Distributed Quickest Detection of Significant Events in Networks

摘要: The problem of quickest detection significant events in networks is studied. A distributed setting investigated, where there no fusion center, and each node only communicates with its neighbors. After an event occurs the network, a number nodes are affected, which changes statistics their observations. may possibly perceive at different times. goal to design sequential rule that can detect when "significant", i.e., has affected less than η nodes, as quickly possible, subject false alarm constraints. algorithm proposed, based on novel combination alternating direction method multipliers (ADMM) average consensus approaches. Numerical results provided demonstrate performance proposed algorithm.