2021 IEEE 21st International Conference on Communication Technology (ICCT) · 2021
Polar codes can approach the Shannon limit with arbitrary binary-input discrete memoryless channels. As one of the most advanced soft-output algorithms, soft cancellation decoding (SCAN) has a better decoding performance than the hard-output algorithm, such as conventional successive-cancellation (SC) decoding. However, it suffers a relative high-complexity and low-throughput due to iterative computation. In this paper, we propose an improved simplified SCAN (ISSCAN) decoding based on frozen bit check (FBC) to further reduce decoding complexity. Two new special nodes are proposed to simplify the traversal of decoding tree. Then the corresponding message passing rules are also derived. Moreover, the FBC method is extended to SCAN decoding to simplify the message passing for ordinary nodes. The soft information can also be updated directly for nodes satisfying the FBC condition. Therefore, based on proposed method, we can obtain an excellent low-complexity SCAN decoding. Simulation results show that proposed ISSCAN decoding can save computational complexity significantly with negligible error performance loss when comparing with traditional SCAN decoding.