Design and analysis of distributed detection networks with uncertainties

A distributed detection network is an acyclic directed graph with detectors as nodes. Each detector has the ability to process input data consisting of external observations and decisions from preceding detectors, to produce a decision regarding an underlying binary hypothesis-testing problem. The local observations are assumed conditionally independent, given either hypothesis. Each local detector has an unknown probability of being jammed or defective and an unknown probability of providing an incorrect decision if it is jammed or defective. The resulting binary hypothesis-testing problem is solved using concepts of Dempster-Shafer theory. Each detector employs Dempster's combining rule to aggregate its input informtion for a decision. Each detector employs a likelihood ratio test, and the thresholds can be expressed as a function of the local performance characteristics and the uncertainty discount rates. The proposed decision rule has very robust behavior, and with few expectations it significantly outperforms the minimax decision rule and the decision rule that is optimum when there are no jammed or defective detectors.