CONCEAL: A strategy composition for resilient cyber deception-framework, metrics and deployment

Cyber deception is a key proactive cyber resilience technique to reverse the current asymmetry that favors adversaries in cyber warfare by creating a significant confusion in discovering and targeting cyber assets. One of the key objectives for cyber deception is to hide the true identity of the cyber assets in order to effectively deflect adversaries away from critical targets, and detect their activities early in the killchain.Although many cyber deception techniques were proposed including using honeypots to represent fake targets, and mutating IP addresses to frequently change the ground truth of the network configuration [12], none of these deception techniques is resilient enough to provide high confidence of concealing the identity of the network assets, particularly against sophisticated attackers. In fact, in this paper our analytical and experimental work showed that highly resilient cyber deception is unlikely attainable using a single technique, but it requires an optimal composition of various concealment techniques to maximize the deception utility. We, therefore, present a new cyber deception framework, called CONCEAL, which is a composition of mutation, anonymity, and diversity to maximize key deception objectives, namely, concealability, detectability and deterrence, while constraining the overall deployment cost. We formally define the CONCEAL metrics for concealability, detectability, and deterrence to measure the effectiveness of CONCEAL. Finally, we present the deployment of CONCEAL as a service to achieve manageability and costeffectiveness by automatically generating the optimal deception proxy configuration based on existing host/network configuration, risk constraints of network services, and budget constraints. Our evaluation experiments measure both the deception effectiveness based on the above metrics, as well as the scalability of the CONCEAL framework.