非预知物理层密钥的无线网络抗敌意干扰技术研究

Jamming attacks aim at disrupting the ongoing legitimate communications and can lead to Denial of Service (DoS) attacks, throwing serious threats to wireless networks. Jamming cannot be addressed by the higher-layer security techniques such as authentication and encryption. Based on the spread-spectrum techniques, such as direct sequence spread spectrum and frequency hopping, the traditional anti-jamming techniques require the pre-shared PHY-layer keys at the receivers. The PHY-layer key is the spreading pattern of the transmission, such as frequency hopping pattern and spreading sequence. The requirement on the pre-shared keys makes wireless networks vulnerable to smart jammers with strong jamming strength. As the impacts of jamming attacks increase with the development of wireless networks and Internet of Things, it is critical to investigate the anti-jamming techniques in wireless networks. .Therefore, in this project, we will first develop the anti-jamming pair-wise communication strategy based on uncoordinated spread-spectrum techniques, and investigate the anti-jamming broadcast strategy with node cooperations. Then we will propose an anti-jamming scheme for wireless networks, removing the dependency on pre-shared keys. In addition, we will also explore the impacts of channel fading and node synchronization errors, and propose the corresponding solutions. Another research issue is to implement the proposed schemes on universal software radio peripherals (USRP), and build the testbed of wireless anti-jamming techniques. Finally, this project also involves performance evaluation of the wireless anti-jamming techniques against various types of jammers, based on theoretical analysis, simulations over MATLAB and NS2, and prototype implementations. .This work can significantly enhance jamming resistance of wireless networks such as wireless sensor networks and wireless local area networks, and thus improve wireless security. This project is especially important for the development of security sensitive applications in civilian and military networks, such as anti-terrorism, public safety, and emergent communications.

敌意干扰阻塞当前正常通信，可引发拒绝服务攻击，严重威胁无线网络安全。它实施难度低，且无法通过认证和加密等高层安全机制解决。传统抗敌意干扰方案基于扩频技术，要求接收机预知物理层密钥，因而可造成严重的安全隐患。随着无线网络和物联网的发展，以及干扰力强大的智能干扰机的出现，敌意干扰的影响范围和危害日益扩大，因此有关无线抗敌意干扰研究刻不容缓。本研究内容包括：基于非协调扩频技术，优化设计非预知物理层密钥的点对点通信和广播通信的抗干扰技术，进而提出通用的协作型无线网络抗敌意干扰技术方案；研究无线衰落和网络同步误差等因素的影响，并提出对策；在USRP设备上实现该方案，搭建无线抗干扰实验平台；采用理论分析、仿真和现场实测等方法，全面评估其抵御各类干扰机的性能。课题将有助于增强无线局域网和传感网等无线网络抵御敌意干扰的性能，提高其安全性，特别对防恐、维稳和公安应急通信等安全敏感型业务意义重大。

本项目旨在研究非预知物理层密钥的无线网络抗敌意干扰技术，用以抵御无线网络各类敌意干扰，弥补现有抗敌意干扰技术的安全漏洞和不足，从而增强无线网络的安全性，并在确保其抗干扰性能的前提下，有效提高无线网络的通信效率。.本项目主要研究内容包括：基于非协调跳频协作广播技术，在通用软件无线电平台上实现预先设定物理层密钥的抗敌意干扰协作广播方案，增强系统在认知无线网络中的抗敌意干扰性能。实验结果表明，在给定测试环境中，相对于传统的跳频广播系统，该方案使得广播时延和总能耗节约67%和58%；针对合作认知无线电网络和异构认知无线电网络，分别提出相应的功率控制方案，提高抗干扰性能。实验结果表明,当存在1个智能干扰机时，源节点和中继节点均采用所提方案，接收机收到的信噪比提高48.6%；提出智能网络攻击下的移动云计算安全防御方法，通过设计移动云计算中终端用户的安全数据卸载策略，有效抵御包括敌意干扰攻击在内的多种攻击模式；提出基于强化学习的水下传感器网络的抗敌意干扰方案，设计发送功率控制策略，提高通信信噪比和水下传感器网络的通信效率；采用博弈论对无线网络的敌意干扰等攻击行为进行建模，研究抗敌意干扰的通信技术，可有效抑制干扰机的攻击动机；针对智能干扰攻击，在GNURadio/USRP平台上设计实现基于强化学习的无线抗敌意干扰通信系统，提高认知无线网络的信干比和通信效率。.课题发表文章34篇，包括著作1项，期刊论文16篇，会议论文18篇，硕士毕业论文7篇（SCI收录12篇，EI收录30篇）；获得6项软件著作权，2项发明专利获授权；毕业7名研究生；开展国际学术交流10人次。