TelcoAgent: A Scalable 5G Multi-KPM Forecasting With 3GPP-Grounded Explainability

TelcoAgent is a newly proposed framework for Key Performance Measurement (KPM) forecasting in 5G networks, addressing scalability and explainability limitations of existing machine learning methods. It features a three-agent pipeline that constructs a 3GPP knowledge graph, a time-series foundation model (TSFM) for zero-shot forecasting, and a reasoning pipeline for actionable diagnostics. Evaluated on a real-world dataset from a U.S. network operator, TelcoAgent achieved high accuracy across seven KPMs in 200 network cells, making it significant for practitioners focused on proactive network management and performance optimization in telecom environments.