Energy Optimization of Online Tracker for Mobile Devices

Wei Zhao,Renjie Lu,Zhenyu Cui,Haihua Shen

DOI: https://doi.org/10.1109/cscwd.2019.8791851

2019-01-01

Abstract:Nowadays, it is common that both mobile applications and websites log users' visits and keep record of users browser path, even every time user click and every word user type. Each tracking information will be respectively sent to the third-party server via HTTP request. Previous studies have shown that combining multiple small HTTP requests can significantly decrease energy consumption, but not have considered a typical scenario where massive tracking records can be efficiently compressed and transferred with a delay. In this paper, we propose a comprehensive approach to reduce the energy consumption of online tracker by bundling multiple HTTP requests. Our approach first intercepts all tracking related HTTP requests, logging and compressing the requests for bundling, and then uses a proxy-based technique to combine these requests at runtime. Through a set of real-world experiment, the evaluation demonstrates that our approach can achieve an average energy reduction of 10% for mobile devices and make apparent performance and security improvement.