| 研究課題/領域番号 |
23K16870
|
|---|
| 研究種目 |
若手研究
|
| 配分区分 | 基金 |
|---|
| 審査区分 |
小区分60060:情報ネットワーク関連
|
|---|
| 研究機関 | 奈良先端科学技術大学院大学 |
|---|
研究代表者 |
Chen Na 奈良先端科学技術大学院大学, 先端科学技術研究科, 客員准教授 (80838342)
|
|---|
| 研究期間 (年度) |
2023-04-01 – 2026-03-31
|
| 研究課題ステータス |
交付 (2024年度)
|
| 配分額 *注記 |
4,810千円 (直接経費: 3,700千円、間接経費: 1,110千円)
2025年度: 1,170千円 (直接経費: 900千円、間接経費: 270千円)
2024年度: 1,170千円 (直接経費: 900千円、間接経費: 270千円)
2023年度: 2,470千円 (直接経費: 1,900千円、間接経費: 570千円)
|
|---|
| キーワード | Beam Management / Deep Learning / Massive MIMO / RIS/IRS / IRS / Heterogeneous Network / Cell-Free Communication |
|---|
| 研究開始時の研究の概要 |
The B5G/6G systems are confronted with high transmission requirements and a harsh wireless communication environment. This research considers a radio over fiber (RoF) supported multi-layer cell-free heterogeneous network (HetNet) architecture to achieve an efficient transmission in complex scenarios. We first propose the HetNet model and deep learning (DL) method for cooperative beam management considering the nonlinear optic fiber channel and the cell-free wireless channel, providing a possible solution for future wireless communication networks with high throughput and robustness.
|
| 研究実績の概要 |
During the previous year, with the support of this project, we made the following research achievements:
1. We researched on intelligent reflecting surface (IRS)-assisted massive multiple-input multiple-output (mMIMO) system with cell-free (CF) architecture. The IRS-mMIMO-CF system enabled users to access to all the base stations (BSs) via all the IRSs, which enhanced the spectrum efficency (SE) and saved the operation cost.
2. We developed distributed deep learning algorithm, named split federated learning (SFL) for the IRS-mMIMO-CF system. Specifically, the proposed self-enhanced multi-task SFL (SM-SFL) approach could simultaneously tackle wireless channel semantic reconstruction and cooperated beamforming with shared knowledge. The proposed model could achieve sufficient SE with imperfect channel state information (CSI), multi-device cooperation, and low labeling overhead.
3. We developed new hardware design for IRS metasurfaces. Specifically, we proposed a varactor diode mounted large-via and multi-via mushroom-type (VDLM/VDMM) structure for beam control at multiple frequency band for heterogeneous networks (HetNets). Furthermore, we propose a conformal IRS model for more flexible reflection control and stronger coverage ability.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
For the system model consideration, we developed the system to IRS-mMIMO-CF system and finished the theoretical analysis. For the deep learning algorithm design, we developed SM-SFL algorithm for distributed device cooperation with high efficiency on multiple task excecutation. For hardware implementation, we developed multiple IRS demos for HetNets and for more flexible applications that are capable for 5G/6G systems.
Overall, the research is progressing smoothly.
|
| 今後の研究の推進方策 |
Based on the current achievements, in the next step, we will continue with the HetNet coverage enhancement assisted by IRS.
1. We will research on the wireless propagation environment sensing and the influence on beam management.
2. We will also improve the IRS design into more practical multi-task applications. Specifically, we will develop an IRS controller for adaptive beam control as well as sensing integrated communication.
3. We will continue the international cooperation and hold international discussion routines and seminars to further improve our study.
|
