Collaborative safe Bayesian optimization
| Reference number | |
| Coordinator | Ericsson AB |
| Funding from Vinnova | SEK 100 000 |
| Project duration | January 2025 - August 2025 |
| Status | Ongoing |
| Venture | 6G - Competence supply |
| Call | 6G - Supervision of degree work |
Purpose and goal
This project addresses optimizing network configurations through artificial intelligence (AI). As 6G networks are expected to be more complex and densely deployed, manual setting of base station parameters will be unfeasible. This project aims to enable AI-native networks where nodes can collaborate to optimize their performance without compromising user experience. The work investigates extensions of Safe Bayesian Optimization techniques to multi-agent systems in telecommunications.
Expected effects and result
The expected outcome is a novel algorithm leveraging 6G network densification for safe and distributed optimization of network parameters. The work investigates extensions of Safe Bayesian Optimization techniques to multi-agent systems in telecommunications. It will involve implementing such an optimization algorithm and evaluating its performance in a simulated network.
Planned approach and implementation
The project starts with a review of safe optimization methods. The student will formulate the 6G-specific collaborative optimization problem with the supervisors. The student will implement and extend existing algorithms and evaluate their performance on domain-specific metrics using a simulator. Weekly meetings will ensure progress, provide feedback and support. The outcomes will be documented in a final report including the literature review, problem formulation, and numerical results.