Truthful Multi-Combinatorial Double Auction Mechanism for Cloud Resource Allocation

Abstract

Efficient resource allocation in cloud computing environments is crucial for ensuring platform performance and service quality. To address the heterogeneity of cloud resources and the diverse demands of cloud users, this paper proposes a truthful multi-combinatorial double auction (TMCDA) mechanism for cloud resource allocation. The mechanism introduces resource weight ratios to reflect the relative importance and proportional relationships among different resource types, thereby enabling a more precise evaluation of heterogeneous resource combinations. Meanwhile, cloud users are allowed to submit multiple substitutable combinatorial resource requests, which increases allocation flexibility and improves transaction success rates. In the payment design, a critical price calculation method similar to the second-price auction is adopted to guarantee truthfulness and economic efficiency. By integrating resource bundle matching with weighted pricing, the proposed mechanism improves resource utilization and overall social welfare. Simulation results demonstrate that, compared with Truthful Combinatorial/Multi-unit multi-item Double Auction for Cloud computing (TCMDAC) and Combinatorial Double Auction Resource Allocation (CDARA), the proposed approach achieves at least 15% and 30% improvements in social welfare, respectively, highlighting its advantages in transaction fairness, incentive compatibility, and allocation efficiency.