Research On Resource Allocation Based On Double Auction In Cloud Environment

Cloud computing is a new type of computing method based on the Internet.Cloud computing abstracts massive computing resources into services through the virtualization technology such as KVM and the resource scheduling framework such as OpenStack.Cloud computing resource allocation is the focus of cloud computing research.In the research of cloud computing resource allocation,resource allocation theory based on market mechanism can effectively mobilize and motivate individual enthusiasm and rationally realize resource allocation,which is an important direction for cloud resource allocation research.There are various resource allocation models based on market mechanism.The combinatorial double auction model has high resource allocation efficiency and can allocate multiple types of resources at the same time,so that it is very suitable for cloud computing environments with large-scale,dynamic,heterogeneous and other characteristics,so that it become the research focus of cloud computing resource allocation.The bidder drop problem refers to the problem that in the combinatorial double auction model,the buyer with lower bid or seller with higher bid price withdraws from the auction market after the multi-turn auction fails,resulting in the auction market being monopolized by a few traders.In addition,the traditional combinatorial double auction model lacks security considerations and cannot satisfy the traders' requirements for quality of service of cloud computing.For the above two issues,the specific work of this paper is as follows:1.A combinatorial double auction algorithm(F-CDA)based on fair function is proposed.The F-CDA algorithm combines the fairness mechanism with the traditional combinatorial double auction algorithm(CDA),and it dynamically adjusting the bidder or asking price of the cloud trader through the fair function,so that the probability of the trader exiting the auction market is reduced,thereby maintaining the number of traders in the entire cloud market and avoid the cloud market being monopolized by a few traders.Experiments show that the F-CDA algorithm can effectively solve the bidder drop problem in the combinatorial double auction market.2.A combinatorial double auction algorithm(TF-CDA)based on trust and fairness mechanism is proposed.Aiming at the low security problem of the combinatorial double auction cloud resource allocation model based on price mechanism,the trust factor is added to the F-CDA algorithm.The TF-CDA algorithm considers the cloud resource allocation from the two dimensions of trust and price.TF-CDA algorithm eliminate malicious cloud traders by calculating the trust threshold of the cloud market,thereby improving the security of the entire cloud market,and TF-CDA algorithm calculate the equivalent price based on the trust function,positively motivating traders with higher trust in the cloud market,thereby improving the whole Quality of service in the cloud computing market.Experiments show that the TF-CDA algorithm can effectively eliminate malicious traders in the cloud market and encourage cloud traders with higher trust.