Network Size Research Articles

Identification of network interactions from time series data: An iterative approach.

The first step toward advancing our understanding of complex networks involves determining their connectivity structures from the time series data. These networks are often high-dimensional, and in practice, only a limited amount of data can be collected. In this work, we formulate the network inference task as a bilinear optimization problem and propose an iterative algorithm with sequential initialization to solve this bilinear program. We demonstrate the scalability of our approach to network size and its robustness against measurement noise, hyper-parameter variation, and deviations from the network model. Results across experimental and simulated datasets, comprising oscillatory, non-oscillatory, and chaotic dynamics, showcase the superior inference accuracy of our technique compared to existing methods.

“Clustering Protocol Based on Territorial Energy for Random Deployment of Wireless Sensor Networks"

Wireless Sensor Networks (WSNs) have become increasingly essential for monitoring and gathering data in diverse environments. The efficient utilization of energy resources is crucial for prolonging the network lifetime in WSNs. This paper presents a comparative analysis of three prominent clustering protocols: Low Energy Adaptive Clustering Hierarchy (LEACH), Residual Energy Aware Clustering (REAC), and Territorial Energy-Driven Clustering Protocol (TEDC). TEDC introduces a novel approach where cluster heads are elected based on territorial dominance and energy levels, aiming to enhance energy efficiency and prolong network lifetime. The performance of LEACH, REAC, and TEDC is evaluated under various scenarios, including node density, network size, and communication range. Key performance metrics such as energy consumption, network lifetime, and packet delivery ratio are analyzed to assess the effectiveness of each protocol. The results provide valuable insights into the comparative strengths and weaknesses of these protocols, aiding network designers and researchers in selecting the most suitable protocol for specific WSN deployments. Additionally, the paper discusses potential optimizations and future research directions to further enhance the performance of clustering protocols in WSNs.

Short-Term Charging Load Prediction of Electric Vehicles with Dynamic Traffic Information Based on a Support Vector Machine

This study proposes a charging demand forecasting model for electric vehicles (EVs) that takes into consideration the characteristics of EVs with transportation and mobile load. The model utilizes traffic information to evaluate the influence of traffic systems on driving and charging behavior, specifically focusing on the characteristics of EVs with transportation and mobile load. Additionally, it evaluates the effect of widespread charging on the distribution network. An urban traffic network model is constructed based on the multi-intersection features, and a traffic network–distribution network interaction model is determined according to the size of the urban road network. Type classification simplifies the charging and discharging characteristics of EVs, enabling efficient aggregation of EVs. The authors have built a singular EV transportation model and an EV charging queue model is established. The EV charging demand is forecasted and then used as an input in the support vector machine (SVM) model. The final projection value for EV charging load is determined by taking into account many influencing elements. Compared to the real load, the proposed method’s feasibility and effectiveness are confirmed.

Enhancing security in wireless sensor networks: A fusion of deep learning and energy-efficient routing

The development of tiny sensing nodes efficient for wireless communication in Wireless Sensor Networks (WSNs) can be attributed to the rapid advancements in processors and radio technology. Data transmission occurs through multi-hop routing in WSN, which relies on nodes’ cooperation. The collaboration between nodes has rendered these networks susceptible to various attacks. It is imperative to employ a security scheme to evaluate the dependability of nodes in distinctive malicious nodes from non-malicious nodes. In recent years, there has been a growing significance placed on security-based routing protocols with energy constraints as valuable mechanisms for enhancing the security and performance of WSNs. A novel solution called the Deep Learning-based Hybrid Energy Efficient and Security System (DL-HE2S2) is introduced to address these challenges. The research workflow encompasses various essential stages, namely the deployment of nodes, the creation of clusters, the selection of cluster heads, the detection of malevolent nodes within each group, and the determination of optimal paths intra- and inter-clusters employing the routing algorithm for efficient packet transmission. The design of the algorithm is focused on achieving energy efficiency and enhancing network security while also taking into account various performance metrics, including a mean network lifetime of 187.244 hours, a throughput of 59.88 kilobits per second, an end-to-end latency of 11.939 milliseconds, a packet loss of 14.9%, a packet delivery ratio of 99.194%, network security at 92.026%, and energy usage of 19.424 J. This research examines the algorithm’s scalability and efficiency across various network sizes using a Network Simulator (NS-2). DL-HE2S2 offers valuable insights that can be applied to practical implementations in multiple applications.

