MACC: MEC-Assisted Collaborative Caching for Adaptive Bitrate Videos in Dense Cell Networks

Haojia He, Songtao Guo, Lu Yang, Ying Wang

Abstract: Caching adaptive bitrate video at edge nodes (ENs) can provide multi-version video-on-demand (VoD) services to end users (EUs) with better experience. However, due to the limited cache capacity of ENs, it is important to decide which video content and corresponding bitrate version to be cached in the EN. In this paper, we first propose a user request hit profit (RHP) model, and then based on the RHP model we envision a mobile edge computing (MEC)-assisted collaborative caching scheme (MACC). Specifically, we model the communication links between ENs and EUs as a bipartite graph to employ the collaborative caching among ENs; and we consider the transcoding relationship between different versions to effectively utilize the processing capacity of ENs. Due to the NP-completeness of the cache placement problem, we prove it is a monotone submodular function maximization problem, and propose the proactive cache placement based on maximum RHP increment (PCP-MRI) algorithm and the reactive cache replacement based on maximum RHP increment (RCR-MRI) algorithm. Extensive simulation results show that, compared with existing methods, the proposed MACC has significant performance improvements in cache hit ratio, initial waiting delay and backhaul traffic load.