Research On Service Composition Algorithm Based On Intention In Internet Of Things

With the development of Io T business,the types of Io T services are becoming more and more complex.Io T service composition is the process,method and technique of combining various services into a value-added service in a specific way and according to a given logic.However,the service composition process will face two problems:(1)in the service discovery process,the demand description of the service requester does not match with the function description of the service provider,while the traditional matching method relying on the service function description is used,and the accuracy of service discovery is low;(2)the existing service composition algorithm is generic and not designed for specific Io T scenarios,resulting in the service composition results cannot meet the actual needs of users.This thesis addresses the above problems,as follows:(1)Intent-based algorithm for Io T service description and discovery is proposed.The intent class is introduced in the Ontology Web Language for Services(OWL-S),an Io T service description document,to solve the problem of mismatch between service description and service requestor’s demand description.The Context and Qo S Intentional Service Model(CQISM)is designed and stored in the OWL-S files,and matching service information with service requester demand information by designing service discovery algorithms based on intent,context and Qo S.The experimental results show that the proposed service description and discovery algorithm can improve the accuracy rate by 6.7% compared with the traditional service discovery method.(2)A multi-objective service composition model is designed for emergency handling services in the Internet of Vehicles scenario,and a set of service composition solutions satisfying service requestors’ needs is obtained by the Enhanced Multi-Objective Harris Hawks Optimizer(EMOHHO).Simulation results show that the proposed method in this thesis can obtain service composition solutions with better performance in a shorter execution time with specific performance metrics including generational distance,spread,inverted generational distance,and search time.(3)Based on the method in(1)above,the OWL-S semantic service description file is generated after extending the intentional service ontology,and service discovery is performed based on the intentional information to select a set of service instances with higher similarity for each service class in the service composition;the mathematical model proposed in(2)and the EMOHHO algorithm are used for service composition and optimization selection to generate the final composite service solution.In the service composition,the not extended OWL-S files and the extended OWL-S files are used as service description files in the service pool for service matching,and the service composition and optimization selection are completed by EMOHHO or other multi-objective optimization algorithms,and the experimental results show that the combination of the extended OWL-S files and EMOHHO algorithm can improve the convergence and the diversity of service composition results.