QnAs with H. Vincent Poor

The proliferation of wireless devices in the last three decades has been heralded as a technological revolution. H. Vincent Poor, the Michael Henry Strater University Professor at Princeton University, has been studying these devices and the networks on which they run since their infancy, and has helped usher in the revolution. With a background in information theory, electrical engineering, and computer science, Poor has led formative research on signal processing, social networks, multiuser communications, and smart grids. PNAS recently spoke to Poor, who was elected to the National Academy of Sciences in 2011, about his current research.

• Download figure
• Open in new tab
• Download powerpoint

H. Vincent Poor. Image courtesy of David Kelly Crow (photographer).

PNAS:Your recent work examines security in the physical layer of wireless transmission. What is the physical layer, and why is security in this layer crucial?

Poor:Networks operate through multiple layers of infrastructure that encode, transmit, route, regulate, and decode information. Layering is a very important part of network design, and it is one of the reasons why networks are ubiquitous. The advantage of having layers is that you can innovate in one layer without having to redesign the whole network every time. The physical layer refers to the actual medium through which signals travel and carry information. In the case of wireless, this is basically the ether through which radio signals are transmitted.

Layered infrastructures help secure data transmission, but not all networks are created equally. Higher-layer security, which is used on networks such as the cellular network, simply isn’t practical in networks like mobile ad hoc networks, in which messages are passed from mobile terminal to mobile terminal until they reach their destination. In this situation, it is difficult and inefficient to manage conventional methods of network security. Similarly, the kind of devices that are envisioned for the so-called Internet of Things won't have a lot of computing power to implement encryption algorithms. Security in the physical layer ensures that information transmitted in these kinds of networks can be protected.

PNAS:What are some of the issues associated with incorporating physical security into the wireless layer?

Poor:First, we must understand that security is a whole portfolio of different issues. It’s not just the confidentiality of messages, but also user authentication, message integrity, and so forth. My Inaugural Article (1), coauthored with Rafael Schaefer of the Technical University of Berlin, summarizes the history and recent advances in wireless physical layer security, and also discusses some of the challenges, including practical security-code design, which is far from mature. Authentication is also a major practical issue. A set of questions also arises in how physical layer secrecy interacts with higher network layers. We’ve done some work along those lines, but there is a lot remaining. This is partly why I wanted to write this article for PNAS: so that others may be motivated to join this research effort.

PNAS:How can you use properties inherent in radio transmission, diffusion and superposition for example, to enable security in the physical layer?

Poor:The concept relates back to research conducted by Aaron Wyner at Bell Labs in the 1970s on the “wiretap channel.” Wyner was thinking about security in telephone lines, but the principles apply to wireless architectures as well. The combination of diffusion, which is essentially broadcast, and superposition, which is the addition of multiple signals in a medium, means that when you transmit a radio signal terrestrially it’s going to reflect off of intervening objects, and then when the reflections get to a receiver, they add constructively and destructively; this results in self-interference or fading. Fading depends on the geometry of the radio channel and so will differ from receiver to receiver. Fading fluctuations allow an eavesdropper’s receiver to be degraded over time, which provides opportunities for secure transmissions.

Another way that you can take advantage of the superposition property is a classical technique for degrading receivers, namely jamming, although we tend to use more sophisticated and surgical techniques these days. The use of multiple antennas also allows you to transmit in ways so that you simultaneously favor the intended receiver and degrade an eavesdropper.

PNAS:Are there systems that currently take advantage of the physical layer for security?

Poor:This is primarily a research field now, although it is beginning to move into practice. There are prototypes, and the concept has received attention from industry and start-ups.

PNAS:You began working on wireless networks in the early 1980s. Did you envision how ubiquitous they would become?

Poor:When I started in this field, the modern era of wireless was just beginning, but the cellular concept goes back to the 1940s. It wasn’t until the advent of smaller, more powerful and energy-efficient semiconductor devices that it was possible to think about developing the personal wireless communication devices we have today. For a 30-year-old technology, wireless technology has grown incredibly. It’s one of the most rapidly adopted technologies of our time or maybe of any time. It has become so much more than what people envisioned when it first emerged, and it continues to evolve at a very rapid pace.