In this paper, an agile smart attacker model in spectrum sensing of cognitive radio network (CRN) is introduced. This smart attacker does not make the channel busy all the time, instead it senses spectrum and when a primary user leaves, it occupies the spectrum by mimicking the signal characteristics of the primary users. To model such a smart attacker, we use a dependent Markov chain to model the primary user and primary user emulation attacker activities, simultaneously. We derive the transition probabilities for the assumed dependent Markov model. Then, the effect of the primary user and attacker traffic on a secondary user's throughput under Rayleigh flat fading channel is investigated and closed form expressions are derived for the average probability of detection and false alarm. Furthermore, the impact of this attacker on the performance of a CRN and the throughput of the secondary user network is studied analytically. We also derive a test rule based on the generalized likelihood ratio test, to detect the legitimate user from smart illegitimate user. In addition, launching primary user emulation attacker in the conventional and proposed smart attacking procedures are compared, where the results demonstrate that by using smart attacking, the deterioration in throughput of the secondary user's network is considerable. In fact, it is shown that under proper selection of the traffic parameters, the secondary user's network throughput may tend toward zero. Finally, the accuracy of the obtained results are analyzed and verified by the simulation results.

中文翻译：

---

瑞利平板衰落信道下智能流量感知的主要用户仿真攻击及其对次要用户吞吐量的影响

本文介绍了一种用于认知无线电网络（CRN）频谱感知的敏捷智能攻击者模型。这个聪明的攻击者不会一直使信道繁忙，而是会感知频谱，并且当主要用户离开时，它会模仿主要用户的信号特征来占用频谱。为了对这种聪明的攻击者进行建模，我们使用了相关的马尔可夫链来同时对主要用户和主要用户仿真攻击者的活动进行建模。我们推导了假设的依赖马尔可夫模型的转移概率。然后，研究了瑞利平坦衰落信道下主要用户和攻击者流量对次要用户吞吐量的影响，并针对检测和虚警的平均概率推导了封闭形式的表达式。此外，分析地研究了此攻击者对CRN性能和辅助用户网络吞吐量的影响。我们还基于广义似然比检验得出检验规则，以从智能非法用户中检测合法用户。此外，比较了在常规和建议的智能攻击过程中启动主用户仿真攻击者的结果，结果表明，通过使用智能攻击，辅助用户网络的吞吐量会显着下降。事实上，它被示出，根据业务参数的适当选择，次要用户的网络吞吐量可能趋向于零。最后，通过仿真结果对所获得结果的准确性进行了分析和验证。