One-Step Asynchronous Data Fusion DLMS Algorithm

Abstract

In recent years distributed estimation has attracted much attention. In traditional distributed algorithms, each node performs data fusion over synchronous data, which causes lots of time consumptions in the actual situations and estimation performance degradation. To deal with this problem, we propose a new one-step asynchronous data fusion strategy in distributed estimation algorithms. Moreover, the proposed algorithms with or without measurement data sharing are studied to provide different asynchronous cooperation strategies. In particular, the convergence behavior of the proposed asynchronous fusion algorithms is analyzed, and why asynchronous fusion can improve estimation performance and reduce time consumptions are also analyzed. The effectiveness of the proposed algorithms is demonstrated through some illustrative examples. Simulation results show that the proposed algorithms considerably outperform the traditional DLMS algorithms and LMS algorithm.