Research On Common Control Channel Establishment And Spectrum Mobility In Cognitive Radio Networks

With the rapid increasing of wireless communication traffic, the spectrum resource scarcity has been a serious problem of restricting wireless communication technologies. Cognitive radio(CR), which is known as an intelligent promising technology for resolving spectrum resource scarcity in future, is gradually mature with the concerted efforts of researchers based on the development for more than ten years. Since the definition of CR is proposed, researchers focus on how to detect and utilize the idle licensed spectrum accurately in order to access to the licensed spectrum without interfering primary users. Although CR technology has made a lot of breakthrough progress in different protocol layer, there are still many issues to be further addressed. This paper focuses on two key technology problem common control channel design and spectrum mobility based on support vector machine and maximal independent set in graph theory. And solving schemes are presented accordingly.Firstly, in order to enhance ATTR performace for common control channel rendezvous algorithms, an indefinite length(IL) channel rendezvous algorithm is proposed this paper based on the analysis and comparison of the existing channel hopping schemes. The applicability of the new algorithm is strong, which makes IL algorithm can be used to setup a new link for users under the symmetric model and the asymmetric model. In addition, we prove IL algorithm can guarantee rendezvous for two cognitive users in finite time. And IL algorithm is not restricted by its application model. Simulation results show that the proposed IL algorithm not only enhances ATTR for CUs under the symmetric model, but also obtains better ATTR performace than that of other classic schemes under the asymmetric model especially when the common available channel number of CUs is big relatively.Secondly, in order to enhance spectrum availability prediction performace in mobile CR network, a support vector machine based spectrum availability prediction algorithm is presented based on the analysis of the existing spectrum mobility literatures. The proposed prediction algorithm analyzes the working activities of primary users and the mobility of cognitive users together. And the joint feature vector of spectrum availability, which is the input vector for support vector machine, is extracted based on the theoretical analysis. Thus, the ultimate prediction expression of spectrum availability is obtained. The new algorithm with a fast convergence speed gains better short-time prediction accuracy rate and miss prediction rate performance than the two algorithms just depending on the location and speed information, which are very vital to the real CR network deployment. Additionally, a rational parameter design can remedy the prediction performance degradation caused by high speed SUs with strong randomness movements.Lastly, in order to take the time-space domain spectrum mobility characteristics and the cognitive user mobility into consideration simultaneously for the mobile CR network, a prediction based backup channel spectrum handoff scheme is proposed. Based on the analysis of the interference restriction for different cognitive users, the new handoff algorithm utilizes maximum independent set theory to achieve fairness by assigning the same channel to different cognitive users during the backup channel selection. Simulation results validate that the presented spectrum handoff algorithm obtains higher throughout and fairness performances and lower collision rate than the traditional handoff schemes. In addition, considering cognitive users should be allocated an available idle channel before spectrum handoff, this paper designs a prediction based spectrum allocation mechanism. According to spectrum utilization maximization rule and fairness rule, PGA and PFA are presented. Simulation results show PGA and PFA gets better fairnes performances than GA and FA, respectively.