Diffusion Complex Total Least Mean M-Estimate Adaptive Algorithm for Distributed Networks and Impulsive Noise

Complex domain adaptive filtering algorithms (CD-AFA) have attracted the attention of researchers. However, there are very few studies on AFA for complex-valued errors-in-variables (CEIV) models. Additionally, considering the advantages of distributed methods in various environments such as wireless and sensor networks, this brief proposes the diffusion complex total least mean M-Estimate adaptive filtering algorithm (DCTLMM-AFA). This algorithm extends the total least squares (TLS) method to complex domain and applies it to distributed networks for the first time. In addition, the DCTLMM-AFA also robust to impulsive noise is attributed to the characteristics of the M estimation function (MF). Finally, this brief provides an exhaustive analysis of the algorithm’s theoretical performance in the presence of impulsive noise. Simulation results substantiate both the advantages of the proposed algorithm and the accuracy of the theoretical analysis presented in this brief.