系统工程与电子技术

• 通信与网络 • 上一篇    下一篇

薛睿, 魏强, 徐锡超   

  1. 哈尔滨工程大学信息与通信工程学院, 黑龙江 哈尔滨 150001
  • 出版日期:2015-01-13 发布日期:2010-01-03

Dynamic iteration stopping algorithm for Q-ary LDPC coded CPM

XUE Rui, WEI Qiang, XU Xi-chao   

  1. College of Information and Communication Engineering, Harbin Engineering University, Harbin 150001,China
  • Online:2015-01-13 Published:2010-01-03

摘要:

针对正反馈存在于短帧多元低密度奇偶校验码(low density parity check, LDPC)与高阶连续相位调制(continuous phase modulation, CPM)迭代系统中的问题,提出一种在低信噪比采用加权联合平均外信息交换、在中高信噪比采用加权外信息交换的外信息处理方式。该方式对解调器与译码器之间传递的外信息进行相应处理,可以有效控制波动过大的位概率。同时将交叉熵(cross entropy, CE)和辅助硬判决(hard decision aided, HDA)两种迭代停止准则与加权外信息处理方式结合进行动态迭代译码。理论分析和仿真结果表明,对外信息交换进行加权操作并联合动态迭代停止算法能够在有效减少正反馈的同时使迭代的收敛性得到优化,同时迭代停止算法可减少平均迭代次数,提高系统的实时性。

Abstract:

In order to decrease the positive feedback in the iterative detection of the short frame Q-ary low density parity check (LDPC) code and the high-order continuous phase modulation (CPM) system, the combination average and weighted extrinsic information exchange for the low SNR, and the weighted extrinsic information exchange for the medium and high signal-to-noise ratio are proposed. In this information exchange, the extrinsic information between the demodulator and the decoder is correspondingly processed, which can control excessive fluctuations bit probability. Meanwhile, it performs decoding with two stopping criteria of dynamic iterative decoding: cross entropy (CE) and hard decision aided (HDA). Theoretical analysis and simulation results show that the combination of weighted extrinsic information exchange and the dynamic iterative stopping algorithm can not only resist the positive feedback effectively, and improve the convergence of iterative detection and the system bit error rate performance, but also reduce the average number of iterations, and keep the system in real-time.