An Energy Management Mechanism for Smartphones

現在的智慧型手機通常擁有許多的無線介面，例如: GPRS, 、EDGE、, WCDMA、UMTS、WLAN和藍芽等等。也提供多樣化及個人化的應用程式，例如: 收發郵件瀏覽網頁觀看網路電視和遊戲等等。然而，這些多媒體功能對於手機是相當耗電的，因此，手機的使用時間也成為一個相當大的課題。 為了了解手機的耗電問題，我們首先在不同的情境下去量測整隻手機的耗電量，並發現應用程式的耗電量是可以從整隻手機的耗電量推估出來。進而，我們計算單位位元組的耗電量去評估網路應用程式的耗電效率。而我們發現收發郵件的每個單位位元組的耗電量是會隨著信件的大小去改變的，當信件越大，則單位位元組的耗電量也就越小。根據這個特性，我們提出了一個針對智慧型手機的即時郵件(Push Mail)的省電管理機制，而在我們的解決方案下，手機的使用時間可以提升至二倍之多。 最後，我們也針對智慧型手機提出一個系統性的電力管理機制，而系統性的電力管理機制根據使用者的使用特性跟使用者所使用的網路來推估手機的使用時間。Current designs of smartphones usually provide several communication interfaces such as GPRS, EDGE, WCDMA, UMTS, WLAN, and Bluetooth,...etc, and support various personalized applications such as mails, internet surfing, IPTV, and games,...etc. However, battery life time for smartphones is becoming a major drawback because those media-rich functionalities are so diversified that they can bring high power consumption to the smartphones. In order to resolve these power-related problems, we measure the total power consumption of smartphones under a variety of scenarios. In fact, the power consumption of applications can be split from the total power consumption of the smartphones. In our measurement experiments, the power consumption per byte is obtained after distinguishing the energy-efficiency of applications, themselves, the backlight power, and the power consumption per byte of receiving and transmitting data over wireless interface. Based on this fact, we propose a mechanism which focuses on the optimization of Mobile Phone''s Push Mail. According to the simulation results, our proposed mechanism can improve the expected battery life. Finally, we propose a systematic approach to optimize the power-consumption for smartphone applications. The systematic approach can predict the battery life time based on the previous users'' behavior and the condition of the selected network