Abstract
The performance of classical security authentication models can be severely affected by imperfect channel estimation as well as time-varying communication links. The commonly used approach of statistical decisions for the physical layer authenticator faces significant challenges in a dynamically changing, non-stationary environment. To address this problem, this paper introduces a deep learning-based authentication approach to learn and track the variations of channel characteristics, and thus improving the adaptability and convergence of the physical layer authentication. Specifically, an intelligent detection framework based on a Convolutional-Long Short-Term Memory (Convolutional-LSTM) network is designed to deal with channel differences without knowing the statistical properties of the channel. Both the robustness and the detection performance of the learning authentication scheme are analyzed, and extensive simulations and experiments show that the detection accuracy in time-varying environments is significantly improved.
Highlights
College of Information Management, Beijing Information Science and Technology University, Department of Electrical and Computer Engineering, George Mason University, Fairfax, VA 22030, USA; Abstract: The performance of classical security authentication models can be severely affected by imperfect channel estimation as well as time-varying communication links
Once the Convolutional-LSTM network is trained, it is evaluated on the test set, and the authentication performance is measured in terms of its detection accuracy and false alarm rate, which are the following: Accuracy = Pr(Alice|Alice ∪ Eve)
The advantage of the Convolutional-LSTM network is that it does not require the statistical characteristics of the channel, and intelligent training can make the model adapt to changes in the environment
Summary
We consider the following channel model in a time-varying communication system, where Alice, Bob and Eve are at geographically different locations. It is reasonable to assume that the initial communication transfer between Alice and Bob is established, using the upper-layer protocol before the spoofer arrives, which allows the authenticator to estimate Alice’s physical properties. The main task of the authenticator, Bob, is to evaluate whether or not the source of a newly estimated message is from Alice by looking at the difference between the vectors. It is assumed that physical layer authentication starts at time t = 1, and the newly estimated vector HQ,t is appended to H A , and becomes the ( M + 1)st channel estimate. By learning the properties of complex time-varying channels, the reliability and security of the intelligent authentication can be provided for legitimate communications
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
