A Novel “Single-Path” vs. “Few-Path” Test Based on Higher Order Statistics to Possibly Start-Up Coherent Combining

Abstract

This paper focuses on an innovative hypothesis test for discrimination of wireless mobile channels based on higher order statistics to possibly start-up coherent combining. We have devised a new testing procedure, namely the Rakeness test, that statistically measures how much the series under investigation (amplitude samples of matched filter receiver's output) fits Rice vs. non-Rice models. This is equivalent to discriminate between the cases of a channel with one single dominant path (strong propagation), or with few dominant paths (weak propagation). Then, mathematical expressions for the bias and variance of the new testing variable are derived, by a 3-D reduced Taylor’s expansion up to the second order. The achieved results, obtained throughout theory and simulations, evidence the robustness of this innovative test. Our test can hence be used as a preliminary signal processing method to decide if simpler choices (i.e. matched filters) or cumbersome coherent combining strategies (i.e. Rake receivers) can be effectively implemented at the receiver’s side.