Intelligent Attacks Research Articles

Defensive deception against reactive jamming attacks in remote state estimation

This paper considers a synthetic counter-measure, combining transmission scheduling and defensive deception, to defend against jamming attacks in remote state estimation. In the setup studied, an attacker sabotages packet transmissions from a sensor to a remote estimator by congesting the communication channel between them. In order to efficiently degrade the estimation accuracy, the intelligent attacker tailors its jamming strategy by reacting to the real-time information it collects. In response to the jamming attacks, the sensor with a long-term goal will select the transmission power level at each stage. In addition, by modifying the real-time information intentionally, the sensor creates asymmetric uncertainty to mislead the attacker and thus mitigate attacks. Considering the dynamic nature of the process, we model the strategic interaction between the sensor and the attacker by a general stochastic game with asymmetric information structure. To obtain stationary optimal strategies for each player, we convert this game into a belief-based dynamic game and analyze the existence of its optimal solution. For a tractable implementation, we present an algorithm that finds equilibrium strategies based on multi-agent reinforcement learning for symmetric-information stochastic games. Numerical examples illustrate properties of the proposed algorithm.

A Linear Game Framework for Defending Power Systems Against Intelligent Physical Attacks

Game theory has recently been applied for analyzing the security of power systems under intelligent attack threats. This letter presents a linear game framework for optimal allocation of a defense budget to confront intelligent attacks on power systems. The presented framework offers the advantage that it can be applied in large power systems with a large number of components under attack threat without too much complexity. The presented linear framework is successfully applied on test systems.

A game theoretic approach to multi-channel transmission scheduling for multiple linear systems under DoS attacks

Recent years have illustrated the soaring importance of security issues of cyber-physical systems (CPSs). In this paper, we consider the remote state estimation of a CPS while subject to denial-of-service (DoS) attacks. A number of sensors monitor different linear dynamical systems and transmit the measurements to a remote estimator through a multi-channel network that may suffer corruption due to an intelligent attacker. Both sensors and the attacker have energy constraints. At each time-step, sensors select which channel to transmit over while the attacker chooses which channel to congest. Both sides select strategies based on the current state. In order to model the infinite horizon decision problem, we construct a two player zero-sum game and propose a Nash Q-learning algorithm to solve for the optimal strategies of both players. Finally, simulations are provided to illustrate our results.

Intelligent attack defense scheme based on DQL algorithm in mobile fog computing

Fog computing is a technology that can expands the network computing mode of cloud computing and extends network computing from the network center to the network edge. It adds fog layer between cloud data center layer and Internet of Things (IoT) device layer, and provides data storage, processing, forwarding and other functions for devices using the network edge. In mobile fog computing (MFC) networks, fog nodes communicate with end users through wireless networks. Malicious users can choose different attack modes to attack legitimate users. There is a lack of research on the subjective choice of attack modes for malicious users in current work. To solve this problem, an intelligent attack defense scheme based on Double Q-learning (DQL) algorithm in MFC is proposed. Firstly, the security model involving malicious users in MFC is described. Based on Prospect Theory (PT), a static method of subjective zero-sum game between malicious users and legitimate users is constructed. Secondly, a dynamic subjective game scheme based on DQL algorithm is proposed to resist intelligent attacks. The simulation results show that compared with the Q-learning-based method for resisting intelligent attacks, the proposed method can enhance the security of MFC network and enhance the protection performance.

Physical layer authentication under intelligent spoofing in wireless sensor networks

Location based access in wireless sensor networks (WSN) are vulnerable to location spoofing attacks. In this paper, we investigate the physical layer (PHY-layer) authentication in the threat of an intelligent location spoofing attack. The intelligent attack can emulate the legitimate channel information and maximize its long-term cumulative reward. First, we analyze the feasibility of this intelligent attack and investigate how it threats to the networks. Specifically, we derive the optimal transmit power allocation and find the worst case for the defenders, namely optimal intelligent attack, in which the attacker can learn the intelligent attack action based on the beamforming with optimal transmit power allocation. To defend against such an intelligent attack with high accuracy and low overhead, we develop a cooperative PHY-layer authentication scheme. Then, we provide an in-depth analysis on the belief and derive the belief bounds and the closed-form expression for the belief threshold. Furthermore, considering the whole computation complexity and the double counting problem in a loopy graph, we propose the cooperative neighbour selection algorithm to accelerate belief convergence and reduce the overhead. Finally, the simulation results reveal that the proposed method can significantly improve the defense performance compared with the state-of-art methods.

