极化码中关键信息比特集合的非均匀奇偶校验级联方案

在中短码长条件下极化码信道极化不完全,在奇偶校验级联码的译码过程中容易发生错误传播影响译码算法性能.为了降低错误传播对奇偶校验级联性能的影响,设计了一种新型奇偶校验级联方法.该方法通过高斯估计选取部分关键易错信息比特进行非均匀分段校验,能够有效降低错误传播对奇偶校验性能的影响,同时与循环冗余校验级联选择正确路径,可以提升译码算法在大列表和高信噪比条件下的译码性能.仿真表明应用新型级联码相比于CA-SCL(Cyclic-redundancy-check Aided Successive Cancellation List)平均能提升0.1～0.15 dB译码性能.此外,新型级联码结合自适应算法,可以利用译码算法性能的提升使自适应算法在更小列表下译码成功,降低自适应算法在较低信噪比下6％～25％的译码复杂度.

Nonuniform parity-check concatenated Polar codes for key information bit set

Under the condition of short and medium code length,the polarization of the Polar code channel is incomplete,and error propagation is likely to occur in the decoding process of the parity-check concatenated code,which affects the performance of the decoding algorithm.In order to reduce the impact of error propagation on performance,a novel parity-check concatenation method is designed.This method uses Gaussian approximation(GA) to select some key error-prone information bits for non-uniform segmentation check,which can effectively reduce the impact of error propagation on parity-check performance,at the same time,it is concatenated with cyclic redundancy check(CRC) to select the correct path,which can improve the performance of decoding algorithm under the condition of large list and high signal-to-noise ratio(SNR).Simulation results show that compared with cyclic redundancy check aided successful cancellation list(CA-SCL) the decoding performance is improved by 0.1~0.15 dB on average by applying the proposed concatenated code.In addition,the new concatenated code combined with adaptive algorithm can make the adaptive algorithm decode successfully in a smaller list,and reduce the decoding complexity of the adaptive algorithm from 6% to 25% at low SNR.