Research On Resource Collaborative Management And Allocation Mechanism In Mobile Cloud Environment

With the rapid development of Internet technology,the number of mobile users has grown rapidly.To meet the resource demand of the resource-limited mobile users,Mobile Cloud Computing came into being.The mobile users are often far away from cloud data centers,so the service quality of the delay sensitive mobile cloud computing tasks often can not be met.In this thesis,the widely distributed idle resource nodes with limited capacity in the network are considered to be integrated and provide mobile cloud computing services by forming a crowd-funding cooperative resource pool.In order to realize effective collaborative management of these idle resources,the following three key problems need to be solved:how to effectively motivate resource nodes in the network to actively participate in the crowd-funding system and contribute resources sustainably;how to customize the resource cooperation coalition and allocate resources for different mobile cloud tasks according to the specific demands;how to realize an optimized resource scheduling for the mobile cloud tasks to reduce the power consumption and improve the task execution stability.To solve these problems,full consideration was given on many related aspects such as resource revenue,service quality,stability of task execution,and power consumption.The main research contents and innovations of this thesis include:(1)A reasonable resource pricing mechanism has been proposed in this thesis.To encourage the resource nodes contribute their resources sustainably,the influence of future income is considered in this mechanism for the first time.Compared to the fixed pricing scheme,our resource pricing mechanism can realize the price adjustment in different task cycles.In addition,this resource pricing mechanism can motivate more resource nodes to participate in the crowd-funding system and provide resource service by allowing different resource price for different resource nodes with different service capabilities.(2)Two resource coalition customization mechanisms are proposed separately according to the benefits of resource providers and resource users.For the resource nodes are usually ability-limited,resource cooperative coalition need to be formed to meet the resource demand of mobile cloud services.Firstly,according to the benefit of the resource provider,a coalition customization mechanism for maximizing the resource revenue is proposed.This mechanism can increase the resource revenue of crowd-funding system and improve the resource utilization rate.Secondly,according to the benefit of the resource users,a coalition customization mechanism for maximizing the service level is proposed.This mechanism can effectively guarantee the service quality requirements of mobile cloud tasks.(3)A resource scheduling algorithm is proposed to optimize the power consumption and task execution stability.As the resource nodes in the crowd-funding system are always heterogeneous,there are significant differences in service performance among different nodes.The resource scheduling algorithm proposed in this thesis creatively can reduce the overall power consumption and improve the task execution stability at the same time,and this algorithm can adjust the degree of preference for different optimization goals as needed.In addition,we improved the traditional roulette wheel selection operator,which can effectively accelerate the convergence speed and reduce the algorithm execution time.