Research
Latest academic paper digests, data charts, and research insights from the 6G and AI research community.
62 papers
Token-Free Language Models for Efficient Telecom Log Analysis
Dr. Chen Li, Dr. Marco Fiore — IMDEA Networks / NEC Laboratories Europe
Traditional LLMs struggle with telecom network logs due to their technical vocabulary and structured format not aligning well with standard tokenization. We propose a byte-level token-free language model specifically designed for telecom log analysis. Our model processes raw byte sequences directly, avoiding out-of-vocabulary issues common with standard tokenizers on network log data. On a benchmark of 1M real operator logs, our approach achieves 91% fault classification accuracy and generates root cause explanations that experts rate as helpful 85% of the time.
Reward Shaping for Safe Reinforcement Learning in Network Control
Dr. Tianyu Wang, Prof. Robert Schober — University of Erlangen-Nuremberg
Deploying RL agents in live networks carries the risk of unsafe actions that degrade service. We propose a reward shaping framework that incorporates safety constraints from network SLAs directly into the RL training process. Our constrained RL approach guarantees that QoS violations remain below 0.1% while still achieving 90% of the throughput optimality of unconstrained agents. We validate on a commercial 5G testbed with 50 active users.
Vision Transformers for RF Spectrum Monitoring and Classification
Dr. Tim O'Shea, Dr. Nathan West — DeepSig Inc.
We apply Vision Transformers (ViT) to RF spectrum monitoring by treating spectrograms as images. Our ViT-RF model classifies 24 modulation types with 98.5% accuracy at 10 dB SNR, outperforming CNN baselines by 3.2%. The attention mechanism provides interpretable visualization of which time-frequency regions drive classification decisions. The model runs at 2ms per spectrogram on edge GPU hardware, enabling real-time spectrum monitoring for 6G cognitive radio applications.
Efficient On-Device Training for Adaptive 6G Receivers
Dr. Junmo Kim, Prof. Youngchul Sung — KAIST
We propose an efficient on-device training framework that enables 6G receiver neural networks to continuously adapt to changing channel conditions without cloud connectivity. Using a combination of pruned backpropagation and mixed-precision training, our approach reduces on-device training memory by 8x and energy by 5x compared to standard backpropagation, while maintaining adaptation quality. This enables continuous learning on mobile device chipsets with limited resources.
Sub-THz Wideband Channel Sounder: Design and 140 GHz Measurements
Dr. Theodore Rappaport, Dr. Yunchou Xing — NYU WIRELESS
We present the design and operation of a sub-THz wideband channel sounder operating at 140 GHz with 2 GHz instantaneous bandwidth. Extensive measurement campaigns in indoor office, shopping mall, and outdoor urban micro environments yield over 50,000 channel impulse responses. We develop a 3GPP-compatible statistical channel model for the 140 GHz band and provide open-source access to the measurement dataset to accelerate 6G research worldwide.
STAR-RIS: Simultaneous Transmitting and Reflecting for Full-Space Coverage
Dr. Jiayi Zhang, Prof. Derrick Wing Kwan Ng — University of New South Wales
We present the first comprehensive study of Simultaneous Transmitting And Reflecting RIS (STAR-RIS) for 6G full-space coverage. Unlike conventional RIS that only reflects signals, STAR-RIS serves users on both sides of the surface. We develop optimal beamforming algorithms for three STAR-RIS protocols (energy splitting, mode switching, time switching) and show that STAR-RIS improves sum-rate by 40-70% over conventional RIS in multi-user scenarios where users are distributed on both sides.
Orbital Angular Momentum Multiplexing for 6G Capacity Enhancement
Dr. Bo Thide, Prof. Fabrizio Tamburini — Uppsala University / University of Padova
We demonstrate orbital angular momentum (OAM) multiplexing as an additional degree of freedom for increasing wireless channel capacity in 6G systems. By generating and detecting multiple OAM modes simultaneously, we achieve 4x spectral efficiency improvement over conventional MIMO at 28 GHz over a 100-meter link. Our custom antenna array generates 8 orthogonal OAM modes, and we develop a neural network-based detector that separates modes with less than 1% inter-mode crosstalk.
6G Positioning: Centimeter Accuracy with Joint Communication and Sensing
Dr. Henk Wymeersch, Dr. Gonzalo Seco-Granados — Chalmers University / Universitat Autonoma de Barcelona
We present a 6G positioning system achieving centimeter-level accuracy by jointly exploiting communication signals and sensing returns. Our multi-band approach combines sub-6 GHz for coarse positioning with mmWave for fine refinement, using a deep learning fusion framework. Field experiments in an indoor environment demonstrate 2.3 cm median positioning error, a 10x improvement over 5G NR positioning. The system simultaneously provides communication service with negligible throughput degradation.
Fluid Antenna Systems: A New Paradigm for 6G MIMO
Dr. Kai-Kit Wong, Prof. Kin-Fai Tong — University College London
We introduce fluid antenna systems (FAS) as a new paradigm for 6G MIMO where antenna positions can be dynamically reconfigured in real time. Unlike fixed antenna arrays, FAS uses liquid metal or electronically switchable pixel antennas to optimize antenna placement for current channel conditions. Our prototype demonstrates that FAS achieves 50% higher capacity than fixed arrays with the same number of elements by exploiting spatial diversity through position optimization.