透過您的圖書館登入
IP:18.224.214.215
  • 會議論文
  • OpenAccess

在SDN/NFV環境下運用機器學習偵測DDoS攻擊之研究

A Study of Machine Learning based DDoS Detection Mechanism in SDN/NFV Environment

摘要


網路技術的快速發展使駭客的攻擊方式趨於成熟化與多樣化,傳統網路的架構已無法因應多元化的攻擊來源。軟體定義網路(Software Defined Network, SDN)與網路功能虛擬化(Network Functions Virtualization, NFV)技術的出現,為網路業者帶來了創新的變革,未來網路安全的架構設計將朝向可程式化與虛擬化的方向演進,不僅具備了彈性管理與降低硬體建置成本等優勢外,也有利於發展網路安措施以抵禦惡意攻擊的威脅。本論文使用OpenStack雲端運算平台來實踐NFV開發環境,並結合SDN技術架設SDN/NFV網路安全技術實驗環境。在實驗環境中,模擬分散式阻斷服務攻擊(Distributed denial-of-service attack, DDoS),並使用機器學習演算法訓練出的訊務分類器,以達到即時性網路流量監控與攻擊的偵測的效果。

並列摘要


The rapid development of network technology has made hackers' attack methods mature and diversified. The traditional network architecture has been unable to respond to diversified sources of attacks. The emergence of Software Defined Network (SDN) and Network Functions Virtualization (NFV) technology has brought about innovative changes for network operators. The future architecture of network security will be developed towards programmable and virtualization. It not only has the advantages of flexible management and low cost of hardware construction, but also helps to develop network security measures to protect against the threat of malicious attacks. This paper uses OpenStack cloud computing platform to practice NFV development environment, and combines SDN technology to set up SDN/NFV network security technology experimental environment. It achieves real-time network traffic monitoring and attack detection through the establishment of distributed denial-of-service attack (DDoS) simulation and attack monitoring technology.

延伸閱讀