A double auction method for resource allocation on computational grids

Considering dynamic, heterogeneous and autonomous characteristics of computing resources in the computational grid environment and the flexibility and effectivity of economics methods applied to solve the problem of resource management, a modified double auction method for resource allocation on computational grids is presented. Firstly, a double auction framework for resource allocation in the computational grid environment is described, which consists of buyers, sellers and one market maker. Secondly, a modified double auction mechanism is presented, where the uniform auction is adopted aiming at CPU resources, and the transaction fee can be adjusted flexibly. Thirdly, the property of the presented double auction mechanism is discussed, and the presented double auction mechanism proves to satisfy dominant strategy incentive compatibility, budget balance, and individual rationale, in addition the efficiency of the double auction mechanism is defined. Finally, the efficiency of the presented double auction mechanism is analyzed through experiments, and experimental results show that the efficiency of the presented double auction mechanism increases as the number of buyers and sellers increases, and indicate that the presented double auction mechanism is suitable for large-scale grid computing systems. Moreover, the difference between the number of sellers and the number of buyers can also have influence on the efficiency.