Characterizing the Power of Moving Target Defense via Cyber Epidemic Dynamics

Yujuan Han, Wenlian Lu, Shouhuai Xu

Abstract

Moving Target Defense (MTD) can enhance the resilience of cyber systems against attacks. Although there have been many MTD techniques, there is no systematic understanding and quantitative characterization of the power of MTD. In this paper, we propose to use a cyber epidemic dynamics approach to characterize the power of MTD. We define and investigate two complementary measures that are applicable when the defender aims to deploy MTD to achieve a certain security goal. One measure emphasizes the maximum portion of time during which the system can afford to stay in an undesired configuration (or posture), without considering the cost of deploying MTD. The other measure emphasizes the minimum cost of deploying MTD, while accommodating that the system has to stay in an undesired configuration (or posture) for a given portion of time. Our analytic studies lead to algorithms for optimally deploying MTD.

Introduction

Moving Target Defense (MTD) is a proactive security paradigm designed to increase the uncertainty and complexity for attackers by dynamically changing a system's configuration. While intuitively effective, MTD has historically lacked a rigorous, quantitative framework to measure its efficacy. Most existing evaluations are limited to specific simulations or localized views. This research addresses this gap by applying cyber epidemic dynamics to provide a global, analytic characterization of MTD's "power."

Methodology: Cyber Epidemic Dynamics

The core of the proposed framework is the modeling of malware or attack propagation as a biological epidemic on a network graph. In this model, nodes represent system components and edges represent potential attack vectors.

Configuration Switching and Stability

The system is modeled as switching between various configurations $C_1, C_2, \dots, C_N$. Typically, $C_1$ represents an "undesired" or insecure configuration (often necessary for performance or legacy reasons), while $C_j$ ($j \ge 2$) are MTD-induced secure configurations. The stability of the system—its ability to converge to a "clean state"—is determined by an epidemic threshold. Specifically, global security is maintained if the cure rate ($\beta$) and infection rate ($\gamma$) satisfy a condition relative to the largest eigenvalue of the network's adjacency matrix, $\lambda_1(A)$.

Configuration switching timeline between insecure ($C_1$) and secure ($C_j$) states.

Quantifying the Power of MTD

The research introduces two primary measures to quantify MTD power:

  1. The Insecurity Budget ($\pi_1^*$): This measure defines the maximum fraction of time a system can reside in an insecure configuration $C_1$ without triggering a global epidemic. The study proves that a system does not need to be in a secure state 100% of the time to remain globally stable.
  2. Minimum Deployment Cost: This measure focuses on minimizing the operational overhead of switching between MTD configurations while still satisfying the global stability condition for a given "insecurity budget."

Optimal MTD Deployment

The study reveals that the optimal strategy for deploying MTD depends significantly on the nature of the cost function associated with defense configurations.

  • Convex Cost Functions: When the cost of defense is convex, the optimal strategy involves switching between the two "closest" secure configurations to the insecure state.
  • Concave Cost Functions: For concave cost functions, the defender should ideally utilize only the single most effective (and likely most expensive) configuration in conjunction with the insecure state.
Comparison of optimal MTD strategies for convex (a) and concave (b) cost functions.

Conclusion

By providing analytic formulas and optimization algorithms, this work moves MTD from empirical observation to theoretical predictability. It demonstrates that global stability can be achieved even when a system is periodically insecure, provided the MTD strategy is properly orchestrated. Future work will address the challenges of unknown attack parameters, strategic attackers, and network heterogeneity.

Summary updated

Abstract

Moving Target Defense (MTD) can enhance the resilience of cyber systems against attacks. Although there have been many MTD techniques, there is no systematic understanding and {\em quantitative} characterization of the power of MTD. In this paper, we propose to use a cyber epidemic dynamics approach to characterize the power of MTD. We define and investigate two complementary measures that are applicable when the defender aims to deploy MTD to achieve a certain security goal. One measure emphasizes the maximum portion of time during which the system can afford to stay in an undesired configuration (or posture), without considering the cost of deploying MTD. The other measure emphasizes the minimum cost of deploying MTD, while accommodating that the system has to stay in an undesired configuration (or posture) for a given portion of time. Our analytic studies lead to algorithms for optimally deploying MTD.
BibTeX
@inproceedings{han2014characterizing,
  author = {Yujuan Han and Wenlian Lu and Shouhuai Xu},
  title = {{Characterizing the Power of Moving Target Defense via Cyber Epidemic Dynamics}},
  booktitle = {HotSoS},
  year = {2014},
  url = {https://openreview.net/forum?id=reAAD0idTO},
}
Figure