Anomaly detection adaptable to attack changes based on decision making of human brain
| Project/Area Number |
18K18040
|
|---|
| Research Category |
Grant-in-Aid for Early-Career Scientists
|
| Allocation Type | Multi-year Fund |
|---|
| Review Section |
Basic Section 60060:Information network-related
|
|---|
| Research Institution | Osaka University |
|---|
Principal Investigator |
Kuze Naomi 大阪大学, 基礎工学研究科, 助教 (20778071)
|
|---|
| Project Period (FY) |
2018-04-01 – 2020-03-31
|
| Project Status |
Completed (Fiscal Year 2019)
|
| Budget Amount *help |
¥2,470,000 (Direct Cost: ¥1,900,000、Indirect Cost: ¥570,000)
Fiscal Year 2019: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2018: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
|
|---|
| Keywords | セキュリティ / 集団的行動選択 / 強化学習 / 移動制御 / アンチジャミング / 情報ネットワーク / サイバーセキュリティ / 意思決定 / ネットワークセキュリティ |
|---|
| Outline of Final Research Achievements |
Because of the rapid diversification of cyber attacks, it is important to consider novel approaches, which have high adaptability to attack changes, based on decision making and learning of brains.
First, we introduced collective decision making in animal groups to self-organizing systems. In detail, we adapted the concept of flexible leadership to channel selection in wireless sensor networks, and demonstrated that with the proposed mechanism, appropriate channels could be selected according to environmental changes. Next, we focused on a mobile robot with wireless communication, proposed QoS-based anti-jamming mobile control mechanism based on reinforcement learning. We revealed that with the proposed mechanism, the agent could learn mobile scheme that avoid jamming attacks.
|
| Academic Significance and Societal Importance of the Research Achievements |
申請者は、ネットワークの大規模化、多様化に伴う、サイバー攻撃の多様化への対策のため、脳の意思決定、学習の仕組みの応用について研究を実施した。
申請者は、ヒトのグループにおける集団的行動選択においてみられるflexible leadershipの概念を無線センサネットワークにおけるチャネル選択手法に適用し、不確実な情報化で適切なチャネル選択が可能であることを示した。また、移動ロボットを対象として、QoSに基づいた強化学習を利用した無線通信を妨害するジャミング攻撃にロバストな移動制御手法を提案し、ジャミングを回避するよう移動方策の学習が可能であることを示した。
|
Report
(3 results)
Research Products
(8 results)
