Analysis of cognitive radio spectrum access with finite primary users and infinite secondary users

Abstract

In the cognitive radio networks, the users' spectrum access behavior determines the system capacity and spectrum utilization. There are three classical schemes for cognitive radio spectrum access: the random channel assignment, the reservation based assignment and the non-random channel assignment. In this paper, we study the performance of these schemes in a new scenario: a cognitive radio network with finite primary users and infinite secondary users. We model the users' behavior in these schemes by Markov chains. Moreover, the blocking, transition and dropping probabilities of the secondary users based on the user connection rate are derived as the performance metrics. Furthermore, we evaluate the secondary system capacity of these schemes. Numerical results show that the random and non-random channel assignments can get the lowest blocking probability, the reservation based assignment can get the lowest dropping probability, and the non-random channel assignment can provide the most system capacity and avoid the channel handoff for the secondary users.