Data-driven Adaptive Filtering for Impulsive Noise Mitigation in Power Line Communication System in Industrial Application

Data-driven adaptive filters, in which a deep neural network is incorporated as a core architectural component, are applied to noise mitigation in industrial power line communication. Noise is modeled using an alpha-stable distribution, which is well suited to asynchronous impulses in industrial power line communication. The results indicate that data-driven adaptive filters can be effectively used for noise mitigation in industrial settings, ensuring stability and outperforming traditional least mean square filters and purely data-driven filters.

Keywords: Power Line Communication, Adaptive filtering, Data-driven signal processing, Noise Mitigation, Asynchronous impulse, Industrial communication systems.