Optimal Fairness Scheduling for Coded Caching in Multi-AP Wireless Local Area Networks
share
Coded caching schemes exploit the cumulative cache memory of the users by using simple linear encoders, outperforming uncoded schemes where cache contents are only used locally. Considering multi-AP WLANs and video-on-demand (VoD) applications where users stream videos by sequentially requesting video “chunks", we apply existing coded caching techniques with reduced subpacketization order, and obtain a computational method to determine the theoretical throughput region of the users' content delivery rates, calculated as the number of chunks delivered per unit of time per user. We then solve the fairness scheduling problem by maximizing the desired fairness metric over the throughput region. We also provide two heuristic methods with reduced complexity, where one of them maximizes the desired fairness metric over a smaller region than the throughput region, and the other uses a greedy algorithmic approach to associate users with APs in a fair way.
READ FULL TEXT