Distributed Detection for Macrodiversity Enhanced Cellular Communication Systems

This paper considers the design of soft decision distributed detection schemes for achieving macrodiversity in uplink cellular communication systems with and without channel knowledge at the fusion center. Distributed detection macrodiversity (DDM) techniques are proposed where the quatizer thresholds of local detectors at base stations (radio access points) are optimized using a numerical global optimization algorithm called adaptive simulated annealing (ASA). The performance of such DDM schemes is studied for systems with and without channel coding, where the base stations employ multiple receive antennas over quasi-static Rayleigh fading channels. It is shown that properly designed quantizers for the local detectors, with a few resolution bits can make the performance very close to optimal macrodiversity that requires a much larger bandwidth from the connection links to the fusion center.