Disaster recovery strategies for cyber-physical systems considering the Interdependence

Improving the resilience of Cyber-Physical Systems (CPS) against extreme events is vital. Current research, overly focused on power system restoration, often misses the complex interplay between the physical and cyber aspects of these systems, leading to potential vulnerabilities. This paper first proposes a framework for CPS considering interdependence. The framework accounts for cyber-physical collaborative recovery and the connectivity between cyber nodes and control center in the cyber system. The model is built considering factors such as repair crew, repair time, and network reconstruction, aimed at rapidly restoring the structural and functional resilience of CPS in the shortest possible time. Next, it integrates complex network theory to propose different repair strategies based on structural and functional indicators. Finally, the repair efficiency of the proposed strategies is comprehensively evaluated using structural and functional metrics. Simulation results show that the repair efficiency of CPS is superior to traditional power grids, and moderate reduction in the repair time of cyber nodes and a decrease in network size both contribute to enhancing repair efficiency.

Exploring the associations between structural and functional aspects of social relationships and cognition in very old age

Objectives Few studies have examined the associations between social relationships and cognition in very old adults (80+ years). Moreover, limited research has considered structural and functional aspects of social relationships together when exploring their associations with cognition. Therefore, the present study aimed to investigate the associations between structural and functional social aspects and cognition in very old age. Method The sample comprised 518 two-wave participants of the NRW80+ Study, who showed no indication of dementia and an average age of 85 years at baseline (range: 80–97 years). They were assessed twice over an approximate 2-year interval. Multilevel linear growth curve analysis was employed to model cognitive development over very old age and to examine the associations between structural and functional aspects of social relationships and cognitive development. It was controlled for demographic and health characteristics. Results A larger social network size (ß = 0.05, 95% CI [0.02, 0.08], p = 0.002) and frequent compared with occasional, seldom, and no contact with others (ß = 0.16, 95% CI [0.03, 0.28], p = 0.013) were associated with higher global cognition in the fully-controlled model. Model comparisons further revealed that structural aspects better predicted global cognition in very old age compared to functional aspects (χ2(2) = 16.17, p < 0.001). Conclusion The findings underline the positive association between structural aspects of social relationships and global cognition in very old age. They also highlight the need for interventions promoting a socially active lifestyle to prevent cognitive decline in this vulnerable age group.

Socio-economic inequities in mental health problems and wellbeing among women working in the apparel and floriculture sectors: testing the mediating role of psychological capital, social support and tangible assets

BackgroundIt is still unknown whether the mechanisms proposed by the Reserve Capacity Model (RCM) explaining socio-economic health and wellbeing inequities in high income countries can be applied to low-income countries. This study investigates whether different reserve capacities (intra-, inter-personal, and tangible) can explain the association between relative socio-economic position (SEP) and wellbeing outcome measures among Ethiopian women working in Foreign Direct Investment (FDI).MethodUsing a cross-sectional design, we collected quantitative survey data among 2,515 women working in the apparel and floriculture sectors in Ethiopia, measuring GHQ-12 mental health problems, multi-dimensional wellbeing, relative SEP, psychological capital (PsyCap), social support (emotional and financial social support network), and tangible assets (e.g., owning mobile phone, having access to toilet facilities). We used cluster-adjusted structural equation modelling to test whether PsyCap, social support, and/or tangible assets mediate the association between relative SEP (IV) and GHQ-12 mental health problems and multi-dimensional wellbeing (DVs).ResultsPsyCap and the size of the financial support network significantly mediate the socio-economic gradient in both wellbeing outcomes. The size of the emotional social support network shows no association with multi-dimensional wellbeing and shows an unexpected negative association with GHQ-12 mental health problems scores, including a significant mediation effect. Tangible assets show no association with the wellbeing outcome measures and do not mediate socio-economic mental health problems and wellbeing inequities.ConclusionsThe RCM can be applied in low-income countries, although in unexpected ways. Similar to findings from high-income countries, PsyCap and size of the financial social support network show significant mediation effects in explaining mental health problems and wellbeing inequities in Ethiopia. These reserves could therefore serve as a buffer for socio-economic inequities in mental health and wellbeing and can therefore assist in decreasing these inequities for women working in FDI sectors in Ethiopia.

