| Summary: | 碩士 === 國立清華大學 === 通訊工程研究所 === 102 === Recent researches indicate that radio tail is one of the main causes for energy consumption of mobile devices in 3G/4G mobile networks. However, applications (apps) on a mobile device, such as e-mail, instant messaging, keep awaking the network interface to retrieve the latest information, resulting in numerous radio tails and thus consuming lots of energy. In this thesis, a power-saving scheme based on user behaviors is proposed to reduce the occurrence of the radio tails. In the proposed scheme, while a user is not using the device, application traffics are merged together by switching on/off the network interface, and thus power consumption caused by radio tails is reduced. Unlike other existing schemes, the proposed scheme does not need to monitor network traffic, which requires the root permission of mobile devices; it only collects app usage and screen status to learn user behaviors, and controls wake up/sleep timers of the network interface according to the collected contexts for the mobile user. The proposed scheme has been implemented as an Android app, named as Automatic Battery Savior (ABS), in this thesis. Our evaluation results show that devices installed with ABS outperform those with the baseline method.
|
|---|
