Biblio

This paper studies the technology of generating DDPG (deep deterministic policy gradient) by using the deep dual network and experience pool network structure, and puts forward the sampling strategy gradient algorithm to randomly select actions according to the learned strategies (action distribution) in the continuous action space, based on the dispatching control system of the power dispatching control center of a super city power grid, According to the actual characteristics and operation needs of urban power grid, The developed refined artificial intelligence on-line security analysis and emergency response plan intelligent generation function realize the emergency response auxiliary decision-making intelligent generation function. According to the hidden danger of overload and overload found in the online safety analysis, the relevant load lines of the equipment are searched automatically. Through the topology automatic analysis, the load transfer mode is searched to eliminate or reduce the overload or overload of the equipment. For a variety of load transfer modes, the evaluation index of the scheme is established, and the optimal load transfer mode is intelligently selected. Based on the D5000 system of Metropolitan power grid, a multi-objective and multi resource coordinated security risk decision-making assistant system is implemented, which provides integrated security early warning and decision support for the main network and distribution network of city power grid. The intelligent level of power grid dispatching management and dispatching operation is improved. The state reality network can analyze the joint state observations from the action reality network, and the state estimation network uses the actor action as the input. In the continuous action space task, DDPG is better than dqn and its convergence speed is faster.

Cyber-attacks toward cyber-physical systems are one of the main concerns of smart grid's operators. However, many of these cyber-attacks, are toward unmanned substations where the cyber-attackers needs to be close enough to substation to malfunction protection and control systems in substations, using Electromagnetic signals. Therefore, in this paper, a new threat detection algorithm is proposed to prevent possible cyber-attacks toward unmanned substations. Using surveillance camera's streams and based on You Only Look Once (YOLO) V3, suspicious objects in the image are detected. Then, using Intersection over Union (IOU) and Generalized Intersection Over Union (GIOU), threat distance is estimated. Finally, the estimated threats are categorized into three categories using color codes red, orange and green. The deep network used for detection consists of 106 convolutional layers and three output prediction with different resolutions for different distances. The pre-trained network is transferred from Darknet-53 weights trained on 80 classes.

Information security has become a crucial issue not only from the technical standpoint, but also from the managerial standpoint. The necessity for organizations to understand and manage cyber risk has led to the rise of a plethora of risk assessment methods and tools. These approaches are often difficult to interpret and complex to manage for organizations. In this paper, we propose a simple and quantitative method for the estimation of the likelihood of occurrence of a cyber incident. Our approach uses a generalized logistic function and a cumulative geometric distribution to combine the maturity and the complexity of the technical infrastructure of an organization with its attractiveness towards cyber criminals.

New services and products are increasingly becoming integral parts of our daily lives rising our technological dependence, as well as our exposure to risks from cyber. Critical sectors such as transport are progressively depending on digital technologies to run their core operations and develop novel solutions to exploit the economic strengths of the European Union. However, despite the fact that the continuously increasing number of visitors, entering the European Union through land-border crossing points or seaports, brings tremendous economic benefits, novel border control solutions, such as mobile devices for passenger identification for land and sea border control, are essential to accurately identify passengers ``on the fly'' while ensuring their comfort. However, the highly confidential personal data managed by these devices makes them an attractive target for cyberattacks. Therefore, novel secure and usable user authentication mechanisms are required to increase the level of security of this kind of devices without interrupting border control activities. Towards this direction, we, firstly, discuss risk-based and adaptive authentication for mobile devices as a suitable approach to deal with the security vs. usability challenge. Besides that, a novel risk-based adaptive user authentication mechanism is proposed for mobile passenger identification devices used by border control officers at land and sea borders.

In this paper, we present a theoretical approach concerning the econometric modelling for the estimation of cyber-security risk, with the use of time-series analysis methods and alternatively with Machine Learning (ML) based, deep learning methodology. Also we present work performed in the framework of SAINT H2020 Project [1], concerning innovative data mining techniques, based on automated web scrapping, for the retrieving of the relevant time-series data. We conclude with a review of emerging challenges in cyber-risk assessment brought by the rapid development of adversarial AI.

