Efficient Physical Layer Spoofing Detection with an Autoregressive Model

The broadcast nature of the wireless medium makes it vulnerable to impersonation attacks. Traditionally, cryptographic techniques are used against these threats. However, these methods are computationally complex and demanding for lesser equipped devices. The randomness of the wireless channels can be exploited to achieve authentication at the physical layer. In this paper, we propose an efficient channel-based authentication rule to detect the presence of an illegal transmitter. We also build a hypothesis test and then develop an effective likelihood test. The performance of this method is analyzed by studying miss detection and false alarm rates under various parameters. We conclude that a spoofing detection strategy can be developed using adjacent channel coefficients.