实现单通道MPSK信号盲分离的MCMC新算法  

Blind Separation Algorithm of Single Channel MPSK Signals based on MCMC

作　　者：汤俊杰[1] 李辉[1] 戴旭初[1] 

机构地区：[1]中国科学技术大学电子工程与信息科学系,安徽合肥230027

出　　处：《信号处理》2014年第11期1321-1328,共8页Journal of Signal Processing

基　　金：安徽省高校自然科学研究项目(KJ2011A275);国家自然科学基金资助项目(61471335);国家科技重大专项基金资助项目(2013ZX03003007)

摘　　要：本文根据单通道接收两路MPSK混合信号在过采样下的基本模型,针对粒子滤波算法在单通道信号盲分离中的性能瓶颈以及高复杂度问题,提出了基于MCMC方法的新算法。该算法对接收信号进行过采样处理,能够利用更多的波形信息,从而有效抑制噪声的影响。新算法利用Gibbs采样估计MPSK调制符号的后验概率,近似实现了贝叶斯最优估计,并利用最小二乘法实现参数的迭代估计。理论分析与仿真实验表明,相对粒子滤波算法,本文提出的新算法在误码率性能以及复杂度方面具有良好的表现。In order to cope with the performance bottleneck and the time complexity of particle filters algorithm for the blind separation of two MPSK mixture signals obtained by a single channel receiver,a novel algorithm based on Markov Chain Monte Carlo（MCMC） method is established which is based on the over-sampling signal model. By using over-sampling of the received signal,more information of waveform is utilized,so noise suppression is more effective. In the new algorithm the Gibbs sampling method is used to estimate the posterior probabilities of the symbols modulated in MPSK,so that the optimized Bayesian estimation can be obtained approximately. Furthermore,the unknown parameters in the signal model are iteratively estimated by the least Squares（LS） method. Theoretical analysis and the simulation results show the good performance and low complexity of the new algorithm.

关 键 词：信号处理 单通道盲分离 贝叶斯估计 马尔可夫链蒙特卡罗 

分 类 号：TN911.72[电子电信—通信与信息系统]