稳定分布噪声下基于粒子滤波的多径时变信道盲均衡算法

Particle filter-based multi-path time-varying channel blind equalization in stable noise

提出了一种基于粒子滤波的多径时变信道盲均衡算法,并在此基础上进行扩展,提出了一种基于延迟抽样的盲均衡算法。新算法的贡献可总结为:推导出对称α稳定分布(SαS)噪声下对传输码元进行最大后验估计的盲贯序算法;对SαS分布噪声进行高斯近似并递推出信道及噪声未知参数的联合后验分布。仿真结果表明,所提出的算法是有效的,特别是在较强脉冲噪声情况下要优于其他算法。

A particle filtering （PF） based blind equalization algorithm for the multi-path time-varying channel was pre- sented and a delay sampling blind equalization algorithm was proposed. The contribution of the novel algorithm can be summarized as follows： the blind sequential algorithm was derived which performs the maximum a posteriori （MAP） symbol detection in symmetric-alpha-stable （SaS） distribution noise; and the joint posterior distribution of the Gaussian approximation for SaS distribution noise and the joint posterior distribution of the unknown channel and noise parameters were derived and presented. The simulation results demonstrate that the proposed method is valid and outperforms the existing algorithms, especially in the case of strong impulsive noise.