Research On Competition And Interference Avoidance For Cognitive Radio Dynamic Spectrum Access

In the past decade, the concept of cognitive radio has offered a revolutionary per-spective in the development of modern communication. Cognitive radio, which can be implemented through software-defined radio, is aware of channel conditions, energy effi-ciency, bandwidth availability, locations, spectrum usage, the connectivity of a network, etc. And cognitive radio can achieve optimal frequency spectrum usage by adaptively changing its operating parameters on the basis of interactions with the environment and users. Also, the concepts of cognitive radio has been introduced to enhance the efficiency of frequency spectrum usage in next generation wireless and mobile computing systems.Dynamic spectrum access is the key approach in a cognitive radio network which is adopted by a cognitive radio user to access the radio spectrum opportunistically. The in-terference caused by cognitive user in spectrum accessing process is a problem which can be studied from three aspects.1) Rules can be made by primary user to manage the spec-trum access approach for cognitive users, and reduce the interference caused by cognitive users.2) Spectrum sensing is used by cognitive user to avoid collisions with primary user or the other cognitive user.3) Spectrum Resources can be shared between cognitive user-s through competing and cooperating. According to the aspects mentioned above, this dissertation mainly analyzes the spectrum competition and collision avoidance problem. According to game theory and optimal algorithm, methods of solution are provided.First of all, a cognitive period optimization algorithm based on interference com-pensation pricing and a hybrid spectrum access management algorithm is proposed for the primary user, in order to correct the interference caused by cognitive users to share the primary spectrum. The target of traditional cognitive period optimization algorithm is to maximize the average throughput of cognitive user and the demands of primary user are neglected. The cognitive period optimization algorithm based on interference compensa-tion pricing enables primary user to adjust the cognitive period according to the spectrum utilization of primary user. Moreover, the tradeoff between the interference probability for cognitive user and the channel idle probability is considered. The hybrid spectrum access management algorithm is provided to manage the spectrum access approaches for cognitive users. Since the spectrum overlay approach and spectrum underlay approach interferes the primary user in different way. The spectrum prices can be adjust by the hybrid spectrum access management algorithm according to the spectrum utilization of primary user. Then, the cognitive users will choose a better spectrum access approach.Secondly, a time-weighted spectrum sensing optimization algorithm is proposed for cognitive user to reduce the interference to primary user. And an asymmetric collision avoidance strategy based on spectrum sensing game model is proposed for cognitive us-er to reduce the interference to the other cognitive user. The time-weighted spectrum sensing optimization algorithm is proposed to improve the performance of cognitive user when primary user accesses the current channel suddenly. The signal samples are divid-ed into several groups and performed linear weighting time-sequentially. Data Fusion and decision Fusion is used to calculate the data groups. The results show that the time-weighted optimization decision fusion algorithm achieves better detection performance in the case of the same number of groups. Meanwhile, time-weighted optimization algo-rithm will significantly improve the detection probability when primary user accesses the current channel suddenly. Spectrum sensing game model is used to simulate the selfish behavior of cognitive user in a throughput competition. Based on the spectrum sensing model, an asymmetric collision avoidance strategy is provided to optimize the spectrum access probability, and share the channel with this selfish cognitive user.Finally, an exit strategy and an asymmetric active cooperation strategy is proposed to optimize the long-term profit of cognitive user in the spectrum sharing model. In spectrum sharing model, the independent cognitive users compete for spectrum. In the equilibri-um, both cognitive users will spread their power over the entire spectrum. Therefore, the independent cognitive users interfere with each other. The exit strategy specifies an exit time with different spectrum sensing result. Comparing with "interfere-exit" model, according to the exit strategy, cognitive user reduce the number of Channel switching ob-viously. Otherwise, an asymmetric active cooperation strategy is provided for cognitive user to optimize the long-term profit. One cognitive user can take the cooperative behav-ior actively, and the other cognitive user take the cooperative behavior later, then mutual cooperation can be builded. The asymmetric active cooperation strategy is a collusion which is implicit in which it can be enforced without the use of protocol.