基于改进粒子滤波的PSK信号时延和码元联合估计算法  

Improved particle filtering based algorithm for time delay and symbols joint estimation of PSK signals

作　　者：夏楠[1] 邱天爽[1] 

机构地区：[1]大连理工大学电子信息与电气工程学部,辽宁大连116024

出　　处：《通信学报》2012年第4期129-135,共7页Journal on Communications

基　　金：国家自然科学基金资助项目(60872122)~~

摘　　要：提出了一种基于自适应重采样的粒子滤波算法用于对PSK信号的时间延迟进行估计,可以消除由于状态噪声方差设置过小而产生不准确的后验概率分布和设置过大引起的估计误差增大的问题。同时,考虑已有算法无法实现较小时间延迟准确估计的问题,提出了一种码元正向与反向检测相结合的算法,可实现一个码元周期内任意时间延迟的准确估计。另外,对载频偏差进行精确估计并补偿。仿真结果表明这种新方法与原算法相比能够实现更精确的时间延迟估计与更低码元检测误码率。An adaptive particle filtering algorithm based on resampling was proposed for time delay estimation of PSK signals,which could eliminate the influence of inaccurate posterior possibility distribution caused by the small noise variance and the influence of the increase of estimation error caused by the large noise variance.Taking into account the small time delay estimation being invalid in the existing algorithms,a symbol forward and backward detection technique was applied to achieve random time delay accurate estimation in a symbol period.Otherwise,the carrier frequency offset of the received signal was estimated and compensated.Simulation results verify the novel algorithm can achieve more accurate time delay estimation and lower symbol error rate than the existing algorithm.

关 键 词：粒子滤波 自适应重采样 时延估计 码元正向和反向检测 

分 类 号：TN911.72[电子电信—通信与信息系统]