A real or potential threat to various large and small security objects, which comes from both internal and external attackers, determines one or another activities to ensure internal and external security. These actions depend on the spheres of life of state and society, which are targeted by the security threats. These threats can be conveniently classified into political threats (or threats to the existing constitutional order), economic, military, informational, technogenic, environmental, corporate, and other threats. The article discusses a model of an information system, which main criterion is the system security based on the concept of risk. When considering the model, it was determined that it possess multi-criteria aspects. Therefore the establishing the quantitative and qualitative characteristics is a complex and dynamic task. The paper proposes to use the mathematical apparatus of the teletraffic theory in one of the elements of the protected system, namely, in the end-to-end security subsystem.

Data generated from various dynamic applications of Internet of Things (IoT) based healthcare technology is effectively used for decision-making, providing reliable and smart healthcare services to the elderly and patients with chronic diseases. Since these precious data are susceptible to various security attacks, continuous monitoring of the system's compliance and identification of security risks in IoT data propagation is essential through potentially several layers of applications. This paper pinpoints how security-aware data provenance graphs can support compliance checking and risk estimation by including sufficient information on security controls and other security-relevant evidence. Real-time analysis of these security evidence to enable a step-wise validation and providing the evidence of this validation to end-users is currently not possible with the available data. This paper analyzes the security concerns in different phases of data propagation in a designed IoT-health scenario and promotes step-wise validation of security evidence. It proposes a system model with a novel protocol that documents and verifies evidence for security controls for data-object relations in data provenance graphs to assist compliance checking of security regulation of healthcare systems. With this regard, this paper discusses the proposed system model design with the requirements for technical safeguards of the Health Insurance Portability and Accountability Act (HIPAA). Based on the verification output at each phase, the proposed protocol reports this chain of verification by creating certain security tokens. Finally, the paper provides a formal security validation and security design analysis to show the applicability of this step-wise validation within the proposed system model.

Campus Companies and network operators perform risk assessment to inform policy-making, guide infrastructure investments or to comply with security standards such as ISO 27001. Due to the size and complexity of these networks, risk assessment techniques such as attack graphs or trees describe the attacker with a finite set of rules. This characterization of the attacker can easily miss attack vectors or overstate them, potentially leading to incorrect risk estimation. In this work, we propose the first methodology to justify a rule-based attacker model. Conceptually, we add another layer of abstraction on top of the symbolic model of cryptography, which reasons about protocols and abstracts cryptographic primitives. This new layer reasons about Internet-scale networks and abstracts protocols.We show, in general, how the soundness and completeness of a rule-based model can be ensured by verifying trace properties, linking soundness to safety properties and completeness to liveness properties. We then demonstrate the approach for a recently proposed threat model that quantifies the confidentiality of email communication on the Internet, including DNS, DNSSEC, and SMTP. Using off-the-shelf protocol verification tools, we discover two flaws in their threat model. After fixing them, we show that it provides symbolic soundness.

In the federated learning system, parameter gradients are shared among participants and the central modulator, while the original data never leave their protected source domain. However, the gradient itself might carry enough information for precise inference of the original data. By reporting their parameter gradients to the central server, client datasets are exposed to inference attacks from adversaries. In this paper, we propose a quantitative metric based on mutual information for clients to evaluate the potential risk of information leakage in their gradients. Mutual information has received increasing attention in the machine learning and data mining community over the past few years. However, existing mutual information estimation methods cannot handle high-dimensional variables. In this paper, we propose a novel method to approximate the mutual information between the high-dimensional gradients and batched input data. Experimental results show that the proposed metric reliably reflect the extent of information leakage in federated learning. In addition, using the proposed metric, we investigate the influential factors of risk level. It is proven that, the risk of information leakage is related to the status of the task model, as well as the inherent data distribution.

Probabilistic security metrics estimate the vulnerability of a network in terms of the likelihood of an attacker reaching the goal states (of a network) by exploiting the attack graph paths. The probability computation depends upon several assumptions regarding the possible attack scenarios. In this paper, we extend the existing security metric to model networks with intrusion detection systems and their associated uncertainties and time latencies. We consider learning capabilities of attackers as well as detection systems. Estimation of risk is obtained by using the attack paths that are undetectable owing to the latency of the detection system. Thus, we define the overall vulnerability (of a network) as a function of the time window available to an attacker for repeated exploring (via learning) and exploitation of a network, before the attack is mitigated by the detection system. Finally, we consider the realistic scenario where an attacker explores and abandons various partial paths in the attack graph before the actual exploitation. A dynamic programming formulation of the vulnerability computation methodology is proposed for this scenario. The nature of these metrics are explained using a case study showing the vulnerability spectrum from the case of zero detection latency to a no detection scenario.

