一种新的盲自适应频移滤波器

A novel adaptative frequency-shift filter

介绍了基于谱相关理论的频移滤波器,在此基础上提出了采用两种自适应算法的FRESH滤波器:LMS算法和CMA算法,并通过计算机仿真比较了两种算法的性能.结果表明,采用CMA算法的FRESH滤波器不需要训练序列,在性能上比较接近于传统的LMS算法,是一种实用的FRESH滤波器.

This paper introduces the frequency-shift filter which is based on the spectrum correlation theory. Based on this model, two adaptive FRESH filters are brought forward, one of which adopts the LMS algorithm and the other adopts the CMA algorithm. The simulation results demonstrate that CMA-FRESH filter which doesn't require training sequence has similar performance with conventional LMS-FRESH filter and is a kind of practical FRESH filter.