An access optimization algorithm for cognitive networks with imperfect spectrum sensing

Spectrum sensing has a major effect on spectrum access of cognitive networks. In the paper, considering imperfect spectrum sensing, opportunistic spectrum access (OSA) optimization is researched. Spectrum occupancy and spectrum sensing activity are modeled and analyzed with Markov process first, by which the no-collision probability between primary and secondary users at each channel is derived. Then a cross-layer performance index termed effective transmission capacity (ETC) is introduced by combining the no-collision probability at the MAC layer with the channel capacity at the physical layer. Under the criteria to maximize the ETC of secondary networks subject to a collision probability constraint, an access optimization model is proposed, which leads to an optimal OSA algorithm. The maximum ETC obtained by secondary users is therefore calculated. Simulation results indicate that the proposed OSA algorithm can enhance the achievable ETC of secondary networks.