基于在线异常感知的无线电网络抗SSDF攻击协作频谱感知框架  

Cooperative spectrum sensing schema against SSDF attack using online abnormal perception in cognitive radio networks

作　　者：杨震 齐兴斌 Yang Zhen;Qi Xingbin(Laboratory of Sichuan Equipment Manufacturing Industry Robot Application Technology Engineering,Deyang Sichuan 618000,China;School of Software,Shanxi University,Taiyuan 030013,China)

机构地区：[1]四川省装备制造业机器人应用技术工程实验室,四川德阳618000 [2]山西大学软件学院,太原030013

出　　处：《计算机应用研究》2018年第12期3791-3794,共4页Application Research of Computers

基　　金：2014年中国科学院"西部之光"项目;四川省科技厅科技计划资助项目(2014GPT0011)

摘　　要：针对认知无线电网络(CRN)通信可能会因为恶意次级用户(SU)的存在或信道损伤而遭到破坏等问题,提出了一种广义鲁棒的协作频谱感知(CSS)框架,频谱感知和恶意用户识别问题联合构成一种优化问题。提出的方案通过利用节点位置信息获得了性能的提升,利用在线异常感进行实时识别和跟踪,不仅能识别恶意用户和频谱机会,而且能处理不准确或丢失的距离测量和噪声SU报告。仿真实验使用蜂窝场景,在一个特定半径内包含一个主用户(PU)和多个SU。实验结果表明,提出的方案在协作感知中能够有效克服感知数据错误化(SSDF)攻击。此外,与基于主成分分析法(PCA)的鲁棒CSS相比,对距离估计误差不敏感,对攻击强度和恶意用户数的鲁棒性更强,同时可以灵活抵御多种网络攻击。Concerning that the communication of cognitive radio network（ CRN） may be failed because of the existence of a malicious secondary users（ SU） or channel impairments,this paper proposed the framework of a generalized robust cooperative spectrum sensing（ CSS）,in which spectrum sensing and malicious user identification problem are combined an optimization problem. The proposed scheme could improve the performance by utilizing the location information of the nodes,and used the online abnormal perception to real-timely identification and tracking,which could not only identify the malicious users and spectrum opportunities,but also could deal with inaccurate or missing distance measurement and noise SU reports. Simulation experiments using cellular scenarios,within a certain radius containing a primary user（ PU） and multiple SU. The experimental results show that,the proposed schema can effectively overcome the SSDF attack. And compared with robust CSS method based on principal component analysis（ PCA）,the distance estimation error of the proposed method is not sensitive,and the robustness of attack strength and malicious users is much stronger. Meanwhile the proposed method can be flexible to resist a variety of network attacks.

关 键 词：认知无线电网络 感知数据错误化攻击 在线异常感知 协作频谱感知 恶意次级用户 联合优化 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]