- Ri Kwang Il, Ri Chol Ho & Jenni J
- DOI: 10.5281/zenodo.22987208
- GAS Journal of Engineering and Technology (GASJET)
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.
