Maximising communication reliability in vehicular visible light communication systems: a novel Allan variance—based adaptive Kalman filtering approach

The intelligent transportation systems (ITS) using radio frequency (RF) face challenges like interference and limited bandwidth compared to visible light communication (VLC). Integrating vehicular VLC (V-VLC) utilizing the existing light-emitting diodes (LEDs) in automobiles could provide a complementary approach to mitigate RF constraints. Performance degradation by background noise from sunlight can undermine signal quality. So, improving performance and managing interference are the keys to better V-VLC operation. This work proposes a solar noise suppression method for a V-VLC system, leveraging Allan variance and an extended Kalman filter (EKF). Based on the obtained Allan variance, updates are made to the observation model, state vector, and observation noise covariance matrix to achieve superior performance. Minimum mean square error (MMSE) equalization is used. The proposed filter efficiently minimizes the notable autocorrelation triggered by the solar noise, resulting in a bit error rate (BER) reduction of approximately 88% compared to the system without the adaptive filter. This increased convergence highlights the possibility of higher communication performance and reliability. Hence, this work provides a novel adaptive EKF design and offers valuable perspectives on the importance of Allan variance in enhancing the V-VLC system.