Weather and Environment aware Backhaul using SATellite for 6G
Communications
Funded by: SNS-JU Open Call 3 6G-PATH
Summary on 6G Sandbox webpage: Coming soon ...
Total funding: 60K EUR
Project Objectives:
Develop energy-aware mechanisms for battery-powered 5G RAN.
Correlate weather data with Starlink NTN KPIs to enable adaptive scheduling.
Deliver reproducible assets: gNB scheduler configs and datasets combining network, power, and weather metrics.
Quantify energy-QoS trade-offs to extend autonomy while safeguarding critical services.
Key Activities
Establish baseline measurements of power, KPIs, and weather impact.
Implement RAN-side adaptations on 5G gNB (srsRAN).
Develop ML-based weather-aware NTN scheduling for Starlink.
Integrate RAN and NTN methods into a unified control framework.
Release open dataset, configs, and deployment guidelines.
Expected Results
TRL7 prototype of an energy-aware 5G RAN with NTN backhaul, both powered by battery.
Validation of RAN mechanisms that reduce gNB energy use and extend battery life in off-grid settings, while maintaining acceptable QoS.
Exploratory demonstration of weather-aware NTN scheduling and its potential to mitigate QoS degradation under adverse conditions.
Validation of the proposed mechanisms on multiple open-source platforms (srsRAN and OpenAirInterface), ensuring replicability and broad adoption potential.
Open-source tools and datasets to enable replication.
My role (PI):
As the Principle Investigator (PI) of the project, I lead the technical vision and hands-on experimentation, reproducible trials, and critical analysis.
Defined the key KPIs, test scenarios, and overall evaluation plan on the UMA (6G AnyWhere) testbed
Lead baseline measurements covering gNB power, 5G performance KPIs, and Starlink/weather impact
Implemented energy-aware RAN controls to reduce gNB power while maintaining QoS
Designed and integrated weather-aware NTN backhaul scheduling (correlation + lightweight ML + traffic policies)
Built automation for controlled experiments, synchronized logging, and dataset generation
Consolidated results into reports and supported dissemination/demonstration activities
Related Publications:
Coming soon...