Capacity Maximizing Power Allocation for Multi-Channel Visible Light Communications

Advances in visible light communication (VLC) technology and its smart-city applications have gained great interest, e.g., indoor light-fidelity (LiFi) and outdoor vehicle to vehicle (V2V). In order to fully explore the benefits of VLC under diverse channel/noise environments, it is important to schedule power efficiently so that the capacity is maximized or the error rate is minimized. We propose efficient optical power allocation algorithms for multi-channel single-user VLC capacity bounds maximization and error rate minimization for multi-user VLC with the existence of signal-dependent noise (SDN) for potential city applications. It is pointed out in this paper that the wellknown water-filling algorithm and its generalizations do not apply straightforwardly to the capacity bound optimization for VLC and other communication systems based on the intensity modulation and direct detection (IM/DD).