Anomaly traceback using software defined networking

While the threats in Internet are still increasing and evolving (like intra multi-tenant data center attacks), protection and detection mechanisms are not fully accurate. Therefore, forensics is vital for recovering from an attack but also to identify the responsible entities. Therefore, this paper focuses on tracing back to the sources of an anomaly in the network. In this paper, we propose a method leveraging the Software Defined Networking (SDN) paradigm to passively identify switches composing the network path of an anomaly. As SDN technologies tend to be deployed in the next generation of networks including in data centers, they provide a helpful framework to implement our proposal without developing dedicated routers like usual IP traceback techniques. We evaluated our scheme with different network topologies (Internet and data centers) by considering distributed attacks with numerous hosts.