Dynamic HTTP download scheduling with respect to energy consumption

Abstract

Mobile devices, in particular smartphones, are battery powered. As their small size limits their energy capacity, an intelligent and energy-saving use of resources is crucial for the mobile user experience. Especially energy saving while sending and receiving data becomes important due to the alwayson nature of smartphones. One solution for a sophisticated resource usage in wireless access networks is to tailor resource management to applications. The advantage of such mechanisms is to be able to address both the perceived quality of the users and the energy consumption of mobile devices at the same time. We propose several traffic scheduling algorithms for HTTP file download in wireless networks based on this application-aware paradigm. The core idea is to use download scheduling that avoids parallel transmissions and, in contrast, favors sequential transmission. We have implemented and measured the algorithms in a wireless mesh network. The evaluation results quantify the gain of an application-aware resource management with respect to quality of experience and energy consumption.

Kraken.me Mobile: The Energy Footprint of Mobile Tracking

“…Recently, people have started considering energy-efficient continuous context sensing algorithms [15], [17], [18], [5]. Others have focused on optimizing the network power consumption as the second largest consumer after the display [19], [20]. Based on some of this work, Kansal et al [2] analyze the trade-off between accurate sensing and the power consumption of the smartphone, from which they derive a sensor abstraction layer considering the energy consumption.…”

Catching the download train: Energy-efficient file downloading on smartphones

“…Sharma et al presents an energy-efficient architecture, Cool-Tether [10] in order to utilize the available smartphones with MH's existing Internet access. Blenk et al studied the tradeoff; while single TCP flows typically do not utilize the available bandwidth on a cellular link due to variable capacities, parallel transmissions yield an increased average transmission duration [11]. In this paper, we consider both the ground truth power measurements and the download durations in the evaluation of energy consumption.…”