| 電子情報通信学会 研究会発表申込システム 講演論文 詳細 |
技報閲覧サービス [ログイン] 技報アーカイブ |
| トップに戻る | 前のページに戻る | [Japanese] / [English] |
| 講演抄録／キーワード | ||
| 講演名 | 2022-03-02 10:25 Investigation on Beamforming for IRS-Assisted MIMO-OFDM Communication using Machine Learning ○Julian Webber・Kazuto Yano・Norisato Suga・Yoshinori Suzuki（ATR）  SR2021-86 |
|
| 抄録 | （和） | There has recently been considerable interest in intelligent reflective surface (IRS) which can improve the capacity of communications links by facilitating creation of additional communication paths. The performance of an IRS system depends on the accuracy of estimating the channel and hence ability to compute accurate weights which degrade in the presence of interference. Computing the beamformer weights requires high complexity that scales with the array size. Machine learning is a promising technique for learning the multipath environment and computing the optimized weights that achieve almost the same achievable rates as when the channel is known perfectly at the IRS. In this work we investigate the factors affecting IRS performance for an array size of up to 28X28 using software simulation and ray-tracing channel data. |
| （英） | There has recently been considerable interest in intelligent reflective surface (IRS) which can improve the capacity of communications links by facilitating creation of additional communication paths. The performance of an IRS system depends on the accuracy of estimating the channel and hence ability to compute accurate weights which degrade in the presence of interference. Computing the beamformer weights requires high complexity that scales with the array size. Machine learning is a promising technique for learning the multipath environment and computing the optimized weights that achieve almost the same achievable rates as when the channel is known perfectly at the IRS. In this work we investigate the factors affecting IRS performance for an array size of up to 28X28 using software simulation and ray-tracing channel data. | |
| キーワード | （和） | MIMO-OFDM / intelligent reflective surface (IRS) / ray-tracing / machine-learning / neural network / multi-layer perceptron / / |
| （英） | MIMO-OFDM / intelligent reflective surface (IRS) / ray-tracing / machine-learning / neural network / multi-layer perceptron / / | |
| 文献情報 | 信学技報, vol. 121, no. 392, SR2021-86, pp. 6-13, 2022年3月. | |
| 資料番号 | SR2021-86 | |
| 発行日 | 2022-02-23 (SR) | |
| ISSN | Online edition: ISSN 2432-6380 | |
| 著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します．(許諾番号：10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) | |
| PDFダウンロード | SR2021-86 |
|
[研究会発表申込システムのトップページに戻る]