Abstract: To improve the persistence of spectrum allocations in cognitive radio networks, a primary user spectrum-usage model based on the alternative renewal process is given, and the spectrum utilizable rate and stability are defined as parameters to evaluate the quality of the usable spectrums, on the basis of which a novel Proactive Spectrum Selection Algorithm (PSSA) is proposed. By estimating the spectrum availability metric and the probability of a spectrum opportunity satisfying application requirements of cognitive radio users, the PSSA selects a spectrum with higher availability and opportunity probability to access. It achieves a cross-layer optimization between PHY/MAC and application layers. Simulation results show that the PSSA significantly reduces the collision rate and that it also has been proved to be a well self-adaptive method by adjusting weights of spectrum parameters and the spectrum usability threshold.

Key words: cognitive radio network, proactive spectrum selection, spectrum availability, renewal theory, collision rate