Sorbonne University is a world-class academic institution that was created on January 1st, 2018, as a result of the merger of two first-class research-intensive universities, UPMC (University Pierre and Marie Curie) and Paris-Sorbonne. The university is multidisciplinary and research-intensive, with three faculties: humanities, medicine, and science. Each faculty has the autonomy necessary to conduct its ambitious programs in both research and education. Sorbonne University has a student population of 53,500, with 3,400 professor-researchers and 3,600 administrative and technical staff members.

LIP6 is a computer science laboratory that is jointly operated by Sorbonne University (SU) and the French National Center for Scientific Research (CNRS). The laboratory is internationally recognized as a leading research institute. The LIP6 Networks and Performance Analysis (NPA) group focuses on developing an autonomous, reliable, high-performance, and secure digital infrastructure for the future internet. LIP6 is involved in many activities related to advanced wireless, smart mobility, optimization, virtual networks and slicing, internet governance and regulation, monitoring, and large-scale EU-funded testbed projects such as PlanetLab Europe, Onelab, OpenLab, F-Interop, and Armour. The NPA group conducts research in networking through basic research and transfer activities in strong cooperation with worldwide academic partners and industrial leaders. Transfer is measurable through contributions to standardization bodies such as IETF: MLDv2 or RTP XR, the creation of start-ups, as well as numerous industrial contracts. The group activity is supported by permanent researchers, Post-Doc researchers, international visitors, engineers, and PhD students.

Role In the project

SU will provide access to the OneLab/SLICES testbed in Paris, which offers IoT/5G/B5G/6G infrastructure, including SDR/USRP devices, Kubernetes clusters, and over 2700 wireless static and mobile IoT nodes. Additionally, SU will contribute to the development of cybersecure federated deep reinforcement learning methods and techniques. These methods will securely combine local machine learning training to update a global model without compromising scalability, fault-tolerance, and security against poisoning attacks. Finally, SU will contribute to the monetization and consensus-based accountability of resources by leveraging interoperable blockchain ledgers to track provenance and enforce secure negotiation and transaction of resources.