一种改进的短极化码球形译码算法

Improved Sphere Decoding Algorithm for Short Polar Codes

球形译码(Sphere Decoding,SD)能够有效地实现极化码最大似然译码的误码性能,文章提出基于比特翻转的多重球形译码树搜索(Multiple Sphere Decoding Tree Searches Based on Bit-Flipping,BF-MSDTS)算法进一步提升短极化码的性能,该算法通过扩大半径搜索和翻转不可靠比特提升误码性能,同时利用搜索终止准则和固定可靠比特降低复杂度,有效地解决了多重球形译码树搜索(Multiple Sphere Decoding Tree Searches,MSDTS)算法低信噪比下误码性能不如循环冗余校验(Cyclic Redundancy Check,CRC)辅助的串行消除列表(CRC Aided Successive Cancellation List,CA-SCL)算法的问题.仿真结果表明,在低信噪比时,当误码率为10-4时,与CA-SCL算法比较,提高了0.41dB,与MSDTS算法比较,提高了1.17dB.提出的BF-MSDTS算法的误码性能在较差的信道环境下优于现有的MSDTS算法以及CA-SCL算法.

Sphere decoding(SD)can effectively achieve the block error rate performance of maximum likelihood decoding of polar codes.In this paper,multiple sphere decoding tree searches based on bit-flipping(BF-MSDTS)algorithm is proposed to further im-prove the performance of short polar codes.The algorithm improves block error rate performance by expanding radius search and flip-ping unreliable bits,while reducing complexity by using search termination criteria and fixed reliable bits.It effectively solves the prob-lem that the block error rate performance of the multiple sphere decoding tree searches(MSDTS)algorithm is worse than the Cyclic re-dundancy check(CRC)aided successive cancellation list(CA-SCL)algorithm when signal-to-noise ratio is low.The simulation results show that when the signal-to-noise ratio is low and the block error rate is 10^(-4),compared with the CA-SCL algorithm,the 0.41dB is improved,and compared with the MSDTS algorithm,the 1.17dB is improved.The block error rate performance of the proposed BF-MSDTS algorithm is better than the existing MSDTS algorithm and CA-SCL algorithm in poor channel environment.