Novel enabling components for physical layer security powered by deep learning algorithms and metamaterial antennas
Citation & Export
Hide
Simple citation
Nooraiepour, Alireza.
Novel enabling components for physical layer security powered by deep learning algorithms and metamaterial antennas. Retrieved from
https://doi.org/doi:10.7282/t3-kjjp-4695
Export
Description
TitleNovel enabling components for physical layer security powered by deep learning algorithms and metamaterial antennas
Date Created2023
Other Date2023-01 (degree)
Extent185 pages : illustrations
DescriptionThe last few years have witnessed the release of a dizzying array of “smart and connected” devices by technology companies that seems to suggest that the internet-of-things (IoT)-driven revolution is just around the corner. One important challenge to tackle in this realm is providing efficient scalable security solutions for such devices. Physical layer (PHY) security has been put forth as an alternative to traditional higher-layer security approaches like cryptography methods that work based on the burdensome tasks of distribution and management of secret keys. In this work, we tackle the problem of securing wireless systems from the physical layer by fending off passive/active attacks. To this end, we devise novel secure methods and hardware prototypes where the former is powered by the state-of-the-art learning/statistical algorithms while the latter is built upon metamaterial antennas. The first part of our work revolves around physical layer spoofing attacks, where an adversary's goal is to impersonate a legitimate transmitter in the eyes of a legitimate receiver. These attacks which aim at injecting spurious data into the receiver, involve inferring transmission parameters and finding PHY characteristics of the transmitted signals so as to spoof the received signal. We focus on the non-contiguous (NC) orthogonal frequency division multiplexing (OFDM) systems that have been argued to have low probability of exploitation (LPE) characteristics against classic attacks based on cyclostationary analysis, and the corresponding PHY has been deemed to be secure. However, with the advent of machine learning (ML) algorithms, adversaries can devise data-driven attacks to compromise such systems. It is in this vein that PHY spoofing performance of the adversaries equipped with supervised and unsupervised ML tools is investigated in the first part. The supervised ML approach is based on estimation/classification utilizing deep neural networks (DNN) while the unsupervised one employs variational autoencoders (VAEs). Simulation results demonstrate that the performance of the spoofing adversaries highly depends on the subcarriers' allocation patterns used at the transmitter. In particular, it is shown that utilizing a random subcarrier occupancy pattern precludes the adversary from spoofing and secures NC-OFDM systems against ML-based attacks. In the second part, we propose novel transmission schemes, enabled by recent advances in the fields of metamaterial (MTM), leaky-wave antenna (LWA). MTM-LWAs, which offer compact, integrated, and cost-effective alternatives to the classic phased-array architectures, are particularly of interest for emerging wireless communication systems including IoT. The proposed secure schemes are devised to accomplish the functionalities of directional modulation (DM) to enhance physical layer security for OFDM and NC-OFDM transmissions, while enjoying the implementation benefits of MTM-LWAs. Specifically, transmitter architectures based on the idea of time-modulated MTM-LWA have been put forth as a promising solution for PHY security for the first time. To further reduce the implementation cost for achieving DM, we propose a space-time-modulated digitally-coded MTM-LWA that can enable PHY security. From the theoretical perspective, we first show how the proposed space-time MTM antenna architecture can achieve DM through both the spatial and spectral manipulation of the orthogonal frequency division multiplexing (OFDM) signal received by a user equipment (UE). Simulation results are then provided as proof-of-principle, demonstrating the applicability of our approach for achieving DM in various communication settings. To further validate our simulation results, we realize a prototype of the proposed architecture controlled by a field-programmable gate array (FPGA), which achieves DM via an optimized coding sequence carried out by the branch-and-bound algorithm corresponding to the states of the MTM LWA's unit cells. Experimental results confirm the theory behind the space-time-modulated MTM LWA in achieving DM, which is observed via both the spectral harmonic patterns and bit error rate (BER) measurements.
Finally, we investigate how the paucity of available training data affects the learning-based physical layer security solutions. We consider the fundamental task of classification, which has numerous applications for PHY security including spoofing detection, given a limited number of training data samples. We propose a hybrid classification framework---termed HyPhyLearn---is proposed that fuses the strength of both the physics-based statistical models and the learning-based classifiers to achieve superior classification performance. The proposed hybrid approach first estimates the unobservable model parameters using the available (suboptimal) statistical estimation procedures, and subsequently use the physics-based statistical models to generate synthetic data. Then, the training data samples are incorporated with the synthetic data in a learning-based classifier that is based on domain-adversarial training of neural networks. Numerical results demonstrate that the proposed approach leads to major classification improvements in comparison to the existing standalone or hybrid classification methods.
NotePh.D.
NoteIncludes bibliographical references
Genretheses
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.