Multistage Attack Graph Security Games: Heuristic Strategies, with Empirical Game-Theoretic Analysis

Nguyen, Thanh H.; Wright, Mason; Wellman, Michael P.; Singh, Satinder

doi:10.1155/2018/2864873

Cited by 24 publications

(31 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While these complex economic problems continue to be a primary application area of these methods [73][74][75][76], the general techniques have been applied in many different settings. These include analysis of interactions among heuristic meta-strategies in poker [77], network protocol compliance [78], collision avoidance in robotics [79], and security games [80][81][82].…”

Section: Most Closely Related Workmentioning

confidence: 99%

α-Rank: Multi-Agent Evaluation by Evolution

Omidshafiei¹,

Papadimitriou

Piliouras

et al. 2019

Sci Rep

122

View full text Add to dashboard Cite

We introduce α - Rank , a principled evolutionary dynamics methodology, for the evaluation and ranking of agents in large-scale multi-agent interactions, grounded in a novel dynamical game-theoretic solution concept called Markov - Conley chains (MCCs). The approach leverages continuous-time and discrete-time evolutionary dynamical systems applied to empirical games, and scales tractably in the number of agents, in the type of interactions (beyond dyadic), and the type of empirical games (symmetric and asymmetric). Current models are fundamentally limited in one or more of these dimensions, and are not guaranteed to converge to the desired game-theoretic solution concept (typically the Nash equilibrium). α -Rank automatically provides a ranking over the set of agents under evaluation and provides insights into their strengths, weaknesses, and long-term dynamics in terms of basins of attraction and sink components. This is a direct consequence of the correspondence we establish to the dynamical MCC solution concept when the underlying evolutionary model’s ranking-intensity parameter, α , is chosen to be large, which exactly forms the basis of α -Rank. In contrast to the Nash equilibrium, which is a static solution concept based solely on fixed points, MCCs are a dynamical solution concept based on the Markov chain formalism, Conley’s Fundamental Theorem of Dynamical Systems, and the core ingredients of dynamical systems: fixed points, recurrent sets, periodic orbits, and limit cycles. Our α -Rank method runs in polynomial time with respect to the total number of pure strategy profiles, whereas computing a Nash equilibrium for a general-sum game is known to be intractable. We introduce mathematical proofs that not only provide an overarching and unifying perspective of existing continuous- and discrete-time evolutionary evaluation models, but also reveal the formal underpinnings of the α -Rank methodology. We illustrate the method in canonical games and empirically validate it in several domains, including AlphaGo, AlphaZero, MuJoCo Soccer, and Poker.

show abstract

Section: Most Closely Related Workmentioning

confidence: 99%

α-Rank: Multi-Agent Evaluation by Evolution

Omidshafiei¹,

Papadimitriou

Piliouras

et al. 2019

Sci Rep

122

View full text Add to dashboard Cite

show abstract

“…While these complex economic problems continue to be a primary application area of these methods [5,37,38,41], the general technique has been applied in many different settings. These include analysis interaction among heuristic meta-strategies in poker [24], network protocol compliance [43], collision avoidance in robotics [11], and security games [20,25,48]. Research that followed on Walsh's [39] initial work branched off in two directions: the first strand of work focused on strategic reasoning for simulation-based games [44], while the second strand focused on the evolutionary dynamical analysis of agent behavior inspired by evolutionary game theory [31,33].…”

Section: Related Workmentioning

confidence: 99%

Bounds and dynamics for empirical game theoretic analysis

Tuyls¹,

Pérolat²,

Lanctot³

et al. 2019

Auton Agent Multi-Agent Syst

View full text Add to dashboard Cite

This paper provides several theoretical results for empirical game theory. Specifically, we introduce bounds for empirical game theoretical analysis of complex multi-agent interactions. In doing so we provide insights in the empirical meta game showing that a Nash equilibrium of the estimated meta-game is an approximate Nash equilibrium of the true underlying metagame. We investigate and show how many data samples are required to obtain a close enough approximation of the underlying game. Additionally, we extend the evolutionary dynamics analysis of meta-games using heuristic payoff tables (HPTs) to asymmetric games. The stateof-the-art has only considered evolutionary dynamics of symmetric HPTs in which agents have access to the same strategy sets and the payoff structure is symmetric, implying that agents are interchangeable. Finally, we carry out an empirical illustration of the generalised method in several domains, illustrating the theory and evolutionary dynamics of several versions of the AlphaGo algorithm (symmetric), the dynamics of the Colonel Blotto game played by human players on Facebook (symmetric), the dynamics of several teams of players in the capture the flag game (symmetric), and an example of a meta-game in Leduc Poker (asymmetric), generated by the policy-space response oracle multi-agent learning algorithm. Keywords Empirical games • Asymmetric games • Replicator dynamics 1 Introduction Using game theory to examine multi-agent interactions in complex systems is a non-trivial task, especially when a payoff table or normal form representation is not directly available. Works by Walsh et al. [39,40], Wellman et al. [43,44], and Phelps et al. [23], have shown the great potential of using heuristic strategies and empirical game theory to examine such B Karl Tuyls

