Distributed Detection Over Multiple Access Channels

We address the design of binary local sensor quantizers for decentralized detection over multiple access channels. Our goal is to minimize the error probability at the fusion center using a single snapshot of local observations. Considering both the synchronized and asynchronous transmissions among sensors, we establish the optimality of a likelihood ratio test for both cases. For the case of asynchronous transmissions, to compensate for the unknown fading channel parameters and transmission delays, we propose a structure consisting of a RAKE receiver and a square-law detector. Simulations results are presented to demonstrate effectiveness of the design procedure.