一种基于频移滤波器的混合信号盲恢复算法  

A Signal Blind Recovery Algorithm for Mixture Signals Based on FRESH Filter

作　　者：徐彬[1] 芮国胜[1] 陈必然 

机构地区：[1]海军航空工程学院电子信息工程系,山东烟台264001 [2]海军驻成都地区军事代表室,成都610036

出　　处：《电讯技术》2011年第11期31-36,共6页Telecommunication Engineering

基　　金：泰山学者建设专项基金~~

摘　　要：针对单天线接收的频谱混叠的混合信号盲恢复问题,在频移滤波器结构上,提出了一种基于相关函数误差准则的自适应频移滤波信号盲恢复算法。该算法利用滤波器输出信号和参考信号以及混合信号与参考信号之间的相关函数误差来调整自适应滤波器输出权值。分析了该算法的稳态性能。仿真结果表明:在信噪比大于-5 dB的条件下,该算法对混合信号分量的恢复相似度达80%以上,并且该算法仅需已知混合信号分量的循环频率,无需其它先验信息,具有计算量小、敛速度快的优点。This paper studies the bhnd signal recovery problem for frequency-overlapped mixture signals received by single antenna. A signal blind recovery algorithm is proposed based on FRESH （Frequency - shift） filter structure by using correlation function error between estimation signal and reference signal, reference signal and mixing signal as a criterion to update the weights of adaptive filters and its stable property is also discussed. Simulation results show that the recovery similarity of mixture components is higher than 80% when SNR（Signal- to - Noise Ratio） is greater than - 5 dB. Not any prior information but only cyclic frequencies of mixture components are needed for the algorithm, so it is featured by fast convergence speed and less computing complexity.

关 键 词：单天线混合信号 频移滤波器 盲恢复 相关函数 自适应算法 

分 类 号：TN911.7[电子电信—通信与信息系统]