Creating Usage Models to Identify Misbehaving Applications on Mobile Devices

Automated monitoring;user modeling;context-based computing;data analytics;Android;battery;mobile device

Jiang, Qiushiadvisor:Ward, Paul ; affiliation1:Faculty of Engineering ; Ward, Paul ;

University of Waterloo

关键词: mobile device;    data analytics;    Master Thesis;    user modeling;    battery;    context-based computing;    Automated monitoring;    Android;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/14894/3/Jiang_Qiushi.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

Limited battery capacity is currently a major pain point for mobile users. The problemis made worse when poorly designed applications consume a significant amount of power inthe background when they are not actively used by the user. To combat this problem, wepropose an automated monitoring system that can detect misbehaving applications runningon mobile devices. Our system does not require any prior knowledge about the monitoredapplications. Instead, it collects the user’s usage records and builds models to encapsulatethe contexts when the user is likely to use each application. From those models, oursystem can identify misbehaving applications that are consuming system resources whileproviding no useful service to the end user. In this dissertation, we demonstrate the overalldesign for our system. This design allows us to collect detailed usage records while keepingour system’s power consumption at a minimum. We also introduce the steps we take toconstruct our usage models and the rationale behind each key decisions. In the end, weevaluate the effectiveness of our system by running it on a real Android device during atwo month period. From the experiment, we show the misbehaving applications identifiedby our system have a significant impact on the battery life, and misbehaving applicationswith high network usage is the main cause of fast battery drain.