Research On Context Modeling And Reasoning In Pervasive Computing Environments

Pervasive computing, as a new pattern of computing, implements the convergence of the cyber space and the physical space. In this fused world, people can transparently get digital services at anytime and anywhere. Context-aware computing is one of the important issues of pervasive computing, and is an important way to achieve the intelligence of pervasive computing. However, context-aware computing environment is very complex and complex, in which, context information comes from varied data sources with different structures, and there exists conflicts and uncertainties among these contexts. Therefore, we need to represent these contexts by the use of unified data structure or model to realize the sharing between different devices and applications. Furthermore, to acquire more valued high context, we also need an effective method to reason and fuse uncertainty context.The main works of this thesis are as follows:(1) After the analysis of existing frameworks of context-aware computing systems, this thesis uses the ontology to represent and model the context information, and builds a framework of context-aware computing system based on ontology. It has some advantages: a unified description of context knowledge, knowledge-sharing between different services and a common understanding of domain knowledge. And it finally realizes the purposes of distributed context-aware system development and software reuse. (2) As to the uncertainty context reasoning, we combine the domain ontology model and propose a method of uncertainty context reasoning based on the evidence theory, by mapping the ontology model to the trust structure. For the high-conflict evidences, we calculate the credibility of the evidence as a weight and use the weight to average all the evidences. After that, we use the combination rules of evidence to attain the trust extent for each hypothesis, and finally get the deduced results. The experiment proves the validity and accuracy of this reasoning method. (3) In the last section, we have designed and implemented a self-service travelling system within the campus environment. Inspired by the thinking of ontology modeling and reasoning, we detailedly analyzed the process of building campus ontology, described the use of reasoning based on ontology and rules, and presented the implementation of the context-aware application in hardware platform and software architecture. At last, we demonstrated the application available and feasible through the relevant scene.