2022 Fiscal Year Research-status Report
A Study on Federated Learning for Efficient Communication Resource Allocation in 6G Heterogeneous Networks
| Project/Area Number |
22K17877
|
|---|
| Research Institution | Tohoku University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2022-04-01 – 2026-03-31
|
| Keywords | Machine Learning / Satellite Networks / 6G Networks / Network Management / Cloud Systems / Digital Twin |
|---|
| Outline of Annual Research Achievements |
The main result was a protocol for managing data processing from multiple sensors. It determines, based on the location of the source sensor and the state of the network, whether to collect and process the data using Low Earth Orbit satellites, or use a central server. Such protocol is useful for assisting collection and training of Distributed and Federated Learning models in heterogeneous networks, increasing training efficiency by 60% . Additionally, there were results on satellite beamforming, task offloading in cloud networks, and digital twins for network logging andanalysis for a total of 5 journal papers and 3 international conferences.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
I was planning to initially only survey the literature and make concrete plans during year 1. However, I managed to produce and publish some results. Additionally, due to the help of Graduate Students, other topics beyond Machine Learning in Satellite Networks were able to be studied, with corresponding results being published.
|
| Strategy for Future Research Activity |
The plan for the next years is to investigate how to use Machine Learning in network management and changes in the network affect trained models (and what are the best strategies to handle this issue). Additionally, as a secondary goal, I plan to study Machine Learning solutions for network medium selection in vehicular networks.
|
