Channel Modeling and Signal Processing for Array-Based Visible Light Communication System Under Link Misalignment

We address the lens optimization and signal processing for array-based multiple-input multiple-output (MIMO) visible light communication, especially under link misalignment. We employ one concave lens and one convex lens to separate the lights from different light-emitting diodes (LEDs), and optimize the lens structure to minimize the condition number of channel gain matrix. In this way, the light emitted by different LEDs can be separated well from each other. To further mitigate the interference under transmitter/receiver mobility, we propose two signal processing approaches based on successive interference cancellation (SIC) and multi-layer coding (MC), along with maximum ratio combining (MRC) for interference/signal decoding. It is demonstrated that the interference can be reconstructed and eliminated through SIC, and MC can partially decode the interference to better reconstruct the desirable message. The proposed signal processing approaches can significantly improve the sum rate in both link alignment and misalignment cases. Moreover, MC can achieve higher sum rate, especially in the case of link misalignment.