Edge Assisted DASH Video Caching Mechanism for Multi-access Edge Computing

Abstract

Proliferation in mobile devices and the increase in video data consumption on these devices has led to an unprecedented surge of data usage in mobile networks. It is both challenging and expensive for network operators to scale up the network capacity and tackle this ever increasing data demand. Cellular network operators require alternative solutions, like in-network-caching, to solve this problem. Popular streaming services like YouTube use Dynamic Adaptive Streaming over HTTP (DASH) for video streaming where videos are divided into several small segments, and multiple bit-rate versions of each segment are stored in the server. Using store and forward caching method, in the network, may not help as the video segments cached in one session might not be usable for other users. This problem of unusability emerges as different users request different bit-rates of the same video segment. Also, it is not efficient to cache all versions of the video segments at the edge of the network, due to limited storage at the edge. In this paper, we propose a Multi-access Edge Computing (MEC) based video caching mechanism, where only the highest available bit-rate video is cached and by using the processing power available at the MEC it is transcoded to the requested lower bit-rate version. We develop a test-bed to evaluate the performance of the proposed caching mechanism in real time. Through various experimental results, we demonstrate that the proposed method reduces the backhaul traffic load and video load time and increases the cache hit-rate as compared to traditional store and forward caching mechanism.