چارچوبی برای یادگیری عمیق به منظور افزایش امنیت شبکه
Abstract
I- Introduction
II- Software-Defined Networks and Related Work
III- Autoencoders
IV- Detection Framework
V- Conclusion
References
Abstract
Software-Defined Networks (SDN) initiates a novel networking model. SDN proposes the separation of forward and control planes by introducing a new independent plane called network controller. The architecture enhances the network resilient, decompose management complexity, and support more straightforward network policies enforcement. However, the model suffers from severe security threats. Specifically, a centralized network controller is a precious target for two reasons. First, the controller is located at a central point between the application and data planes. Second, a controller is software which prone to vulnerabilities, e.g., buffer and stack overflow. Hence, providing security measures is a crucial procedure towards the fully unleash of the new model capabilities. Intrusion detection is an option to enhance the networking security. Several approaches were proposed, for instance, signature-based, and anomaly detection. Anomaly detection is a broad approach deployed by various methods, e.g., machine learning. For many decades intrusion detection solution suffers performance and accuracy deficiencies. This paper revisits network anomalies detection as recent advances in machine learning particularly deep learning proofed success in many areas like computer vision and speech recognition. The study proposes an intrusion detection framework based on unsupervised deep learning algorithms.
Introduction
The conventional communication networking model consists of three planes. i.e., management, control, and forward or data. The management plane supports network monitoring and configuration. The control plane populates forwarding tables on the physical devices. Consecutively, the forward plane switches packets to ingress and egress ports based on the forwarding tables. For decades, both the Control and the forward planes are integrated into the same networking devices, for instance. Switches or routers. The conventional model provided efficiency from a performance perspective. However, current networks became excessively complicated, and there is a necessity to adopt a more resilient architecture [1]. This paper introduces a framework to enhance the security deficiencies of SDN.The framework is anomalies detection based on machine learning. The next section discusses SDN model and related security threats. The third section investigates the deep learning and its current anomalies detection solution for network security. The fourth section represents our proposed framework.