Diffusion Filtered-x Inverse Tangent Generalized Maximum Correntropy Criterion with Reduced Complexity for Distributed Active Noise Control

Active noise control (ANC) techniques are an alternative for employing passive noise reduction methods in real-world systems. Designing an ANC model across a network of dispersed acoustic sensor nodes with reduced complexity is a challenging task. Active noise mitigation in the existence of impulsive noise is an important issue in a dispersed scenario. A robust diffusion filtered-x inverse tangent generalized maximum correntropy criterion technique for distributed active noise control (DF-xIGDANC) is built to mitigate impulsive noise. Further the proposed DF-xIGDANC is modified as improved DF-xIGDANC (IDF-xIGDANC) technique to reduce the computational complexity. In the inverse tangent technique, the saturation qualities of error non-linearities are combined with the similarity index of generalized maximum correntropy criterion. The proposed method has been examined for active noise cancellation in a dispersed scenario using chaotic and real noise inputs. The effectiveness of noise cancellation can be seen by a 2–3 dB improvement in average noise reduction over existing methods.