Abstract: We built and analyzed the model of the cognitive radio system with buffering by continuous time Markov.Aimed at the character that the access of overmany users will result in the forced termination of the service, we designed an algorithm to control access probability of the cognitive users, which maximizes system throughput in the precondition of satisfying the forced termination limitation. The simulation result shows that, in the system using the algorithm, the user access probability and system throughput increase with the forced termination limitation. Also the largest access probability under the forced termination limitation is the access probability which can maximize the system throughput.The system satisfies the forced termination limitation at the sacrifice of throughput At the same time, the simulation indicates that the introduction of buffering can improve the user access probability and the system throughput performance.

Key words: Cognitive radio, Buffering, Access control, Markov chain, Forced termination probability