Distributed Macro Calibration via Output Synchronization in Lossy Sensor Networks.

摘要: In this paper a novel distributed algorithm for blind macro calibration in sensor networks based on output synchronization is proposed. The formulated as set of gradient-type recursions estimating parameters functions, starting from local criteria dened weighted sums mean square dierences between the outputs neighboring sensors. It proved, basis an originally developed methodology treating higher-order consensus (or synchronization) schemes, that achieves asymptotic agreement gains and osets, sense with probability one. case additive measurement noise, inter-agent communication outages, modication original instrumental variables proved using stochastic approximation arguments modied