Spatial processing for frequency diversity schemes

Abstract

A novel technique to obtain optimum blind spatial processing for frequency diversity spread spectrum (FDSS) communication systems is introduced. The sufficient statistics for a linear combiner, which prove ineffective due to the interferers frequency characteristics, are modified to yield improved detection under partial jamming in the spectral domain. Robustness to partial time jamming is achieved by extending the notion of replicas over the frequency axis to a repetition over the time variable. Analysis and simulations are provided, showing the advantages of using FDSS with spatial diversity to combat the interference when it is confined to a narrow frequency band or short time interval relative to the desired signal extent in either domain.