Presence Of Malicious Attacks Research Articles

Realization of Robust Formation for Multi-UAV Systems Using Control Barrier Functions

Network robustness is a necessary prerequisite for the effective execution of resilient control in distributed multiple unmanned aerial vehicles (UAVs). However, it remains a challenging task to construct a communication graph that satisfies specific network robustness properties. This paper investigates robust formation control of multi-UAV systems using control barrier functions (CBFs). We first propose a novel formation law to drive groups of UAVs into formations with communication graphs that satisfy the [Formula: see text]-robustness. With such a law, all the normal UAVs in the formation are capable of executing a given resilient consensus protocol, and achieving convergence in the presence of malicious attackers. Then, we present a control law that facilitates the establishment of UAV formations, in which the communication graphs satisfy p-fraction robustness. Finally, simulation examples are given to verify the effectiveness of the proposed formation control laws.

Observer based attack detection and security control for UAVs against attacks on desired trajectory

This paper investigates an attack detection and security control strategy for unmanned aerial vehicles (UAVs) in the presence of malicious attacks. The detection, estimation, and compensation of the attack are the key points of the proposed scheme. This article considers that the desired trajectory sent by the ground control station (GCS) to the UAV is maliciously tampered with. First, a description of the UAV system and the desired trajectory attack is presented. The mathematical model is constructed by analyzing the characteristics of the desired trajectory attack. Secondly, an unknown input observer (UIO) is designed to generate the state estimation error. The attack detection framework is constructed by taking into account the state estimation error and the trajectory tracking error. Thirdly, an attack estimation observer with a prescribed H∞ performance is designed to compensate for the impact of attacks on the control system. A disturbance observer (DO) is used to compensate for the system disturbance. Finally, the simulation results are presented on Matlab and Gazebo to validate the effectiveness of the proposed security control strategy.

Mixed-Trust Computing: Safe and Secure Real-Time Systems

Verifying complex Cyber-Physical Systems (CPS) is increasingly important given the push to deploy safety-critical autonomous features. Unfortunately, traditional verification methods do not scale to the complexity of these systems and do not provide systematic methods to protect verified properties when not all the components can be verified. To address these challenges, this article proposes a real-time mixed-trust computing framework that combines verification and protection. The framework introduces a new task model, where an application task can have both an untrusted and a trusted part. The untrusted part allows complex computations supported by a full OS with a real-time scheduler running in a VM hosted by a trusted hypervisor. The trusted part is executed by another scheduler within the hypervisor and is thus protected from the untrusted part. If the untrusted part fails to finish by a specific time, the trusted part is activated to preserve safety (e.g., prevent a crash) including its timing guarantees. This framework is the first allowing the use of untrusted components for CPS critical functions while preserving logical and timing guarantees, even in the presence of malicious attackers. We present the framework its schedulability analysis and the coordination protocol between the trusted and untrusted parts. Our implementation on a Raspberry Pi 3 is also discussed along with experiments showing the behavior of the system under failures of untrusted components, and a drone application to demonstrate its practicality.

Secure Estimation, Attack Isolation, and Reconstruction Based on Zonotopic Unknown Input Observer

In this paper, the issues of secure estimation, attack reconstruction and isolation are addressed for cyber-physical systems (CPSs) in the presence of malicious attacks. It is assumed that disturbances and noises are unknown-but-bounded (UBB). Based on different rank constraints, both zonotopic completely unknown input observer (ZCUIO) and zonotopic partially unknown input observer (ZPUIO) are designed to estimate system states. Then, malicious attacks are reconstructed and isolated based on the proposed zonotopic observers. By utilizing switching technique, the corresponding feedback controllers are designed, which can guarantee that the closed-loop system is stable. Finally, numerical simulations are provided to illustrate the validity of the presented approach.

Secure formation control for resilient multi‐agent cyber‐physical systems under DoS attacks and faults

AbstractA distributed cyber‐security and fault tolerance formation problem is considered in this paper. This is done though a multi‐agent system in which all agents concentrate on a formation reference function and accomplish a desired time‐varying formation even in the presence of malicious attacks and multiple faults. To achieve the resilient control task, a distributed control protocol using an event‐triggered technique and active reconfigurable strategy is proposed in the face of adverse effects for the overall system. When denial of service is launched by adversaries, the attack parameters are bounded by analyzing the Lyapunov stability theorem and the property of the Laplacian matrix. To improve the capability of fault‐tolerance, an observer‐based fault estimator is used to calculate fault values. The simulation results validate that the presented scheme can perfectly address the distributed formation issue with faults over attack intervals.

