آمار
| تعداد نشریات | 45 |
| تعداد شمارهها | 1,469 |
| تعداد مقالات | 17,958 |
| تعداد مشاهده مقاله | 58,289,404 |
| تعداد دریافت فایل اصل مقاله | 19,748,444 |
Enhancing DDoS Attack Detection in Software Defined Networks with Deep Reinforcement Learning | ||
| Computational Methods for Differential Equations | ||
| مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 10 دی 1404 اصل مقاله (998.34 K) | ||
| نوع مقاله: Research Paper | ||
| شناسه دیجیتال (DOI): 10.22034/cmde.2025.65880.3056 | ||
| نویسندگان | ||
| Khashayar Delavari؛ Mehran Shetabi؛ Sayed Alireza Sadrossadat* | ||
| Department of Computer Engineering, Yazd University, Yazd 8915818411, Iran. | ||
| چکیده | ||
| The rapid growth of Software Defined Networking (SDN) offers significant benefits in network flexibility, management, and scalability. However, the centralization of control in SDN poses substantial security risks, especially from Distributed Denial of Service (DDoS) attacks. Traditional detection mechanisms often fall short due to the evolving nature of these threats. This paper introduces a novel Deep Reinforcement Learning (DRL) technique to enhance DDoS attack detection and mitigation in SDN environments. By leveraging DRL’s adaptive learning capabilities, the proposed model continuously learns and adapts to new attack patterns, providing robust defense. The model employs a combination of Autoencoder (AE) and Bidirectional Gated Recurrent Unit (BGRU) to analyze traffic patterns and detect anomalies effectively. Experimental results, using a comprehensive dataset from real network traffic, demonstrate the model’s superior accuracy, higher detection rate, and reduced false-positive rates compared to existing methods. Additionally, the proposed technique incorporates a trust value mechanism to mitigate detected attacks, ensuring enhanced security and reliability for SDN networks. | ||
| کلیدواژهها | ||
| DDOS attacks؛ Deep learning؛ Deep reinforcement learning؛ machine learning؛ Software-Defined Networks | ||
|
آمار تعداد مشاهده مقاله: 32 تعداد دریافت فایل اصل مقاله: 37 |
||