Network Vulnerability Research Articles

Vulnerability assessment and evolution analysis of Beijing's Urban Rail Transit Network

With the development of urban rail transit system, the complexity of the network intensifies, and its vulnerability shifts correspondingly. Understanding the characteristics and evolution of network vulnerability, as well as identifying developmental patterns, enables more scientific network planning. Current researches on network vulnerability predominantly focus on the static vulnerability assessment of existing networks, with limited studies on vulnerability evolution. This paper divides the topological evolution of the Beijing Urban Rail Transit Network (BURTN) from 2000 to 2020 into Formation and Improvement stages using the K-means++ method. By constructing a multidimensional vulnerability assessment model that considers node degree uniformity, network efficiency, and connectivity, the vulnerability evolution characteristics and patterns of BURTN are explored in cases of both Single-station failures and Multi-station consecutive failures (including random and intentional failures). Furthermore, the evolutionary relationship between network vulnerability and network structure is explored using the Ridge regression model. Calculations reveal that in the case of Single-station failures, during the Formation stage, the proportion of highly vulnerable stations (HVS) and the impact of each station failure on network performance decrease significantly, by 69.36 % and 67.67 %, respectively. During the Improvement stage, the proportion of HVS decreases significantly, while the impact of each station failure on network performance decreases slightly, by 79.10 % and 37.04 %, respectively. In the case of Multi-station consecutive failures, during the Formation stage, the network’s ability to cope with both random and intentional failures decreases, with the percentage of network nodes removed at the collapse state decreasing by 24.95 % and 11.12 %, respectively. During the Improvement stage, the network’s ability to cope with random failures remains stable, while its ability to cope with intentional failures decreases. This study helps to understand vulnerability from an evolutionary perspective and provides practical strategies for reducing vulnerability.

Analysis of Agricultural Water Security Based on Network Invulnerability: A Case Study in China's Virtual Water Trade Networks

Abstract“Invulnerability” of complex network was firstly introduced to virtual water (VW) research, aiming to broaden the scope of studies on water use and management. Beginning with the construction of China's virtual water trade networks (VWTNs) of major grain crops, Node Degree (K) and Betweenness Centrality (B) are employed to evaluate and rank the importance of China's 31 regions. Regions with high values for both indicators are identified as playing pivotal roles in the VWTNs: Jiangsu (ranking 1st for both K and B), Hubei (2nd for K, 3rd for B), Henan (3rd for K, 6th for B), Hebei (4th for K, 4th for B), Hunan (4th for K, 5th for B). Using this ranking to simulate the invulnerability of VWTNs under random and intentional attacks. The results reveal a rapid decrease in both Network Efficiency (E) and Maximum Connectivity (C) under intentional attack. In comparison to seven random attacks, E falls below 0.1 and C drops below 0.5 after only three intentional attacks, and the network completely collapsed after 10 intentional attacks. This highlights the VWTN's vulnerability in maintaining food supply and agricultural water security when key regions are subjected to man‐made destruction, such as military blockades or occupations. Future work should include integrating climate change models, crops yield models, and water resource allocation models to protect the key areas. Furthermore, interdisciplinary approaches are crucial for overcoming the limitations of VW research and these findings will provide valuable insights to enhance the optimal regulation of VWTNs.

Securing industrial control systems: Developing a SCADA/IoT test bench and evaluating lightweight cipher performance on hardware simulator

This paper addresses the critical need for enhancing security in Supervisory Control and Data Acquisition (SCADA) networks within Industrial Control Systems (ICSs) to protect the industrial processes from cyber-attacks. The purpose of our work is to propose and evaluate lightweight security measures to safeguard critical infrastructure resources. The scope of our effort involves simulating a secure SCADA/IoT-based hardware test bench for ICSs, utilizing Modbus and MQTT communication protocols. Through case studies in remote servo motor control, water distribution systems, and power system voltage level indicators, vulnerabilities such as Denial of Service (DoS) and Man-in-The-Middle (MiTM) attacks are identified, and security recommendations are provided. To execute our work, we deploy lightweight ciphers such as Prime Counter & Hash Chaining (PCHC) and Ascon algorithm with Compression Rate (ACR) for secure information exchange between the plant floor and the control center. Evaluation of these ciphers on Raspberry Pi focuses on execution speed and memory utilization. Additionally, a comparison with the AGA-12 protocol standard for SCADA networks is conducted to underscore the efficacy of the proposed security measures. Our findings include the identification of SCADA network vulnerabilities and the proposal of lightweight security measures to mitigate risks. Performance evaluation of the proposed ciphers on Raspberry Pi demonstrates their effectiveness, emphasizing the importance of deploying such measures to ensure resilience against cyber threats in SCADA environments.

