Protecting the Quality-of-Service (QoS) of the primary user is very important in cognitive radio networks.In this paper

a practical constraint is considered that the available spectrum for the secondary users is limited in opportunistic spectrum sharing.An optimal pricing function is derived

which can provide QoS-guaranteed for the primary users and more revenue for the primary service provider.Moreover

some practical environment parameters (such as channel quality

traffic fluctuation) are taken into account in spectrum pricing.In dynamic scenarios

the price can converge to the optimal value by iteration.Simulation results show that the QoS of the primary user can be guaranteed in the proposed pricing

while the profit is not less than that in market-equilibrium-based price.