Functional Safety standards like ISO 26262 require a detailed analysis of the dependability of components subjected to perturbations. Radiation testing or even much more abstract RTL fault injection campaigns are costly and complex to set up especially for SoCs and Cyber Physical Systems (CPSs) comprising intertwined hardware and software. Moreover, some approaches are only applicable at the very end of the development cycle, making potential iterations difficult when market pressure and cost reduction are paramount. In this tutorial, we present a summary of classical state-of-the-art approaches, then alternative approaches for the dependability analysis that can give an early yet accurate estimation of the safety or security characteristics of HW-SW systems. Designers can rely on these tools to identify issues in their design to be addressed by protection mechanisms, ensuring that system dependability constraints are met with limited risk when subjected later to usual fault injections and to e.g., radiation testing or laser attacks for certification.

Digital video has various types of entities, which are utilized as embedding domains to hide messages in steganography. However, nearly all video steganography uses only one type of embedding domain, resulting in limited embedding capacity and potential security risks. In this paper, we firstly propose to embed in multi-domains for video steganography by combining partition modes (PMs) and motion vectors (MVs). The multi-domain embedding (MDE) aims to spread the modifications to different embedding domains for achieving higher undetectability. The key issue of MDE is the interactions of entities across domains. To this end, we design two MDE strategies, which hide data in PM domain and MV domain by sequential embedding and simultaneous embedding respectively. These two strategies can be applied to existing steganography within a distortion-minimization framework. Experiments show that the MDE strategies achieve a significant improvement in security performance against targeted steganalysis and fusion based steganalysis.

The denial-of-service (DoS) attack is a very common type of cyber attack that can affect critical cyber-physical systems, such as smart grids, by hampering the monitoring and control of the system, for example, creating unavailability of data from the attacked zone. While developing countermeasures can help reduce such risks, it is essential to develop techniques to recover from such scenarios if they occur by estimating the state of the system. Considering the continuous data-stream from the PMUs as time series, this work exploits the bus-to-bus cross-correlations to estimate the state of the system's components under attack using the PMU data of the rest of the buses. By applying this technique, the state of the power system can be estimated under various DoS attack sizes with great accuracy. The estimation accuracy in terms of the mean squared error (MSE) has been used to identify the relative vulnerability of the PMUs of the grid and the most vulnerable time for the DoS attack.

Intelligent connected vehicle system tightly integrates computing, communication, and control strategy. It can increase the traffic throughput, minimize the risk of accidents and reduce the energy consumption. However, because of the openness of the vehicular ad hoc network, the system is vulnerable to cyber-attacks and may result in disastrous consequences. Hence, it is interesting in design of the connected vehicular systems to be resilient to the sensor attacks. The paper focuses on the estimation and control of the intelligent connected vehicle systems when the sensors or the wireless channels of the system are attacked by attackers. We give the upper bound of the corrupted sensors that can be corrected and design the state estimator to reconstruct the initial state by designing a closed-loop controller. Finally, we verify the algorithm for the connected vehicle system by some classical simulations.

Nowadays the role of Information systems in our life has greatly increased, which has become one of the biggest challenges for citizens, organizations and governments. Every single day we are becoming more and more dependent on information and communication technology (ICT). A major goal of information security is to find the best ways to mitigate the risks. The context-role and perimeter protection approaches can reduce and prevent an unauthorized penetration to protected zones and information systems inside the zones. The result of this work can be useful for the security system analysis and optimization of their organizations.