A Unique Perspective on Vulnerability and Route Dynamics in the Global Liner Shipping Network

A Unique Perspective on Vulnerability and Route Dynamics in the Global Liner Shipping Network

Network disruption via continuous batch removal: The case of Sicilian Mafia.

Network disruption is pivotal in understanding the robustness and vulnerability of complex networks, which is instrumental in devising strategies for infrastructure protection, epidemic control, cybersecurity, and combating crime. In this paper, with a particular focus on disrupting criminal networks, we proposed to impose a within-the-largest-connected-component constraint in a continuous batch removal disruption process. Through a series of experiments on a recently released Sicilian Mafia network, we revealed that the constraint would enhance degree-based methods while weakening betweenness-based approaches. Moreover, based on the findings from the experiments using various disruption strategies, we propose a structurally-filtered greedy disruption strategy that integrates the effectiveness of greedy-like methods with the efficiency of structural-metric-based approaches. The proposed strategy significantly outperforms the longstanding state-of-the-art method of betweenness centrality while maintaining the same time complexity.

Tight isolated toughness bound for fractional (a,b,m)-deleted graphs

A graph [Formula: see text] is a fractional [Formula: see text]-deleted graph if deleting any [Formula: see text] edges from [Formula: see text], the resulting subgraph still admits a fractional [Formula: see text]-factor. As an exclusive graph-based parameter, isolated toughness and its variant are utilized to measure the vulnerability of networks which are modelled by graph topology structures. Early studies have shown the inherent connection between isolated toughness and the existence of fractional factors in specific settings. This paper provides a theoretical perspective on the sufficient conditions for fractional [Formula: see text]-deleted graphs with respect to isolated toughness and its variant. The sharpness of the given isolated toughness bounds is explained by counterexamples.

Towards robust neural networks: Exploring counterfactual causality-based repair

Despite the widespread applications of neural networks, increasing concerns arise regarding their reliability and security. This paper introduces a novel repair approach integrating counterfactual causality with a multi-objective algorithm to mitigate critical security property issues, including backdoor attacks, safety violations, and unfairness. Firstly, we abstract the neural network into a counterfactual structure causal model, utilizing a causal trace to allocate contributions to each hidden layer neuron for output prediction. Secondly, ”faulty” neurons are identified through fault location, utilizing counterfactual causality between positive and negative example-guided attributes and neuronal contribution. Finally, an ablation study determines the number of neurons for repair. The multi-objective algorithm NSGA-III is then applied to optimize the faulty neurons, adjusting weight parameters to rectify erroneous behavior and eliminate vulnerabilities while maximizing the model’s accuracy. This method achieves notable Pareto optimization and effectively eliminates vulnerabilities in neural networks. In comparative experiments against CARE (CAusality-based REpair) techniques, our Counterfactual Causality-Based Repair (CCBR) method significantly reduces backdoor attack success rates and security violations to less than 1%, while maintaining model accuracy and ensuring security for normal inputs. In fairness repair tasks, CCBR improves fairness by 92.56%, a substantial 19.56% increase over CARE. These results affirm the robustness and practicality of CCBR, establishing it as a valuable tool for enhancing neural network safety and fairness.

