Toward Self-Optimizing 6G Networks Through Network Digital Twin Intelligence: A Real Network Traffic Evaluation

Driven by advancing digital ecosystems, the next-generation network vision promises unprecedented levels of intelligence, adaptability, and automation. Key 6G properties such as native artificial intelligence integration, a highly cognitive architecture, ultra-low latency, high-dense connectivity, and self-optimization require automated algorithms and intelligent enabler technologies. Network Digital Twins (NDTs) are a key enabler for mobile networks, due to their capability in replicating physical network operations. In this letter, we discuss how NDTs enable AI-native and self-optimizing networks. Leveraging real network traffic data, our framework achieved up to 98.3% voice/SMS traffic prediction accuracy, self-optimized load reduction, improved energy efficiency, and anomaly detection.