Numerical solutions for optimum distributed detection of known signals in dependent t-distributed noise-the two sensor problem

We examine distributed two-sensor detection of known signals in t-distributed noise which is dependent from sensor to sensor. A Gauss-Seidel algorithm which attempts to minimize the Bayes risk is used to obtain the rules for the decision regions. The best nonrandomized fusion rules are sought. It it shown that the properties of the decision regions can be predicted based on the problem's parameters. In some specific cases the optimum distributed detection sensor rules are shown to be better than likelihood ratio tests by Monte Carlo simulations.