Design of distributed consensus kalman filter based on energy optimization

As to the problem of limited node power in wireless sensor networks, a new energy efficient distributed kalman filtering algorithm is proposed based on distributed consensus kalman filtering algorithm given by Olfati-Saber and site percolation model- a kind of random graph. In this algorithm, each cluster contains with a cluster-head and some member nodes where the cluster-head is used to fuse data which come from member nodes and consensus process between neighbor cluster-head. Then, the convergence of this new algorithm is proved by applying to Lyapunov second method and matrix theory. At last, a simulation example is given to show that compared with the Saber´s algorithm, the algorithm proposed in this paper not only has the approximate estimation performance, but also decrease node energy consumption effectively, and then greatly prolongs the life of wireless sensor networks.