Presentation | 2023-01-19 Multi-Input RNN Based Proactive Prediction of Path Loss using Building Information in UMa Environments Motoharu Sasaki, Naoki Shibuya, Kenichi Kawamura, Nobuaki Kuno, Minoru Inomata, Wataru Yamada, Takatsune Moriyama, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | We report a multi-input RNN model that predicts path loss after 5 seconds using GRU (Gated Recurrent Unit), which is one of RNN (Recurrent Neural Network), as deep learning. As the input information for the multi-input RNN model, we use the time-series data of the latest path loss of the mobile terminal, the surrounding building information of the current position, and the surrounding building information of the prediction target position. The training data and validation data are the path loss measured in Yokosuka City, Kanagawa Prefecture, and the measurement frequencies are 2.2 GHz, 4.7 GHz, and 26.4 GHz. With the proposed model, the RMSE of the prediction results for the validation data was 3.6 dB, 3.8 dB, and 3.7 dB at 2.2 GHz, 4.7 GHz, and 26.4 GHz, respectively. The proposed model can improve the prediction accuracy compared to DNN (deep neural network) model using only building information and the RNN model using only the latest path loss. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Deep learning / RNN / GRU / path loss / proactive prediction / connected car |
Paper # | AP2022-180 |
Date of Issue | 2023-01-12 (AP) |
Conference Information | |
Committee | AP / WPT |
---|---|
Conference Date | 2023/1/19(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Hiroshima Institute of Technology |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Radio propagation, Wireless transmission technology, Antennas and Propagation |
Chair | Hiroshi Yamada(Niigata Univ.) / Kenjiro Nishikaa(Kagoshima Univ.) |
Vice Chair | Mitoshi Fujimoto(Fukui Univ) / Hiroshi Hirayama(Nagoya Inst. of Tech.) |
Secretary | Mitoshi Fujimoto(National Defense Academy) / Hiroshi Hirayama(Mitsubishi Electric) |
Assistant | Tomoki Murakami(NTT) / Asako Suzuki(Fujiwaves) / Yuki Tanaka(Panasonic) |
Paper Information | |
Registration To | Technical Committee on Antennas and Propagation / Technical Committee on Wireless Power Transfer |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Multi-Input RNN Based Proactive Prediction of Path Loss using Building Information in UMa Environments |
Sub Title (in English) | |
Keyword(1) | Deep learning |
Keyword(2) | RNN |
Keyword(3) | GRU |
Keyword(4) | path loss |
Keyword(5) | proactive prediction |
Keyword(6) | connected car |
1st Author's Name | Motoharu Sasaki |
1st Author's Affiliation | NTT(NTT) |
2nd Author's Name | Naoki Shibuya |
2nd Author's Affiliation | NTT(NTT) |
3rd Author's Name | Kenichi Kawamura |
3rd Author's Affiliation | NTT(NTT) |
4th Author's Name | Nobuaki Kuno |
4th Author's Affiliation | NTT(NTT) |
5th Author's Name | Minoru Inomata |
5th Author's Affiliation | NTT(NTT) |
6th Author's Name | Wataru Yamada |
6th Author's Affiliation | NTT(NTT) |
7th Author's Name | Takatsune Moriyama |
7th Author's Affiliation | NTT(NTT) |
Date | 2023-01-19 |
Paper # | AP2022-180 |
Volume (vol) | vol.122 |
Number (no) | AP-339 |
Page | pp.pp.18-23(AP), |
#Pages | 6 |
Date of Issue | 2023-01-12 (AP) |