Charles Kamhoua

Stephanie Milani, Zhicheng Zhang, Nicholay Topin et al.

Many recent breakthroughs in multi-agent reinforcement learning (MARL) require the use of deep neural networks, which are challenging for human experts to interpret and understand. On the other hand, existing work on interpretable reinforcement learning (RL) has shown promise in extracting more interpretable decision tree-based policies from neural networks, but only in the single-agent setting. To fill this gap, we propose the first set of algorithms that extract interpretable decision-tree policies from neural networks trained with MARL. The first algorithm, IVIPER, extends VIPER, a recent method for single-agent interpretable RL, to the multi-agent setting. We demonstrate that IVIPER learns high-quality decision-tree policies for each agent. To better capture coordination between agents, we propose a novel centralized decision-tree training algorithm, MAVIPER. MAVIPER jointly grows the trees of each agent by predicting the behavior of the other agents using their anticipated trees, and uses resampling to focus on states that are critical for its interactions with other agents. We show that both algorithms generally outperform the baselines and that MAVIPER-trained agents achieve better-coordinated performance than IVIPER-trained agents on three different multi-agent particle-world environments.

Volviane Saphir Mfogo, Alain Zemkoho, Laurent Njilla et al.

The proliferation of the Internet of Things (IoT) has raised concerns about the security of connected devices. There is a need to develop suitable and cost-efficient methods to identify vulnerabilities in IoT devices in order to address them before attackers seize opportunities to compromise them. The deception technique is a prominent approach to improving the security posture of IoT systems. Honeypot is a popular deception technique that mimics interaction in real fashion and encourages unauthorised users (attackers) to launch attacks. Due to the large number and the heterogeneity of IoT devices, manually crafting the low and high-interaction honeypots is not affordable. This has forced researchers to seek innovative ways to build honeypots for IoT devices. In this paper, we propose a honeypot for IoT devices that uses machine learning techniques to learn and interact with attackers automatically. The evaluation of the proposed model indicates that our system can improve the session length with attackers and capture more attacks on the IoT network.

Xinyi Wei, Shuo Han, Ahmed H. Hemida et al.

This paper investigates backdoor attack planning in stochastic control systems modeled as Markov Decision Processes (MDPs). A backdoor attack involves an adversary deploying a policy that performs well in the original MDP to pass testing, but behaves maliciously at runtime when combined with a trigger that perturbs system dynamics. We consider a sophisticated attacker capable of jointly optimizing the backdoor policy and its trigger using only a blackbox simulator. During execution, the attacker has access only to partial observations of the system state and is restricted to introduce small perturbations to the system's transition dynamics. We formulate the attack planning problem as a constrained Markov game with an augmented state space and two players: Player 0 learns a backdoor policy that maximizes attack rewards when the trigger is active. However, when the trigger is inactive, the backdoor policy behaves near-optimally in the original MDP; Player 1 designs a finite-memory, observation-based trigger to activate the attack. We propose a switching gradient-based optimization algorithm to jointly solve for the backdoor policy and trigger. Experiments on a case study demonstrate the effectiveness of our method in achieving stealthy and successful backdoor attacks, and how the attack performance varies under different parameters related to the stealthiness of the backdoor attack.

Zelin Wan, Jin-Hee Cho, Mu Zhu et al.

This paper introduces a novel approach, Decision Theory-guided Deep Reinforcement Learning (DT-guided DRL), to address the inherent cold start problem in DRL. By integrating decision theory principles, DT-guided DRL enhances agents' initial performance and robustness in complex environments, enabling more efficient and reliable convergence during learning. Our investigation encompasses two primary problem contexts: the cart pole and maze navigation challenges. Experimental results demonstrate that the integration of decision theory not only facilitates effective initial guidance for DRL agents but also promotes a more structured and informed exploration strategy, particularly in environments characterized by large and intricate state spaces. The results of experiment demonstrate that DT-guided DRL can provide significantly higher rewards compared to regular DRL. Specifically, during the initial phase of training, the DT-guided DRL yields up to an 184% increase in accumulated reward. Moreover, even after reaching convergence, it maintains a superior performance, ending with up to 53% more reward than standard DRL in large maze problems. DT-guided DRL represents an advancement in mitigating a fundamental challenge of DRL by leveraging functions informed by human (designer) knowledge, setting a foundation for further research in this promising interdisciplinary domain.

Joseph Bao, Murat Kantarcioglu, Yevgeniy Vorobeychik et al.

Over the years, honeypots emerged as an important security tool to understand attacker intent and deceive attackers to spend time and resources. Recently, honeypots are being deployed for Internet of things (IoT) devices to lure attackers, and learn their behavior. However, most of the existing IoT honeypots, even the high interaction ones, are easily detected by an attacker who can observe honeypot traffic due to lack of real network traffic originating from the honeypot. This implies that, to build better honeypots and enhance cyber deception capabilities, IoT honeypots need to generate realistic network traffic flows. To achieve this goal, we propose a novel deep learning based approach for generating traffic flows that mimic real network traffic due to user and IoT device interactions. A key technical challenge that our approach overcomes is scarcity of device-specific IoT traffic data to effectively train a generator. We address this challenge by leveraging a core generative adversarial learning algorithm for sequences along with domain specific knowledge common to IoT devices. Through an extensive experimental evaluation with 18 IoT devices, we demonstrate that the proposed synthetic IoT traffic generation tool significantly outperforms state of the art sequence and packet generators in remaining indistinguishable from real traffic even to an adaptive attacker.