The most significant threats to networks usually originate from external entities. As such, the Northbound interface of SDN networks which ensures communication with external applications requires particularly close attention. In this paper we propose the Risk Assessment and Management approach to SEcure SDN (RAMSES). This novel solution is able to estimate the risk associated with traffic demand requests received via the Northbound-API in SDN networks. RAMSES quantifies the impact on network cost incurred by expected traffic demands and specifies the likelihood of adverse requests estimated using the reputation system. Accurate risk estimation allows SDN network administrators to make the right decisions and mitigate potential threat scenarios. This can be observed using extensive numerical verification based on an network optimization tool and several scenarios related to the reputation of the sender of the request. The verification of RAMSES confirmed the usefulness of its risk assessment approach to protecting SDN networks against threats associated with the Northbound-API.

The use of risk information can help software engineers identify software components that are likely vulnerable or require extra attention when testing. Some studies have shown that the requirements risk-based approaches can be effective in improving the effectiveness of regression testing techniques. However, the risk estimation processes used in such approaches can be subjective, time-consuming, and costly. In this research, we introduce a fuzzy expert system that emulates human thinking to address the subjectivity related issues in the risk estimation process in a systematic and an efficient way and thus further improve the effectiveness of test case prioritization. Further, the required data for our approach was gathered by employing a semi-automated process that made the risk estimation process less subjective. The empirical results indicate that the new prioritization approach can improve the rate of fault detection over several existing test case prioritization techniques, while reducing threats to subjective risk estimation.

Regulatory efforts such as the General Data Protection Regulation (GDPR) embody a notion of privacy risk that is centered around the fundamental rights of data subjects. This is, however, a fundamentally different notion of privacy risk than the one commonly used in threat modeling which is largely agnostic of involved data subjects. This mismatch hampers the applicability of privacy threat modeling approaches such as LINDDUN in a Data Protection by Design (DPbD) context. In this paper, we present a data subject-aware privacy risk assessment model in specific support of privacy threat modeling activities. This model allows the threat modeler to draw upon a more holistic understanding of privacy risk while assessing the relevance of specific privacy threats to the system under design. Additionally, we propose a number of improvements to privacy threat modeling, such as enriching Data Flow Diagram (DFD) system models with appropriate risk inputs (e.g., information on data types and involved data subjects). Incorporation of these risk inputs in DFDs, in combination with a risk estimation approach using Monte Carlo simulations, leads to a more comprehensive assessment of privacy risk. The proposed risk model has been integrated in threat modeling tool prototype and validated in the context of a realistic eHealth application.

Measures and methods used in financial sector to quantify risk, have been recently applied to cyber world. The aim is to help organizations to improve risk management strategies and to wisely plan investments in cyber security. On the other hand, they are useful instruments for insurance companies in pricing cyber insurance contracts and setting the minimum capital requirements defined by the regulators. In this paper we propose an estimation of Value at Risk (VaR), referred to as Cyber Value at Risk in cyber security domain, and Tail Value at risk (TVaR). The data breach information we use is obtained from the “Chronology of data breaches” compiled by the Privacy Rights Clearinghouse.

Several operational and economic factors impact the patching decisions of critical infrastructures. The constraints imposed by such factors could prevent organizations from fully remedying all of the vulnerabilities that expose their (critical) assets to risk. Therefore, an involved decision maker (e.g. security officer) has to strategically decide on the allocation of possible remediation efforts towards minimizing the inherent security risk. This, however, involves the use of comparative judgments to prioritize risks and remediation actions. Throughout this work, the security risk is quantified using the security metric Time-To-Compromise (TTC). Our main contribution is to provide a generic TTC estimator to comparatively assess the security posture of computer networks taking into account interdependencies between the network components, different adversary skill levels, and characteristics of (known and zero-day) vulnerabilities. The presented estimator relies on a stochastic TTC model and Monte Carlo simulation (MCS) techniques to account for the input data variability and inherent prediction uncertainties.

The method of choice the control parameters of a complex system based on estimates of the risks is proposed. The procedure of calculating the estimates of risks intended for a choice of rational managing directors of influences by an allocation of the group of the operating factors for the set criteria factor is considered. The purpose of choice of control parameters of the complex system is the minimization of an estimate of the risk of the functioning of the system by mean of a solution of a problem of search of an extremum of the function of many variables. The example of a choice of the operating factors in the sphere of intangible assets is given.

