Physical layer authentication based on channel information and machine learning

摘要

Most of the physical-layer authentication methods are based on the pre-knowledge of the channel information and the threshold is fixed before the communication process. In this paper, we propose a physical layer authentication method based on channel information and machine learning with a dynamic threshold in wireless communication environment. In this method, the threshold of the channel differences is determined by the latest system state according to the ε-greedy strategy. The simulation results illustrate the efficiency of this method.