Machine Learning Based Adaptive Filtering Of Mixed Noises In Visible Light Communication

Abstract

Visible Light Communication (VLC) has become a secure, high speed and cost-effective technology completing the traditional radio frequency wireless communication. However, noise interference from indoor light sources and susceptibility to ambient light of sunlight, flashlight, LED lamp etc. becomes challenge to achieve higher data rates in VLC. The receivers in VLC system are affected with short, amplifier and thermal noises. The noise analysis at receiver becomes difficult due to the combined effect of these noises. Also, with variance in distance, designing filter becomes a challenge. The mitigation of these noises is possible only with an adaptive filter. In this work, a machine learning based adaptive filter is designed to mitigate the mixed noises due to combined effects of short, amplifier and thermal noises and variance in distances.