Maximal-Ratio Eigen-Combining for Smarter Antenna Arrays

In typical mobile wireless scenarios, signals are received with power azimuth angle spectrum (p.a.s.) of variable azimuth angle spread (AS). Therefore, conventional maximum average signal-to-noise ratio beamforming (BF) or maximal-ratio combining (MRC) may not necessarily be effective in terms of performance or signal processing complexity. A newer, more flexible, approach, called maximal-ratio eigen-combining (MREC) is analyzed and found to generalize BF and MRC. For imperfectly-known channels we study both suboptimal and optimal eigen-/combining. MREC-based analysis is shown to simplify MRC performance investigations for correlated channel gains. For MPSK signals, we present average error probability (AEP) expressions for MREC, BF, and MRC, that are new or generalizations of our previous work. Furthermore, we propose a performance-complexity tradeoff criterion (PCTC) for MREC-receiver adaptation to changing AS. Numerical evaluations for typical urban scenarios with realistic Laplacian p.a.s. of random AS demonstrate that PCTC-based MREC is an interesting alternative to BF and MRC, for smart antenna arrays