Junlin Wu, Charles Kamhoua, Murat Kantarcioglu et al.

Deception is a crucial tool in the cyberdefence repertoire, enabling defenders to leverage their informational advantage to reduce the likelihood of successful attacks. One way deception can be employed is through obscuring, or masking, some of the information about how systems are configured, increasing attacker's uncertainty about their targets. We present a novel game-theoretic model of the resulting defender-attacker interaction, where the defender chooses a subset of attributes to mask, while the attacker responds by choosing an exploit to execute. The strategies of both players have combinatorial structure with complex informational dependencies, and therefore even representing these strategies is not trivial. First, we show that the problem of computing an equilibrium of the resulting zero-sum defender-attacker game can be represented as a linear program with a combinatorial number of system configuration variables and constraints, and develop a constraint generation approach for solving this problem. Next, we present a novel highly scalable approach for approximately solving such games by representing the strategies of both players as neural networks. The key idea is to represent the defender's mixed strategy using a deep neural network generator, and then using alternating gradient-descent-ascent algorithm, analogous to the training of Generative Adversarial Networks. Our experiments, as well as a case study, demonstrate the efficacy of the proposed approach.

Juan Shu, Bowei Xi, Charles Kamhoua

Deep neural network (DNN) is a popular model implemented in many systems to handle complex tasks such as image classification, object recognition, natural language processing etc. Consequently DNN structural vulnerabilities become part of the security vulnerabilities in those systems. In this paper we study the root cause of DNN adversarial examples. We examine the DNN response surface to understand its classification boundary. Our study reveals the structural problem of DNN classification boundary that leads to the adversarial examples. Existing attack algorithms can generate from a handful to a few hundred adversarial examples given one clean image. We show there are infinitely many adversarial images given one clean sample, all within a small neighborhood of the clean sample. We then define DNN uncertainty regions and show transferability of adversarial examples is not universal. We also argue that generalization error, the large sample theoretical guarantee established for DNN, cannot adequately capture the phenomenon of adversarial examples. We need new theory to measure DNN robustness.

Abhishek N. Kulkarni, Shuo Han, Nandi O. Leslie et al.

We consider the probabilistic planning problem where the agent (called Player 1, or P1) can jointly plan the control actions and sensor queries in a sensor network and an attacker (called player 2, or P2) can carry out attacks on the sensors. We model such an adversarial interaction using a formal model -- a reachability game with partially controllable observation functions. The main contribution of this paper is to assess the cost of P1's unawareness: Suppose P1 misinterprets the sensor failures as probabilistic node failures due to unreliable network communication, and P2 is aware of P1's misinterpretation in addition to her partial observability. Then, from which states can P2 carry out sensor attacks to ensure, with probability one, that P1 will not be able to complete her reachability task even though, due to misinterpretation, P1 believes that she can almost-surely achieve her task. We develop an algorithm to solve the almost-sure winning sensor-attack strategy given P1's observation-based strategy. Our attack analysis could be used for attack detection in wireless communication networks and the design of provably secured attack-aware sensor allocation in decision-theoretic models for cyber-physical systems.

Todd Huster, Jeremy E. J. Cohen, Zinan Lin et al.

Generative adversarial networks (GANs) are often billed as "universal distribution learners", but precisely what distributions they can represent and learn is still an open question. Heavy-tailed distributions are prevalent in many different domains such as financial risk-assessment, physics, and epidemiology. We observe that existing GAN architectures do a poor job of matching the asymptotic behavior of heavy-tailed distributions, a problem that we show stems from their construction. Additionally, when faced with the infinite moments and large distances between outlier points that are characteristic of heavy-tailed distributions, common loss functions produce unstable or near-zero gradients. We address these problems with the Pareto GAN. A Pareto GAN leverages extreme value theory and the functional properties of neural networks to learn a distribution that matches the asymptotic behavior of the marginal distributions of the features. We identify issues with standard loss functions and propose the use of alternative metric spaces that enable stable and efficient learning. Finally, we evaluate our proposed approach on a variety of heavy-tailed datasets.

Mu Zhu, Ahmed H. Anwar, Zelin Wan et al.

Defensive deception is a promising approach for cyber defense. Via defensive deception, the defender can anticipate attacker actions; it can mislead or lure attacker, or hide real resources. Although defensive deception is increasingly popular in the research community, there has not been a systematic investigation of its key components, the underlying principles, and its tradeoffs in various problem settings. This survey paper focuses on defensive deception research centered on game theory and machine learning, since these are prominent families of artificial intelligence approaches that are widely employed in defensive deception. This paper brings forth insights, lessons, and limitations from prior work. It closes with an outline of some research directions to tackle major gaps in current defensive deception research.

