Detection performance limits of channel impaired distributed sensor networks

Traditional parallel distributed detection systems assume perfect channels between local sensors and a fusion center. However, for the emerging wireless sensor networks with inexpensive sensors and stringent power/delay constraints, error-free transmission is not guaranteed. In this paper, we study the performance degradation due to imperfect channels between the sensors and the fusion center under both asymptotic and non-asymptotic regimes. For asymptotic analysis, we compute the error exponents of the underlying hypothesis testing problem in the presence of discrete transmission channels; while for cases with a finite number of sensors, we determine channel conditions under which the distributed detection systems become useless observing the channel output might not help in reducing the error probability performance at the fusion center. Numerical examples are provided to illustrate the concepts.