Quantitative Cyber Risk Assessment for Distributed Communication Systems Using ALM Distributions

The paper analyzes the probability of a successful cyberattack on a computer network during the inter-patch period. The risk of compromise continuously increases from the moment the network is launched, but sharply decreases after the installation of a security update (patch) as the means available to the attacker become ineffective. This phenomenon, referred to as ”partial forgetting” of vulnerabilities, is proposed to be modeled using Almost Lack of Memory (ALM) distributions. It is shown that such distributions adequately describe the non-monotonic dynamics of network security under periodic system updates.