eff-HAS: Achieve Higher Efficiency in Data and Energy Usage on Dynamic Adaptive Streaming

With the limited battery life of mobile devices and the full usage of shared network resources, mobile users are required to devise canny and amicable plans for energy and data consumption. In this paper, we propose an efficient chunk-scheduling strategy for adaptive video streaming which enhances per-power data utility. We first reveal that currently available chunk-scheduling strategies for adaptive video streaming do not consider the general viewing patterns of streaming users, causing a severe inefficiency in terms of not only data usage but also energy consumption. We then propose a energy- and data-efficient chunk scheduling algorithm that increases the per-power data utility for single bitrate encoded videos. We further extend the basic algorithm customizing it to dynamic adaptive streaming over HTTP (DASH) formatted video streaming systems, the most popular video streaming technique. We implemented the proposed scheme and the experiment results show that, while consuming 2% less energy, the scheme reduces the overall data wastage by 55% compared to a prior competing scheme.