改进型TPC译码算法研究

Research on modified TPC decoding algorithm

Turbo乘积码(Turbo Product Code,TPC)能够在接近Shannon极限的低信噪比情况下,获得较低的误码率。但是TPC译码算法的硬件电路复杂,虽然最小和译码算法能够简化TPC译码算法,但它是以牺牲译码性能为代价的。为了让译码算法在复杂度和译码性能之间取得较好的折衷,根据似然概率,通过同时引入偏移量参数和归一化参数,提出一种改进型TPC译码算法,并且利用最小均方误差准则来计算相关参数。仿真结果表明,与TPC译码算法、最小和译码算法相比,改进型TPC译码算法能够在降低算法复杂度的同时保持良好的译码性能。

TPC can obtain lower bit error rate at low SNR close to the Shannon limit.But the hardware circuit of TPC decoding algorithm is complicated.Minimum-sum decoding algorithm can simplify TPC decoding algorithm,but it is achieved by sacrificing decoding performance.According to likelihood probability,a modified TPC decoding algorithm is proposed by introducing offset parameter and normalization parameter simultaneously in order to have a good tradeoff between complexity and decoding performance.Moreover,the related parameters are calculated by minimum mean square error rule.The simulation results show that,compared with TPC decoding algorithm and minimum-sum decoding algorithm,the modified TPC decoding algorithm can decrease algorithm complexity and keep good decoding performance.