Research
Latest academic paper digests, data charts, and research insights from the 6G and AI research community.
16 papers
AI-Native Air Interface Design: End-to-End Learning for 6G Physical Layer
Dr. Sebastian Dorner, Prof. Stephan ten Brink — University of Stuttgart
We propose an AI-native air interface where the entire physical layer — from modulation to coding to equalization — is learned end-to-end using deep neural networks. Unlike traditional block-based PHY design, our autoencoder-based system jointly optimizes all components for the actual channel conditions. Over-the-air experiments using software-defined radios demonstrate that the learned air interface outperforms 5G NR baseline by 3 dB at 10^-3 BER in frequency-selective fading channels.
Digital Twin Networks: AI-Driven Real-Time Network Simulation for 6G
Dr. Razvan Beuran, Dr. Tarik Taleb — Oulu University / Ruhr University Bochum
This paper presents a comprehensive framework for AI-driven digital twin networks (DTN) that create real-time virtual replicas of physical 6G networks. Our DTN framework uses graph neural networks to model network behavior and reinforcement learning to optimize network configurations in the digital twin before deploying changes to the physical network. Results show that DTN-guided optimization reduces service degradation incidents by 65% and speeds up new service deployment by 4x.
Intent-Based Network Management with Large Language Models
Dr. Carlos Bernardos, Prof. Antonio de la Oliva — Universidad Carlos III de Madrid
We propose an intent-based network management system powered by large language models that translates operator intents expressed in natural language into network configurations. The system uses a chain of specialized LLM agents for intent parsing, conflict resolution, and configuration generation. Evaluation on 500 real operator intents shows 89% end-to-end accuracy in producing valid configurations, with the system handling complex multi-domain intents spanning RAN, transport, and core networks.
AI-Driven Energy Optimization in 6G Networks: A Multi-Objective Approach
Dr. Hanna Bogucka, Prof. H. Vincent Poor — Poznan University of Technology / Princeton University
This paper addresses the critical challenge of energy efficiency in 6G networks through a multi-objective AI optimization framework. Our approach uses evolutionary neural networks to simultaneously optimize energy consumption, spectral efficiency, and user QoS. Deployed on a testbed with 20 base stations, the system achieves 40% energy reduction during low-traffic periods while maintaining 99.9% QoS satisfaction, representing a significant step toward carbon-neutral 6G networks.
Explainable AI for Network Anomaly Detection: Trust in Autonomous Networks
Dr. Thomas Bonald, Dr. Aline Carneiro Viana — Telecom Paris / Inria
We address the critical need for explainability in AI-driven network anomaly detection systems. Our framework combines a high-accuracy anomaly detector with a post-hoc explanation module that provides human-understandable reasons for each alert. The explanation module uses SHAP values adapted for time-series network data, achieving 94% anomaly detection accuracy while providing explanations that network operators rate as helpful 87% of the time.
AI-RAN: Architecture Design and Performance Evaluation for 5G-Advanced
Dr. Mischa Dohler, Dr. Harish Viswanathan — Ericsson Research
This paper presents the architecture design and comprehensive performance evaluation of AI-RAN for 5G-Advanced networks. We define three deployment models — centralized, distributed, and hybrid — and evaluate each across 15 KPIs including throughput, latency, energy efficiency, and operational complexity. Large-scale simulations with 500 base stations and 50,000 users show that the hybrid model achieves the best balance, improving throughput by 35% and reducing energy by 28% compared to traditional RAN.
Generative AI for Network Planning: From Text Descriptions to Optimal Deployments
Dr. Rui Li, Dr. Xu Chen et al. — NTU Singapore
We propose a generative AI system that converts natural language descriptions of coverage requirements into optimal network deployment plans. The system uses a fine-tuned vision-language model to understand geographic constraints from satellite imagery and a diffusion model to generate base station placement configurations. Comparison with human network planners shows that the AI system produces plans with 12% better coverage uniformity and 18% lower cost while reducing planning time from weeks to hours.
Continual Learning for Adaptive Beamforming in Dynamic 6G Environments
Dr. Osvaldo Simeone, Dr. Joonhyuk Kang — King's College London / KAIST
This paper addresses the challenge of maintaining AI beamforming performance in dynamic 6G environments where channel statistics change over time. We propose a continual learning framework that adapts beamforming models to new environments without catastrophic forgetting of previously learned conditions. Our approach uses elastic weight consolidation combined with experience replay, maintaining beamforming performance within 1 dB of separately trained models across 10 different deployment scenarios.
Foundation Models for Wireless: Pre-Training on Network Data at Scale
Dr. Ahmed Alkhateeb, Dr. Umut Demirhan — Arizona State University
We introduce WirelessFM, a foundation model pre-trained on 100TB of diverse wireless network data including channel measurements, traffic patterns, KPIs, and configuration parameters from 50 operators worldwide. WirelessFM can be fine-tuned for any downstream wireless task with minimal data, achieving state-of-the-art results on 12 benchmark tasks including channel estimation, traffic prediction, and anomaly detection. The model reduces the data requirement for new task adaptation by 20x.