Study On Cloud Services Composition Based On Unified Ontology

This dissertation studies Cloud Services Composition based on a unified ontology.By using the proposed Unified Ontology,many approaches and algorithms are presented for Cloud Services Composition,Selection and Verification.Although there are many existing works carried on this range,there is still a need to provide efficient approaches for Cloud Services Composition and Selection.In addition,the result of these approaches,which are obtained in design time,need to be verified before execution in order to avoid serious problems,including the inconsistency of functionality and QoS Criteria.The dissertation studies the problems related to Cloud Services Composition,and the main contributions are listed as follows:(1)Proposing a Unified Ontology based Approach for Cloud Service Selection.This approach includes a Service-Clustering algorithm,a Functional matching and a Non-functional methods.Service-Clustering algorithm is designed based on the Unified Ontology.This algorithm groups similar functional services into Cloud Service clusters to facilitate functional matching process.The Functional Matching and Non-Functional Methods are proposed to select the suitable Cloud Service,which meets Cloud Services consumer's requirements.In details,Cloud Services are matched based on functionality,then the candidate services are filtered based on Cloud Service characteristics,and they are evaluated and ranked using a QoS based Local Selection Method.A case study is provided to illustrate the implementation of the proposed approach.The proposed approach is evaluated and compared with similar works to prove its efficiency and performance.The related results and analysis showed that our approach present better performance comparing with existing similar approaches,which means that our approach is suitable for Cloud Services Selection.(2)Presenting a QoS Combinational Approach for Cloud Services Composition by combining QoS Global Optimization and QoS Local Selection Methods.In the former method,the number of the candidate cloud services of eachabstract services are reduced into a specific number of candidate services using a local services selection process.Then,composite services are generated and the global optimization process is performed to find the optimal composite service that satisfies the global constraints and has the maximum Utility Value.A case study is provided to implement this approach.In addition,the results of experiments showed that the proposed approach significantly outperformed existing works and reduced consuming time while finding Near-to-Optimal composite service and fits for Cloud Services QoS based Composition.(3)Presenting a Semantic-Aware Approach for automatic Cloud Services Composition.This approach automatically finds the most suitable Composite Service for satisfying a given request,which includes requested inputs,requested outputs.The proposed approach contains three steps.Firstly,an Inter-Connected Network of semantic cloud services is built using an efficient approach of the similarity(Outputs-Inputs similarity)between semantic concepts on the Unified Ontology.Secondly,a Forward-Chaining method is adopted in a composition algorithm in order to create composite services that functionally satisfy the request of the cloud service consumer(inputs and outputs).Finally,as several composite services may produce from the composition algorithm,a hybrid ranking method based on similarity quality and QoS Criteria is defined and used to select “ the most suitable ” composite service which is fed back to the Cloud Service Consumer.A case study is provided to implement the methods and algorithms in the proposed approach.The results of conducted experiments using our approach show better performance comparing to similar works.(4)Proposing a Colored Petri Nets based Algorithms for Verifying Services Composition.Colored Petri Nets model is defined in which,component services and their composition construction types in composite services were transferred from OWL-S to CPN using a tool called CPN tool.Two algorithms are proposed to verify the Functionality and QoS Consistency of services composition.The works of services composition in the chapter 3 and 4 are verified using the proposed algorithms.The verification process takes advantage of the proposed unified ontology as the functionality and QoS Criteria are verified using the values of QoS Criteria and inputsand outputs concepts,which are stored in the unified ontology.Two case studies are provided to implement the proposed algorithms.