The Research On Security Mechanism And Key Technology In Cognitive Wireless Network

Cognitive wireless Network (CWN) is dedicated to improve the spectrum resources utilization efficiency by ultilizing the idynamic spectrum resource sharing, to motivate heterogeneous network integration to enhance the network end to end performance, and to promote the network throughput, which is becoming one of the key technologies of the fifth generation communication system. CWN is a revolutionary technology to solve the problem of the shortage of spectrum resources, which is highly praised by scholars at home and abroad, has become the research hotspots in recent years. However, everything has two sides, cognitive wireless network brings new challenges while achieves the implementation of network capacity and the utilization rate of spectrum resources. The new facing security challenges are presented in cognitive features only. Cognitive wireless network not only has the inherent threat of traditional wireless network problems, but also has the characteristics of their own threats, such as cognitive ability, reconfiguration according to the related threats etc..Network security is one of the core problems of wireless network, and is the basic condition for the success of the promotion and application of cognitive wireless networks. Although scholars have studied lots of research in cognitive network security, but the systematic study of cognitive wireless network security is still facing many challenges. This dissertation mainly aims at the influence of security threats on the performance of cognitive radio networks, intrusion detection method, and the secondary user control measures after the attack is detected are studied, the main work and innovation are as follows.The first part is the analysis of the security capacity of cognitive wireless network. This conribution first analyzes the relationship between the network capacity and the node density in the network environment, and derives the transmission capacity domain in cognitive wireless network. Secondly, it studies the secrecy capacity while there are eavesdropping attacks in the network links, as well as the outage capacity while there are interference of attackers. Finally, considering the two kinds of active and passive attacks, it puts forward the security capacity of cognitive wireless network. Without considering the interference of the traditional network attack, the dissertation derive the closed form solution of the security capacity, and prove the relationship between different nodes under the threats and capacity performance.The second part is the research on attack detection technology of cognitive wireless network. The dissertation proposes a distribution collaborative detection scheme according to the characteristics of the attack in CWN. Firstly, the dissertation analyze and study the user misbehavior detection method from the physical layer, and design a novel frame format of spectrum sensing period. In order to improve detection probability, the dissertation present the double threshold decision shceme for cooperative sensing, and the simulation results show that the mechanism can significantly improve the detection performance. The dissertation also propose a method of attack detection based on channel parameter estimation to against the primary user emulation attack through utilizing the time-invariant characteristic of the shadow fading channel parameters which the attackers are difficult to imitate. Using the EM (Expectation Maximization, EM) algorithm that has lower complexity and better performance close to the maximum likelihood estimation, the dissertation compute the channel parameter and get the decision attribute. The simulation results show that the method with low computational cost for the detection accuracy, greatly enhance the detection probability.The third part is security management mechanism. The solution is how to deal with the attack users (or misbehaving users) while they are detected, hence, the dissertation put forward the security management mechanism based on trust judgment in this contribution. The first, the dissertation study the confidence modeling, and give the general attributes of confidence, and detailed describe the process of trust, which contains generation, characterization, measurement, interaction and update. Then, penalty mechanism based on trust value is given, which is used to control the misbehaving users access network. The dissertation divide the misbehaving users into three categories with different penalties according to the extent of the damage to the network, in this case, the dissertation can achieve effective management of misbehaving users. In addition, the dissertation also consider the network architecture and design the hierarchical structure of central control and clustering network, which reduce the burden of Data Fusion Center in the update process of trust value. Finally, the dissertation gives the process of security management mechanism. The simulation results verify the production rule of trust value, attenuation characteristic, cluster head selection criterion, punishment mechanism, safety performance test, and the network overhead. And the simulations prove that it can effectively protect the safe operation of the network under the security management mechanism.As the last, the dissertation summarizes our work and analyzes the prospect of the developing tendency aiming to further improve the performance of CWN security.