Food-web complexity, consumer behavior, and diet specialism: impacts on ecosystem stability

AbstractEcological stability is a fundamental aspect of food web dynamics. In this study, we explore the factors influencing stability in complex ecological networks, characterizing it through biomass oscillations and species persistence. Using an Extended Niche model, we generate diverse food web structures and investigate the effects of intraspecific consumer interference, network size, connectance, and diet specialism on stability. Our findings reveal that intraspecific consumer interference plays a pivotal role in shaping stability. Higher interference results in stable dynamics, reducing oscillations and extinctions. Additionally, differences emerge between food webs comprised of invertebrate consumers and those of ectotherm vertebrates, with the latter showing higher oscillations. Network size and connectance also influence stability, where larger and more connected webs tend to exhibit reduced oscillations. Overall, our study sheds light on the complex interplay of factors affecting ecological stability in food webs. Understanding these dynamics is crucial for biodiversity conservation and ecosystem management.

Pre-yielding and necking process of double network hydrogels revealed by sample geometry effects

Understanding the yielding and necking mechanisms of double network (DN) materials is crucial for establishing structure-property correlations. While previous studies have primarily focused on how macro-yielding behavior changes with network structure and swelling, in this work, we study the local internal damage in the pre-yielding process to unveil the yielding and necking mechanisms for typical DN gels. Through birefringence retardation imaging during tensile deformation on DN gels of various sample geometries, our findings reveal the following key points: 1) Initiation and Growth of Damage Zones: Prior to macroscopic yielding, damage zones initiate from the sample edges and gradually grow towards the sample center with elongation. 2) Rapid Propagation and Yielding: Beyond a certain stress threshold, damage zones rapidly propagate under constant loading. Two damage zones eventually merge at the sample center, resulting in yielding. 3) Intrinsic Stress Determination: The stress at this point is intrinsic and determined by the structure of the two networks. 4) Pseudo Size-Dependency: Samples with insufficient width exhibit a pseudo size-dependency of yielding stress, as yielding occurs before reaching the critical stress. To explain the origin of intrinsic yielding stress, we introduce an intrinsic effective crack length (cI) as a measure of stress concentration around the internal crack tip of the damaged zone in the first network. Beyond this length, the influence on internal stress concentration around the damage zone is effectively screened by the load-bearing effect of the stretchable second network. The estimated cI, dependent on the microstructure of the two networks, was approximately 10 times the mesh size of the first network for typical DN gels.

Variability of Social Inclusion Patterns Involving Personal, Family and Social Characteristics in Latino Migrant Families in Spain

ABSTRACTMigration entails important changes for families. Despite existing evidence on the process of adaptation, little is known about the degree of social inclusion in a context of high cultural affinity. We examine the variability of social inclusion patterns, indexed by personal, family and social conditions, in 263 Latin American families, with at least one child under 12 years, migrating to Spain. Three profiles of social inclusion were identified using cluster analysis: High inclusion (32%), characterized by a good financial and job situation, low perception of economic difficulty, good social support and medium scores in attraction to neighbourhood and community belonging and high scores in neighbourly relationship. Partial inclusion (35%) was characterized by ratings of extreme economic difficulties and lack of job, high scores in perceived economic hardship, high scores in all types of social support (instrumental, emotional and affectionate support and positive social interaction) and high scores in attraction to neighbourhood and community belonging, and medium scores in neighbourly relationship. Finally, Low inclusion cluster (33%) was characterized by moderate economic difficulties and lack of a job, high scores in perceived economic hardship, low scores in all types of social support and low neighbourhood cohesion in all the factors. Multinomial logistic regression analysis showed that a model comprising family composition, size of social networks and years of residence in the community increased the probability of belonging to the Partial or Low inclusion profiles compared with the High inclusion one. In conclusion, the heterogeneity in the patterns of social inclusion and their intersection with family conditions suggests the need for targeted intervention actions to reduce the risk of social exclusion.

