Large Deviations Analysis to the Performance of Distributed Detection

The book delves into the performance of distributed detection systems using large deviation techniques across two models. The first model examines error performance as the number of sensors increases, highlighting the effects of quantized sensor data on binary hypothesis testing. Key findings reveal relationships between Neyman-Pearson and Bayes error exponents. The second model explores spatially correlated Gaussian distributions and challenges the assumption of optimality in contiguous marginal likelihood ratio quantizers, presenting a sufficient condition for optimality with single observations.