SaTC: CORE: Small: Robust Physical Layer Security with Channel Knowledge Uncertainty

University Of Arizona, Tucson AZ

Investigators

Abstract

Wireless communication infrastructure is poised to make a significant impact beyond cellular networks, such as in emerging internet-of-things (IoT) networks, device-to-device communications, and autonomous systems. Such systems may not have the full-fledged infrastructure or capabilities to achieve confidentiality via traditional cryptographic approaches. Furthermore, the broadcast nature of the wireless transmission medium can make them prone to information leakage through eavesdropping. This project develops new approaches to achieve provable wireless security in the presence of uncertainty about the wireless channel. The results of this project can have a wide reaching societal impact through mechanisms that can enable light-weight, and provably secure communication over the physical layer, and contribute to trustworthy operation of wireless infrastructure for various emerging applications. In addition to the research impact, the project will have an educational impact through curriculum development and the training of undergraduate and graduate students, in addition to outreach activities through the University of Arizona Science, Engineering and Mathematics (ASEM) scholars program and the STEM learning center. Over the last decade, several significant and fascinating insights have emerged on how to leverage knowledge of wireless channels to achieve secrecy. Despite these advances, the practical viability of physical layer security faces obstacles, primarily due to the fact that channel knowledge is available imperfectly, and is almost always unavailable from adversarial entities. This project will lay the foundations for achieving secrecy in scenarios where the precise knowledge about the wireless channel is minimal or non-existent. The key underpinnings behind making wireless security robust to channel knowledge uncertainty are argued to be rooted in the fundamental properties of wireless channels: path loss, fading, and inter-symbol interference. Instead of viewing these as negative attributes of the wireless channel, this project will devise approaches that systematically leverage statistical heterogeneity in these properties to gain advantage against external adversaries, even without the precise knowledge of the wireless channel. In particular, the specific goals of the project are: (i) to leverage heterogeneity of inter-symbol interference to achieve secrecy with only statistical knowledge about the channel; (ii) to develop principled encoding/decoding mechanisms that guarantee secrecy with only topological knowledge about the network; (iii) to generalize these concepts to large-scale single-hop and multi-hop networks with untrusted nodes; and (iv) to obtain information-theoretic bounds on the corresponding secrecy capacities for the various models with statistical channel knowledge.