A Novel Spectrum Allocation Mechanism Based on Graph Coloring and Bidding Theory

Cognitive radios (CRs) have a potential to improve spectrum utilization by enabling users to access the spectrum dynamically without disturbing licensed primary users (PUs). In this paper, we consider the spectrum allocation in wireless networks with multiple selfish secondary users (SUs) and a spectrum pool formed by underutilized spectrum of primary users. A novel spectrum allocation mechanism is proposed to obtain the overall utility of both primary and secondary users, in which graph coloring and bidding theory play an important role. Moreover, the traditional mechanisms such as distributed random and greedy algorithms are presented in this paper. The simulation results show that our proposed scheme not only have a similar performance to the optimal allocation algorithm, but also can obtain the profit of primary users.