Mitigation of Side-Effect Modulation in Optical OFDM VLC Systems

• Main Authors: Mohammed M. A. Mohammed, Cuiwei He, Jean Armstrong
• Format: Article
• Language: English
• Published: IEEE 2018-01-01
• Series: IEEE Access
• Subjects: ACO-OFDM; DCO-OFDM; estimation; intensity modulated direct-detection OFDM; interference; VLC
• Online Access: https://ieeexplore.ieee.org/document/8485363/

Side-effect modulation (SEM) has the potential to be a significant source of interference in future visible light communication (VLC) systems. SEM is a variation in the intensity of the light emitted by a luminaire and is usually a side effect caused by the power supply used to drive the luminaires. For LED luminaires powered by a switched mode power supply, the SEM can be at much higher frequencies than that emitted by conventional incandescent or fluorescent lighting. It has been shown that the SEM caused by commercially available LED luminaires is often periodic and of low power. In this paper, we investigate the impact of typical forms of SEM on the performance of optical OFDM VLC systems; both ACO-OFDM and DCO-OFDM are considered. Our results show that even low levels of SEM power can significantly degrade the bit-error-rate performance. To solve this problem, an SEM mitigation scheme is described. The mitigation scheme is decision-directed and is based on estimating and subtracting the fundamental component of the SEM from the received signal. We describe two forms of the algorithm; one uses blind estimation, while the other uses pilot-assisted estimation based on a training sequence. Decision errors, resulting in decision noise, limit the performance of the blind estimator even when estimation is based on very long signals. However, the pilot system can achieve more accurate estimations, and thus a better performance. Results are first presented for typical SEM waveforms for the case where the fundamental frequency of the SEM is known. The algorithms are then extended to include a frequency estimation step and the mitigation algorithm is shown also to be effective in this case.