show abstract

“…Lastly, we will highlight works which reason about both these factors that will affect the overall Quality of Service (QoS) when an MTD is deployed. a) Considers only performance of individual defense actions: Similar to the case of security metrics, when we look at performance considerations for an individual defense, most game-theoretic works model these as a part of the defender's utility function [62], [78], [79], [80], [95]. Various equilibrium concepts in these games yield movement strategies for the defender that gives priority to constituent defenses that have low performance costs while ensuring that the security is not impacted by a lot.…”

Section: A) Considers Only Security Of Individual Defenses: Mostmentioning

confidence: 99%

“…Various equilibrium concepts in these games yield movement strategies for the defender that gives priority to constituent defenses that have low performance costs while ensuring that the security is not impacted by a lot. In [62] and [95], the reward functions are defined at an abstract level and the authors point out that they can be used to consider the performance cost of constituent defenses. In [78] and [79], the authors consider the impact of placing Network-based Intrusion Detection Systems (NIDS) on the latency of the network and use centrality based measures as heuristic guidance for it.…”

Section: A) Considers Only Security Of Individual Defenses: Mostmentioning

confidence: 99%

“…On the other hand, modeling this as a decision problem for the defender with partial observability results in a POMDP where calculating strategies, for even small-sized cloud networks, becomes unrealistic [94]. As an academic exercise, researchers have even gone one step further and modeled the interaction as a Partial Observable Stochastic Game (POSG) [95] giving up on any hope of being useful in realistic settings.…”

Section: How To Movementioning

confidence: 99%

See 1 more Smart Citation

A Survey of Moving Target Defenses for Network Security

Sengupta

Chowdhary

Sabur

et al. 2020

IEEE Commun. Surv. Tutorials

182

View full text Add to dashboard Cite

Network defense techniques based on traditional tools, techniques, and procedures (TTP) fail to account for the attacker's inherent advantage present due to static nature of network services and configurations. Moving Target Defense (MTD), on the other hand, provides an intelligent countermeasure by dynamically re-configuring the underlying systems, thereby reducing the effectiveness of cyber attacks. In this survey, we analyze the recent advancements made in the development of MTD tools, techniques and procedures (TTP) and highlight how these defenses can be made more effective with the use of artificial intelligence techniques for decision making. We first define a unified formal notation for MTDs that can capture different aspects of such defenses. We then categorize these defenses into different sub-classes depending on how they answer the three questions-what to move, when to move and how to moveshowcasing how game theoretic strategies can effectively answer the latter question. To understand the usefulness of these defense methods, we study the implementation of such MTD techniques. We find that (relatively) new networking technologies such as Software Defined Networking (SDN) and Network Function Virtualization (NFV) provide effective means for implementing these dynamic defense methods. To encourage researchers and industry experts in using such defenses, we highlight industry use-cases and discuss the practicality and maturity of these defenses. To aid readers who want to test or deploy MTD techniques, we highlight existing MTD test-beds. Our survey then performs both a qualitative and quantitative analysis to better understand the effectiveness of these MTDs in terms of security and performance. To that extent, we use well-defined metrics such as risk analysis, performance costs for qualitative evaluation and metrics based on Confidentiality, Integrity, Availability (CIA), attack representation, QoS impact, targeted threat models and defense cost for quantitative evaluation. Finally, we conclude by summarizing research opportunities that our survey elucidates.

show abstract

Multistage Attack Graph Security Games: Heuristic Strategies, with Empirical Game-Theoretic Analysis

Cited by 24 publications

References 33 publications

α-Rank: Multi-Agent Evaluation by Evolution

α-Rank: Multi-Agent Evaluation by Evolution

Bounds and dynamics for empirical game theoretic analysis

A Survey of Moving Target Defenses for Network Security

Contact Info

Product

Resources

About