EAGER: Utilizing Contextual Channel Characteristics for Secure Communications in Wireless Ad Hoc Networks

设备之间的安全引导是在敏感和任务关键型应用中部署无线自组织网络的关键一步。大多数现有的安全引导解决方案都假定存在公共资源，例如附加硬件、辅助安全带外(OOB)通道或设备之间预先存在的信任。然而，在普遍存在的环境中，这些假设并不总是实用的。考虑到当今数十亿可用的无线设备，设备可能是异类的，硬件功能或用户界面高度多样化，并且公共资源有限。本课题对无线自组网中的自举安全问题进行了探索性的研究，不假设任何额外的硬件、OOB信道或预先存在的信任。为此，只有可自动用于任何无线设备的物理层通道特征用于安全引导。研究了设备移动性和位置等上下文因素对信道特性的影响；设计了基于上下文信道特性的认证密钥建立算法；通过理论分析和实际测试验证了安全自举算法和技术的有效性和效率；该项目的成功为在不同的自组织网络上下文中利用信道特征进行安全自举的可行性提供了指导；该项目生成的算法提供了非常轻量级的安全自举解决方案，可以部署在普适环境中的任何商业现成(COTS)设备上。该项目的研究成果将纳入教育，并以教程、讲座、出版物和在线软件工具包的形式传播。

Bootstrapping security among devices is a crucial step for deploying wireless ad hoc networks in sensitive and mission-critical applications. Most existing solutions to security bootstrapping assume the existence of common resources such as additional hardware, auxiliary secure out-of-band (OOB) channels, or pre-existed trust among devices. In pervasive environments, however, these assumptions are not always practical. Given billions of available wireless devices nowadays, devices can be heterogeneous and highly diversified in hardware capabilities or user interfaces, and have limited common resources. This project conducts an exploratory study on bootstrapping security in wireless ad hoc networks without assuming any additional hardware, OOB channel or pre-existed trust. To this end, only physical-layer channel characteristics, which are automatically available to any wireless device, are used for security bootstrapping. Contextual factors such as device mobility and location are studied regarding their impacts on channel characteristics; algorithms for authenticated secret key establishment based on the contextual channel characteristics are designed; the effectiveness and efficiency of the security bootstrapping algorithms and techniques are evaluated through both theoretical analysis and experiments on a real testbed.The success of this project provides a guideline on feasibility of utilizing channel characteristics for security bootstrapping in different ad hoc network contexts; the algorithms generated by this project provide extremely lightweight security bootstrapping solution that can be deployed on any commercial-off-the-shelf (COTS) devices in pervasive environments. The research results of this project will be integrated into education, and disseminated in the forms of tutorials, talks, publications, and online software toolkit.