Inter-Heterogeneous-Objects Dynamic Relationship Discovery Driven Agent Service Mechanism In Internet Of Things

Heterogeneity exists in Internet of Things(IoT),in which heterogeneous objects and inter-objects relationships are the most common.It causes the conflicts of inter-objects interaction in IoT,which is one of the inherent contradictions in IoT,and seriously hampers the future development of IoT.To solve this problem,based on characteristics analysis to inter-objects interaction data,a novel method of multi-agent based inter-objects relationships discovery in IoT and its application model are proposed.Using this method,not only the hidden physical inter-objects relationships surrounding human individuals can be mined,but also their niche structure can be generated combining with physical inter-objects relationships and their raleted individuals.To shield object heterogeneity in Niche,an abstract agent is correspondingly assigned to each object in IoT,is controled by the discovered inter-objects relationships,and finally provides dynamic services to the related human individual in IoT.In terms of implementation process,the propoed application model consists of three layers.Physical object layer:By inference analysis,individual-centric Niche finds the inter-objects relationships from individuals in itself,and uploads context data generated from object interaction scenario to IoT through sensors.System layer:Get and analyze uploaded context data,compute with the proposed inter-objects dynamic relationship calculation formula to obtain individual-centric dynamic relationships,and store it(rules)in rule base and ontology base to form service logics.Agent service layer:Assign an agent to its paired object in physical object layer,use similar isomorphic agent to interact with physical objects,map agent execution actions to its paired physical objects by constructed service logics,and then serve targeted individuals in IoT.Finally,a simulation experiment of the above-proposed service model is carried out in the context of six scenarios.To solve object heterogeneity,under each scenario and six-mixed scenarios,a comparison to traditional semantic techniques is done.The results show that the above-proposed method not only can meet user service demands in IoT,but also its service providing accuracy and conflict resolution in service process are all abbviously better than its compaired methods.