Research And Application Of Context-Aware Modeling Based On Pervasive Computing

The rapid development of the internet of things makes the pervasive computing andcontext-aware to be promoted and applied in more and more areas. Context-aware means theprocess that in the constant interaction between humans and computers, pervasive systemacquire the context information related to the user’s needs and provide them the particularservice. And context modeling plays a very important role of information management incontext-aware technology. Context modeling helps pervasive systems acquire, integrate andprocess context information. However, there is no unified standard modeling techniques ofcontext-aware, and different pervasive systems also have different context modelingframeworks. In this paper, context-aware technique at home and abroad is deeply researched,as well as the context modeling framework and technical characteristics in pervasive system.And a scalable context-aware modeling framework with high flexibility is proposed anddesigned.The work done by the paper is divided into three main areas. At first, a generic softwareframework for context-aware pervasive computing-based modeling is proposed. Theframework is divided into context perceiving layer, context management layer, contextinference layer and context application layer, and each layer has the basic architecture. Context perceiving layer is mainly responsible for collecting various types of contextinformation. Context management layer reconstructs context information to specific contextmodels for improving query performance. Context reference layer analyzes and infers contextinformation and calculates the valuable information which users concern. Context applicationlayer provide particular information service to users by application models. Secondly, thepaper proposes object-oriented and ontology modeling based on XML (Extensible MarkupLanguage), this modeling method combines the relevance of the object-oriented, the clarity ofontology and storage of XML language. So this method has better performance in timing,storage, query and reference. Thirdly, this paper proposes Location and path modeling basedon LBS (Location-based Service), which is special for the pervasive system whose core islocation and path. This method can quickly obtain user’s position information and pathinformation, and then calculate their frequent path.At first, the paper describes the research background, including the concept of pervasivecomputing, context information, context-aware and context modeling. Secondly the currentachievement of pervasive system and context modeling technique is analyzed, and the currentdomestic and international research of context modeling technique is divided into four types:targeted context modeling, relevant context modeling, referential context modeling and hybridcontext modeling. And their characteristics, advantages and disadvantages are researched.Thirdly, a common set of context-aware modeling framework is proposed and designedcombined with the current pervasive system’s demand, as well as two context modelingmethods: object-oriented and ontology modeling based on XML, location and path modelingbased on LBS. Finally, the implementation solution of these two modeling techniques isproposed, as well as the performance analysis and validation of the common modelingframework and modeling techniques by an intelligent services push system based on futuremarket.