| 1 Introduction |
| 2 Related |
| 2.1 Existing Work on Power Saving |
| 2.2 Thin-Client Based Approach |
| 2.3 Proxy Based Approach |
| 2.4 Cloud Based |
| 3 Motivation |
| 3.1 Delay Issues in Web Browsing |
| 3.2 Power Consumption of Web Browsing in 3G |
| 4 Virtual Machine Based Proxy |
| 4.1 System |
| 4.2 Communication |
| 4.2.1 Features for Opening Web Pages |
| 4.2.2 Supporting User Interactions |
| 4.2.3 Compressions |
| 4.2.4 Adaptation Techniques |
| 4.3 Design of VMP at the Service Provider |
| 4.3.1 VM Management |
| 4.3.2 Managing VM Image |
| 4.3.3 Scalability Issues |
| 4.3.4 Implementation Issues |
| 4.4 Security and Privacy Issues |
| 4.4.1 Security-Enhanced Model for the VMP Architecture |
| 4.4.2 Asymmetric Encryption and Authentication System |
| 4.4.3 Trusted Computing Module |
| 4.4.4 Antivirus Scanning Service |
| 4.4.5 System Overhead |
| 4.5 Discussions |
| 4.5.1 Supporting Interactive Applications |
| 4.5.2 Deployment Issues |
| 5 Performance |
| 5.1 The Experimental Setup |
| 5.2 The Web Page Benchmark |
| 5.3 Experimental Results on Delay |
| 5.3.1 Average Delay for Opening Benchmark Web Pages |
| 5.3.2 Effects of Transmitting Multiple Screens of Data |
| 5.3.3 Effects of Data Compression |
| 5.4 Experimental Results on Power Consumption |
| 5.4.1 Power Consumption of the Smartphone Browser |
| 5.4.2 Power Consumption with Web Page |
| 5.4.3 The Benefit of Performing Antivirus Scanning with VMP |
| 6 Conclusions and Future Research |