How do social networks, perception of social isolation, and loneliness affect depressive symptoms among Japanese adults?

This study aims to elucidate the complex relationship among social isolation, loneliness, and perception of social isolation and its influence on depressive symptoms by evaluating a hypothetical model. This understanding is essential for the formulation of effective intervention strategies. We conducted an online survey on Japanese adults (N = 3,315) and used the six-item Lubben Social Network Scale to assess the size of their social networks. We employed a single question to gauge their perception of social isolation. Loneliness was assessed using the three-item UCLA Loneliness Scale, and depressive symptoms were examined using the Patient Health Questionnaire-9. Structural equation modeling was employed to test the hypothesized model. The final model demonstrated satisfactory fit with data (χ2 (1) = 3.73; not significant; RMSEA = 0.03; CFI = 1.00; TLI = 1.00). The size of social network demonstrated a weak negative path to loneliness and depressive symptoms (β = -.13 to -.04). Notably, a strong positive association existed between perception of social isolation and loneliness (β = .66) and depressive symptoms (β = .27). Additionally, a significant positive relationship was found between loneliness and depressive symptoms (β = .40). Mediation analysis indicated that perception of social isolation and loneliness significantly intensified the relationships between social networks and depressive symptoms. Results indicate that interventions of psychological approaches, such as cognitive-behavioral therapy, are effective in reducing the perception of social isolation and loneliness, which may lead to the prevention of depressive symptoms. Future longitudinal studies are expected to refine and strengthen the proposed model.

Extractor-attention-predictor network for quantitative photoacoustic tomography

Quantitative photoacoustic tomography (qPAT) holds great potential in estimating chromophore concentrations, whereas the involved optical inverse problem, aiming to recover absorption coefficient distributions from photoacoustic images, remains challenging. To address this problem, we propose an extractor-attention-predictor network architecture (EAPNet), which employs a contracting–expanding structure to capture contextual information alongside a multilayer perceptron to enhance nonlinear modeling capability. A spatial attention module is introduced to facilitate the utilization of important information. We also use a balanced loss function to prevent network parameter updates from being biased towards specific regions. Our method obtains satisfactory quantitative metrics in simulated and real-world validations. Moreover, it demonstrates superior robustness to target properties and yields reliable results for targets with small size, deep location, or relatively low absorption intensity, indicating its broader applicability. The EAPNet, compared to the conventional UNet, exhibits improved efficiency, which significantly enhances performance while maintaining similar network size and computational complexity.

Loose bolt localization and torque prediction in a bolted joint using lamb waves and explainable artificial intelligence

With the increasing application of artificial intelligence (AI) techniques in the field of structural health monitoring (SHM), there is a growing interest in explaining the decision-making of the black-box models in deep learning-based SHM methods. In this work, we take explainability a step further by using it to improve the performance of AI models. In this work, the results of explainable artificial intelligence (XAI) algorithms are used to reduce the input size of a one-dimensional convolutional neural network (1D-CNN), hence simplifying the CNN structure. To select the most accurate XAI algorithm for this purpose, we propose a new evaluation method, feature sensitivity (FS). Utilizing XAI and FS, a reduced dimension 1D-CNN regression model (FS-X1D-CNN) is proposed to locate and predict the torque of loose bolts in a 16-bolt connected aluminum plate under varying temperature conditions. The results were compared with 1D CNN with raw input vector (RI-1D-CNN) and deep autoencoders-1D-CNN (DAE-1D-CNN). It is shown that FS-X1D-CNN achieves the highest prediction accuracy with 5.95 mm in localization and 0.54 Nm in torque prediction, and converges 10 times faster than RI-1D-CNN and 15 times faster than DAE-1D-CNN, while only using a single lamb wave signal path.

