An Intelligent SDN-Enabled Framework for Real-Time IPv6 IP Spoofing Detection and Prevention
TL;DR
This paper provides a critical review of the available IPv6 spoofing prevention methods and proposes an enhanced multi-layered framework that uses improved ingress and egress filtering, Artificial Intelligence (AI)-based anomaly detection, and Software-Defined Networking (SDN) to automatically mitigate threats in real time.
Abstract
With the increase in the adoption of Internet Protocol version 6 (IPv6), the issue of Internet Protocol (IP) spoofing has remained a significant risk that damages the integrity of networks. Although some countermeasures, like Secure Neighbor Discovery (SEND) and Router Advertisement (RA) Guard, are available, these solutions are complex to deploy, consume high resources, and can be bypassed by advanced attacks. This paper provides a critical review of the available IPv6 spoofing prevention methods and proposes an enhanced multi-layered framework that uses improved ingress and egress filtering, Artificial Intelligence (AI)-based anomaly detection, and Software-Defined Networking (SDN) to automatically mitigate threats in real time. The enhanced framework is precisely identifying spoofed traffic, whereas SDN dynamically applies policies in the network to isolate and block malicious sources without affecting legitimate traffic. The framework is implemented over a Mininet and Docker-based setup in which it is tested under controlled and real-world traffic conditions. It had an average precision and recall of 0.98 with an F1-score of 0.97 and a prevention latency averaging less than 0.4 seconds. Through the centralized traffic visibility provided by SDN and the pattern recognition capabilities of Machine Learning (ML), the proposed framework adjusts smoothly to traffic changes and new spoofing techniques. Its modular architecture allows it to be integrated into existing systems and deployed in a manner that does not interfere with the flow of operations. The proposed solution is better in robustness, scalability, and adaptability than any traditional static filtering approach, and therefore it is a practical approach to enhance security in both dual-stack and IPv6-only network environments. This research provides practical implications and further highlights the possibility of ML and SDN contributions in the creation of proactive IPv6 security systems