Improvement of indoor VLC network downlink scheduling and resource allocation.

Indoor visible light communications (VLC) combines illumination and communication by utilizing the high-modulation-speed of LEDs. VLC is anticipated to be complementary to radio frequency communications and an important part of next generation heterogeneous networks. In order to make the maximum use of VLC technology in a networking environment, we need to expand existing research from studies of traditional point-to-point links to encompass scheduling and resource allocation related to multi-user scenarios. This work aims to maximize the downlink throughput of an indoor VLC network, while taking both user fairness and time latency into consideration. Inter-user interference is eliminated by appropriately allocating LEDs to users with the aid of graph theory. A three-term priority factor model is derived and is shown to improve the throughput performance of the network scheduling scheme over those previously reported. Simulations of VLC downlink scheduling have been performed under proportional fairness scheduling principles where our newly formulated priority factor model has been applied. The downlink throughput is improved by 19.6% compared to previous two-term priority models, while achieving similar fairness and latency performance. When the number of users grows larger, the three-term priority model indicates an improvement in Fairness performance compared to two-term priority model scheduling.