SDN-based Firewall using Machine Learning

Due to the widespread use of the Internet, the traditional destination-based forwarding which includes two steps: (1) match and (2) action, has caused much stress on the network due to its traffic and security issues. With SDN, network control has become simpler. In addition to simplification, network security has always been a critical aspect of networking, where a firewall helps to secure traffic by enforcing some security policies. Based on the packet headers, this firewall can detect malicious packets and then drop or block them. In this paper, we present a firewall that uses an intuitive machine-learning algorithm to detect malicious packets based on the header. The firewall has been implemented in two SDN controllers: (1) RYU controller and (2) POX, for a simple mesh topology and compared based on packet loss and delay.

Paper

Full text

PDF

SDN-based Firewall using Machine Learning

Semantic Scholar · Computer Science · 2023

Abstract

Due to the widespread use of the Internet, the traditional destination-based forwarding which includes two steps: (1) match and (2) action, has caused much stress on the network due to its traffic and security issues. With SDN, network control has become simpler. In addition to simplification, network security has always been a critical aspect of networking, where a firewall helps to secure traffic by enforcing some security policies. Based on the packet headers, this firewall can detect malicious packets and then drop or block them. In this paper, we present a firewall that uses an intuitive machine-learning algorithm to detect malicious packets based on the header. The firewall has been implemented in two SDN controllers: (1) RYU controller and (2) POX, for a simple mesh topology and compared based on packet loss and delay.

Similar papers

© 2026 NYSGPT2525 LLC