| 講演抄録/キーワード |
| 講演名 |
2018-10-31 13:25
[ポスター講演]Neural Network-based Channel Occupancy Rate Prediction ○Hiroki Iwata・Kenta Umebayashi(TUAT)・Ahmed Al-Tahmeesschi・Miguel Lopez-Benitez(UoL) SR2018-88 |
| 抄録 |
(和) |
A spectrum occupancy prediction can provide useful information to spectrum sharing-based wireless networks. In this paper, we investigate a neural network-based channel occupancy rate prediction. Specifically, based on real occupancy rate data obtained by our spectrum measurement system, we reveal a relationship between the prediction performance and several neural network parameters, such as the number of hidden layers and the number of units in each hidden layer. Numerical evaluation shows an advantage of the neural network-based prediction by comparing with other prediction methods, such as autoregressive model. |
| (英) |
A spectrum occupancy prediction can provide useful information to spectrum sharing-based wireless networks. In this paper, we investigate a neural network-based channel occupancy rate prediction. Specifically, based on real occupancy rate data obtained by our spectrum measurement system, we reveal a relationship between the prediction performance and several neural network parameters, such as the number of hidden layers and the number of units in each hidden layer. Numerical evaluation shows an advantage of the neural network-based prediction by comparing with other prediction methods, such as autoregressive model. |
| キーワード |
(和) |
COR Prediction / Dynamic Spectrum Access / Smart Spectrum Access / Spectrum Measurement / / / / |
| (英) |
COR Prediction / Dynamic Spectrum Access / Smart Spectrum Access / Spectrum Measurement / / / / |
| 文献情報 |
信学技報, vol. 118, no. 274, SR2018-88, pp. 75-82, 2018年10月. |
| 資料番号 |
SR2018-88 |
| 発行日 |
2018-10-23 (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 |
SR2018-88 |
|