Machine learning in an SDN network environment for DoS attacks

Abstract

Denial of service (DoS) attacks in Software-Defined Network (SDN) environments are increasing despite the capabilities and benefits of SDN. Software-based traffic analysis, centralized control and automatic information forwarding offered by SDN, makes it easier to detect and react effectively to DoS attacks; however, the security of the SDN itself has not yet been resolved, and there are a number of potential vulnerabilities not only of the DoS type, on the SDN platforms. In this document, we review some applications and defense mechanisms to mitigate these types of attacks in an SDN network environment. This work could help us to understand the advantages of SDNs compared to current network architectures, without leaving aside the security issue that will continue to be maintained over the years.