A transmission scheme for secure estimation of cyber-physical systems against malicious attack

This paper investigates the secure state estimation problem of the cyber-physical systems (CPSs) in the presence of malicious attacks. Considering the destructiveness and stealthiness of the attacker, a novel stochastic transmission scheme, in which the repeater generates transmission data by randomly combining some measurement values, is proposed to ensure the effectiveness of the attack detector. A filter equipped with the χ 2 detector and the stochastic transmission scheme is proposed to conduct the secure state estimation. The attack probability condition of the estimation error being mean-square bounded is given, which indicates that stochastic transmission scheme has a limited effect on the filter. Some numerical simulations are given to illustrate the effectiveness of the proposed transmission scheme.

Detecting Distance-Spoofing Attacks in Arbitrarily-Deployed Wireless Networks

The present work addresses localization problem in arbitrarily-deployed wireless networks in the presence of malicious attackers that can manipulate (spoof) their distance measurements. This problem is very important for the forthcoming applications that depend on accurate location information, since malicious attacks can potentially lead to disastrous consequences without the consideration of the security aspect. Therefore, this work proposes a new scheme for attacker detection based on generalized likelihood ratio test (GLRT) applied after a preliminary location estimate is obtained based on a simple geometric approach. The detected attacker is then removed from the localization process, which is converted into a generalized trust region sub-problem (GTRS) framework by applying the law of cosines (LC) and solved efficiently via bisectioning. The work also introduces theoretical bounds on the probability of detection in closed-form and studies the performance of the proposed algorithm from computational complexity, success in attacker detection, and localization accuracy perspectives. Our simulations corroborate the effectiveness of the proposed scheme, offering good localization accuracy and detection performance even in noisy environments

Toward Secure Localization in Randomly Deployed Wireless Networks

Being able to accurately locate wireless devices, while guaranteeing high-level of security against spoofing attacks, benefits all participants in the localization chain (e.g., end users, network operators, and location service providers). On the one hand, most of existing localization systems are designed for innocuous environments, where no malicious adversaries are present. This makes them highly susceptible to security threats coming from interferers, attacks, or even unintentional errors (malfunctions), and thus, practically futile in hostile settings. On the other hand, existing secure localization solutions make certain (favorable) assumptions regarding the network topology (e.g., that the target device lies within a convex hull formed by reference points), which restrict their applicability. Therefore, this work addresses the problem of target localization in randomly deployed wireless networks in the presence of malicious attackers, whose goal is to manipulate (spoof) the estimation process and disable accurate localization. We propose a low-complex solution based on clustering and weighted central mass to detect attackers, using only the bare minimum of reference points, after which we solve the localization problem by a bisection procedure. The proposed method is studied from both localization accuracy and success in attacker detection point of views, where closed-form expressions for upper and lower bounds on the probability of attacker detection are derived. Its performance is validated through computer simulations, which corroborate the effectiveness of the proposed scheme, outperforming the state-of-the-art methods.

Adversarial attack and defense strategies for deep speaker recognition systems

Robust speaker recognition, including in the presence of malicious attacks, is becoming increasingly important and essential, especially due to the proliferation of smart speakers and personal agents that interact with an individual’s voice commands to perform diverse and even sensitive tasks. Adversarial attack is a recently revived domain which is shown to be effective in breaking deep neural network-based classifiers, specifically, by forcing them to change their posterior distribution by only perturbing the input samples by a very small amount. Although, significant progress in this realm has been made in the computer vision domain, advances within speaker recognition is still limited. We present an expository paper that considers several adversarial attacks to a deep speaker recognition system, employs strong defense methods as countermeasures, and reports a comprehensive set of ablation studies to better understand the problem. The experiments show that the speaker recognition systems are vulnerable to adversarial attacks, and the strongest attacks can reduce the accuracy of the system from 94% to even 0%. The study also compares the performances of the employed defense methods in detail, and finds adversarial training based on Projected Gradient Descent (PGD) to be the best defense method in our setting. We hope that the experiments presented in this paper provide baselines that can be useful for the research community interested in further studying adversarial robustness of speaker recognition systems.