Cybersecurity Assessment and Vulnerability Modelling of Networks and Web Services in Nigerian Colleges of Education

Cybersecurity threats are among the most significant risks facing organizations and government today, and administrative boards have now been held accountable. This is an experimental research activity conducted to perform a holistic cybersecurity assessment and vulnerability modelling on the Information and Communication Technology (ICT) infrastructure and services of Colleges of Education in the six geopolitical zones. The study adopts an integrated bi-modal threat modelling and assessment (IBTMA) method by combining assessment and modelling approaches, which involves mixed-methods, along with computer-based experimentation to comprehensively evaluate and model cybersecurity threats, identify vulnerabilities, and propose effective mitigation strategies. Logistic regression data analysis was used to model the relationship between dependent variables (e.g., presence or absence of vulnerabilities or threats) and independent variables (e.g., cybersecurity practices, system configurations, policies, and staff training programs). This cybersecurity assessment provides the initial understanding of the security landscape and practices. The next step involves using the Microsoft Threat Modeling tool on the assets to identify specific threats. These threats are then prioritized based on their potential impact and likelihood. Assessment result of the vulnerability exposure is supported by the threat modelling report, which shows several threats: tampering, elevation of privilege, denial of service, privilege escalation, information disclosure, and spoofing. Findings from the study indicate that colleges face critical network and web vulnerabilities that need holistic solution.

A framework for using event evolutionary graphs to rapidly assess the vulnerability of urban flood cascade compound disaster event networks

Increasingly frequent flood disasters have caused great losses in recent years. Urban floods induce not only natural geological disasters but also social accidents. These disaster events, called urban flood cascade compound disaster events (UFCCDEs) in this study, have significant cascade, superposition, and amplification effects. However, conventional data sources and processing methods make it difficult to analyze the detailed course of disaster events caused by urban floods, thereby hindering the vulnerability assessment of UFCCDE networks (UFCCDENs). Herein, we propose a framework considering the interactions between disaster events caused by urban floods for rapidly and comprehensively assessing the vulnerability of the UFCCDEN. First, social media data (Sina Weibo) are processed to analyze the spatio-temporal distribution of UFCCDEs and construct a UFCCDEN based on an event evolutionary graph. Second, complex network theory is applied to evaluate the importance of disaster events and the vulnerability of disaster causal chains in the constructed UFCCDEN. Finally, the global efficiency of the network is calculated to assess the propagation efficiency of the UFCCDEN before and after implementing disaster mitigation strategies based on the assessment results to demonstrate the performance of the assessment framework. The coastal megacity Guangzhou was selected as an example. The results showed that, social media data can provide detailed and valid information about UFCCDEs, which can be used to construct the UFCCDEN based on the event evolutionary graph. Waterlogging is found to be the most important disaster event in the UFCCDEN. Furthermore, power facilities, drainage facilities, and roads should be given top priority in the prevention and mitigation of urban floods because of their significant cascading amplification effects. The proposed framework can make the propagation efficiency of the UFCCDEN markedly decrease by 37–62% and 44%, based on the assessment results of disaster events and causal chains, respectively.

An Intelligent Attack Detection Framework for the Internet of Autonomous Vehicles with Imbalanced Car Hacking Data

The modern Internet of Autonomous Vehicles (IoVs) has enabled the development of autonomous vehicles that can interact with each other and their surroundings, facilitating real-time data exchange and communication between vehicles, infrastructure, and the external environment. The lack of security procedures in vehicular networks and Controller Area Network (CAN) protocol leaves vehicles exposed to intrusions. One common attack type is the message injection attack, which inserts fake messages into original Electronic Control Units (ECUs) to trick them or create failures. Therefore, this paper tackles the pressing issue of cyber-attack detection in modern IoV systems, where the increasing connectivity of vehicles to the external world and each other creates a vast attack surface. The vulnerability of in-vehicle networks, particularly the CAN protocol, makes them susceptible to attacks such as message injection, which can have severe consequences. To address this, we propose an intelligent Intrusion detection system (IDS) to detect a wide range of threats utilizing machine learning techniques. However, a significant challenge lies in the inherent imbalance of car-hacking datasets, which can lead to misclassification of attack types. To overcome this, we employ various imbalanced pre-processing techniques, including NearMiss, Random over-sampling (ROS), and TomLinks, to pre-process and handle imbalanced data. Then, various Machine Learning (ML) techniques, including Logistic Regression (LR), Linear Discriminant Analysis (LDA), Naive Bayes (NB), and K-Nearest Neighbors (k-NN), are employed in detecting and predicting attack types on balanced data. We evaluate the performance and efficacy of these techniques using a comprehensive set of evaluation metrics, including accuracy, precision, F1_Score, and recall. This demonstrates how well the suggested IDS detects cyberattacks in external and intra-vehicle vehicular networks using unbalanced data on vehicle hacking. Using k-NN with various resampling techniques, the results show that the proposed system achieves 100% detection rates in testing on the Car-Hacking dataset in comparison with existing work, demonstrating the effectiveness of our approach in protecting modern vehicle systems from advanced threats.

