Date of Award
1-1-2025
Document Type
Dissertation
Degree Name
Ph.D. in Engineering Science
First Advisor
Ramanarayanan Viswanathan
Second Advisor
Lei Cao
Third Advisor
Mustafa Matalgah
School
University of Mississippi
Relational Format
dissertation/thesis
Abstract
With the promising use of Intelligent Reflecting Surface (IRS) for enhancing the security in wireless communication, this dissertation explores secure wireless communication facilitated by IRS to improve the secrecy rate of the system. We consider a multiple-input single-output (MISO) communication system supported by the IRS in the presence of a single legitimate user (Bob) and an illegitimate user (Eve).
First, we focus on maximizing the secrecy rate of the system through joint optimization of transmit beamforming and the phase shifts of the IRS elements. We examine two methods: (1) Method 1, which directly optimizes the secrecy rate as the objective function under the assumption that all channel state information (CSI) is available, and (2) Method 2, which aims in maximizing the rate for Bob when Eve's CSI is unavailable at the base station (BS) and eventually computes the secrecy rate. Method 2 is more practical compared to Method 1, as in practice, eavesdroppers typically operate covertly and do not disclose their information to the system they intend to intercept. For both methods, the block coordinate descent method is used to solve the respective optimization problems. We aim to investigate if the secrecy rate from the latter method, with only the CSI of the legitimate user, would perform similarly to the former case where all the CSI of the legitimate and illegitimate users are available. Simulation results show comparable performance primarily at low transmit power but as transmit power increases, the performance gap is observed, and Method 1 outperforms Method 2. We also examine the impact of IRS deployment strategy and Eve's position on secrecy performance.
Next, we explore the IRS-assisted system with Artificial Noise (AN) for the improvement of secrecy rate when Eve's CSI is not available for optimization, i.e., Method 2. AN is generated by the BS using a portion of the total transmit power to strategically degrade the signal reception at the illegitimate user while preserving Bob's signal reception quality. Since eavesdropper's CSI is unknown for AN design, we propose projecting AN onto the null space of Bob's channel. A fixed power splitting strategy is used for dividing the total power between signal transmission and AN generation. Two scenarios are examined: one with Bob and Eve positioned close to each other, and another with Eve situated farther away from Bob. Simulation results show that AN can either enhance or degrade the secrecy rate depending on the scenario and power levels at BS. In some cases, the reduction in power allocated to Bob's signal outweighs the benefit of added interference to Eve causing lower secrecy rate.
Lastly, we examine the phase direction of the optimized transmit beamforming and IRS elements. For this, we first compare the phase of the beamforming vector and the phase of the IRS elements for two different power levels, and then, show the result in terms of Bob's rate and secrecy rate by using the optimized phase values obtained with a specific power level. Our findings indicate that when the position of the legitimate user remains constant, the extensive optimization process does not need to be repeated with variation in transmit power. Simulation outcomes validate that the communication rate remains unchanged even when optimized parameters derived from one power setting are applied to another.
Recommended Citation
Pradhan, Babita, "Secure Wireless Communication with Intelligent Reflecting Surfaces" (2025). Electronic Theses and Dissertations. 3363.
https://egrove.olemiss.edu/etd/3363