Secure Localization and Velocity Estimation in Mobile IoT Networks With Malicious Attacks

Secure localization and velocity estimation are of great importance in Internet-of-Things (IoT) applications and are particularly challenging in the presence of malicious attacks. The problem becomes even more challenging in practical scenarios in which attack information is unknown and anchor node location uncertainties occur due to node mobility and falsification of malicious nodes. This challenging problem is investigated in this article. With reasonable assumptions on the attack model and uncertainties, the secure localization and velocity estimation problem is formulated as an intractable maximum a posterior (MAP) problem. A variational-message-passing (VMP)-based algorithm is proposed to approximate the true posterior distribution iteratively and find the closed-form estimates of the location and velocity securely. The identification of malicious nodes is also achieved in the meantime. The convergence of the proposed VMP-based algorithm is also discussed. Numerical simulations are finally conducted and the results show the VMP-based joint localization and velocity estimation algorithm can approach the Bayesian Cramer Rao bound and is superior to other secure algorithms.

Security Control of Multiagent Systems Under Denial-of-Service Attacks.

Security control aims to guarantee the consensus of multiagent systems (MASs) in the presence of the denial-to-service attacker. Most of the existing distributed controllers are invalid during the attack interval due to the paralyzed communication channels. In order to overcome this difficulty, a novel hybrid distributed control protocol is designed. Here, the controller uses the latest information saved in the buffers in the presence of malicious attacks, which will further enhance the security of MASs. Some sufficient conditions on the coupled strength and attack parameters are derived to achieve the leader-following consensus of MASs. Furthermore, we estimate the upper bounds of denial-of-service (DoS) frequency and DoS duration which the MASs can tolerate before losing consensus. Notice that we also reduce the computational complexity via the property of the Kronecker product. Besides, an observer-based model is proposed and the corresponding consensus criterion is established to reduce the effects of attackers on the controller. Finally, the efficiency of our theoretical results is illustrated by a numerical example.

Probabilistic analysis of trust based decision making in hostile environments

Trust and reputation systems are one of the most important tools for decision making which are able to predict the reliability of an agent, taking into account the past experience and received recommendations. However, the effectiveness of these systems may be greatly reduced in hostile environments with the presence of malicious attackers. Despite the increasing use of these systems and the importance of their robustness in hostile environments, So far the most efforts on reinforcing of these systems have been focused on trust computation models, not on the trust based decision making. This paper discusses the importance and impact of decision-making on the robustness of trust systems in hostile environments and shows its great impact in this regard. In the paper, the most important trust based decision-making methods are investigated using formal verification methods in different attack scenarios (including maliciousness and selfishness) and the results are analyzed to find the strengths and weaknesses of each one. Considering the obtained results, the paper also proposed a new approach for trust-based decision making. The case studies show that the proposed approach is able to improve the robustness of trust and reputation models in all different attack scenarios.

Towards a framework of key technologies for drones

The potential applications for drones, especially those in manned areas or into non-segregated airspace, are currently not possible without the development and validation of certain key enabling technologies: “detect and avoid”, “air traffic management”, “command and control (C2) link”, and “security and cyber resilience”. This issue has a high impact on European innovation as identified by SESAR JU and demands R&D investments and incentives for the convergence of shared technologies and markets. In this paper, we discuss the vision of COMP4DRONES project to complement SESAR JU efforts with a particular focus on safe software and hardware drone architectures. COMP4DRONES will bear a holistically designed ecosystem ranging from application to electronic components. The ecosystem aims at supporting (1) efficient customization and incremental assurance of drone-embedded platforms, (2) safe autonomous decision making concerning individual or cooperative missions, (3) trustworthy drone-to-drone and drone-to- ground communications even in presence of malicious attackers and under the intrinsic platform constraints, and (4) agile and cost-effective design and assurance of drone modules and systems. Lead applications driving ecosystem development and benchmarking on the fields of transport, inspection, logistic, precision agriculture, and parcel delivery will be produced.

Resilient and Robust Synchronization of Multiagent Systems Under Attacks on Sensors and Actuators.

Resilient and robust distributed control protocols for multiagent systems under attacks on sensors and actuators are designed. A distributed H∞ control protocol is designed to attenuate the disturbance or attack effects. However, the H∞ controller is too conservative in the presence of attacks. Therefore, it is augmented with a distributed adaptive compensator to mitigate the adverse effects of attacks. The proposed controller can make the synchronization error arbitrarily small in the presence of faulty attacks, and satisfy global L2 -gain performance in the presence of malicious attacks or disturbances. A significant advantage of the proposed method is that it requires no restriction on the number of agents or agents' neighbors under attacks on sensors and/or actuators, and it recovers even compromised agents under attacks on actuators. Simulation examples verify the effectiveness of the proposed method.