Estimating the vulnerability of industrial network infrastructure in Central and Eastern Europe

Industrial infrastructure has suffered an unprecedented number of attacks in Central and Eastern Europe (CEE). This situation can be attributed to many geopolitical factors, including hybrid military conflicts and criminal activity. Industrial networks belonging to the countries that were once under Soviet influence suffer from an elevated risk of cyberattacks. The goal of this work is to propose an easy way to estimate the vulnerabilities of industrial networks to cyber threats on a national level. Since analysis of the industrial vulnerability landscape is difficult, this study proposes an assessment based on the popularity of vulnerable technologies—VPc. This metric is composed of search volume data on keywords related to industrial network technologies and reported security vulnerabilities associated with these words. Data on 116 keywords was analysed and a country-specific VPc index was calculated for twenty states in CEE. The analysis of the popularity of industrial technologies and vendors in CEE reveals interesting information about the industrial security and vulnerability landscape. The results show that some countries (e.g. Estonia) have more resilient industrial infrastructure than others (e.g. Belarus). The results presented in this study are not in conflict with other data and estimation attempts, including the National Cyber Security Index (NCSI). As new vulnerabilities are noted every day, the industrial security landscape changes rapidly. Therefore, a new easy-to-use metric (VPc) can be successfully used for general estimations. This work shows that the VPc score agrees with other estimates and analyses, but as with any other general estimation tool, it must be used with caution.

Spatial Structure and Vulnerability of Container Shipping Networks: A Case Study in the Beibu Gulf Sea Area

Ports play an important role in maintaining the effectiveness of maritime logistics. When ports encounter congestion, strikes, or natural disasters, the maritime container transportation network might be significantly affected. The Beibu Gulf sea area is a key channel to supporting China’s participation in international economic cooperation in the western region. It is highly susceptible to the influence of the political and economic instability. This study introduces a dual-component framework to analyze the inherent structure and potential vulnerabilities of the container transportation network in the Beibu Gulf Sea areas. The findings show that the core layer of the network exhibited circular solidification characteristics. The entire network heavily relies on some core ports, such as Haiphong Port, Ho Chi Minh Port, and Qinzhou Port, and it highlights the potential increases in vulnerability. The finding shows that deliberate attacks have a greater impact than random attacks on the normal operations of maritime networks. If ports with high intermediary centrality are attacked, the connectivity and transportation efficiency of the Beibu Gulf maritime network will be significantly affected. However, under such circumstances, redistributing cargo transportation through route adjustments can deal with the transmission of cascading failures and maintain the network’s resilience. Based on the existing knowledge and the data collected in a case study, this research stands out as the first to provide a critical examination of the spatial structure and vulnerability of container shipping networks in the Beibu Gulf sea.

Uncovering the structure and evolution of global virtual water and agricultural land network

The effective management and allocation of water and land resources is pivotal for ensuring global food security and facilitating the transition towards sustainable development. Combining multi-regional input-output analysis and complex network modelling, this study examining the spatial-temporal evolution and topological structures of multi-layer virtual water and agricultural land networks, analyzed the network characteristics of different countries/regions, and tested the network vulnerability between virtual water trade network and virtual agricultural land network. From the results, it is revealed that 30.9 %—34.14 % of global water resources and 32.9 %—40 % of global agricultural land are associated with international trade. Meanwhile, the geographical distribution of this trade has shifted from low-income countries to middle and high-income countries, showing three distinct fluctuations in their instability levels. Additionally, quantitative network analysis indicates that resource trade networks exhibit small-world characteristics with increasing trends in network density, clustering coefficient, and average path length values over time. Among the two disturbance scenarios, both virtual water and virtual agricultural land networks were more vulnerable under deliberate assault than under random failure, however, the agricultural land network demonstrates greater toughness compared to the virtual water network. These findings highlight the importance of sustainably managing global water and land resources within a framework for sustainable development which aims at promoting both worldwide food security and environmental protection simultaneously.

