A Supermartingale Theory-Based Random Compensation Algorithm for Visible Light Communication/Radio Frequency Hybrid Networks

Visible light communication (VLC)/radio frequency (RF) hybrid networks exhibit large system capacity, high reliability, and low access latency. To simplify the resource allocation process of such hybrid networks, we devise a random compensation algorithm based on supermartingale theory. Firstly, a network queuing model with random arrival traffic and random compensation service mechanism are established. Then, the delay performance is evaluated based on supermartingale theory, and a tighter delay violation probability bound is derived. A search algorithm is designed to obtain the compensation bandwidth and compensation probability based on the RF. Finally, the devised algorithm is compared with that based on large deviation theory. The results of simulations demonstrate that the proposed algorithm is authentic and could save bandwidth.