Embracing interference in wireless systems

The wireless medium is a shared resource. If nearby devices transmit at the same time, their signals interfere, resulting in a collision. In traditional networks, collisions cause the loss of the transmitted information. For this reason, wireless networks have been designed with the assumption that interference is intrinsically harmful and must be avoided. This dissertation takes an alternate approach: Instead of viewing interference as an inherently counterproductive phenomenon that should to be avoided, we design practical systems that transform interference into a harmless, and even a beneficial phenomenon. To achieve this goal, we consider howwireless signals interact when they interfere, and use this understanding in our system designs. Specifically, when interference occurs, the signals get mixed on the wireless medium. By understanding the parameters of this mixing, we can invert the mixing and decode the interfered packets; thus, making interference harmless. Furthermore, we can control this mixing process to create strategic interference that allow decodability at a particular receiver of interest, but prevent decodability at unintended receivers and adversaries. Hence, we can transform interference into a beneficial phenomenon that provides security. Building on this approach, we make four main contributions: We present the first WiFi receiver that can successfully reconstruct the transmitted information in the presence of packet collisions. Next, we introduce a WiFi receiver design that can decode in the presence of high-power cross-technology interference from devices like baby monitors, cordless phones, microwave ovens, or even unknown technologies. We then show howwe can harness interference to improve security. In particular, we develop the first system that secures an insecure medical implant without any modification to the implant itself. Finally, we present a solution that establishes secure connections between any two WiFi devices, without having users enter passwords or use pre-shared secret keys. Thesis Supervisor: Dina Katabi Title: Professor of Computer Science and Engineering