Date of Completion

5-24-2016

Embargo Period

12-24-2016

Keywords

Physical Layer Security, Underwater Acoustic Communicatons, OFDM, Sparse Channel Estimation, Secret Key Generation, Jamming

Major Advisor

Rajeev Bansal

Associate Advisor

Peter Willett

Associate Advisor

Shalabh Gupta

Field of Study

Electrical Engineering

Degree

Doctor of Philosophy

Open Access

Open Access

Abstract

Physical layer security has been under extensive investigation in recent years in wireless radio communications. However, its study in the context of underwater acoustic (UWA) communications is very limited. This dissertation will explore the fundamental properties of UWA channels to achieve physical layer security. It includes three research topics: 1) Channel estimation in UWA systems leveraging the inherent channel sparsity; 2) Secret key generation through the reciprocity of UWA channels; 3) Self-protection jamming in half-duplex systems leveraging large propagation delays.

The first part of the dissertation deals with sparse channel estimation in UWA orthogonal frequency division multiplexing (OFDM) systems. By exploiting the sparse nature of UWA channels, compressed sensing (CS) based channel estimation methods have demonstrated superior performance compared to conventional least-squares (LS) methods. However, a priori information of channel sparsity level is required to set the regularization parameter properly. We propose a data-driven sparsity learning approach based on a linear minimum mean squared error (LMMSE) equalizer to tune the regularization parameter for OFDM transmissions.

The second part of the dissertation focuses on secret key generation in UWA channels. Predefined secret keys are often used to encrypt information. However, they could be leaked to eavesdroppers. A key generation protocol is presented where secret keys are dynamically generated by quantizing the measured amplitudes on OFDM subcarriers, and then using error correction codes for secret bits extraction according to the Slepian-Wolf coding principle. By analyzing the performance based on collected field data, an improved key generation protocol is proposed by incorporating two modules to increase the channel correlation and deal with channel dynamics.

The last part of the dissertation presents a self-protection jamming approach for block transmissions in half-duplex UWA systems. Different from existing approaches, where additional helpers (e.g., relays) are needed to transmit jamming signals, the proposed protocol does not need any helper but instead relies on the legitimate receiver itself. This approach exploits the half-duplex nature of underwater transceivers and the block-based transmission structure, by taking advantage of the large propagation delays to create interference at the eavesdropper without affecting the reception of the intended user.

COinS