Student name: Muhammad Adnan

RUN-EU institution: Technological University of the Shannon: Midlands Midwest, Ireland

Abstract

Quantum computing (QC) has emerged as a revolutionary technology with the potential to solve complex problems that are difficult for classical computing. Quantum Machine Learning (QML) is a particularly promising field that combines quantum computing with machine learning techniques, aiming to enhance the performance and efficiency of algorithms. This research explores the competitive performance of quantum machine learning with cybersecurity aspects. The integration of Quantum Machine Learning (QML) with cybersecurity presents a novel approach to modern security challenges. Unlike traditional classical machine learning, which often struggles with complex cryptography and large-scale data issues, QML leverages the power of quantum computing to enhance cryptographic protocols, improve anomaly detection, and strengthen threat mitigation. This research will focus on different key areas such as quantum-enhanced anomaly detection, quantum-resistant and enhanced cryptography, cryptographic techniques, adversarial robustness, and the use of quantum algorithms for threat detection using hybrid quantum-classical frameworks and benchmarking and standardisation. By using quantum kernels and variational quantum circuits, QML shows potential in defending networks against advanced cyber threats while evaluating current advancements, limitations, and prospects in this field.

This research will adopt a hybrid research approach that integrates theoretical analysis with quantum simulations. Initially, conducting extensive literature review to evaluate the current state of quantum machine learning (QML) applications in cybersecurity. The research subsequently focuses on the design and evaluation of quantum machine learning models, such as Variational Quantum Circuits (VQCs) and Quantum Support Vector Machines (QSVMs), aimed at security threat detection. Quantum-enhanced anomaly detection techniques are examined through the utilisation of quantum kernels on Noisy Intermediate-Scale Quantum (NISQ) devices. Additionally, the study investigates quantum adversarial robustness by assessing adversarial perturbations within quantum classical hybrid networks. Simulations are carried out using Qulacs, IBM Qiskit and PennyLane to benchmark QML models against traditional security frameworks. The results will evaluate based on accuracy, robustness, and computational efficiency in comparison to conventional cybersecurity models.