Research On Some Key Techniques Of Context-aware Based On RFID

With the rapid development of the information and computing technology, the computing mode is also undergoing profound reform. The computing model has gradually developed from the traditional distributed mobile computing to pervasive computing model. The essential characteristic of pervasive computing is the integration of physical space and information space, which provides transparent and ubiquitous services for people. In this thesis, we mainly concern on context-aware computing? an important branch of pervasive computing. Context-aware computing is crucial to the realization of pervasive computing.In the background of pervasive computing, this thesis firstly summarizes and analyzes the research satus of context-aware computing including context modeling methods, context reasoning, context storing and context-aware system framework.Then this thesis proposes an ontology context model based on RFID. The model uses hierarchical modeling method. Context ontologies are categorized by field and managed using unified format. The experimental results show that the model can rapidly build domain ontology for the context-aware system.On that basis, this thesis gives a RFID based context-aware service framework. It includes underlying communication environment layer, context-aware middleware layer and service layer. The function of each module in the framework and the relationships between modules are elaborated. Then two examples are given to illustrate the context model and service model further.Based on the analysis and research of the existing caching replacement algorithms under wireless environment, this thesis proposes a context caching replacement algorithm RCRA which is suitable for pervasive computing environment. In the algorithm, the value of the context is determined by its access probability, history access frequency and freshness. When new context is to be added into the cache or the cache is full, the algorithm removes context from the cache based on the value of the context. Finally, experiments are carried out to compare RCRA with LRU. The result shows that the RCRA improves the hit rate of the cache.