The competitive price model is applied in the analysis of spectrum sharing in cognitive radio networks. As the prices of shared spectrum are determined independently by primary users in practical application, to deal with the solution of equilibrium price without acknowledgement of current price strategies from other primary users, the iterative method is used to calculate price of the shared spectrum. In order to achieve better characteristics, an exponential iterative method for the competitive price based spectrum sharing model is proposed. It can achieve the desired spectrum price only with the prices strategies during last iteration. Simulation results show that the proposed exponential gradient iterative scheme can achieve the equilibrium price that maximizes the profits of primary users, and have better convergence performance than linear gradient iterative scheme.