A Fuzzy Order Statistics Fusion Rule for Distributed Detection System
Qu Fuyong · Information and Electronic Engineering · 2007
A new fusion rule,Fuzzy Order Statistics(FOS) fusion rule,is presented to improve the detection performance of binary local decision in multisensor distributed detection systems.Analytical expressions are given for the false alarm probability.Simulation results based on local sensors in a distributed system indicate that in homogeneous environment,order statistics fusion rule based on fuzzy membership function has better performance than that the binary local decision rule has,and order statistics fusion rule is also superior to thesumfusion rule based on fuzzy membership function.The order statistics fusion rule is robust in multi-targets.