认知无线电网络中的合作频谱感知方法及安全性研究

发布时间：2019-05-17 01:28

【摘要】：为了有效缓解当下日益紧张的频谱资源稀缺现状,认知无线电技术已经成为了一项至关重要的科学技术。它通过对授权频谱资源进行重复再接入操作,高效缓解了不断快速膨胀的无线接入业务和频谱资源紧缺现状这二者之间的矛盾。频谱感知是实现认知无线电技术的基础核心部分,是保证主用户免受有害干扰、提高可用频谱利用率的重要手段。但由于大前提是既要尽可能地提高频谱利用率,又要保证主用户能正常通信、免受其它有害干扰。所以如何优化频谱感知参数,使得花最小的系统开销获得最大的系统检测增益,是我们要研究的一种重点部分;同时在无线通信环境中,信号的传输容易受到路径损耗、阴影效应和多径效应的影响,而且感知节点本身可能主动发动对感知信息的篡改、伪造、窃听等多种攻击行为,从而导致错误的判决。故本文针对在一个更为实际的、不安全的感知环境下的合作频谱感知问题进行了系统的研究工作,取得的主要成果为:1.基于带有恶意节点的更为实际的频谱感知环境,研究了基于合作感知的频谱共享网络模型,次级用户将会根据合作感知结果动态地调整其发射功率。为了防止恶意节点对感知系统的感知性能造成严重影响,研究了如何进行合作感知以提高感知性能。在一定的检测概率和相关功率约束下,建立了一个以最大化次级网络的吞吐量为目标函数的优化问题。仿真实验首先突出说明了恶意节点数目对频谱感知的影响重大,同时还表明无论是否存在恶意节点,提出的算法均可有效地计算出最优的感知时间和发射功率,且在降低最大干扰功率限制和最大发射功率限制时,网络的吞吐量是增大的。2.针对一种拟人化的“智能”攻击模式,提出了一种基于信任值的有效防御网络攻击的合作频谱感知模型。通过实验仿真,相比较与其它两种极端情况(网络中无恶意的理想情况和带有恶意节点无防御的最坏情况),从虚警概率和平均吞吐量两个方面验证其检测性能,说明该合作方案通过动态调整各个感知节点的信任值,可以达到有效的防御网络中的恶意节点,保证整个网络的检测性能。
[Abstract]:In order to effectively alleviate the increasing shortage of spectrum resources, cognitive radio technology has become an important science and technology. By repeated reaccess of authorized spectrum resources, it efficiently alleviates the contradiction between the rapidly expanding wireless access service and the shortage of spectrum resources. Spectrum sensing is the basic core of cognitive radio technology, and it is an important means to ensure that the main user is protected from harmful interference and to improve the available spectrum efficiency. However, the premise is not only to improve the spectrum efficiency as much as possible, but also to ensure that the main user can communicate normally and be free from other harmful interference. Therefore, how to optimize the spectrum sensing parameters so that the minimum system overhead can achieve the maximum system detection gain is a key part of our research. At the same time, in wireless communication environment, the transmission of signal is easily affected by path loss, shadow effect and multi-path effect, and the perceptual node itself may actively initiate tampering with perceptual information, forgery, eavesdropping and other attacks. This leads to the wrong decision. Therefore, this paper systematically studies the cooperative spectrum sensing problem in a more practical and insecure sensing environment, and the main results are as follows: 1. Based on the more practical spectrum sensing environment with malicious nodes, the spectrum sharing network model based on cooperative sensing is studied. The secondary users will dynamically adjust their transmission power according to the cooperative sensing results. In order to prevent malicious nodes from seriously affecting the perceptual performance of the perceptual system, this paper studies how to carry out cooperative sensing to improve the perceptual performance. Under certain detection probability and correlation power constraints, an optimization problem is established, which takes maximizing the throughput of the secondary network as the objective function. The simulation results first highlight the great influence of the number of malicious nodes on spectrum sensing, and also show that the proposed algorithm can effectively calculate the optimal sensing time and transmission power regardless of the existence of malicious nodes. When the maximum interference power limit and the maximum transmit power limit are reduced, the throughput of the network is increased. 2. Aiming at a humanized "intelligent" attack mode, a cooperative spectrum sensing model based on trust value is proposed to effectively defend against network attacks. Compared with the other two extreme cases (the ideal case without malice in the network and the worst case with malicious nodes without defense), the detection performance is verified from two aspects of false alarm probability and average throughput, compared with the other two extreme cases (the ideal case without malice in the network and the worst case with malicious nodes without defense). It is shown that by dynamically adjusting the trust value of each perceptual node, the cooperation scheme can effectively defend against malicious nodes in the network and ensure the detection performance of the whole network.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925

【相似文献】