Blind Detection: Advanced Techniques for WiFi-Based Drone Surveillance

The great availability of commercial drones has raised growing interest among people, since remotely piloted vehicles can be employed in numerous applications. The pervasive use of these devices has created many privacy and safety concerns that need to be addressed by means of proper surveillance systems able to cope with such threats. In this paper, we propose a WiFi statistical fingerprint-based drone detection approach, which is capable of identifying nearby drone threats, even in the presence of malicious attacks. We present a performance analysis carried out through experimental tests, where our solution is able to achieve very good results in terms of correct recognitions in many real-life scenarios, with a peak true positive rate of 96 %.

Robust Cooperative Spectrum Sensing Based on Adaptive Reputation and Evidential Reasoning Theory in Cognitive Radio Network

In this paper we investigate the problem of cooperative spectrum sensing (CSS) in cognitive radio networks (CRNs) in the presence of malicious attacks from secondary users (SUs). We propose a novel CSS approach based on a new adaptive reputation and evidential reasoning theory (ERT) to detect the malicious SUs, to strengthen the sensing ability, and to keep good stability and generalization ability. Different from the previous work, we assume that the fusion center has no prior information of the radios in CRNs except the condition that the honest SUs are in majority. Specifically, to counter different attack strategies, we present an adaptive reputation which fully considers the signal-to-noise ratio, historical sensing data information, and current local sensing results via an energy detection method. The proposed adaptive reputation can effectively identify the malicious SUs when the CRNs are with single-attack or multiple-attack models. Furthermore, we present a CSS scheme via the proposed adaptive reputation and ERT on the basis of local sensing results derived from the energy detection method. The new CSS scheme has good robustness and generality and can decrease the computational complexity and overcome the evidence conflict problems of Dempster–Shafer theory-based CSS schemes. Extensive simulations are carried out to demonstrate the performance and the superiority of the proposed approach in contrast with previous methods.

Bluff Forwarding: A Practical Protocol for Delivering Refreshed Symmetric Keys on a Multi-Path Big Data Ingestion System

In this paper, we present a clean-slate design of a novel and practical protocol for transporting refreshed symmetric keys for multi-path big data ingestion systems. Our objective is to securely and reliably deliver the refreshed keys from data sources to the data collection servers even in the presence of malicious attackers. To satisfy this objective, we first adapt the secret sharing algorithm. We split symmetric keys into multiple secrets in such a way that partial retrieval of the secrets does not warrant the reconstruction of the original key. Then, we hide the updated secret keys by shuffling them into a group of randomly generated fake keys. These keys are encapsulated in a message container called a bluff. We develop a protocol that makes it computationally infeasible for the attackers to distinguish between bluffs and normal data. We implement and test this protocol on Apache Flume, which is a widely used, state-of-the-art data ingestion system. We analyze the security aspects of our protocol and observe the effects of various configuration settings on the data ingestion performance.

TRAST: Trust-Based Distributed Topology Management for Wireless Multimedia Sensor Networks

A distributed topology management scheme in wireless multimedia sensor networks (WMSNs) ensures coverage of an event, prolongs network lifetime, and maintains connectivity between camera sensor (CS) nodes. However, the deployment of WMSNs in unattended environments makes the nodes vulnerable to security attacks. Hence, security issues should be considered, along with topology management, in WMSNs. In this work, we propose a trust-based distributed topology management scheme, named TRAST, for use in WMSNs. TRAST exploits the received signal strength of the control packets, which are then used to construct the distributed topology. In a non-secure distributed topology, the use of trust helps in providing coverage of an event, and maintaining connectivity, even in the presence of malicious attacks. The proposed topology management scheme achieves higher average coverage ratio and average packet delivery ratio than those corresponding to the LDTS and T-Must schemes, in the presence of malicious attacks.

