An area based physical layer authentication framework to detect spoofing attacks

Abstract

In this paper, we propose an area-oriented authentication framework, which aims to provide a light-weight first authentication by reducing the complexity in acquiring and maintaining many different reference vectors as in the traditional one-by-one authentication framework. Under the proposed framework, we first derive the missing detection probability and the false alarm probability, respectively. Then we quantitatively evaluate the average risks that a spoofer is successfully detected or a legitimate user is falsely alarmed, at any position in a certain area. And correspondingly three kinds of areas are defined as the clear area where the spoofers prefer not to attack, the danger area where the spoofers have pretty high probabilities to attack successfully, and the warning area where the legitimate users are much likely to be falsely reported as attackers. These results depict the security situation distribution of a region, and provide useful insights for network operators to take proper following strategies. Finally, simulations are given to verify our analytical derivations and show the impacts of system parameters.

Physical Layer Authentication in Wireless Networks-Based Machine Learning Approaches

“…Recently, many limited-capacity devices and users have demanded latency-sensitive services that require lightweight authentication techniques [ 9 ]. Meanwhile, physical layer security techniques have appeared that provide lightweight security solutions.…”