Turbo乘积码构造方法的研究

A Study on the Construction of Turbo Product Code

Turbo乘积码因其具有接近香农限的译码性能和适合高速译码的并行结构,已成为纠错编码领域的研究热点。Turbo乘积码的分量码一般由扩展汉明码构造而成,所以该类码字编码和译码的硬件实现比较简单。当Turbo乘积码采用扩展汉明码作为子码时,随着信噪比的提高,码字的最小码重对误帧率的影响会逐步增大。文中改进了Turbo乘积码编码结构,在只增加较小的编译码复杂度和时延的情况下,提高了码字的最小码重,并减少了最小码重码字在码字空间所占的比例。通过仿真和分析,比较了这种码和TPC码在误帧率性能、码字的最小码重分布以及最小码间距估计上的差异。

Turbo Product Codes （TPC） are now one of intensely studied areas of error correction coding due to their close to Shannon limit performance and parallel decoding architecture. The minimum Hamming weight of TPC is increased and the rate of minimum Hamming weight codes in code space is decreased. Meanwhile it only increases a little of decoding complexity and delay. Performance comparisons between our codes and TPC on BLER, the distributions of minimum Hamming weight of codes and minimum distance estimation are carried out by simulation and analysis.