Research On Methods Of Data Access And Service Provision In Internet Of Things

The Internet of Things(IoT)aims to connect everything,provide ubiquitous services,support smart cities,smart homes,industrial control and other fields.Massive IoT data access and distributed service provision are the core technologies of its wide application.Aiming at the problems of the high complexity of IoT data access and the difficulty of providing service portfolios,such as the differentiation of massive IoT devices,the diversification of communication protocols,the privatization of data formats,and the combination of service provision,this thesis focuses on the rapid access of heterogeneous IoT data,the provision of multi-domain IoT service combination,designs and implements the IoT platform data access and service provision system.Aiming at the problems of inconsistent data representation of massive heterogeneous devices,inconsistent resource descriptions and information expressions of specific scenarios and devices,this thesis proposes an ontology-based resource description method for the IoT,and designs a resource model that characterizes IoT data from five aspects,including attributes,status,functions,interfaces,and security,to achieve accurate and general description of heterogeneous device resources.Aiming at the problem of diversified ways of providing IoT services,it is impossible to provide a consistent view for application development,this thesis proposes a GraphQL-based approach to the service of IoT resources,designs a service call interface based on the device model,makes the IoT services have the advantages of platform independence,high integration,need to offer,supports the combination of IoT services.Experiments show that compared with the one-to-one adaptation and SDK access scheme,the time overhead of this scheme is reduced by 59.09% and 32.43%,respectively,reduces the complexity of multi-source data access and improves the capabilities of IoT services.Aiming at the diversification and increase in the number of IoT services,unable to provide the complex combination of service requirements that guarantee QoS,this thesis studies the IoT service provision method based on MPGA.Aiming at the non-functional needs of users,comprehensively considering QoS attributes such as service cost,response time,availability,geographic distance,and reputation,this thesis proposes a QoS-based IoT service quality model,according to the user's preferences for non-functional attributes,an objective function is established,and the optimal combined service is found through solving.Aiming at the problem of service composition time constraints,an Improved Multiple Population Genetic algorithm(IMPGA)method is proposed,improves the immigration operation of the Standard Multiple Population Genetic Algorithm(SMPGA).it further reduces the probability of SMPGA entering the local optimal solution,and improves the convergence speed of the population.Based on the above methods,a case of IoT service combination containing complex workflows and multiple non-functional requirements is designed.Experiments show that the IMPGA scheme proposed in this thesis improves the average convergence efficiency proposed by SMPGA and Y-MPGA 30.61% and 35.49%,improved the quality of IoT service provision.Based on the method presented,this thesis designs and implements the IoT data access and service provision system,and integrates it into the "Micro Things" IoT platform.The developed system reduces the workload of IoT data access by 20%,provided complex IoT combined services according to user needs,realized the sharing of IoT resources,verified the feasibility of the proposed method,and had been applied to "Smart Building","Smart Park","Smart Manufacturing" and other projects.