Unmasking SDN flow table saturation: fingerprinting, attacks and defenses

Abstract

Software-Defined Networking stands as a pivotal technology in attaining the essential levels of flexibility and scalability demanded by pervasive and high-performance network infrastructure required for digital connected services. Nonetheless, its disaggregated and layered architecture makes it open to the time-based fingerprinting attacks. Besides, limited flow table capacity of the switches alleviates table saturation attacks. In this paper, an automated attacker tool called TASOS is proposed to infer flow table utilization rate, size and replacement algorithm. With this set of information, the attacker can conduct intelligent saturation attacks. Furthermore, a lightweight defense mechanism (LIDISA) for proactively deleting flow rules is described. A comprehensive simulation setup with different network conditions shows that the proposed techniques achieve superior success rate in diverse settings.