Cloud security game theory scoring from predation models in simulation

The economics of cloud computing result from amassing computation resources while being able to distribute workload in space and time. The backbone of this ability is virtualization, which abstracts the host hardware, sharing it through virtual machines. This means of interface is also a primary vehicle and target for attackers. The counter-measures to this threat consider the costs and benefits to the cloud’s essential functions. Where the future development of the cloud is also considered, this competition between attackers and victims can be modeled in extended game theory. Yet, the attacker and victim costs and benefits, expressed as measures of expense and utility, necessary for game-theory methods are elusive. This paper establishes such a game as a predator–prey contest played out on a data-center environment. A set of contestant parameters are applied at the threshold of a viable model to the characteristic boundaries. Measurement of system health is extracted in relief with individual cost and benefit then contrasted to risk. An examination of metrics capable of validating extended interaction is found to demonstrate variation on three orders of magnitude.