A Strategy-Proof Combinatorial Double Auction Mechanism for Cloud Resource Allocation

The resource allocation optimization problem based on combination auctions, in which cloud service providers provide multiple types of resources and cloud users request different combinations of resources, is still a challenge in cloud computing and edge computing. The diversity of demands and the complexity of resource combinations make the direct search for optimal solutions extremely difficult, and the mechanism design in the combination double auction scenario also faces more challenges. We propose a strategy-proof combinatorial double auction mechanism for cloud resource allocation (CBAA) model, which utilizes a greedy mechanism to solve the dynamic allocation of multidimensional cloud resources and designs a payment algorithm based on critical value theory, which ensures the truthfulness of all the participants and encourages the bidders to offer their prices in a truthful manner. This mechanism takes into account the utility of the cloud user and the cloud service provider and maximizes social welfare. The experimental results show that compared to other algorithms, CBAA improves social welfare by approximately 17%-20%, the utility of cloud users and cloud service providers by approximately 4.09%, and the cloud service provider's resource utilization by approximately 10%. The CBAA is shown to be incentive compatible and individually rational and to achieve budget equilibrium and economic efficiency.