GMM Based Semi-Supervised Learning for Channel-Based Authentication Scheme

Abstract

Abstract-Authentication schemes based on wireless physical layer channel information have gained significant attention in recent years. It has been shown in recent studies, that the channel based authentication can either cooperate with existing higher layer security protocols or provide some degree of security to networks without central authority such as sensor networks. We propose a Gaussian Mixture Model based semi-supervised learning technique to identify intruders in the network by building a probabilistic model of the wireless channel of the network users. We show that even without having a complete apriori knowledge of the statistics of intruders and users in the network, our technique can learn and update the model in an online fashion while maintaining high detection rate. We experimentally demonstrate our proposed technique leveraging pattern diversity and show using measured channels that miss detection rates as low as 0.1% for false alarm rate of 0.4% can be achieved.

Deep Learning Approaches for Open Set Wireless Transmitter Authorization

“…1, for rejecting samples from new classes. Unlike feature based approaches [4]- [7] which use well separated features like CSI, our approach leverages the power of deep learning to learn the features that separate the seen classes from new ones.…”

“…For feature-based PLA, existing works have considered using transmitter fingerprints due to hardware imperfections [3] or channel state information (CSI) [4]. Learning approaches based on extracted features rejecting new transmitters have used Gaussian mixture models [4]- [7].…”

Supervised Learning for Physical Layer Based Message Authentication in URLLC Scenarios

“…In [12], two approaches based on Support Vector Machines and Linear Fisher Discriminant Analysis respectively, are presented that yielded acceptable results. A Gaussian Mixture Model based technique in combination with exploitation of the channel responses for different antenna modes is considered in [13]. Due to these promising results, the goal of our work is to further investigate ML based identification methods, especially such using supervised learning.…”

Security Solutions for Local Wireless Networks in Control Applications based on Physical Layer Security

“…Message Authentication based on Physical Layer Metadata Hypothesis Testing (NPHT), e.g. used in Xiao et al (2008), or more sophisticated machine learning algorithms such Gaussian Mixture Models introduced in Gulati et al (2013) can be utilized. Every detection algorithm has basically two main performance indicators, the detection probability P D and the false alarm rate P FA .…”