2020 Fiscal Year Annual Research Report
強化学習を用いた広帯域ミリ波最適スペクトル制御と先験知識を予測する技術の開発
| Project/Area Number |
20J12528
|
|---|
| Research Institution | Keio University |
|---|
Principal Investigator |
曹 誉文 慶應義塾大学, 理工学部, 特別研究員(PD)
|
|---|
| Project Period (FY) |
2020-04-24 – 2022-03-31
|
| Keywords | Analog beamforming / mmWave communications / deep learning / routing / caching / dense networks / federated learning |
|---|
| Outline of Annual Research Achievements |
In this research year, my research focuses mainly on the analog beamformer designs and it's application in multiuser millimeter-wave (mmWave) communications using deep learning. In addition, the transmit beamforming incorporating the spatial modulation technology has been studied. On the other hand, mobility-aware routing and caching for dense networks by using federated learning have been investigated as well. Related research results have been published in 3 high-quality journal papers and 3 conference papers, respectively. The experiments manifest that the proposed scheme can accurately estimate beam qualities and give high probability of optimal beam selections with low overhead. In addition, the theoretical and numerical analyses show that the proposed caching approach can provide high average cache efficiency (CE) in relative to the state-of-the-art caching schemes.
|
| Research Progress Status |
翌年度、交付申請を辞退するため、記入しない。
|
| Strategy for Future Research Activity |
翌年度、交付申請を辞退するため、記入しない。
|
Research Products
(8 results)
