Maximizing the Effectiveness of an Advanced Persistent Threat.

Xiaofan Yang,Tianrui Zhang,Luxing Yang,Luosheng Wen,Yuan Yan Tang

2017-01-01

Abstract:To achieve an intended objective, an adversary may conduct an advanced persistent threat (APT) campaign against a targeted cyber network. Before an APT attack is launched, the attacker must maximize the effectiveness of the attack by properly allocating available APT resource. This paper addresses the APT effectiveness maximization problem. First, an APT-related cyber attack-defense process is modeled as an individual-level dynamical system, and the APT effectiveness maximization problem is modeled as a constrained optimization problem. Second, a type of good APT resource allocation schemes, which are known as Genetic-Algorithm-Based (GAB) schemes, are derived by solving the established optimization problem with a well-designed genetic algorithm. Next, the influences of different factors, including the available APT resource per unit time, the attack duration and the network heterogeneity, on the cost effectiveness of a GAB scheme are concluded through computer simulations. Finally, five types of heuristic APT resource allocation schemes are considered, and an experimental comparison among the cost effectiveness of these schemes and GAB schemes is conducted. This work helps understand the pros and cons of APTs.