Secure Data Sharing Framework Based on Supervised Machine Learning Detection System for Future SDN-Based Networks

Securing Data-sharing mechanism between Software Defined Networks (SDN) nodes represent one of the biggest challenges in SDN context. In fact, attackers may steal or perturb flows in SDN by performing several types of attacks such as address resolution protocol poisoning, main in the middle and rogue nodes attacks. These attacks are very harm full to SDN networks as they can be performed easily and passively at all SDN layers. Furthermore, data-sharing permit to an attacker to gather all sensitive flows and data from SDN architecture. In this chapter, we will propose a framework for secure data sharing that detect and stop intrusions in SDN context while ensuring authentication and privacy. To do so, we propose a defense mechanism that detect and reduce the risk of attacks based on advanced machine learning techniques. The learning and data pre-processing steps was performed by using a constructed data set dedicated to SDN context. The simulation results show that our framework can effectively and efficiently address sniffing attacks that can be detected and stopped quickly. Finally, we observe high accuracy with a low false-positive for attack detection. © 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG.