EAGER: Physical Layer Security for the Internet of Things

网络安全是当今技术开发和部署中最紧迫的问题之一，也是一个严重的社会问题。无线网络在这方面尤其具有挑战性。物联网是无线通信的一个新兴方面，预计将互联数千亿台设备，覆盖家庭、车辆和工业环境。这些设备将用于从自动驾驶汽车到医疗保健的各种应用。物联网应用的复杂性、广度和范围使其特别容易受到网络攻击，同时也使得使用传统方法保护其免受此类攻击变得特别困难。此外，这些设备的数量庞大、可用的功率低、组成这些设备的硬件有限，以及缺乏连接这些设备的传统基础设施，这也给在这种情况下使用传统的网络安全方法带来了严峻的技术挑战。这项研究旨在开发一种新的物联网安全范式，其中利用射频环境的物理属性来增强设备之间通信的安全性。这项工作代表了一种全新的物联网通信安全方法，对这些技术在更多安全敏感应用中的部署能力具有深远影响。因此，这项研究有可能在未来几年成为我们信息基础设施的主要组成部分的环境中改变网络安全。拟议的工作将解决物联网安全中的关键问题，这些问题基于物联网的定义方面：短包通信、大规模和广泛分布的部署以及大规模数据收集。物理层安全方法利用传输介质中的资源来保证安全通信，防止窃听者，是解决物联网安全所带来的挑战的有前途的解决方案。这项探索性研究将考虑物理层安全原则在物联网中应用的潜力。设想了三个主要目标：短包的安全传输；安全函数计算；以及物联网网络中保密容量的伸缩律。短包是物联网应用的关键部分，例如车载通信、警报系统和传感器网络。物理层安全方面的工作主要集中在经典的无限块长度的香农体制上，这种体制不适合这种应用。因此，了解短数据块长度机制中物理层安全的基本原理是将此类方法应用于物联网的关键一步。物联网还与非常大的分布式数据应用相关联，其中物联网终端是产生大量空间分布数据的传感器。在这种情况下，从这些数据进行可靠和安全的计算是物联网应用中网络安全的一个重要方面。因此，开发这样的技术是开发安全的空间分布式传感系统的关键一步。此外，伸缩规律一直是理解大规模无线网络(如物联网)功能的重要组成部分。确定此类网络中的保密容量如何扩展将有助于更好地理解物联网支持安全通信的基本能力，从而可以指导物联网应用的安全协议和编码方案的开发。

Cybersecurity is a one of the most pressing issues in technology development and deployment today, and is a serious societal concern. Wireless networks are particularly challenging in this regard. The Internet of Things is an emerging aspect of wireless communications, which is expected to interconnect hundreds of billions of devices, spanning home, vehicular, and industrial environments. These devices will be used for applications ranging from autonomous vehicles to health care. The complexity, extent and range of applications envisioned for the Internet of Things make it especially vulnerable to cyber-attack, while at the same time making it particularly difficult to protect from such attacks using traditional methods. Furthermore, the massive number of these devices, the low power available to them, the limited hardware of which they will be comprised, and the lack of traditional infrastructure to connect them, also pose severe technical challenges to the use of traditional methods of cyber security in this setting. This study aims to develop a new security paradigm for the Internet of Things, in which the physical properties of the radiofrequency environment are used to enhance the security of communications between devices. This work represents a completely new approach to communications security in the Internet of Things, which has far-reaching implications on the ability of these technologies to be deployed in a greater number of security-sensitive applications. Thus, this research has the potential to transform cybersecurity in an environment that is sure to become a major part of our information infrastructure in the coming years.The proposed work will address critical issues in Internet of Things security, which are based on the defining aspects of the Internet of Things: short-packet communication, massive and widely distributed deployment, and large-scale data collection. Physical-layer security methods, which exploit resources in the transmission medium to guarantee secure communication against eavesdroppers, are promising solutions to address the challenges posed in securing the Internet of Things. This exploratory study will consider the potential of physical-layer-security principles for application in the Internet of Things. Three main thrusts are envisioned: secure transmission of short packets; secure function computation; and scaling laws for secrecy capacity in Internet of Things networks. Short packets are a critical part of Internet of Things applications such as vehicle-to-vehicle communications, alerting systems, and sensor networks. Much work on physical layer security has focused on the classical Shannon regime of infinite block-length, which is not suitable for such applications. Thus, developing an understanding of the fundamentals of physical layer security in the short block-length regime is a critical step in applying such methods to the Internet of Things. The Internet of Things is also associated with very large distributed data applications, in which the Internet of Things terminals are sensors generating large amount of spatially distributed data. In such situations, reliable and secure computation from such data is an important aspect of cyber-security in Internet of Things applications. Therefore, developing techniques to do so is a critical step in the development of secure spatially distributed sensing systems. Moreover, scaling laws have been an important part of the understanding of the capabilities of large-scale wireless networks, such as the Internet of Things. Determining how secrecy capacity scales in such networks will lead to a greater understanding of the fundamental ability of Internet of Things to support secure communication, and can thereby guide the development of secure protocols and coding schemes for Internet of Things applications.