Abstract: A distribution game-theoretic algorithm was proposed to improve the rationality and effectiveness of game strategy. First, a new price function was applied to the rare spectrum resource allocation, considering the quadratic-cost of primary users. Second, a dynamic trust value mechanism was designed to optimize spectrum sharing condition, punishing the secondary users by decreasing the trust value when they disturb the communications of primary users. Then, a hierarchical multi-period algorithm with price feedback was presented for the primary users and secondary users to get the maximum value and the spectrum utilization rate. Numerical simulation results show the rationality and effectiveness of the proposed algorithm.