FURI | Spring 2026

Machine Learning-Based Detection and Defense Against Attacks in Delay Tolerant Networks

Delay Tolerant Networks (DTNs) enable communication in environments with limited or disrupted infrastructure, including disaster zones, rural areas, and space missions. This project investigates whether machine learning can detect malicious routing behaviors such as blackhole, grayhole, and flooding attacks more effectively than traditional heuristic methods. Using The ONE simulator, the research will generate DTN traffic datasets, train attack-detection models, and compare their performance with existing approaches using precision, recall, and F1-score.