Towards Secure Monitoring and Control Systems: Diversify!

This work discusses the role of diversity as a mean towards secure monitoring and control. The intuition underlying the proposal is that diversity can be leveraged to raise the effort it takes to conduct a successful attack (in terms of attack resources and time) to such a level so as to make it pointless to attempt an attack at all. For example, let us consider an attack that requires compromising two machines in order to be successful. If the machines are identical, it suffices to compromise one machine and then repeating the exploit for the other, i.e., the chance of a successful attack PSA to the system is related to the chance of compromising just one machine (PSA≈PM). When the machines are different, PSA is smaller because it becomes somewhat related to chance of compromising each machine separately (i.e., PSA≈PM1×PM2): succeeding is harder and time-consuming. Diversity is not used here to replicate components. We claim that a monitoring and control system, when possible, can smartly combine diverse technologies to significantly increase the effort to conduct a successful attack. Key aspects, issues and future research directions are briefly discussed in the following.