| Summary: | 碩士 === 國立臺灣大學 === 電信工程學研究所 === 100 === As localization and wireless communication technologies become mature, mobile devices can locate themselves by new network-based techniques such as Wi-Fi and cell-tower based localization rather than Global Position System (GPS). These new techniques expand the outdoor operating environment of GPS to almost everywhere.
Selecting the best localization scheme for location-based service (LBS) is not a trivial task. First, di erent localization schemes consume di erent amount of energy. For example, network-based localization requires mobile devices to access a remote location server via Internet, which will consume extra power. Second, user operating environment changes with time. As a result, a good localization scheme for now may not be the best choice later at a di erent location. Finally, di erent LBSs have very di erent requirements. For example, navigation would require much faster update of user location. Therefore, a localization scheme that takes a longer time to locate a device may not be appropriate.
In this thesis, we focus on selecting the best power-saving localization technique based on the requirement of an LBS for Android phones. The power consumption of localization techniques available to an Android phone is measured and the corresponding energy models are developed. Based on the energy models, we propose a policy o ering the best power-saving localization technique. An algorithm that dynamically determines localization techniques by detecting the environment is proposed. The proposed algorithm is also implemented in Android phones.
|
|---|
