Securing Public Safety Networks: Machine Learning–Driven Anomaly Detection in Telecom Signaling

Telecom signaling systems like SS7, Diameter, and 5G SBA are mission-critical for public safety. Static rule-based anomaly detection is inadequate against evolving threats. This paper proposes a machine learning (ML) framework combining statistical features, unsupervised learning, and domain-specific supervised models. Drawing on experience from Motorola Solutions, we present architecture, methodology, evaluation, and operational challenges. Experiments on simulated traces show higher accuracy and fewer false positives than rule-based systems, suggesting ML can enhance resilience in public safety telecom.

Cybersecurity Cloud Computing Computer Networks