Web browsing is one of the most common and important services provided by smartphones. However, it currently suffers from long delays and creates a huge drain on the battery lifetime of smartphones. These limitations are caused by complex interactions between the processing flow in mobile web browsers and specialized characteristics of the wireless radio interface, and by the processing limitations of the smartphones. This project addresses these limitations by focusing on three intertwined issues: (i) various techniques to reorganize the computation sequence of the web browser to let the wireless radio interface enter sleep earlier, are designed, implemented and evaluated; (ii) practical data mining based methods are introduced to predict the user viewing time of webpages and determine when the smartphone should switch to low power state, considering the resource limitations of the smartphone and various tradeoffs between delay and power; and (iii) a new architecture is proposed to shift the computing from smartphones to the virtual-machine based proxy to address the computation limitations in smartphones considering scalability issues and bandwidth constraints. This project will make significant theoretical and technical advances on reducing the power consumption and access delay of web browsing in wireless networks. The success of this project will affect many people's daily life, making smartphones last longer and web browsing faster. The results of the project will be disseminated widely through high quality publications and presentations. The proposed research will also be integrated with the education curricula at the Pennsylvania State University.
