Mohamed Elmahallawy

Peer-reviewed papers

JSAC

IEEE journal

Best Paper

Runner-up · PerCom 2024

Office: 144B Elson S. Floyd Building, 2710 Crimson Way, Richland, WA 99354.

Mohamed Elmahallawy is an Assistant Professor in the Department of Computer Science and Cybersecurity at Washington State University. He is also a former Postdoctoral Fellow in the Department of Computer Science at Missouri University of Science and Technology collaborating with Dr. Sanjay Madria. His academic focus includes Cybersecurity, Cryptography, and Trustworthy Artificial Intelligence (AI), with a particular focus on Federated Learning, Large Language Models (LLMs), and their secure and ethical deployment in distributed systems. Specifcally, his research aims to develop trustworthy and explainable AI solutions for diverse applications, including Space AI, underground mining, medical and healthcare systems, and the Internet of Things (IoT). During his M.Sc. under the supervision of Prof.Dr-Ing. Christian Haubelt and his Ph.D. journey under the guidance of Dr. "Thomas" Tie Luo, Mohamed gained valuable experience by working at several international institutions, including Rostock University in Germany, Tennessee Technological University, and Missouri University of Science and Technology in the USA. He has published over 38 papers in leading journals (e.g., JSAC, IEEE IoT) and conferences (e.g., PerCom, MDM, Big Data, GlobeCom, ICC). Notably, his recent paper presented at the IEEE PerCom conference was awarded with the Best Paper Runner-up Award.

His innovative work on developing secure and trustworthy AI in many applications has garnered significant attention, resulting in many international talks. In addition to his research contributions, Mohamed actively participates in the academic community by serving as a reviewer for several journals, including JSAC, TMC, TSUSC, TII, and EAAI.

Research Interest

Dedicated to advancing cutting-edge research to ensure privacy, security, and trustworthiness in Artificial Intelligence (AI) for Distributed Systems and Large Language Models (LLMs). My work encompasses:

Federated Learning & PrivacyPrivacy-preserving distributed training, secure aggregation, and Byzantine-robust FL systems.

Space AI & Satellite NetworksML on LEO satellites, satellite edge computing, and NOMA-based FL for space communications.

Trustworthy LLMs & IoT SecuritySecure and ethical LLM deployment; anomaly detection and safety in distributed systems.

WE ARE HIRING

Seeking highly motivated Ph.D. students interested in secure and trustworthy AI for distributed systems. Include your CV, transcripts, and English proficiency scores in your message. Subject: Potential_PhD_student_Your_Name.

Apply via email

News

Oct 2026	Congratulations to my Ph.D. student Mohamed Shaaban! His paper "Tokenized Key-Gated Adapter Routing: A Secure Access Control Mechanism Against Private Data Leakage in LLMs" was accepted in the IEEE Transactions on Artificial Intelligence (TAI)! 🎉
Sep 2026	Congratulations to my Ph.D. student Ahmed Abdelnaby! His paper "CLIPGuard: Region-Level Black-Box Defense Against Stealthy Embedding-Space Backdoors in CLIP" was accepted at the 30th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2026)! 🎉
Mar 2026	Congratulations to my Ph.D. student Nur Afsa Syeda! Her paper "Learning What Can Be Picked: Active Reachability Estimation for Efficient Robotic Fruit Harvesting" was accepted at PAKDD 2026! 🎉
Feb 2026	Congratulations to my Ph.D. student Mohamed Shaaban! His paper "SecureGate: Learning When to Reveal PII Safely via Token-Gated Dual-Adapters for Federated LLMs" was accepted at the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2026)! 🎉
Mar 2026	Congratulations to Ph.D. student collaborator Shreen Gul! Her paper "Prototype Fusion: A Training-Free Multi-Layer Approach to OOD Detection" is available on arXiv. 🎉
Nov 20, 2024	Congratulations! My paper titled "Predicting Battery Levels of Sensor Nodes Using Reinforcement Learning in Harsh Underground Mining Environments" was accepted by The 40th ACM/SIGAPP Symposium On Applied Computing (SAC 2025) 🎉
Oct 26, 2024	Congratulations! Four of my papers have been accepted for presentation at the prestigious IEEE BigData 2024 Conference! 🎉
Jun 24, 2024	Congratulations! My paper titled CAV-AD: A Robust Framework for Detection of Anomalous Data and Malicious Sensors in CAV Networks was accepted by The 21st IEEE International Conference on Mobile Ad-Hoc and Smart Systems (MASS 2024) 🎉
Jun 05, 2024	Spreading Research Around the World: Developing Machine Learning Models on Satellites
Apr 30, 2024	I have honored the Dean's Ph.D. Scholar Award for 2024

