From 5G to Satellite: AI Approaches for Energy-Efficient, High-Performance Network Infrastructures

The exponential growth in data traffic, driven by 5G/6G rollout, cloud computing, real-time applications, and massive IoT deployment, places increasing strain on high-speed network infrastructures. Conventional, rule-based network management cannot adapt quickly enough to dynamic traffic patterns, resulting in inefficient energy use and degraded performance. Artificial intelligence, including deep reinforcement learning, federated learning, graph neural networks, and large language models, offers a path toward networks that self-optimize, predict demand, and conserve energy without sacrificing throughput or latency. This Research Topic establishes a dedicated scholarly platform for cutting-edge research at the intersection of artificial intelligence and high-speed networking, with a dual emphasis on energy conservation and performance enhancement. It aims to consolidate and disseminate novel AI-driven methodologies applied across diverse high-speed network architectures, including optical backbone networks, software-defined networks, edge and fog computing systems, and non-terrestrial satellite networks. By bringing together contributions from academia, industry, and research institutions worldwide, this collection bridges the gap between theoretical AI advances and practical network deployments, and aligns networking research with global sustainability objectives by promoting green, energy-proportional network designs guided by intelligent automation. We invite original research articles, reviews, and perspectives from researchers working on: - AI and machine learning methods for energy-efficient resource management in 5G/6G radio access networks - Intelligent traffic engineering, routing, and scheduling in optical backbone and metro networks - Reinforcement learning and autonomous control for software-defined networking and network function virtualisation - Federated learning and distributed AI for privacy-preserving network optimisation - Energy-aware management of edge, fog, and data center computing infrastructures - AI-driven QoS optimisation, self-healing mechanisms, and anomaly detection in high-speed networks - Digital twin frameworks for network simulation, planning, and green performance benchmarking - Semantic communications and large language model applications for intelligent network control - AI-assisted resource allocation in satellite and non-terrestrial networks - Interdisciplinary work bridging communication engineering, machine learning, and sustainable computingBoth simulation-based and real-world experimental contributions are welcome. Work from academic institutions and industry practitioners is encouraged, provided it offers original, rigorous, and reproducible insights. Topic editors: Shahzad Ashraf, Huma Jamshed, Furqan Memon Source: https://www.frontiersin.org/research-topics/82997/from-5g-to-satellite-ai-approaches-for-energy-efficient-high-performance-network-infrastructures