Research On Resource Selecting And Scheduling Strategy In Cloud Computing

Cloud computing is a new business model and has got widespread concern in academia and industry since it was proposed.The core concept of cloud computing is resource pool known as "cloud".Commercial characters of cloud computing make the problem of resources selecting and scheduling more complex which has become the core issue and the hot topic of cloud computing.In this paper,the existing research of resource selecting and scheduling strategy in cloud computing is classified and analyzed in terms of optimization goals,scheduling model and technical methods firstly.Then analyzed deficiencies of the prior research status of performance-and-quality-oriented study and market-oriented economic mechanism study with a large number of references:1.The existing performance-and-quality-oriented study mostly inherited from related issues of grid computing research.Some of them consider task execution time only without concern of other attributes such as cost which is indispensable for cloud users.Some take both time and cost attributes into account,but the efficiency of the algorithm still needs to be improved.2.In recent years,with the rise of more and more cloud computing provider,market-oriented economic mechanism model and algorithm began to be proposed.This studies indicate that using simple economic model with fixed-price and direct supply is not conducive to cloud resources rational and equitable distribution.There is a lack of effective trading mechanism for two-way selection supported among cloud users and cloud resource providers leading to trade imbalance in cloud resources market.Compared with the former,research in this direction started quite late and authoritative.achievement is less.To solve these problems,this paper carry out depth study of cloud resource selecting and scheduling strategy.The main contributes and innovations are as follows:1.With time and cost objectives considered,a performance-and-quality-oriented strategy for cloud resource selecting and scheduling is proposed based on improved genetic algorithm.This strategy considerecd both time and cost optimization objectives.And to solve the premature problem of traditional genetic algorithm,improved genetic operators is introduced in this strategy.Finally,to verify the effectiveness of this strategy,CloudSim is.extended.Experiment is done to compare the proposed strategy with basic genetic cloud resource scheduling policy.The result shows that our strategy perform better in terms of task execution time and cost.In particular,when the number of tasks is 100,the decrease rate of execution time and cost is about 7.97%and 12.85%.2.A market-oriented economic mechanism for resource selecting and scheduling based on combinatorial double auction is proposed aimed to promote fairness of cloud resources trading and improve the user and provider's overall effectiveness.This auction model includes not only the price factor,but also the quality of service of cloud resources.Besides,a penalty mechanism is proposed in case of that cloud resource provider offer false QoS which is lower than promised to encourage them to provide more high-quality resources.Finally,to achieve and evaluate the proposed mechanism,CloudSim is extended.And comparative experiment with[33]is done to verify the effectiveness and superiority of the mechanism proposed in this paper.