Abstract: On account of the characteristics of cloud resources， a cloud resource allocation method and bidding strategy were proposed on the basis of the microeconomic method and support vector regression method. Firstly， the block ability was used to describe buyer demand and seller resources. Secondly， the system architecture of cloud resource was constructed， and according to the resource transaction history of each buyer， their bidding price was predicted by using support vector regression method. Based on the above， according to Nash equilibrium theory of incomplete information， simultaneous upward bidding auction was designed， and the final transaction price was formed. What’s more， cloud resources were allocated by the final transaction price. Simulation results showed that the proposed method and strategy were feasible and effective.

Key words: simultaneous upward bidding auction, price prediction, support vector regression, cloud computing, Nash equilibrium