Cognitive radio (CR) is a key technology that can enable opportunistic spectrum access, which enables secondary users (SUs) to dynamically exploit the under-utilized channels in the licensed spectrum, owned by primary radio networks (PRNs), referred to as dominant firms. Such sharing is subject to interference, SU QoS and cost constraints, in which SUs should not introduce harmful interference to PR users, achieve QoS rate demand and pay a price for using the licensed PR spectrum. The price of accessing idle PR channels depends on the level of channel utilization and price paid by PRs to access the channels, while the amount of needed spectrum to serve the rate demand of each SU heavily depends on the link-quality of the various channels. In this paper, the spectrum assignment problem in a CR network (CRN), referred to as follower firm, is investigated with the target of serving the largest possible number of SUs with the least possible total price paid to the PRNs (highest CRN profit) while being aware of the time-varying achieved transmission rate and level of utilization of the various PR channels. The problem is mathematically expressed as an optimization problem with the goal of maximizing the number of served SUs and the profit made by the CRN, which has been shown to be a binary linear programming (BLP) problem. Due to the high complexity of solving such optimization, we use the well-known sequential-fixing optimization method to obtain sub-optimal solutions. Simulation results indicate that our channel-assignment optimization significantly increases the CRN profit by reducing the price paid to the PRNs while achieving comparable performance offered by previous price-unaware protocols.