Selected Publications [See More]

BigData'25

Decentralized Trust for Space AI: Blockchain-Based Federated Learning Across Multi-Vendor LEO Satellite Networks

Mohamed Elmahallawy, and Asma Jodeiri Akbarfam

In 13th IEEE International Conference on Big Data (BigData), 2025

HTML
BigData'25

Secure and Privacy-Preserving Federated Learning for Next-Generation Underground Mine Safety

Mohamed Elmahallawy, Sanjay Madria, and Samuel Frimpong

In 13th IEEE International Conference on Big Data (BigData), 2025

HTML
ECAI'25

When Secure Aggregation Falls Short: Achieving Long-Term Privacy in Asynchronous Federated Learning for LEO Satellite Networks

Mohamed Elmahallawy, and Tie Luo

In 27th European Conference on Artificial Intelligence (ECAI), 2025

HTML
MASS'25

Detecting Untargeted Attacks and Mitigating Unreliable Updates in Federated Learning for Underground Mining Operations

Md Sazedur Rahman, Mohamed Elmahallawy, Sanjay Madria, and Samuel Frimpong

In 22nd IEEE International Conference on Mobile Ad-Hoc and Smart Systems (MASS), 2025

HTML
SAC'25

Predicting Battery Levels of Sensor Nodes Using Reinforcement Learning in Harsh Underground Mining Environments

Manish Anand Yadav, Mohamed Elmahallawy, Sanjay Madria , and 1 more author

In ACM/SIGAPP Symposium on Applied Computing (SAC) , 2025

Abs PDF

We propose PPO-LSTM, a deep reinforcement learning approach specifically tailored for underground mining environments. The framework considers miners' movements and communication dynamics to predict sensor battery levels, facilitating timely energy harvesting for sensors nearing depletion at critical locations.
MASS'24

CAV-AD: A Robust Framework for Detection of Anomalous Data and Malicious Sensors in CAV Networks

Md Sazedur Rahman, Mohamed Elmahallawy, Sanjay Madria , and 1 more author

In 2024 IEEE International Conference on Mobile Ad-Hoc and Smart Systems (MASS) , 2024

Abs HTML PDF

CAV-AD proposes a novel framework tailored to CAV networks for distinguishing abnormal readings amidst multiple anomaly data while identifying malicious sensors, achieving an average accuracy of 98% and F1 score of 89%.
PerCom'24

Stitching Satellites to the Edge: Pervasive and Efficient Federated LEO Satellite Learning

Mohamed Elmahallawy, and Tie Luo

In 2024 IEEE International Conference on Pervasive Computing and Communications (PerCom) , 2024

Abs HTML PDF

We propose a novel FL-SEC framework that empowers LEO satellites to execute large-scale ML tasks onboard efficiently. Reduces FL convergence time by nearly 30 times and energy consumption to 1.38 watts, with 96% accuracy. Awarded Best Paper Runner-up at IEEE PerCom 2024.
JSAC'24

Communication-Efficient Federated Learning for LEO Constellations Integrated With HAPs Using Hybrid NOMA-OFDM

Mohamed Elmahallawy, Tie Luo, and Khaled Ramadan

IEEE Journal on Selected Areas in Communications, Vol. 42(5), 2024

Abs HTML PDF

NomaFedHAP introduces NOMA and high-altitude platforms as distributed parameter servers for LEO satellite FL, enabling fast and bandwidth-efficient model transmissions with closed-form outage probability analysis.
GLOBECOM'23

Secure and Efficient Federated Learning in LEO Constellations Using Decentralized Key Generation and On-Orbit Model Aggregation

Mohamed Elmahallawy, Tie Luo, and Mohamed I. Ibrahem

In IEEE Global Communications Conference (GlobeCom) , 2023

Abs HTML PDF

FedSecure uses decentralized key generation with functional encryption and on-orbit model aggregation to protect satellite data privacy, reducing convergence delay from days to hours with 85.35% accuracy.