The Research For Secure Service And Optimal Resource Management Of Mobile Cloud Computing Networks Based On Smdp

With the development of wireless communication technology, the technologies of resource optimization and management have been playing a very important role in the wireless communication network. As the limited bandwidth of wireless spectrum, how to use the existing wireless communication technologies and spectrum bandwidth to increase the mobile user’s quality of service (QoS) and bring the maximal system benefits for the entire wireless communication network has been a very important research topic in the wireless communication technologies.Cloud computing has become a calculation method based on the internet with the development of computing and communication technologies in recent years. Cloud computing resources include computing resources such as cpu and memory, etc, and communication resources. It has become an increasingly prominent research topics need to be solved in the cloud computing network to improve the quality of service (QoS) of cloud computing secure service and the entire system rewards of cloud computing network (including the overall rewards of both the cloud and the cloud computing end-users), thus to improve the revenue expenditure ratio of cloud computing network through the full use of existing cloud computing resources and the optimal allocation of different regions’cloud computing resources. Based on our current research, there are little previous literatures in this area, which is the motivation of this thesis to study the optimal management of mobile cloud computing resources for secure service as well.In this paper, we model and analyze the optimal resource management issues of mobile cloud computing network for secure service based on the Semi-Markov Decision Process (SMDP), and obtain the optimal secure service resource management decision-making strategies. Through our proposed optimal secure service resource management model, we do not only obtain the maximal overall rewards (including both the system and the clients) of mobile cloud computing network, but also improve the quality of service (QoS) of secure service. First of all, we study the dynamic resource allocation of wireless multimedia service in the wireless communication network. Based on the Semi-Markov Decision Process (SMDP), we propose a dynamic resource allocation model of wireless communication channel for elastic wireless multimedia service accrding to dynamic demands. Through the proposed model, we obtain the optimal decision-making strategy for dynamic resource management of wireless multimedia service. This optimal decision-making strategy can obtain the maximal rewards of wireless communication network by taking into account both the performance of wireless communication network and the occupation costs of the wireless communicaton channel resources. In this thesis, we further evaluate and verify the performance of our proposed model by simulation.Secondly, we study the optimal management of cloud computing resources for cloud computing security services, and first propose a Security Service Admission Model (SSAM) based on Semi-Markov Decision Process (SMDP) to optimize the allocation and management for the cloud computing resources of cloud computing security services. This model can consider both the incomes obtained by admitting the cloud computing security service requests by cloud computing network, and the occupation costs of cloud computing resources to provide the cloud computing security services for cloud computing end-users. Our proposed model can not only increase the overall long-term gains of cloud computing network, but also improve the quality of service (QoS) of cloud computing end-users. The experiment results prove the correctness of theoretical analysis.Thereafter, the optimal secure service resource management of single cloud computing service domain in the cloud computing network is investigated in this thesis. Generally, one cloud computing service domain can allocate one or more Virtual Machines (VMs) for cloud computing secure service to improve the computation speed based on the available cloud computing resources in this service domain. Therefore, it has become a very important research issue in cloud computing resource management to effectively allocate cloud computing resources for each cloud computing secure service, while improving both the overall benefits of cloud computing service domain and the quality of service (QoS) of cloud computing secure service. In this thesis, a new cloud computing resource management model is proposed based on Semi-Markov Decision Process (SMDP) to resolve this issue. The proposed model does not only take into account the income and expenditure of the cloud, more importantly, but also consider the income and expenditure of the cloud customer for the first time. The optimal decision-making strategy obtained by the proposed model can improve both the entire long-term rewards of cloud computing service domain and the quality of service (QoS) of cloud computing secure service. Theoretical analysises and experimental results show that the performance of our proposed model has improved significantly compared to that of the conventional Greedy Algorithm.Generally, cloud computing service domains are configured distributedly based on geographic locations. It has become another important research issue in cloud computing resource optimal management to achieve the full use of existing cloud computing resources and to improve both the entire rewards of the whole cloud computing network and the quality of service (QoS) of cloud computing secure service by allocating the cloud computing resources between multiple cloud computing service domains based on the existing cloud computing resources, which is not seen in the previous researches of cloud computing resources optimal management. Thus, in the end of this thesis, based on our previous researches on the optimal cloud computing resource allocation of single cloud computing service domain, we propose a new Semi-Markov Decision Process based (SMDP) model for the optimal management of cloud computing resources to achieve the optimal decision-making strategies of cloud computing resource management between multiple cloud computing service domains in a cloud computing network. Theoretical analysises and experiment results indicate that the optimal decision-making strategy of cloud computing resources management obtained by our proposed cloud computing resource management model can not only substantially increase the overall long-term gains of the cloud computing network which consists of multiple cloud computing service domains, but also significantly reduce the blocking probability of cloud computing secure service thus to substantially increase the quality of service (QoS) of cloud computing secure service, compared with the traditional Greedy Algorithm algorithm.