Wi-Fi operates in the open radio spectrum and is vulnerable to interference from neighbouring networks. This can result in packet loss and retransmissions, which greatly reduces network capacity and user experience.
In the DISCO project, we have built a decentralized system for discovering and grouping Wi-Fi networks to allow them to collaborate. The system is designed to create clusters of nodes that interfere each other the most, for instance neighbours in an apartment building. Wi-Fi users can then automatically coordinate using machine learning or graph-coloring algorithms to avoid interfering each other.
We have several topics available in this area and you may also propose your own ideas. They range from practical implementations on routers to simulations and theoretical work on clustering algorithms.
Possible topics this year:
- Measure impact of group collaboration
- Experimentally measure effects of interference in the latest Wi-Fi standards
- Implement an experimental prototype of DISCO algorithms in rust or C++ that runs on Wi-Fi 7 or Wi-Fi 6 routers
- Evaluate ML algorithms for Wi-Fi collaboration in a simulated Wi-Fi environment
- Improve collaboration with accurate Wi-Fi time synchronisation
- Evaluate new techniques for avoiding interference in Wi-Fi 6 / 7 using new protocol features
Learning outcome
Experience
- in formulating, investigating and answering research questions
- conducting, evaluating and interpreting research results
- in-depth understanding of wireless networks and Wi-Fi
- experience running large scale network simulations
Conditions
We expect that you:
- have been admitted to a master's program in MatNat@UiO
- take this as a long thesis
- are interested in and have some knowledge of C++ or rust programming and computer networks
- include the course IN5060 in the study plan, unless you have already completed a course on classical (non-ML) data analysis