The Internet of Things (IoT) ecosystem can benefit from federated learning (FL) as a decentralized machine learning paradigm that enhances scalability, privacy, and computational efficiency. Modern machine learning models typically require vast amounts of user data to achieve optimal performance, yet acquiring high-quality data remains a significant challenge. Furthermore, collecting data from distributed IoT devices raises critical privacy and security concerns. Blockchain-based consensus mechanisms offer robust security measures that can enhance data protection for both participants and the network. This Ph.D. dissertation explores three fundamental research questions: 1) Can blockchain technology be integrated with distributed machine learning frameworks to enhance security? 2) How can quantum computing contribute to securing distributed computing environments? 3) What methods can be developed to improve the security of machine learning models in real-world applications? To address the first question, we investigate the integration of Proof-of-Work and Proof-of-Elapsed-Time consensus mechanisms into distributed machine learning frameworks. Given that distributed learning techniques are susceptible to data tampering, encryption techniques can provide an additional layer of security. For the second research question, we aim to develop a quantum encryption scheme designed to disrupt inference computations in machine learning-based attacks, such as those utilizing Generative Adversarial Networks (GANs). By introducing controlled inference miscalculations, this approach seeks to fortify the training process against adversarial threats. This research contributes to the intersection of blockchain security, quantum encryption, and federated learning. The goal of third research question is to integrate the developed innovative solutions to enhance the security of decentralized machine learning frameworks in real-world applications, such as critical infrastructure resilience and intelligent transportation systems.

Ervin Moore is a Ph.D. Candidate in the Knight Foundation School of Computing and Information Sciences (KFSCIS) at Florida International University (FIU). He is a member of the solid Lab and works under the co-supervision of Dr. M. Hadi Amini and Dr. Shabnam Rezapour. His research interests include trustworthy distributed machine learning, quantum encryption, blockchain and natural language processing. Ervin received a M.Sc. in Artificial Intelligence and Machine Learning, and a BA in Communication Technologies. He has published in selective journals such as IEEE Transactions on Artificial Intelligence and IEEE IoT Journal. His research has been partly funded by multiple DHS Fellowships, including DHS ADvanced Education and Research for Machine Learning-driven Critical Infrastructure REsilience Center (ADMIRE Center) fellowship, DHS Center for Advancing Education on Critical Infrastructures Resilience Fellowship, and DHS AERC/ADAC. His research has been also partly funded by the USDOT UTC National Center for Transportation Cybersecurity and Resiliency (TraCR). Co-Major Professor: Dr. M. Hadi Amini Co-Major Professor: Dr. Shabnam Rezapour