Research On Truthful Multi-Requirements Auction-based Echanism For Virtual Resource Allocation And Pricing In Clouds

With the development of cloud computing,the market scale of cloud computing services has developed rapidly.It has become a research hotspot in the field of cloud computing to study on how to allocate and price cloud computing resources reasonably and efficiently.Currently,the mechanism based on fixed price method is widely used.However,this kind of pricing method is inflexible,which often causes the cloud resource price too high and cannot attract users to rent.As a result,a large number of virtual resources are idle.Therefore,this paper focuses on the auction-based mechanism for virtual resource allocation and pricing in Clouds The main research contents are as follows:This paper analyzes the current research of cloud computing resource allocation and pricing issues,and proposes a credible off-line multi requirements resource allocation and pricing algorithm(MRAM).The feature of this mechanism is that users can submit multiple requirements in an auction.We have demonstrated that under this mechanism,resource providers can obtain more revenue while ensuring that user bids are truthful.In addition,a monotonic heuristic algorithm is proposed for resource allocation.The core idea of the algorithm is to design the redistribution strategy through the concept of resource rarity,which can ensure the maximum benefit of the cloud resource provider;when designing the payment algorithm.The critical value theory was used to calculate the payment price,thus ensuring the fairness and credibility of the mechanism.The MRAM algorithm was tested and analyzed in many aspects such as social welfare,execution time,and resource utilization,and achieved good results.Because under the offline resource allocation mechanism,users often need to wait for a long time to use virtual resources,thus reduce the enthusiasm of users.Therefore,this paper proposes an online multi-demand resource allocation and pricing algorithm.We designed the heuristic algorithm OMRAM to find a viable allocation scheme.In the experimental part,we use CPLEX to program the optimal mechanism,and experimentally analyze both algorithms in terms of social welfare,execution time,resource utilization and users served.