An Online Network Selection Method for Multi-Authorization Cognitive Radio Networks

The invention provides an online network selection method for multi-licensed cognitive radio networks. Network selection scheme design is carried out for the condition that prior knowledge of licensednetworks is unknown and is sensed imperfectly. A collision limited online network selection scheme is provided, and through limitation of collision probabilities between PU (Primary Users) and SU (Secondary User), the interference of the SUs on the normal transmission of the Pus is reduced. According to the collision limited online network selection scheme, the collision probabilities of the licensed networks can be effectively controlled within an upper limit range. Due to the limitation of a collision probability updating mechanism, idle channels in the licensed networks are ignored by theCRNs (Cognitive Radio Networks), and access opportunities are wasted. In order to further improve the throughput of the CRNs, the invention also provides a greedy online network selection scheme and the online network selection scheme based on the minimum collision probability. A simulation result shows that according to the two schemes, the throughput of a CRN system can be improved, the online network selection scheme based on the minimum collision probability is optimal, and moreover, a relatively low call drop rate is obtained.