On the overflow probability and cross-layer secondary user selection in spectrum leasing of cognitive radio networks

In this paper, we are interested in a cross-layered spectrum scheduling for the secondary network under the spectrum leasing process. The process selects a secondary user to access the released spectrum, independent of the user who has participated in the cooperative primary transmission. At first, a cross-layer analysis to derive a lower bound on the overflow probability of the secondary network is given. It will be the achievable optimal probability of overflow for the secondary network. The result leads us to introduce the cross-layer secondary user selection method which is based on α-algorithms. While physical layer based approaches conventionally take the channel state into account, the cross-layer secondary user selection method considers both the achievable rate and the queue-length of each secondary user. We show that the cross-layer approach can be considered as a sub-optimal solution to achieve the optimal probability of overflow. Simulation results are presented to validate our analysis and evaluate the efficiency of the proposed approach.