Analysis Of Dynamic Behaviors In Network And Applications In Wireless Communication

Complex system, especially intelligent complex system, is widespread in thenature and the human society. Such complex system can be represented as com-plex network with interconnection of individuals. Among the dynamic behaviors inthe complex network, synchronization is particular active. Cognitive radio networkis the special network of complex network in the communication feld. Spectrumsensing and spectrum sharing are the key technologies in the cognitive radio net-work. This thesis is focus on synchronization in the complex network and spectrummanagement in the cognitive radio network.1. The analysis of the synchronization in complex network is frstly from thesynchronization of the inner nodes of a complex network and obtains a series ofsynchronization criteria. The frst part of this thesis discusses the complete syn-chronization in two coupled complex networks. The dynamic information of thenodes in the two complex networks can be shared, due to the two complex networksare coupled in the dynamic evolutionary process. Then, the synchronization crite-ria are presented. Here, the nodes in the two complex networks are diferent, thenumber of the nodes are diferent and the couple methods are diferent.2. The dynamic behavior of the nodes in the complex network is sometimesdriven by a driven system. Here, we study the driven-response network and presentthe corresponding synchronization criteria. Projective synchronization is consid-ered, which is due to the fact that the diferent scale factors gain diversity. Besides,we consider anticipate, exact, delay projective synchronization. Two synchroniza-tion criteria are presented for diferent conditions, and the network model has moreextendibility.3. The data generated by the chaotic nodes in the driven-response networkwith diferent scale factors is more complex and more irregular. These charactersprompt the chaotic data from the driven-response network is applied in the imageencryption. Security analysis show these chaotic data is more subtle, the encrypteddata is more complex and the decryption is more difcult.4. The special formation of complex network in the communication feld—cognitive radio network, one of the key technologies is spectrum sensing. Sensingthe locations of idle spectrum efciently and quickly to improve the spectrum uti-lization. In order to reduce the required sensing information, we adopt the matchingpursuit and orthogonal matching pursuit in the compressed sensing theory. Accord-ing to the specialty of the spectrum sensing, we improve it and present three efcientalgorithms based on the statistical character of the sensing information. Simula-tion results show the three algorithms can obtain better detection performance and faster detection speed. Besides, the convergence analysis of the adaptive orthogonalmatching pursuit algorithm is given.5. Game theory is active in spectrum management of the cognitive radio net-work. We adopt the dynamic evolutionary game to model the dynamic spectrummanagement. The original dynamic equation requires the information of the allnodes, which restrains its actual applications. Here, we frstly improve the originaldynamic equation, and then apply it in the dynamic spectrum sharing and het-erogeneous network selection. Simulation results show that the improved dynamicequation not only requires less information but also accelerates the evolutionaryspeed and converges to the equilibrium state quickly.