NDTAODV: Neighbor Defense Technique for Ad Hoc On-Demand Distance Vector (AODV) to Mitigate Flood Attack in Manets

Mobile Ad Hoc Networks (MANETs) are collections of mobile nodes that can communicate with one another using multihop wireless links. MANETs are often deployed in the environments, where there is no fixed infrastructure and centralized management. The nodes of mobile ad hoc networks are susceptible to compromise. In such a scenario, designing an efficient, reliable and secure routing protocol has been a major challengesue over the last many years. The routing protocol Ad hoc On-demand Distance Vector (AODV) has no security measures in-built in it. It is vulnerable to many types of routing attacks. The flood attack is one of them. In this paper, we propose a simple and effective technique to secure Ad hoc Ondemand Distance Vector (AODV) routing protocol against flood attacks. To deal with a flood attack, we have proposed Neighbor Defense Technique for Ad hoc On-demand Distance Vector (NDTAODV). This makes AODV more robust. The proposed technique has been designed to isolate the flood attacker with the use of timers, peak value and hello alarm technique. We have simulated our work in Network Simulator NS-2.33 (NS-2) with different pause times by way of different number of malicious nodes. We have compared the performance of NDTAODV with the AODV in normal situation as well as in the presence of malicious attacks. We have considered Packet Delivery Fraction (PDF), Average Throughput (AT) and Normalized Routing Load (NRL) for comparing the performance of NDTAODV and AODV.

A Novel Approach for intrusion Detection in Heterogeneous Wireless Sensor Networks using multipath Routing

The key concept of my redundancy management is to exploit the tradeoff between energy consumption vs. to the gain of reliability, timeliness, and also security for maximizing the system useful lifetime in this paper I am proposing to explore more extensive malicious attacks in addition to packet dropping and bad mouthing attacks, each and every with some different implications for energy, security and also reliability, and the investigate intrusion detection and multipath routing based tolerance protocols to react to these attacks. Intrusion detection is a security mechanism which is used to identify those who are trying to break and misuse the system without authorization and also those who have legitimate access but misusing their privileges. Intrusion detection system (IDS) design for CPSs has attracted considerable attention. Detection techniques in general can be classified into three types: the signature based, the anomaly based, and the specification based techniques. In the area of signature based IDS techniques, an IDS for CPSs that tests an automated transform from XML parole to Snort signature in an electricity distribution laboratory. An IDS for CPSs that takes a multi trust hybrid approach using signature based detection and traffic analysis. My work is different from these studies in that I use specification based detection rather than signature based detection to deal with unknown attacker patterns. Wireless sensor networks are deployed in an unattended environment in which energy replenishment is difficult if not impossible. Due to some limited resources, a WSN must not be only satisfied by the application specific QoS requirements such as reliability, timelines and also security, but also minimize the consumption of energy to prolong the system's useful lifetime. Where the tradeoff between the energy consumption vs. reliability gain with that of the goal to maximize the WSN system lifetime has been well explored in the literature. However, here no before work exists to consider that the tradeoff in the presence of malicious attackers. The Intrusion detection system monitors the activities of the system, analyze the activities to determine, any of the activity is violating the security rules. Once an IDS determines that an unusual activity or an event that is known to cause an attack happens, it then generates an Introduction. Thus, very likely the system would employ an intrusion detection system (IDS) with the goal to detect and remove malicious node .Multipath routing is considered an effective mechanism for fault and intrusion tolerance to improve data delivery in WSNs. It satisfies the energy consumption through turn off the sensor nodes for period of time to save energy. The basic idea is that the probability of at least one path reaching the sink node or base station increases as I have more paths doing data delivery. While the most priority prior research focused on using multipath routing to improve the reliability, some attention has to be paid for using multipath routing to tolerate insider attacks. A cyber physical system (CPS) typically comprises sensors, actuators, control units, and physical objects for controlling and protecting a physical infrastructure. Because of the dire consequence of a CPS failure, protecting a CPS from malicious attacks is of paramount importance. In this paper, we address the reliability issue of a CPS designed to sustain malicious attacks over a prolonged mission period without energy replenishment. A CPS often operates in a rough environment wherein energy replenishment is not possible, and nodes may be compromised (or captured) at that times. Thus, an intrusion detection and response system (IDRS) must be detected as malicious nodes without unnecessarily wasting energy to prolong the system lifetime. Multipath routing is considered an effective mechanism for fault and intrusion tolerance to improve data delivery in WSNs. The basic possibility is that the probability of at least one path reaching the sink node or base station increases as I have more paths doing data delivery. While the most priority prior research focused on using multipath routing to improve reliability some attention has been paid to using multipath routing to tolerate insider attacks. Coming to these studies, however, largely (highly) ignored the tradeoff between QoS gain vs. energy consumption with which it can adversely shortens the system lifetime.