Adaptive Reduced-Rank Processing Using a Projection Operator Based on Joint Iterative Optimization of Adaptive Filters For CDMA Interference Suppression

摘要

This paper proposes a novel adaptive reduced-rank filtering scheme based onthe joint iterative optimization of adaptive filters. The proposed schemeconsists of a joint iterative optimization of a bank of full-rank adaptivefilters that constitutes the projection matrix and an adaptive reduced-rankfilter that operates at the output of the bank of filters. We describe minimummean-squared error (MMSE) expressions for the design of the projection matrixand the reduced-rank filter and simple least-mean squares (LMS) adaptivealgorithms for its computationally efficient implementation. Simulation resultsfor a CDMA interference suppression application reveals that the proposedscheme significantly outperforms the state-of-the-art reduced-rank schemes,while requiring a significantly lower computational complexity.