A Resource-aware Distributed Kalman Filter with Stochastic Communication Based on Site-percolation Model
Chunxi Yang, Jie Zhu, Jing Zhang, Lingyun Huang · 2019
This paper deals with a distributed Kalman filtering (DKF) algorithm for wireless sensor networks (WSNs) which focus on the topological structure of WSNs for reducing the communication distance and communication bandwidth. Clustering techniques and the site-percolation model are used to reduce communication distance and communication complexity among nodes. Then, the convergence is analyzed by the matrix theory. Moreover, a hybrid approach by incorporating dichotomy is proposed to obtain the suitable site-occupation probability in order to decrease unnecessary energy consumption for a given filtering accuracy. Finally, a simulation example is given to show that the proposed algorithm decreases the energy consumption of WSNs effectively at the cost of excess estimation performance.