An SMDP-based optimal admission control scheme in cognitive radio networks

Cognitive radio network is expected to solve the spectrum scarcity by enabling unlicensed secondary users (SUs) to opportunistically utilize the licensed spectrum unused by primary users (PUs). In this paper, we propose an optimal admission control scheme which maximizes the profit while guaranteeing QoS of the PUs. The profit maximization problem is formulated as a Semi-Markov Decision Process (SMDP) and its relevant components are derived. A Linear Programming (LP) algorithm is presented to obtain the optimal admission control policy. QoS provisioning for PUs is also considered by adding constraint to the SMDP formulation. Simulation results show that, the proposed optimal admission control scheme achieves more profit than the Complete Sharing and Threshold-based admission control scheme, and the blocking probability of PUs can be strictly constrained.