Low-complexity adaptive successive cancellation list polar decoder based on relaxed sorting

Due to its advantage of performance over other algorithms, successive cancellation list (SCL) decoding has become one of the most favorable algorithms for polar codes. However, it still suffers a lot from the linear increasing complexity with list size l. In this paper, an adaptive SCL polar decoder based relaxed sorting (RS) approach is proposed, which successfully reduces the sorting complexity by simplifying the list size with proposed metrics. First, we partition the candidate list simply by the mean of metrics, which certainly results in performance degradation compared to the conventional SCL decoder. Then a scaling factor is introduced to narrow the performance gap. The scaling factor is determined on the proposed Trade-off Metric between Performance and Complexity (TMPC) via numerical simulations. Comparison results have shown that the proposed adaptive SCL decoder can achieve similar decoding performance as the conventional SCL decoder with at least 10% reduction on the average list size and overall computational complexity.