Cooperative caching for HTTP-based adaptive streaming contents in cache-enabled radio access networks

Abstract

The rapid growth of the number of mobile broadband subscribers has led to an exponential increase of the Internet traffic, of which a major part is video traffic. In 4G/5G mobile networks, the eNBs should cache the popular contents to reduce the transit cost at the backhaul links. Recently, HTTP-based adaptive streaming is widely used to transfer the video contents to the end users in the Internet. With HTTP-based adaptive streaming, each video content is stored in several representations corresponding to different video performances. The representation with a better performance yields a higher user satisfaction, however, it consumes more storage and transit cost. In this paper, we propose a cooperative caching model in which the eNBs collaborate in caching and request routing the video contents. The proposed optimization problem is a large-scale integer linear program which is an NP-hard problem. Based on alternating direction method of multipliers technique, a distributed algorithm converging to the solution to the relaxation problem is then developed. It results to fractional caching and request routing which serve as a performance benchmark of the cooperative caching. Moreover, we propose a nearest-neighbor request routing policy and a lightweight-cooperative eviction algorithm for integral caching. Extensive simulations show that the proposed request routing and eviction algorithms achieve more than 78% of the performance benchmark.