News and Announcements

• Paper Presentation at WMC 2026

  April 15, 2026
  
  Our team member Saswati Pal presented our paper Toward Clinically-Inspired Validation of ML-Driven Source Localization in Molecular Communication at the 10th Workshop on Molecular Communications (WMC 2026), Istanbul, Turkey. Accurate localization of tumor sources in the human circulatory system is essential for precision oncology. In prior work, we developed a machine learning (ML) framework to localize anomaly sources using temporal biomarker profiles measured at receiver sites. This work-in-progress paper extends the framework by validating the ML model on a clinically-inspired dataset that emulates endocrine signaling in a controlled synthetic environment. Preliminary results show an accuracy of 90%, indicating the potential of ML-driven approaches for tumor source localization in clinically relevant molecular communication settings.
  

• Kick-off new BMFTR project NEXT-G

  April 08, 2026
  
  We are happy to announce our new project "NEXT-G: Explainable and Trustworthy AI/ML for 6G and Beyond", funded by BMFTR, and carried out within the Software Campus program in collaboration with Huawei Munich Research HQ. The project is led by our team member Osman Tugay Basaran as Project Manager, together with a three-researcher team. In this project, we investigate explainable, trustworthy, and robust AI/ML methods for AI-native 6G, with a focus on transparent network intelligence, low-latency explainability, adversarial robustness for connected robotics use cases in smart hospitals and factories.
  

• New IEEE Transactions on Networking article

  April 07, 2026
  
  Our article Byzantine-Resilient Federated Learning under Heterogeneity and Heavy Tails has been accepted for publication in IEEE Transactions on Networking. Byzantine resilience is essential in federated learning (FL) to safeguard model training from malicious or faulty participants. However, existing Byzantine-resilient methods struggle when faced with heavy-tailed gradient noise, a common challenge in heterogeneous environments. In this work, we propose a Byzantine-resilient FL framework specifically designed to handle both heterogeneity and heavy-tailed noise. Our approach builds on robust distributed stochastic heavy-ball optimization, incorporating update normalization and gradient/momentum clipping to mitigate the effects of heavy-tailed noise. We establish the first high-probability convergence guarantees for Byzantine-resilient FL under these conditions, showing that our algorithms achieve optimal Byzantine resilience and align with known lower bounds. Additionally, we introduce an efficient variant of the nearest neighbor mixing technique, leveraging random projections to significantly reduce computational costs in high-dimensional settings. Through rigorous theoretical analysis and extensive empirical evaluations, we demonstrate that our methods outperform existing approaches in robustness against both Byzantine failures and heavy-tailed noise.
  (link to more information)

• Mateusz Zakrzewski winning GI/ITG KuVS Best Bachelor Thesis Award

  April 01, 2026
  We congratulate Mateusz Zakrzewski for winning the Best Bachelor Thesis Award of the GI/ITG Communication and Distributed Systems (KuVS) special interest group! The prize will be awarded at NetSys 2027. Mateusz's thesis has already lead to a publications at IEEE/IFIP WONS 2026 (Multi-Link Scheduling with Restricted Target Wake Time in Wi-Fi 7).
  

• CCS on Tour 2026

  March 25, 2026
  
  The Telecommunication Networks Group (TKN) held its annual research retreat "CCS on Tour 2026" from March 24th to 27th. During the four-day event in Freising (near Munich, Germany), we had interesting discussions on PhD life, soft skills, politics, mental health, but certainly also on research topics like distributed learning, attention models, edge computing, wireless sensing, cross-technology communication.
  

• Habilitation degree in Electrical Engineering

  March 12, 2026
  
  Our team member Jorge Torres Gómez has received the certificate for Habilitation in the field of Electrical Engineering. The habilitation certifies the ability to conduct independent research and to teach courses at universities, and it formally qualifies the holder to supervise doctoral students and to pursue a professorship. His habilitation research work focused on developing communication systems that interact with biological processes at the nano- and microscales. This research direction lies at the intersection of telecommunications engineering, molecular communication, and bio-nanotechnology, contributing to the emerging vision of the Internet of Bio-Nano-Things. The habilitation procedure included a cumulative research thesis as well as academic examination components, including a habilitation colloquium and a teaching demonstration. Together, these elements assess both the candidate’s scientific contributions and their ability to teach at the university level.
  

