Detection and Resolution Approaches for Layered Context Inconsistency in Pervasive Context-Aware Systems

• Main Authors: Yuen-kuei Hsueh, 薛元揆
• Other Authors: Been-Chian Chien
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/36393057180565724854

碩士 === 國立臺南大學 === 資訊工程學系碩士班 === 98 === Most of researches on context-aware computing focused on the construction of systems and applications. Nowadays constructing complicated systems for large number of users decreases the correctness and reliability of context-aware systems. For improving the quality of context information, the ability of distinguishing the correct context from the incorrect context is the main resolution for a context-aware system. The reason is that the incorrect context may lead context-aware system abnormal reactions and wrong decisions. Two main reasons of causing incorrect contexts are context inconsistency and context conflict. In this thesis, we concentrate on the issue of context inconsistency. A few researches have proposed to focus on designing detection and resolution of context inconsistency. However, most of them were dealt with specific cases and lack of systematic methodology for resolving context inconsistency on the context of different context levels. In this thesis, a layered context management architecture for the inconsistency resolutions is proposed. The architecture is based on a layered context model including sensor layer, event layer, service layer according to the context layers are designed to improve the performance of resolving context inconsistency. The effects on context-aware systems for the proposed approaches on different context layers are also evaluated and discussed.