Survey on SDN-based Intrusion Detection Systems

With the growing reliance on the Internet and today's interconnected world, cyber threats have been tremendously increased. Thus, Internet is vulnerable to a wide range of attacks that risk its data. Ensuring the security and integrity of information transmitted and stored across networks has become crucial objective. Intrusion Detection Systems (IDS) play a critical role in detecting security breaches within computer networks and mitigating the potential loss of highly valuable data. Recently, Software-Defined Networks (SDN) have revolutionized the traditional networking paradigm by decoupling the control plane from the data plane, thereby enhanced network management and agility. This paper provides a literature survey on SDN-based IDSs, their classifications, and their different implementations using recent machine and deep learning techniques. Exploiting machine and deep learning in IDS contributed significantly in IDSs detection accuracy as well as focused on a new area of data security, including the adopted learning models, attributes, and learning parameters. Not only the detection accuracy impact, but also optimizing and accelerating inference time is a common goal in the deployment of IDSs while implementing machine and deep learning techniques. This paper aims to create awareness and understanding of the importance of implementing robust security measures to protect sensitive information and ensure the integrity and availability of network resources.