| 研究課題/領域番号 |
20K04484
|
|---|
| 研究種目 |
基盤研究(C)
|
| 配分区分 | 基金 |
|---|
| 応募区分 | 一般 |
|---|
| 審査区分 |
小区分21020:通信工学関連
|
|---|
| 研究機関 | 岡山大学 |
|---|
研究代表者 |
侯 亜飛 岡山大学, 自然科学学域, 助教 (60598457)
|
|---|
| 研究分担者 |
上原 一浩 岡山大学, 自然科学研究科, 教授 (10221798)
冨里 繁 岡山大学, 自然科学研究科, 准教授 (60362951)
田野 哲 岡山大学, 自然科学研究科, 教授 (80378835)
|
|---|
| 研究期間 (年度) |
2020-04-01 – 2024-03-31
|
| 研究課題ステータス |
交付 (2022年度)
|
| 配分額 *注記 |
4,420千円 (直接経費: 3,400千円、間接経費: 1,020千円)
2022年度: 910千円 (直接経費: 700千円、間接経費: 210千円)
2021年度: 1,820千円 (直接経費: 1,400千円、間接経費: 420千円)
2020年度: 1,690千円 (直接経費: 1,300千円、間接経費: 390千円)
|
|---|
| キーワード | LCX localization / Particle filter / MIMO capacity / Spatial modulation / 2D/3D localization / MIMO / Neural Network / RSSI prediction / LCX位置推定 / 無線システム / 信号処理 / アンテナ技術 |
|---|
| 研究開始時の研究の概要 |
This research will investigate 2D/3D indoor localization using the combinations of multiple LCXs. We will research channel models related to LCX parameters and user position, and improve the localization accuracy using multiple θs, DTOAs. Finally, a new particle filter like interactive localization method using multiple LCXs will be analyzed and optimized.
|
| 研究実績の概要 |
In this year, a particle filter-based localization method is proposed. The proposed method uses LCXs to receive the signals from the target user. The TOAs of the signals at are utilized as the indicator for filtering in the localization process.
Different from localization methods in the previous work, the proposed particle filter-based method can be applied not only in the 2-D localization scenario but also in the 3-D localization scenario. The results show that the proposed method can achieve promising localization accuracy in both 2-D and 3-D scenarios.
Furthermore, we also investigated the capacity improvement of LCX system using spatial modulation with the on/off the LCX slots. We proposed a solution using Genetic Algorithms (GA) to find an optimal solution for On/off patterns.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
1: 当初の計画以上に進展している
理由
The results for 2D indoor multipath-rich environment showed the mean error of the conventional TDOA method is 1.8 m. However, for the particle filter -based method with the number of particles as 1000, the mean errors of the particle filter method can be 0.69 m. The 90% accuracy of the TDOA method is 2.9 m, and the 90% accuracy of the particle filter method is 0.89 m.
For 3D indoor multipath-rich environment, the mean value of the location error, when the number of particles (N) is set as 1000, is 0.94 m. When N is reduced to 100 and 500 which can have less computational complexity, the location errors are 2.47 m and 1.43 m respectively.
All results show that the localization accuracy of the proposed particle filter-based localization method for both 2D and 3D scenarios is promising.
|
| 今後の研究の推進方策 |
In this year, the particle filter based interactive and iterative scheme has been proposed and evaluated. However, there exists many unknown questions. For example, how to set the optimal parameters for particle filter-based algorithm? How to reduce the number of particle filter and the complexity.
Till now, we only considered single user localization. The simultaneous indoor localization of multiple users is more important due to the coming of IoT era with massive devices. In addition, new antenna technologies such as reconfigurable intelligent surfaces (RIS) are emerging which can be combined for LCX-based indoor scenarios. These new factors will be researched in following years.
|
