Adaptive Detection Strategies for Enhanced BER Performance in MIMO Visible Light Communication Systems

This study investigates the Bit Error Rate (BER) performance of a 4×4 MIMO visible light communication (VLC) system using several detection techniques, including Minimum Mean Square Error (MMSE), Maximum Likelihood Detection (MLD), Adaptive Zero-Forcing (ZF), and Beamforming. The evaluation is conducted across multiple scenarios involving variations in transmitter separation (dtr), receiver separation (drx), field-of-view (FOV), and half-power semi-angle (HPSA). Results reveal that beamforming consistently outperforms other detection methods at higher SNR levels, while at lower SNR levels, MLD yields the lowest BER, followed by Adaptive ZF and MMSE. Optimal BER performance occurs with a receiver separation of 40 cm when LED separation is fixed at 2.5 m. With fixed PD separation, the BER is minimized at a 2 m separation between LEDs. For different FOV settings, each detection method achieves optimal BER at specific FOV angles: pseudo-inverse, MMSE, and Adaptive ZF perform best at 60 degrees, while MLD and beamforming show improved BER at 50 degrees. Additionally, varying the HPSA shows that all detection methods perform best at an HPSA of 40 degrees.