Distributed filtering with stochastically triggered communication

This note investigates the problem of event-based distributed filtering, where a data-scheduler is embedded to regulate the communication among sensor nodes. A novel distributed filtering framework is designed based on the random triggered communication mechanism and matrix intersection fusion mechanism. It is shown that the designed algorithm can preserve the consistency of the estimation, and the estimation error covariances are uniformly bounded by the solutions of harmonic-coupled algebraic Riccati equations (HCRE). Compared with the literature, the design procedure fully exploits the non-transmission information and correspondingly improves the performance of the filter.