• TKN at Unesco House (IASEAI'26)

  March 12, 2026
  
  Our team member Osman Tugay Basaran participated in the Annual Summit of the International Association for Safe and Ethical Artificial Intelligence (IASEAI), held at the historic UNESCO House in Paris. The summit gathered a global cohort of researchers, policymakers, and industry leaders to address the critical challenge of ensuring AI development remains safe, ethical, and aligned with human values. Beyond broad governance, we contributed to the “AIxBio: Safety Evaluations, Governance, and Standards” workshop, organized by the Johns Hopkins Center for Health Security with the European Commission AI Office, to tackle the specific technical risks at the intersection of AI and life sciences. Our involvement focused on bridging the gap between high-level ethical principles and technical implementation, ensuring that as AI scales across diverse technological areas, it remains transparent and firmly aligned with human values.
  

• New IEEE Transactions on Mobile Computing article

  March 07, 2026
  
  Our article Hierarchical Federated Learning in Device-to-Device Networks with Learning-Topology Co-Optimization has been accepted for publication in IEEE Transactions on Mobile Computing. Federated learning (FL) enables collaborative model training across distributed devices while preserving privacy. However, growing heterogeneity in device resources and communication links challenges conventional FL, especially when relying on a single central server. Hierarchical federated learning (HFL) mitigates these issues by organizing devices into clusters coordinated through intermediate aggregators. Yet, the effectiveness of HFL critically depends on how clusters are formed: intra-cluster communication must be efficient, device computational capacities should be balanced to reduce stragglers, and data heterogeneity must be managed to ensure stable convergence. In this work, we propose a learning–topology co-optimization framework for HFL in networks where nodes communicate with each other with links of varying quality (e.g., device-to-device (D2D) or mesh networks). Our method jointly optimizes device connection topology and learning directions, leading to communication-efficient clusters that remain well aligned in optimization space. We provide a convergence analysis under mild assumptions, showing how inter- and intra-cluster divergence affect learning stability. Extensive experiments demonstrate that our approach consistently improves HFL performance, yielding at least a 6% accuracy gain under unbalanced data distributions and over 16% reduction in training time for regression tasks compared with existing clustering algorithms.
  (link to more information)

• TKN at WONS 2026

  March 04, 2026
  
  Our team member Dr. Doganalp Ergenc and guest researcher Elena Tonini presented their works at WONS 2026 in Switzerland. In his paper Multi-Link Scheduling with Restricted Target Wake Time in Wi-Fi 7, Doganalp presents a novel time-sensitive scheduling approach levaraging the new features of Wi-Fi 7. In Leveraging Mutual Information in Stochastic CSI Analysis for Wi-Fi Sensing, Elena investigates the existence of mutual information across different CSI measurements in the context of Wi-Fi sensing. Last but not least, our external researchers Laura Finarelli and Berk Buzcu took part in the conference as web chair and local organizer, respectively. We thank and congratulate all our researchers for their contributions!
  

• New IEEE Transactions on Control Systems Technology article

  March 03, 2026
  
  Our article Decentralized Model Predictive Control for Platooning: Enhancing Human-Driver Collaboration has been accepted for publication in IEEE Transactions on Control Systems Technology. Recent advances in cooperative adaptive cruise control have demonstrated the potential for vehicle platooning to revolutionize road transportation through enhanced safety, reduced congestion, and improved energy efficiency. While autonomous vehicle technology continues to evolve rapidly, current regulatory frameworks and safety considerations necessitate the human driver supervision. This creates a unique challenge in developing control systems that can effectively balance autonomous operation with human intervention. To enhance the human-driver collaboration with autonomous vehicle platooning, in this paper, we present a novel decentralized model predictive control framework that explicitly incorporates human-driver interaction while maintaining desired inter-vehicle distances and velocities in platoon formations. This framework employs a distributed architecture where each vehicle operates independently and exchanges local measurements through vehicle-to-vehicle communication. To overcome the inherent unreliability of wireless communications in real-world scenarios, we develop a robust distributed state estimation strategy. This approach enables each vehicle to combine local sensor measurements with received data to construct accurate estimates of the full platoon state. Based on these estimates, vehicles compute optimal control actions locally while achieving performance comparable to an ideal centralized controller with perfect communication.
  (link to more information)

Load all older news