A Novel Service Model Framework for Signaling Storm Detection in Open RAN

This paper presents a novel service model framework based to detect and mitigate signaling storm attacks in Open Radio Access Network (O-RAN). The framework explores O-RAN’s architecture to enable near real-time detection through the RAN Intelligent Controller (RIC). Our proposed xApp monitors initial access procedures and uses Machine Learning (ML) algorithms to identify anomalous behaviors and determine the most appropriate mitigation strategy. Using E2 interface protocols and key performance measurements, the system implements a control loop that accepts or rejects connection attempts based on ML inference results. We validated our framework on a realistic testbed using commercial hardware and software, demonstrating its effectiveness, particularly in scenarios where multiple simultaneous connection attempts could disrupt network operations. The proposed framework achieved a precision of 0.99, recall of 0.92, and F1-score of 0.99 in detecting attacks and properly mitigated attacks during initial access procedures in a 5G O-RAN network.