Using one class SVM to counter intelligent attacks against an SPRT defense mechanism

A widely used defense mechanism in current wireless sensor networks is based on using a Sequential Probability Ratio Test (SPRT). This test does not require a fixed sample size, and this reduces the number of communications required and the battery consumption, both of which are of capital interest in such networks. However, SPRT assumes that the distributions under test do not change, and this assumption needs not hold for current attackers.We divide this work in two parts, First, we develop an optimal attack strategy against a Bernoulli SPRT mechanism which is used in many defense mechanisms in current literature. The control law for such an attacker turns out to be easy to implement and very effective, thus posing a significant threat for the defense mechanisms that use such SPRT. Second, we make use of One Class Supporting Vector Machines to obtain a modified SPRT test that is able to detect, not only such an attacker, but potentially any other attack mechanism that has not a similar spectrum to the expected signal from normal sensors. Our work is validated via simulations, showing that the attacker we propose is a real threat to SPRT mechanisms, but also, that our proposed defense mechanism can efficiently cope with such an attacker.

Assigning Operating Reserves in Power Systems Under Imminent Intelligent Attack Threat

Operating reserves serve as efficient tools to respond to the sudden outage of power system components caused by either natural events or intentional attacks. Contrary to the natural threats, the intentional attacks exhibit no stochastic behavior and are launched to inflict as much damage as possible to system. Moreover, these attacks are strongly influenced by the measures taken by the system operator. As a consequence, the probabilistic methods of operating reserve procurement and allocation do not prove to be efficient to confront the intelligent attacks. In this paper, the interaction between the system operator as the defender and the intelligent attacker is investigated within a linear two-person zero-sum game model. In this model, the system is scheduled and the operating reserves are allocated and deployed in a limited number of time intervals in such a way that the system incurs the least cost for the worst-case attack. The efficacy of the proposed method is examined by conducting case studies of single- and multiple-target attacks on two test systems. The results indicate that the solutions obtained by the proposed method can result in a considerable reduction in the system costs in comparison to those obtained when the system is scheduled by probabilistic methods and an intelligent attack is launched.

Operational Perspectives Into the Resilience of the U.S. Air Transportation Network Against Intelligent Attacks

Protecting critical infrastructures has been deemed a natural security imperative by the U.S. Government for the past 20 years. The resilience of transportation infrastructures is of particular importance, as they serve a crucial rule in economic development and citizen mobility throughout the world. This paper presents a defender–attacker–defender model to analyze the potential impacts of intelligent attacks and worst case disruptions on the U.S. air transportation network, as well as possible protection steps that could be taken to minimize the negative outcomes of such disruptions. Furthermore, to analyze the effects of intermodal connections on the resilience of the air network, a second layer representing a hypothetical bus network is added to the model and studied. We use these models, supported by publicly available data, to identify routes likely to be attacked by intelligent adversaries and those critical to the resilient operation of the air network in such scenarios. We also demonstrate the potential benefits of intermodal linkages toward maintaining network operations and identify promising research directions for this type of integrated and intelligent transportation system.

A resource-preserving self-regulating Uncoupled MAC algorithm to be applied in incident detection

The connectivity of embedded systems is increasing accompanied with thriving technology such as Internet of Things/Everything (IoT/E), Connected Cars, Smart Cities, Industry 4.0, 5G or Software-Defined Everything. Apart from the benefits of these trends, the continuous networking offers hackers a broad spectrum of attack vectors. The identification of attacks or unknown behavior through Intrusion Detection Systems (IDS) has established itself as a conducive and mandatory mechanism apart from the protection by cryptographic schemes in a holistic security eco-system. In systems where resources are valuable goods and stand in contrast to the ever increasing amount of network traffic, sampling has become a useful utility in order to detect malicious activities on a manageable amount of data. In this work an algorithm – Uncoupled MAC – is presented which secures network communication through a cryptographic scheme by uncoupled Message Authentication Codes (MAC) but as a side effect also provides IDS functionality producing alarms based on the violation of Uncoupled MAC values. Through a novel self-regulation extension, the algorithm adapts it’s sampling parameters based on the detection of malicious actions. The evaluation in a virtualized environment clearly shows that the detection rate increases over runtime for different attack scenarios. Those even cover scenarios in which intelligent attackers try to exploit the downsides of sampling.

Antijamming Game Framework for Secure State Estimation in Power Systems

In this paper, we investigate the secure state estimation (SSE) problem in power systems, where the physical system is measured by meters and mainly focuses on the measurements sent to a remote estimator via wireless networks faced with jamming attacks. Malicious attacks on the transmission paths block the data transmission and deteriorate the performance of estimation. Various works have been proposed to cope with the transmission failure. Few of them have considered the case that a smart attacker could adjust strategies according to the defensive methods with advanced communication techniques. We propose the antijamming game framework for SSE. Under this framework, defensive path selection (DPS) is proposed based on multiagent reinforcement learning to make optimal path selection against the intelligent attacker and improve the transmission performance for SSE. The effectiveness of the proposed method is theoretically proved to improve the robustness of estimation and the capability of DPS is analyzed.