Analyzing risk influencing factors of ship collision accidents: A data-driven Bayesian network model integrating physical knowledge

The interaction and propagation effects among risk-influencing factors (RIFs) often lead to maritime accidents. Therefore, to support maritime risk management in decision-making, it is imperative to conduct a quantitative risk assessment (QRA) of these accidents, which requires a methodology capable of capturing causal relationships in limited cases. To achieve this goal, this study proposes a data-driven Bayesian network (BN) model that integrates physical knowledge with the QRA. Based on the collected data, a combination of domain knowledge, structure learning, and parameter learning is employed to construct the model. In the data-driven phase, three structural learning algorithms including Bayesian search (BS), Greedy thick thinning (GTT), and Peter-Clark (PC) algorithms, were used in combination with the expectation maximization algorithm of parameter learning. Then, using four indices calculated by the confusion matrix, the most fitting algorithm for structure learning was chosen, and the confusion matrix was obtained by five-fold cross-validation. Finally, network propagation impact (NPI) is introduced to prioritize RIFs. Preventive measures and emergency plans are formulated based on the network resilience metric (NRM) and network vulnerability metric (NVM). This approach leverages the objectivity of a data-driven BN and integrates domain expertise to enhance model validity. A case study of collisions was conducted to demonstrate the applicability of the model. The data were sourced from 327 collision accident reports provided by the China Maritime Safety Administration, and the PC algorithm was chosen. Findings indicate that the “Management system”, “Supervision”, and “Crew training” RIFs are prioritized for preventive measures, whereas “Communication” receives more resources for emergency handling. The results suggest the formulation of risk management strategies for practitioners to improve maritime safety.

Aging drives cerebrovascular network remodeling and functional changes in the mouse brain

Aging is frequently associated with compromised cerebrovasculature and pericytes. However, we do not know how normal aging differentially impacts vascular structure and function in different brain areas. Here we utilize mesoscale microscopy methods and in vivo imaging to determine detailed changes in aged murine cerebrovascular networks. Whole-brain vascular tracing shows an overall ~10% decrease in vascular length and branching density with ~7% increase in vascular radii in aged brains. Light sheet imaging with 3D immunolabeling reveals increased arteriole tortuosity of aged brains. Notably, vasculature and pericyte densities show selective and significant reductions in the deep cortical layers, hippocampal network, and basal forebrain areas. We find increased blood extravasation, implying compromised blood-brain barrier function in aged brains. Moreover, in vivo imaging in awake mice demonstrates reduced baseline and on-demand blood oxygenation despite relatively intact neurovascular coupling. Collectively, we uncover regional vulnerabilities of cerebrovascular network and physiological changes that can mediate cognitive decline in normal aging.

Quantifying the vulnerability of road networks to flood-induced closures using traffic simulation

Disruptions in road networks can significantly impact traffic flow, necessitating the identification and mitigation of vulnerable components. This study presents a methodology to assess the vulnerability and robustness of road networks, addressing the research gap in comprehensive approaches that effectively evaluate both controlled intersections and uncontrolled/free-flow highway segments. The proposed framework employs traffic simulation tools to model road closures within designated study areas, focusing on networks in Qatar. Performance measures such as average control delay and volume to capacity ratios are utilized to calculate vulnerability indices for network components. The two-pronged approach quantifies the impact of simulated closures at both the component and network levels, revealing critical intersections and road segments that markedly affect network performance during disruptions. The findings contribute to the field by providing transportation engineers with an effective quantitative tool to detect vulnerabilities in their road network, guiding design plans and actions to mitigate the potential harmful impacts of disruptive events on transportation infrastructure. The methodology's scalability ensures its applicability to various operational contexts, offering significant insights for transportation planning and infrastructure resilience. By establishing a comprehensive vulnerability analysis framework, this study enhances the understanding and management of road network vulnerabilities in complex urban environments. Further research on integrating dynamic traffic assignment and real-time data is recommended to further enhance the predictive accuracy and applicability of the proposed framework, ultimately improving transportation infrastructure resilience.

