Research On QoX Quality System And Algorithms For Context Information Processing

The emergence of ubiquitous computing has changed the traditional computing model.It truly realizes human-centered,frees people from heavy operations,and greatly improves the quality of experience(QoE).As one of key technologies of ubiquitous computing,context-aware technique can automatically"perceive"the context information of the environment,analyze and process the acquired context information,and then make corresponding adjustments based on the processed context information.finally realize preset service objectives and meet the requirements of quality of service(QoS).The existing context-aware systems are mainly oriented to specific application scenarios,and they exist for different functions,but lack a quality assessment system in the system level.On the other hand,in context-aware systems,due to high dynamic environment,network delay and so on,the quality of acquisition devices(QoD)cannot be effectively guaranteed,which results in context inconsistency problem.At the same time,the quality of context(QoC)is reduced,which makes the system unable to achieve correct reasoning and make inappropriate decisions.Therefore,it is of great significance to achieve effective elimination of context inconsistency.Concerning the actual needs of the development of context-aware systems,this thesis studies the QoX quality system for context processing and context inconsistency elimination algorithms.The specific research contents are as follows:(1)Study QoD,QoC,QoS and QoE index in context-aware systems,and define them as a comprehensive quality index QoX.Meanwhile,considering the characteristics of various context-aware systems,this thesis proposes a novel QoX quality system for context information processing,which can intuitively display the performance of context-aware systems.In addition,the intrinsic relationship among the four indexes of QoX is studied,and the realization of system function is transformed into the interaction of quality parameters,which realizes adaptive adjustment of the system(2)From the perspective of realizing effective context inconsistency elimination by improving the accuracy of quality of context information,this thesis quantitatively analyzes the impact of different QoC parameters on the accuracy of inconsistency elimination,and proposes a new overall quality of context(OQoC)indicator by integrating multiple QoC parameters in a reasonable way.The proposed OQoC indicator can reflect the overall quality level of context information from external characteristics of sensors and intrinsic characteristics of the context data,and indirectly reflect the inconsistency degree of context information.(3)Based on the OQoC indicator,this thesis constructs different reliability functions of Dempster-Shafer(D-S)theory for different context information,and proposes a new context inconsistency elimination algorithm based on OQoC indicator and D-S theory.In the personal identity verification scenario,several simulation experiments are conducted.The experiment results verify the effectiveness and stability of the proposed algorithm in eliminating inconsistency from multiple aspects.Compared with other existing algorithms,the proposed algorithm has the advantages of high precision and real-time adjustment.Through the above research contents,they not only effectively solve the problem of context inconsistency,but also improve the quality of context information,which lays a good foundation for the subsequent fusion reasoning.In addition,they provide a comprehensive quality assessment framework and adaptive optimization method for various context-aware systems,which improves the overall performance of context-aware systems.The above researches provide powerful theoretical support for the in-depth development and wide application of context-aware technology.