Development of wireless resources via multi-dimensional radio environment analysis with crowdsensing mobile terminals

2021 Fiscal Year Final Research Report

• Project/Area Number: 19K14988
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 21020:Communication and network engineering-related
• Research Institution: The University of Electro-Communications (2021); Tokyo University of Science (2019-2020)
• Principal Investigator: Sato Koya 電気通信大学, 人工知能先端研究センター, 助教 (60822533)
• Project Period (FY): 2019-04-01 – 2022-03-31
• Keywords: 無線通信 / 電波伝搬 / 周波数共用 / 空間統計 / 機械学習

Outline of Final Research Achievements

We established a multidimensional interpolation method based on big wireless data for received power characteristics. Specifically, the method was extended to joint spatial-frequency interpolation and generalized to any transmission locations, whereas the conventional method only supports spatial interpolation for a fixed base station and single frequency band. Next, a communication scheme utilizing big wireless data was proposed for vehicular communication systems. We showed that the big data-aided communication design improves communication efficiency. Finally, we proposed a fast and accurate data analysis method among terminals using decentralized federated learning as an advanced topic. It was shown that this method efficiently learns wireless channels.

Free Research Field

無線通信、空間統計、機械学習

Academic Significance and Societal Importance of the Research Achievements

無線システムの需要増大に伴い、新規システムへ割り当てる帯域の不足や既存システムの混雑といった問題が年々深刻化している。本研究を通して得られた成果により、限られた無線周波数資源の無駄のない活用の実現が期待される。例として、他者への干渉が小さく済みそうな帯域を積極利用し、かつ通信品質が不安定な箇所は避けることで安定かつ高速な通信を実現するといった使い方が挙げられる。スマートフォンやIoTセンサはもちろん、車車間通信のような信頼度が求められるシステムをサポートする基盤技術となるであろう。