To manage cybersecurity risks in practice, a simple yet effective method to assess suchs risks for individual systems is needed. With time-to-compromise (TTC), McQueen et al. (2005) introduced such a metric that measures the expected time that a system remains uncompromised given a specific threat landscape. Unlike other approaches that require complex system modeling to proceed, TTC combines simplicity with expressiveness and therefore has evolved into one of the most successful cybersecurity metrics in practice. We revisit TTC and identify several mathematical and methodological shortcomings which we address by embedding all aspects of the metric into the continuous domain and the possibility to incorporate information about vulnerability characteristics and other cyber threat intelligence into the model. We propose $\beta$-TTC, a formal extension of TTC which includes information from CVSS vectors as well as a continuous attacker skill based on a $\beta$-distribution. We show that our new metric (1) remains simple enough for practical use and (2) gives more realistic predictions than the original TTC by using data from a modern and productively used vulnerability database of a national CERT.

With the ever-increasing popularity of LiDAR (Light Image Detection and Ranging) sensors, a wide range of applications such as vehicle automation and robot navigation are developed utilizing the 3D LiDAR data. Many of these applications involve remote guidance - either for safety or for the task performance - of these vehicles and robots. Research studies have exposed vulnerabilities of using LiDAR data by considering different security attack scenarios. Considering the security risks associated with the improper behavior of these applications, it has become crucial to authenticate the 3D LiDAR data that highly influence the decision making in such applications. In this paper, we propose a framework, ALERT (Authentication, Localization, and Estimation of Risks and Threats), as a secure layer in the decision support system used in the navigation control of vehicles and robots. To start with, ALERT tamper-proofs 3D LiDAR data by employing an innovative mechanism for creating and extracting a dynamic watermark. Next, when tampering is detected (because of the inability to verify the dynamic watermark), ALERT then carries out cross-modal authentication for localizing the tampered region. Finally, ALERT estimates the level of risk and threat based on the temporal and spatial nature of the attacks on LiDAR data. This estimation of risk and threats can then be incorporated into the decision support system used by ADAS (Advanced Driver Assistance System). We carried out several experiments to evaluate the efficacy of the proposed ALERT for ADAS and the experimental results demonstrate the effectiveness of the proposed approach.

Accurately modeling human decision-making in security is critical to thinking about when, why, and how to recommend that users adopt certain secure behaviors. In this work, we conduct behavioral economics experiments to model the rationality of end-user security decision-making in a realistic online experimental system simulating a bank account. We ask participants to make a financially impactful security choice, in the face of transparent risks of account compromise and benefits offered by an optional security behavior (two-factor authentication). We measure the cost and utility of adopting the security behavior via measurements of time spent executing the behavior and estimates of the participant's wage. We find that more than 50% of our participants made rational (e.g., utility optimal) decisions, and we find that participants are more likely to behave rationally in the face of higher risk. Additionally, we find that users' decisions can be modeled well as a function of past behavior (anchoring effects), knowledge of costs, and to a lesser extent, users' awareness of risks and context (R2=0.61). We also find evidence of endowment effects, as seen in other areas of economic and psychological decision-science literature, in our digital-security setting. Finally, using our data, we show theoretically that a "one-size-fits-all" emphasis on security can lead to market losses, but that adoption by a subset of users with higher risks or lower costs can lead to market gains.

Automotive systems have always been designed with safety in mind. In this regard, the functional safety standard, ISO 26262, was drafted with the intention of minimizing risk due to random hardware faults or systematic failure in design of electrical and electronic components of an automobile. However, growing complexity of a modern car has added another potential point of failure in the form of cyber or sensor attacks. Recently, researchers have demonstrated that vulnerability in vehicle's software or sensing units could enable them to remotely alter the intended operation of the vehicle. As such, in addition to safety, security should be considered as an important design goal. However, designing security solutions without the consideration of safety objectives could result in potential hazards. Consequently, in this paper we propose the notion of security for safety and show that by integrating safety conditions with our system-level security solution, which comprises of a modified Kalman filter and a Chi-squared detector, we can prevent potential hazards that could occur due to violation of safety objectives during an attack. Furthermore, with the help of a car-following case study, where the follower car is equipped with an adaptive-cruise control unit, we show that our proposed system-level security solution preserves the safety constraints and prevent collision between vehicle while under sensor attack.