降秩盲自适应FRESH滤波器的研究

Reduced-Rank Blind Adaptive FRESH Filtering

提出了基于最小均方误差准则(MMSE)的盲自适应FRESH滤波器;给出了基于互谱准则(CSP)、主分量分析(PCA)和多级维纳滤波器(MSWF)等方法的降秩处理,并对不同的方法进行了性能比较。计算机仿真结果表明,这种降秩盲FRESH滤波器能够有效地分离或提取谱重叠信号。

A blind adaptive frequency-shift （BA-FRESH） filter is proposed in the means of MMSE. The reduced-rank implementation of the BA-FRESH filter is then presented based on cross-spectral （CSP） metric, principal component analysis （PCA）, and multistage Wiener filter （MSWF）. The performance of the reduced-rank BA-FRESH filter is outlined by computer simulation. The reduced-rank BA-FRESH filter not only has a lower computational complexity, but is also more efficient in signal extraction as compared to the conventional BA-FRESH filter.