An Empirical Study on the Effect of Training Data Perturbations on Neural Network Robustness.

The vulnerability of modern neural networks to random noise and deliberate attacks has raised concerns about their robustness, particularly as they are increasingly utilized in safety- and security-critical applications. Although recent research efforts were made to enhance robustness through retraining with adversarial examples or employing data augmentation techniques, a comprehensive investigation into the effects of training data perturbations on model robustness remains lacking. This paper presents the first extensive empirical study investigating the influence of data perturbations during model retraining. The experimental analysis focuses on both random and adversarial robustness, following established practices in the field of robustness analysis. Various types of perturbations in different aspects of the dataset are explored, including input, label, and sampling distribution. Single-factor and multi-factor experiments are conducted to assess individual perturbations and their combinations. The findings provide insights into constructing high-quality training datasets for optimizing robustness and recommend the appropriate degree of training set perturbations that balance robustness and correctness, and contribute to understanding model robustness in deep learning and offer practical guidance for enhancing model performance through perturbed retraining, promoting the development of more reliable and trustworthy deep learning systems for safety-critical applications.

TRAFFIC ANOMALY DETECTION IN VEHICLE BUS BY RECURRENT LSTM NEURAL NETWORK

Modern high-end cars have many electronic control units for driving assistance that combine huge amounts of data about the functioning of car components. A significant part of these vehicles use a controller area network for communication between electronic units. Controller area network is a simple and reliable network protocol that due to its simplicity lacks any security mechanisms for data transmission. The problem of controller area network vulnerability is worsening as constantly growing amounts of data between cars, road infrastructure and the Internet. The traffic of attacks on controller area networks can be treated as abnormal that allows using anomaly detection methods for their recognition. In this work we propose the recurrent long short-term memory encoder-decoder neural network for controller area network attacks detection.

К ВОПРОСУ ОБЕСПЕЧЕНИЯ СЕТЕВОЙ БЕЗОПАСНОСТИ

The following are considered:  principles and means of ensuring the security of a corporate network, as well as features of virtual private networks that provide reliable protection and secure data transmission;  principles and measures to ensure the security of the corporate network, as well as the types of information protection provided by these security measures;  virtual private networks (VPNs) that provide a secure encrypted user connection to the network. Using a VPN provides reliable data protection, masking the user's geolocation, access to regional content and secure data transfer;  systems for detecting and preventing network hacking threats, among which the main focus is on the Kali Linux software package, as well as features of the Windows subsystem for Linux and Windows network monitoring tools;  Windows network monitoring tools, including utilities that work in console command-line mode and allow you to check the availability of remote personal computers and diagnose the connection. The Kali Linux software package is a distribution package for the Linux operating system that has more than 600 preinstalled network penetration testing programs. Among the features of the package, the article discusses the search and exploitation of network vulnerabilities, checking the correct configuration of the SSL certificate and open ports, tracing the data transfer route, checking the availability of servers and searching for network problems.

Identification of Key Risk Nodes and Invulnerability Analysis of Construction Supply Chain Networks

The construction supply chain confronts interruption risks that raise significant concerns regarding industry safety and stability. Consequently, exploring risk management strategies from both enterprise and supply chain network perspectives is crucial. This study employs complex network theory and the cascade failure model to propose a methodology tailored to the unique characteristics of the construction supply chain, facilitating the identification of key risk nodes and the conduct of invulnerability analyses. By evaluating the importance of construction enterprise nodes and their risk propagation ability during cascade failures, this method enables the comprehensive identification of key risk node enterprises within the construction supply chain network. Furthermore, this study examines and discusses strategies for enhancing network invulnerability by taking into account node capacity, load, and resilience. Empirical results indicate that the key nodes and risk nodes in the construction supply chain network are mainly located upstream and downstream, displaying specific distribution patterns. In addition to core enterprises, key risk nodes comprise some strong suppliers at the intermediary and lower tiers of the supply chain. Adjustments to node enterprise parameters like capacity, load, and resilience have diverse impacts on the invulnerability of the construction supply chain network. This study clarifies the distribution patterns of key risk nodes within the construction supply chain network and the variations in network invulnerability under particular conditions, providing valuable insights for risk management decision-making.