Research On Cross-layer Utility Optimization And Resource Allocation Of Enterprise Application Cloud Deployment

In the context of the rapid development of cloud computing technology and the acceleration of digital transformation of Chinese enterprises,many enterprises have migrated or will migrate their applications to the cloud.As a new service model,cloud computing has changed the business development of enterprises in the past years and brought benefits to enterprises with related service needs,but the current research does not analyze the the entire utility of cloud computing system from the perspective of the relationships among Software as a Service(SaaS)users,Software as a Service(SaaS)providers and Infrastructure as a Service providers.In this thesis,from the perspective of the mutual utility between SaaS users,SaaS providers and IaaS providers,an optimization model is established for the service quality optimization and resource allocation involved in the cloud deployment process,and is analyzed by using convex optimization theory.The main work of this thesis is as follows:First of all,in the interaction between SaaS users at the service layer and SaaS providers,the goal of the SaaS provider is to minimize the payment to the underlying IaaS provider,and they hope to maximize themselves under the premise of ensuring the quality of service of the SaaS users.The SaaS users pay for SaaS providers under the constraints of maximizing service quality.In this thesis,a service quality optimization model that maximizes utility is established at the service level.The optimal solution of the quality optimization is analyzed and a sub-gradient based algorithm is proposed to solve the convex optimization problem.Then the specific implementation steps of the algorithm are also given.Secondly,in the interaction between the SaaS provider and the IaaS provider at the resource layer,the utility relationship between the SaaS provider and the IaaS provider is used to maximize the applications utility of resource allocation model in the case where the elastic service and the inelastic service coexist.The model is analyzed by applying the convex optimization programming theory,and the non-convex optimization problem is transformed into a convex optimization problem by using Jensen's inequality,and the approximate optimization problem when multi-class services coexist in resource allocation is obtained.At the same time,a sub-gradient-based resource allocation algorithm was proposed,and the specific implementation steps of the algorithm were also presented.Finally,simulations are performed to verify the performance of the proposed cross-layer utility optimization model and resource allocation algorithm for enterprise application cloud deployment.