低复杂度和低译码时延TPC迭代译码算法

现有的TPC串行迭代译码结构复杂度相对较高，译码时延较大，而低译码延时的Argon并行迭代译码结构则与串行结构相比有一定的性能损失。针对这些问题，本文提出了一种并行改进迭代译码结构。使用该改进并行迭代译码结构能够达到和串行结构相同的译码性能，并且译码时延降低为串行结构的一半。为了进一步降低译码复杂度和译码时延，在低可信度码元的搜索，候选码字欧氏距离的计算以及似然码字和竞争码字的搜索方面进一步作了优化。其中在欧氏距离的计算中采用格雷编码的测试图样，较大的减少了译码复杂度。最后完成了TPC译码并行改进结构的硬件实现，实测表明4次迭代的TPC译码器可以达到28Mbps的译码速度。

A Low Complexity and Low Latency TPC Iterative Decoding Algorithm

The TPC serial iterative decoding algorithm in use has the problems of relatively high complexity and large decoding delay. The decoding delay of Argon parallel iterative decoder is lower, but it has some performance losses. In order to solve these problems, an improved parallel iterative decoding structure of TPC is designed. The TPC decoder using the new parallel structure has the same performance with the serial structure and the latency reduces to half of the serial structure. In the application, the optimum method of the search of least reliable bit positions is proposed. The calculation of candidate codes Euclidean distances and the search of ML codeword and competing code word give low computation and decoding latency. The 4 iterations TPC decoder reaches the decoding speed of 28Mbps in application.