Vulnerability analysis of cascading dynamics in smart grids under load redistribution attacks

Smart grids integrate power system engineering with information and communication technology to form a complex system. In this study, a framework is proposed to model the smart grid as interdependent complex networks and investigate the topology vulnerabilities subject to targeted attacks. As a serious targeted cyber-physical threat, the load redistribution attack is a type of false data injection attack. Adversaries can intentionally intercept communication traffic flows or cooperatively manipulate data transmitted over the communication network to mislead the control centre to perform incorrect actions, thereby tripping off lines or breakers, and even triggering cascading failures. Based on the intelligent attack mechanism, a new approach is introduced to identify sequential attack targets according to the node importance, which can in turn collapse the entire power system. Based on the particular characteristics of power systems, the components require achieving a power balance through the redistribution or shedding of loads in each attack. Compared to random removal and block removal, the effectiveness of the proposed strategy was validated on the IEEE 39-bus and the actual provincial backbone network. From the analysis of the experimental results, we find that the topology vulnerability is closely related to the critical node types and locations.

Decentralized Defence of a (Directed) Network Structure

ABSTRACT We model the decentralized defence choice of agents connected in a directed graph and exposed to an external threat. The network allows players to receive goods from one or more producers through directed paths. Each agent is endowed with a finite and divisible defence resource that can be allocated to their own security or to that of their peers. The external threat is represented by either a random attack on one of the nodes or by an intelligent attacker who aims to maximize the flow-disruption by seeking to destroy one node. We show that under certain conditions a decentralized defence allocation is efficient when we assume the attacker to be strategic: a centralized allocation of defence resources which minimizes the flow-disruption coincides with a decentralized equilibrium allocation. On the other hand, when we assume a random attack, the decentralized allocation is likely to diverge from the central planner’s allocation.

Learning attack mechanisms in Wireless Sensor Networks using Markov Decision Processes

In this work, we identify two related problems that arise in many Wireless Sensor Networks defense mechanisms: the problem of ad-hoc defense and the problem of optimality. These problems open the door to attacks that could severely affect the performance of defense mechanisms. In this work, we use Markov Decision Processes as framework to model an attacker that is able to exploit these two problems. This allows us to model a defense mechanism theoretically and to obtain the performance of an attack against it, as well as to obtain the optimal attack against the defense mechanism - i.e., the attack that harms the most the defense mechanism. We also make use of Deep Reinforcement Learning tools, showing that they can be used by an intelligent attacker to successfully exploit a possibly unknown defense mechanism, providing a compromise between attack results and computational cost. We test our approach by thoroughly studying a Cooperative Spectrum Sensing attack, which we use to illustrate the framework proposed and to highlight their strengths and weaknesses.

A Security-Enhanced Cluster Size Adjustment Scheme for Cognitive Radio Networks

Cognitive radio network (CRN) is the next generation wireless network that allows unlicensed users [secondary users (SUs)] to explore and use the underutilised licensed channels (white spaces) owned by licensed users (primary users). The purpose is to increase the spectrum utilization for enhanced network performance. Clustering segregates SUs in a CRN into logical groups (clusters) with each consisting of a leader (cluster head) and member nodes. A budget-based cluster size adjustment scheme is applied to enable each cluster to adjust its number of member nodes in its cluster based on the availability of white spaces in order to improve network scalability. However, cluster size adjustment is prone to attacks by malicious SUs that launch random and intelligent attacks. Hence, we incorporate an artificial intelligence approach called reinforcement learning (RL) into a trust model to countermeasure the random and intelligent attacks. The simulation results show that RL-based trust model increases the utilization of white spaces and cluster size to improve network scalability and enhance network performance despite the presence of RL-based intelligent attacks.

Enhanced Secure Transmission Against Intelligent Attacks

In this paper, we proposed an enhanced secure scheme for the wireless communication system threatened by an intelligent attacker, which can work in eavesdropping, jamming, and spoofing modes. The conventional secure scheme is to apply $Q$ -learning-based algorithm to reach a Nash equilibrium (NE) in the framework of a zero-sum game between the transmitter and attacker, which, however, requires the number of antennas at the transmitter to be much larger than that at the attacker. To overcome this limitation, we first consider the scenario where the attacker can flexibly increase the number of antennas in order to increase the attack rate. By adaptively setting the number of antennas at the transmitter and the legitimate receiver equal to that at the attacker, we then apply the beamforming at the transmitter to suppress the eavesdropping and use the filtering at the receiver to prevent the jamming and spoofing. By incorporating the beamforming and filtering, the benefits of the attacker in this game are efficiently restricted. Furthermore, the $Q$ -learning-based power control strategy is used to reach a new NE. The simulation results have been demonstrated to show that the proposed scheme can suppress the intelligent attack efficiently, which outperforms the conventional scheme in the secrecy performance.

