Spectrum Access Algorithm Of Cognitive Radio Networks

Along with the wireless communication demand's unceasing development, the wireless frequency spectrum is getting scarcer, which is non-renewable resource. However, the FCC statistical data indicates that although certain wireless frequency bands have been crowded, still have the massive wireless frequency spectrum resources to be at the idling condition. Subject to static allocation regulations now, the people are unable to use those idle frequency spectrum resources flexibly. This is the primary problem under the current spectrum management system.Opportunity spectrum access is the tendency of radio technological development now, is breakthrough to the fixed spectrum management rule. The cognition radio (CR) is one kind of opportunity spectrum access, through carrying on the examination, analysis, learn, inference to the periphery wireless communication parameter's history and the current observation, adjusts its own transmission parameter based on the observation foundation, can thus complete the user's transmission duty in the appropriate frequency by the suitable modulation system and the emissive power. The cognition radio technology can raise the spectrum use efficiency effectively.Firstly, this paper gives the origin of cognitive radio technology, development, research status, and the definition and function of cognitive radio. The second chapter describes in detail the spectrum sensing and spectrum access technology and gives the major methods and models individually. In the third chapters according to specialty of cognitive radio, partially observable Markov decision-making process (POMDP) model is developed. Under the POMDP model, based on ML criterion estimation of channel transformation probability is designed; by analyzing the optimal access policy of POMDP, based on limited iterative method, the modified algorithm is proposed which keeps the performance and reduce the computable complexity. The fourth chapter includes a distributed POMDP model of cooperation. By interaction in the neighboring user information based on policy gradient methods to approach gradually global optimal access, the method can reduce the collision between cognitive users and improve the efficiency of spectrum.