Energy-Efficient QoE-Driven Strategies for Context-Aware RAT Selection

Abstract

This paper formulates an optimization problem that maximizes an aggregate utility that captures the “in-context” suitability of available radio access technologies (RATs) to support adaptive video streaming subject to a single-homing constraint. To efficiently solve the considered problem, a novel network-assisted quality-of-experience (QoE)-driven methodology is devised, and its impact on the end-user devices is evaluated. The proposed approach is evaluated and benchmarked against its distributed and centralized counterparts from a cost-benefit perspective. The results reveal that the proposed strategy significantly outperforms its distributed counterpart, and performs differently with respect to its centralized counterpart depending on the number of video clients. At low loads, it performs similarly with much less control overhead. At high loads, the proposed strategy scales up well, while the centralized approach gets overwhelmed by an increasing uplink signaling. A practicality analysis of the proposed strategy for battery-powered devices reveals that its gain in terms of uplink signaling outweighs its cost in terms of processing load, which results in a drastic reduction of the consumed energy. Therefore, the proposed solution provides a win-win situation, where the video clients can sustain good QoE levels at reduced energy consumption, while the network can accommodate more users with existing capacity.