DAST: A VLM-LLM Framework for Cross-Interface Anomaly Detection in O-RAN
Researchers have developed DAST, a novel framework utilizing a VLM-LLM-VLM pipeline for detecting anomalies in Open Radio Access Networks (O-RAN). This system addresses the limitations of traditional methods by converting time-series data into visual representations and leveraging large language models for analysis. DAST achieved a 0.910 F1-Score and 0.843 Accuracy in tests on real network traces, outperforming existing anomaly detection techniques. AI