In this paper, a special class of distributed composite binary hypothesis testing problem with monotonic likelihood ratio is investigated. The sensor observations are assumed to be conditionally independent given a fixed but unknown parameter θ where θ ∈ Θ1 under the H1 hypothesis and θ = θ0 under the H0 hypothesis. The optimal form of sensor decision rule is established under both the Neyman-Pearson and Bayesian criteria. As an illustrative example, the design of an optimal cognitive radio rule for cooperative spectrum sensing is established.
Available at: http://works.bepress.com/hao_chen/16/