DASH based video caching in MEC-assisted heterogeneous networks

Abstract

Recently, the demand for video applications has become more and more pressing. To save backhaul bandwidth and reduce the network delay, Mobile Edge Computing (MEC) has arisen as a promising technique in future mobile networks. It can provide the storage, computing and networking resources in close-proximity to the mobile users through deploying multiple MEC servers at mobile edge. However, how to effectively utilize the limited storage capacity of these MEC servers is still the major challenge, especially when Dynamic Adaptive Streaming over HTTP (DASH) is used for video transmission. Recent studies have shown that layered encoding is very suitable for streaming service using DASH standard. Therefore, in our paper, combining with the feature of DASH-based scalable video transmission, we first design a novel multi-server cooperative scheme with software-defined networking (SDN) architecture for future MEC-assisted heterogeneous networks, to effectively manage and schedule the available resources, thus guaranteeing the requested video delivered to the clients with higher quality and lower delay. And then, based on the above scheme, the end-end delay is carefully analyzed for each real video streaming. With a full consideration of the averag end-to-end delay of video streaming and the provided video quality of MEC servers, we finally propose a Dynamic Programming based Adaptive Caching Algorithm (DPACA) to obtain an optimized caching strategy under the constraint of cache storage capacity at each MEC sever. Simulation results show that, our algorithm can improve the cache hit ratio while significantly reducing the end-to-end delay. Moreover, our algorithm has higher flexibility and adaptability.