系统工程与电子技术 ›› 2018, Vol. 40 ›› Issue (6): 1378-1384.doi: 10.3969/j.issn.1001-506X.2018.06.27

• 通信与网络 • 上一篇    下一篇

单通道混合信号的深度联合分离译码算法

付君, 彭华, 杨勇   

  1. 解放军信息工程大学信息系统工程学院,河南 郑州 450002
  • 出版日期:2018-05-25 发布日期:2018-06-07

Depth joint algorithm on separation and decoding of single-channel mixed signals

FU Jun, PENG Hua, YANG Yong   

  1. Institute Information System Engineering, PLA Information Engineering University, Zhengzhou 450002, China
  • Online:2018-05-25 Published:2018-06-07

摘要: 针对采用低密度奇偶校验(low density parity check, LDPC)码编码的单通道同频混合信号,提出一种深度联合分离译码算法。采用了Gibbs分离过程中的逐符号分离步骤和译码过程中的逐符号译码步骤之间交替更改变量节点似然软信息的方式,在初期译码步骤输出的软信息模值较小,便于分离步骤对译码步骤的结果及时进行纠正,降低整体误比特率,软信息达到译码门限后开始发挥译码作用并最终和分离步骤达到一致收敛。仿真结果表明,本文提出的深度联合算法能够有效避免传统迭代分离译码算法由于分离过程的误比特率超出软输入软输出(soft input soft output, SISO)译码器纠错能力范围导致算法无法收敛的现象,降低联合分离译码算法的门限并提升整体性能,对于8PSK调制的混合信号在误比特率为10-3时有1 dB的性能改善

Abstract: A depth joint algorithm on joint separation-decoding is proposed for low density parity check (LDPC) code single-channel co-frequency modulated signals. The log likehood ratio (LLR) of variable nodes is alternatively updated by the single symbol by symbol steps of decoding or separation, the LLR magnitude is small and it is easy and timely to rectify by separation steps. Decoding will work after the bit error rate below the threshold, and the algorithm convergences finally. Simulation results show that the convergence threshold is declined and the performance is enhanced by uniting a single step of separation and decoding while the traditional algorithm cannot be convergent, and good tolerances of the proposed algorithm towards noise are shown for 8PSK mixing signals when bit error rate is 10-3.