Key Hiding of Multimedia Applications in Multimedia File

The project entitled Effective key Generation for Multimedia Application is the application developed to embed an video file in another video signal. It is concerned with embedding information in an innocuous cover Speech in a secure and robust manner. This system makes the Files more secure by using the concepts Steganography and Cryptography, Steganography, poor cousin of Cryptography is the art of hiding messages inside other messages such that the very existence of the message is unknown to third party. The goal of cryptography is to make data unreadable by a third party, the goal of Steganography is to hide the data from a third party Through the use of advanced computer software, authors of images and software can place a hidden trademark in their product, allowing them to keep a check on piracy. This is commonly known as watermarking. Hiding serial numbers or a set of characters that distinguishes an object from a similar object is known as finger printing. Together, these two are intended to fight piracy. The latter is used to detect copyright violators and the former is used to prosecute them. But these are only examples of the much wider field of Steganography. The cover data should not be significantly degraded by the embedded data, and the embedded data should be as imperceptible as possible. The embedded data should be as immune as possible to modifications from intelligent attacks or anticipated manipulations. Thus it is necessary that the hidden message should be encrypted.

Q-Learning Based Physical-Layer Secure Game Against Multiagent Attacks

In this paper, we consider a Q-learning-based power allocation strategy for a secure physical-layer system under dynamic radio environments. In such a system, the transmitter sends the information to the receiver threatened by M(M ≥ 2) intelligent attackers which have several attack modes and will bring out the severe issue of information security. To safeguard the system security, we formulate the insecure problem as a stochastic game which consists of M+1 players: the transmitter which can flexibly choose its transmit power, and M smart attackers that can determine their attack types. Then, the Nash equilibria (NEs) of the physical-layer secure game are derived, and their existence conditions are taken into account. The simulation results show that the proposed power allocation strategy in the stochastic game can efficiently suppress the attack rate of smart attackers even if there exist multiple smart attackers.

SEUI-64, bits an IPv6 addressing strategy to mitigate reconnaissance attacks

It has been widely assumed by the research community that the network reconnaissance attacks in IPv6 networks are unfeasible because they would take tremendous effort to perform address scanning of 264 hosts in an IPv6 subnet. However, recent research has revealed feasibility of these attacks by investigating a number of native IPv6 networks. The research concluded that an intelligent attacker could easily reduce the target search space by predicting the network host addressing schemes when performing the scanning. This indeed enhances security concerns and undermines the chances of IPv6 being deployed. This paper outlines the IPv6 addressing strategies currently used and proposes a new strategy to mitigate reconnaissance attacks. The new strategy is evaluated against some reconnaissance attack approaches and compared with some recent IPv6 addressing strategies. The experimental results confirm the effectiveness and validation of the addressing strategy in terms of the mitigation of reconnaissance attacks.

Game framework for optimal allocation of spinning reserve to confront intelligent physical attacks on power system

Spinning reserve (SR) serves as an efficient tool to respond to a sudden outage of power system components caused by either natural events or physical attacks. Contrary to natural threats, the intelligent physical attacks are strongly influenced by the measures taken by the system operator to defend the system and exhibit no stochastic behaviour. As a consequence, the probabilistic methods of SR provision and allocation do not prove to be appropriate to confront intelligent attacks. In this study, the interaction between the system operator as the defender and the intelligent attacker is investigated within a game model. In this model, the SR is allocated in such a way that the highest interruption cost that the attacker can impose is minimised. The technical limitations of SR provision together with operational constraints are taken into consideration. The efficacy of the proposed method is examined by conducting case studies of single and multiple-target attacks on components of IEEE 24-bus and IEEE 118-bus test systems. The results show that the proposed scheme produces a considerable reduction in the system interruption cost in comparison with the probabilistic methods.

CCP game: A game theoretical model for improving the scheduling of chemical cluster patrolling

Chemical clusters can be attractive targets for terrorism, due to the extreme importance of them as well as due to the existence of dangerous materials. Patrolling is scheduled for better securing chemical clusters. However, the current patrolling strategies fail on competing with intelligent attackers and therefore can be non-optimal. The so-called chemical cluster patrolling (CCP) game is proposed in this paper. The CCP game employs game theory as a methodology, aiming at randomly but strategically scheduling security patrols in chemical clusters. The patroller and the attacker are modelled as the two rational players in the CCP game. The patroller's strategy is defined as probabilistically traveling within the cluster or patrolling some plants while the attacker's strategy is formulated as a combination of an attack target, the start time of the attack, and the attack scenario to be used. The Stackelberg equilibrium and a robust solution which takes into consideration of the patroller's distribution-free uncertainties on the attacker's parameters are defined for predicting the outcome of the CCP game. Results of the case study indicate that the patrolling strategy suggested by the CCP game outperforms both the fixed patrolling route strategy and the purely randomized patrolling strategy.