AI, ML, and LLM Integration in 5G/6G Networks: A Comprehensive Survey of Architectures, Challenges, and Future Directions
Yusuf Usman, Habeeb Oladipupo, Adegboyega Daniel During, A. Robert, Robin Chataut
Abstract
The transition from 5G to 6G networks demands groundbreaking advances in intelligence, adaptability, and security to support emerging applications such as real-time telemedicine, immersive extended reality (XR), and autonomous systems. This article provides a comprehensive analysis of how artificial intelligence (AI), machine learning (ML), and large language models (LLM) are revolutionizing next-generation telecommunications. We present a structured roadmap for integrating these technologies into 6G infrastructure, emphasizing their transformative roles in intelligent network management, dynamic resource allocation, and proactive threat mitigation. By addressing key challenges such as ultralow latency, heterogeneous data handling, and ethical governance, this study bridges theoretical innovations with practical applications. Notable contributions include novel frameworks for AI-enhanced security, self-healing networks, and privacy-preserving techniques like federated learning. Furthermore, we explore critical ethical considerations, including bias mitigation and transparency in AI decision-making, while highlighting emerging research directions such as adaptive learning systems and hybrid AI architectures. This work underscores the synergistic potential of AI and 6G, equipping researchers and industry stakeholders with actionable insights to develop resilient, user-centric networks that will shape the future of global connectivity.