Radio-telepathy

TLDR

Key establishment is difficult in mobile settings lacking infrastructure, making secure communications challenging. The paper proposes a protocol that lets two users derive a shared cryptographic key from the unique, rapidly decorrelating wireless channel response. The protocol extracts key bits by level‑crossing quantization of correlated stochastic channel responses, evaluated theoretically, numerically, and experimentally on 802.11 hardware. The protocol resists eavesdropping and spoofing attacks without authentication and achieves about 1 bit/s key rates in indoor 802.11 settings, also working with coarse signal‑strength measurements.

Abstract

Securing communications requires the establishment of cryptographic keys, which is challenging in mobile scenarios where a key management infrastructure is not always present. In this paper, we present a protocol that allows two users to establish a common cryptographic key by exploiting special properties of the wireless channel: the underlying channel response between any two parties is unique and decorrelates rapidly in space. The established key can then be used to support security services (such as encryption) between two users. Our algorithm uses level-crossings and quantization to extract bits from correlated stochastic processes. The resulting protocol resists cryptanalysis by an eavesdropping adversary and a spoofing attack by an active adversary without requiring an authenticated channel, as is typically assumed in prior information-theoretic key establishment schemes. We evaluate our algorithm through theoretical and numerical studies, and provide validation through two complementary experimental studies. First, we use an 802.11 development platform with customized logic that extracts raw channel impulse response data from the preamble of a format-compliant 802.11a packet. We show that it is possible to practically achieve key establishment rates of ~ 1 bit/sec in a real, indoor wireless environment. To illustrate the generality of our method, we show that our approach is equally applicable to per-packet coarse signal strength measurements using off-the-shelf 802.11 hardware.