Mengmeng Ge, Jin-Hee Cho, Charles A. Kamhoua et al.

Internet of Things (IoT) devices can be exploited by the attackers as entry points to break into the IoT networks without early detection. Little work has taken hybrid approaches that combine different defense mechanisms in an optimal way to increase the security of the IoT against sophisticated attacks. In this work, we propose a novel approach to generate the strategic deployment of adaptive deception technology and the patch management solution for the IoT under a budget constraint. We use a graphical security model along with three evaluation metrics to measure the effectiveness and efficiency of the proposed defense mechanisms. We apply the multi-objective genetic algorithm (GA) to compute the {\em Pareto optimal} deployments of defense mechanisms to maximize the security and minimize the deployment cost. We present a case study to show the feasibility of the proposed approach and to provide the defenders with various ways to choose optimal deployments of defense mechanisms for the IoT. We compare the GA with the exhaustive search algorithm (ESA) in terms of the runtime complexity and performance accuracy in optimality. Our results show that the GA is much more efficient in computing a good spread of the deployments than the ESA, in proportion to the increase of the IoT devices.

Zheyuan Ryan Shi, Ariel D. Procaccia, Kevin S. Chan et al.

Today's high-stakes adversarial interactions feature attackers who constantly breach the ever-improving security measures. Deception mitigates the defender's loss by misleading the attacker to make suboptimal decisions. In order to formally reason about deception, we introduce the feature deception problem (FDP), a domain-independent model and present a learning and planning framework for finding the optimal deception strategy, taking into account the adversary's preferences which are initially unknown to the defender. We make the following contributions. (1) We show that we can uniformly learn the adversary's preferences using data from a modest number of deception strategies. (2) We propose an approximation algorithm for finding the optimal deception strategy given the learned preferences and show that the problem is NP-hard. (3) We perform extensive experiments to validate our methods and results. In addition, we provide a case study of the credit bureau network to illustrate how FDP implements deception on a real-world problem.

Muhammad Saad, Jeffrey Spaulding, Laurent Njilla et al.

In this paper, we systematically explore the attack surface of the Blockchain technology, with an emphasis on public Blockchains. Towards this goal, we attribute attack viability in the attack surface to 1) the Blockchain cryptographic constructs, 2) the distributed architecture of the systems using Blockchain, and 3) the Blockchain application context. To each of those contributing factors, we outline several attacks, including selfish mining, the 51% attack, Domain Name System (DNS) attacks, distributed denial-of-service (DDoS) attacks, consensus delay (due to selfish behavior or distributed denial-of-service attacks), Blockchain forks, orphaned and stale blocks, block ingestion, wallet thefts, smart contract attacks, and privacy attacks. We also explore the causal relationships between these attacks to demonstrate how various attack vectors are connected to one another. A secondary contribution of this work is outlining effective defense measures taken by the Blockchain technology or proposed by researchers to mitigate the effects of these attacks and patch associated vulnerabilities

Muhammad Saad, Laurent Njilla, Charles Kamhoua et al.

Selfish mining is a well known vulnerability in blockchains exploited by miners to steal block rewards. In this paper, we explore a new form of selfish mining attack that guarantees high rewards with low cost. We show the feasibility of this attack facilitated by recent developments in blockchain technology opening new attack avenues. By outlining the limitations of existing countermeasures, we highlight a need for new defense strategies to counter this attack, and leverage key system parameters in blockchain applications to propose an algorithm that enforces fair mining. We use the expected transaction confirmation height and block publishing height to detect selfish mining behavior and develop a network-wide defense mechanism to disincentivize selfish miners. Our design involves a simple modifications to transactions' data structure in order to obtain a "truth state" used to catch the selfish miners and prevent honest miners from losing block rewards.

AbdelRahman Eldosouky, Walid Saad, Charles Kamhoua et al.

Critical infrastructure protection (CIP) is envisioned to be one of the most challenging security problems in the coming decade. One key challenge in CIP is the ability to allocate resources, either personnel or cyber, to critical infrastructures with different vulnerability and criticality levels. In this work, a contract-theoretic approach is proposed to solve the problem of resource allocation in critical infrastructure with asymmetric information. A control center (CC) is used to design contracts and offer them to infrastructures' owners. A contract can be seen as an agreement between the CC and infrastructures using which the CC allocates resources and gets rewards in return. Contracts are designed in a way to maximize the CC's benefit and motivate each infrastructure to accept a contract and obtain proper resources for its protection. Infrastructures are defined by both vulnerability levels and criticality levels which are unknown to the CC. Therefore, each infrastructure can claim that it is the most vulnerable or critical to gain more resources. A novel mechanism is developed to handle such an asymmetric information while providing the optimal contract that motivates each infrastructure to reveal its actual type. The necessary and sufficient conditions for such resource allocation contracts under asymmetric information are derived. Simulation results show that the proposed contract-theoretic approach maximizes the CC's utility while ensuring that no infrastructure has an incentive to ask for another contract, despite the lack of exact information at the CC.