ERROR-BOUNDS FOR PARALLEL DISTRIBUTED DETECTION UNDER THE NEYMAN-PEARSON CRITERION

The Neyman-Pearson performance of a distributed detection system is co nsidered wherein n independent and identically distributed observation s are quantized locally into M-ary messages and transmitted to a fusio n center. Under fairly general assumptions, it is shown that the type II error probability achieved by the best identical-quantizer system i s at most a fixed (in n) multiple of that achieved by the absolutely o ptimal system.