Robust distributed adaptation under arctangent and maximum correntropy criterion

In this paper, the problem of robust distributed estimation in undirected networks is explored in depth. The Arctangent framework has garnered widespread adoption for adaptive estimation in prior research, with numerous adaptive estimation algorithms stemming from this foundation. However, the integration of the Arctangent framework with the maximum correntropy criterion, and its potential for enhanced estimation results, remains unexplored. In this paper, we have conducted a study in distributed networks and proposed a new robust distributed estimation algorithm using the Arctangent framework and the maximum correntropy criterion. Simulation experiments demonstrate that the proposed algorithm exhibits superior performance compared to the benchmark algorithm in both Gaussian and impulsive noise environments. Finally, a theoretical analysis of the proposed algorithm is conducted.