A diffusion block least-mean square algorithm for distributed adaptive estimation

Distributed adaptive estimation algorithms are known as a solution to the issue of linear estimation over distributed networks. However, as we show in this paper their performance deteriorate considerably when the links between nodes in the network are noisy. To address this problem, in this paper we propose a new distributed diffusion adaptive estimation algorithm which is base on the block adaptive filtering. By block adaptive filtering, the number of communications between nodes reduces to the block length times than the sample data processing. Less communications between nodes in turn decreases the effect of noisy links. The simulation results show that our proposed algorithm outperforms in steady-state estimation error than sample data processing algorithm.