Optimal Cloud Resource Allocation Method Research Based On Internet Of Things

Following the Internet and mobile Internet,the Internet of Things has set off the third wave of information industry revolution.In the era of the IoTs,through the interconnection of all things,the information of the whole physical world is collected,transmitted,stored and processed,and ultimately intellectualized and intelligent through the rich application of the IoTs.As the core of the IoTs world,the cloud platform needs to manage and empower the devices of the IoTs,process the massive data generated by the IoTs,and initiate real-time response to various services and applications of the IoTs.Therefore,cloud platforms faces many challenges,one of which is how to efficiently and effectively allocate cloud resources in order to obtain the quality of service required by IoTs users and communications in the IoTs.Facing the challenge mentioned above,this dissertation establishes three math-ematical models those attack the resource allocation problem from the perspective of cost-performance and quality of service(QoS)respectively,and give corresponding effective algorithm to obtain the corresponding optimal resource allocation strate-gy.The main innovations and contributions of this dissertation are summarized as follows:·Optimizing cloud resource allocation with cost performance trade-off based on IoTs by the Feasible Direction Method.An issue of cost performance tradeoff stems from various limited resources in the IoT systems and the competing services requirements.From this point of view,within the constraints of resources and service demands,considering that the utility function of cloud users in the IoTs is smooth,nondecreasing,and concave,a non-concave non-linear optimization model is established to maximize the cost performance ratio of the IoTs.In fact,the optimization problem can be proved to be a quasi-concave maximization problem.Fur-ther,the original problem can be reduced to a pseudo-concave maximization problem and thereby identify a local solution,which is proved to be exactly the desired global one.We therefore propose a feasible direction method and design its corresponding algorithm to yield the desired solution.Finally,we provide a numerical example to demonstrate the theoretical findings of re-source allocation and the proposed algorithm for the cloud computing based on IoT systems.·Optimizing cloud resource allocation with cost performance trade-off based on IoTs by the primal—dual method.Considering the cost performance tradeoff of the IoTs,a non-concave optimiza-tion model is established to maximize the cost performance ratio of the IoTs in the case that the utility function of IoTs users is smooth,nondecreasing and non-concave under the constraints of resources.Aiming at this optimiza-tion problem,according to the theory of primal-dual,a nonlinear method is proposed to correct the classical function to eliminate the dual gap,and the corresponding duality theory is established.The second-order multiplier iter-ation algorithm(SOMI)is used to obtain the optimal solution of the original optimization problem and the convergence of the sequence generated by the algorithm is further discussed.In fact,this is an optimal resource allocation scheme in a distributed way.We treat the IoTs terminal and cloud link as dis-tributed processors to coordinately compute the solution of the dual problem.In this system,IoTs terminal select the amount of cloud resources to maximize their own cost performance ratio,and cloud link adjust each resource price to coordinate the decision of IoTs terminal,and this process just embodies the primal-dual idea.We also demonstrates the iteration and validate our theoret-ical results by a numerical example.further,we analyses the changing rules of cost,utility and their ratios in the process of interactive iteration calculation.·Optimal Resource Allocation in IoTs Cloud Platforms for QoS AssuranceWhen the IoTs develops to a certain extent,the various resources of cloud platform are enough,the prices of various resources are relative ly low and then the IoTs cloud users will be more concerned about the performance of the application.At this time,the problem to be considered is how to deploy cloud platform and how to allocate cloud resources effectively,which can not only ensure the quality of service(QoS)of IoTs cloud users,but also avoid unnecessary waste of resources.To face this challenge,we firstly establish a non-concave optimization model.According to the theory of Primal-du-al,the duality theory is established based on the nonlinear method,and the optimal resource allocation scheme is realized.We treat the IoTs terminal and cloud link as distributed processors to coordinately compute the solution of the dual problem by using the second-order multiplier iteration algorithm(SOMI).In this system,IoTs terminal selects the amount of cloud resources to maximize its own utility,and the cloud link adjusts each resource's capacity to coordinate with the decision of IoTs terminal,and this interactive process executes until the total utility of the system reaches the maximum.Finally,the numerical results are given and the effects of various parameters on the results are analyzed.The difficulty of solving the problems raised in this dissertation lies not only in the multivariate and non-concave nature of all objective functions,but also in the com-plex coordination among all potential IoTs users due to sharing of heterogeneous cloud resources.The approachs proposed in this dissertation are effective and feasi-ble,and can be believed to be applicable in IoTs cloud platforms directly,to deliver better service and to optimally manage the limited resources.