| 研究課題/領域番号 |
20K11764
|
|---|
| 研究機関 | 茨城大学 |
|---|
研究代表者 |
王 瀟岩 茨城大学, 理工学研究科(工学野), 准教授 (10725667)
|
|---|
| 研究分担者 |
梅比良 正弘 茨城大学, 理工学研究科(工学野), 教授 (00436239)
|
|---|
| 研究期間 (年度) |
2020-04-01 – 2023-03-31
|
| キーワード | wireless access / reinforcement learning |
|---|
| 研究実績の概要 |
In FY2020, we started from the fundamental workload, i.e., optimizing local radio resource management, by
designing distributed deep reinforcement learning based approach. The intelligence is placed at user equipments, who learn their wireless access decisions by relying only on a local set of observations from the wireless environment, such as channel quality and interference levels. We clarified the tradeoff between
allocated radio resource’s granularity and learning algorithm’s convergence speed by simulations on TensorFlow. We also evaluated the performance of the proposed scheme in terms of transmission delay and packet drop rate by comparing baseline schemes.
|
| 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
We consider that the research progresses smoothly. We have published 2 journal papers, 2 internal conference papers and multiple domestic conference papers in FY2020 under the support of this funding.
|
| 今後の研究の推進方策 |
In the following year, we will consider the problems with sophisticated and practical models. Meanwhile, we will perform the experiments by realizing the proposed scheme in testbeds.
|
| 次年度使用額が生じた理由 |
コロナウィルス感染防止のため、多数の国際・国内学会が中止するため、残額が生じてしまう。残りの助成金は２０２１年度の学会の参加費と学術論文の登録費として使用する予定である。
|
