Autonomous attack mitigation through firewall reconfiguration

Packet filtering firewalls represent a main defense line against cyber attacks that target computer networks daily. However, the traditional manual approaches for their configuration are no longer applicable to next-generation networks, which have become much more complex after the introduction of virtualization paradigms. Some automatic strategies have been investigated in the literature to change that old-fashioned configuration approach, but they are not fully autonomous and still require several human interventions. In order to overcome these limitations, this paper proposes an autonomous approach for firewall reconfiguration where all steps are automated, from the derivation of the security requirements coming from the logs of intrusion detection systems to the deployment of the automatically computed configurations. A core component of this process is React-VEREFOO, which models the firewall reconfiguration problem as a Maximum Satisfiability Modulo Theories problem, allowing the combination of full automation, formal verification, and optimization in a single technique. An implementation of this proposal has undergone experimental validation to show its effectiveness and performance.