| Presentation | 2021-11-05 A Study on Faster Rardio Signal Detection Using Deep Learning Taichi Ohtsuji, Taro Abe, Katsuaki Suzuki, Toshiki Takeuchi, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In order to use the spectrum efficiently, it is required to be able to detect signals in real time and with high frequency resolution for broadband radio signals. Conventional radio signal detection methods have problems in automating the power threshold setting, which may lead to false or missed detections. Recently, methods of applying deep-learning-based object detection to radio signal detection has been proposed. However, if the signal detection process based on deep object detection is performed only by the central processing unit (CPU), the required processing time may not be satisfied. In this study, to realize real-time signal detection based on deep object detection using only the CPU, a reduction method of processing time is investigated. Evaluation using measured data shows that the processing frame rate can exceed the required frame rate of 10 fps by using the image partitioning method together with the YOLOv3-tiny model. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Radio Signal Detection / Deep Learning / Object Detection / YOLO |
| Paper # | SR2021-51 |
| Date of Issue | 2021-10-28 (SR) |
| Conference Information | |
| Committee | SR |
|---|---|
| Conference Date | 2021/11/4(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Online |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Software Defined Radio, Cognitive Radio, Spectrum Shareing, etc. |
| Chair | Suguru Kameda(Hiroshima Univ.) |
| Vice Chair | Osamu Takyu(Shinshu Univ.) / Kentaro Ishidu(NICT) / Kazuto Yano(ATR) |
| Secretary | Osamu Takyu(Mie Univ.) / Kentaro Ishidu(Tokai Univ.) / Kazuto Yano(NTT) |
| Assistant | Mai Ohta(Fukuoka Univ.) / Taichi Ohtsuji(NEC) / WANG Xiaoyan(Ibaraki Univ.) / Akemi Tanaka(MathWorks) |
| Paper Information | |
| Registration To | Technical Committee on Smart Radio |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A Study on Faster Rardio Signal Detection Using Deep Learning |
| Sub Title (in English) | |
| Keyword(1) | Radio Signal Detection |
| Keyword(2) | Deep Learning |
| Keyword(3) | Object Detection |
| Keyword(4) | YOLO |
| 1st Author's Name | Taichi Ohtsuji |
| 1st Author's Affiliation | NEC Corporation(NEC) |
| 2nd Author's Name | Taro Abe |
| 2nd Author's Affiliation | NEC Corporation(NEC) |
| 3rd Author's Name | Katsuaki Suzuki |
| 3rd Author's Affiliation | NEC Corporation(NEC) |
| 4th Author's Name | Toshiki Takeuchi |
| 4th Author's Affiliation | NEC Corporation(NEC) |
| Date | 2021-11-05 |
| Paper # | SR2021-51 |
| Volume (vol) | vol.121 |
| Number (no) | SR-227 |
| Page | pp.pp.58-64(SR), |
| #Pages | 7 |
| Date of Issue | 2021-10-28 (SR) |