Cost minimization of multi-class demands groomed over multi-rate OTN interfaces considering WDM-layer constraints

This paper introduces a practical comprehensive approach to optimize the cost of installed Optical Transport Network (OTN) interfaces, as well as the required number of wavelengths in the substrate Wavelength Division Multiplexing (WDM) optical network. The approach assumes grooming variable-bit-rate traffic demands on multi-rate OTN interfaces. We propose an Integer Linear Programming (ILP) formulation to address the combined problem of OTN interface deployment, virtual-network formation, traffic grooming on the established distinct-capacity virtual paths, and virtual network embedding on the WDM physical substrate. The presented model focuses on mitigating the capital expenditure required to deploy OTN interfaces, usually the most costly element in the network, as well as minimizing the required number of wavelengths in the Routing and Wavelength Assignment (RWA) problem, contributing to the overall reduction in network costs and ensuring open capacity for future expansions. As the RWA problem is NP-hard and involves significant computational costs, we also propose a multi-step heuristic approach to tackle larger network instances more efficiently. In this approach, the optimization steps for the OTN design and the RWA solutions are conducted independently. For the OTN-layer planning, the optimum ILP formulation (OTN-ILP) is maintained, so that the optimum solution for the cost of installed interfaces on the OTN layer is guaranteed. The RWA problem is proposed through either an ILP formulation (RWA-ILP) or a heuristic approach. The RWA-ILP model can find the same or slightly superior number of wavelengths compared to the optimum solution of the Integrated ILP formulation, but it can provide results in reasonable times only for small networks. On the other hand, the heuristic RWA is able to run any network size, but it requires more wavelengths compared to the RWA-ILP model.

Microplastics shaped performance, microbial ecology and community assembly in simultaneous nitrification, denitrification and phosphorus removal process

The widespread use of microplastics (MPs) has led to an increase in their discharge to wastewater treatment plants. However, the knowledge of impact of MPs on macro-performance and micro-ecology in simultaneous nitrification, denitrification, and phosphorus removal (SNDPR) systems is limited, hampering the understanding of potential risks posed by MPs. This study firstly comprehensively investigated the performance, species interactions, and community assembly under polystyrene (PS) and polyvinyl chloride (PVC) exposure in SNDPR systems. The results showed under PS (1, 10 mg/L) and PVC (1, 10 mg/L) exposure, total nitrogen removal was reduced by 3.38–10.15 %. PS and PVC restrained the specific rates of nitrite and nitrate reduction (SNIRR, SNRR), as well as the activities of nitrite and nitrate reductase enzymes (NIR, NR). The specific ammonia oxidation rate (SAOR) and activity of ammonia oxidase enzyme (AMO) were reduced only at 10 mg/L PVC. PS and PVC enhanced the size of co-occurrence networks, niche breadth, and number of key species while decreasing microbial cooperation by 5.85–13.48 %. Heterogeneous selection dominated microbial community assembly, and PS and PVC strengthened the contribution of stochastic processes. PICRUSt prediction further revealed some important pathways were blocked by PS and PVC. Together, the reduced TN removal under PS and PVC exposure can be attributed to the inhibition of SAOR, SNRR, and SNIRR, the restrained activities of NIR, NR, and AMO, the changes in species interactions and community assembly mechanisms, and the suppression of some essential metabolic pathways. This paper offers a new perspective on comprehending the effects of MPs on SNDPR systems.

Boosting efficiency in state estimation of power systems by leveraging attention mechanism

Ensuring stability and reliability in power systems requires accurate state estimation, which is challenging due to the growing network size, noisy measurements, and nonlinear power-flow equations. In this paper, we introduce the Graph Attention Estimation Network (GAEN) model to tackle power system state estimation (PSSE) by capitalizing on the inherent graph structure of power grids. This approach facilitates efficient information exchange among interconnected buses, yielding a distributed, computationally efficient architecture that is also resilient to cyber-attacks. We develop a thorough approach by utilizing Graph Convolutional Neural Networks (GCNNs) and attention mechanism in PSSE based on Supervisory Control and Data Acquisition (SCADA) and Phasor Measurement Unit (PMU) measurements, addressing the limitations of previous learning architectures. In accordance with the empirical results obtained from the experiments, the proposed method demonstrates superior performance and scalability compared to existing techniques. Furthermore, the amalgamation of local topological configurations with nodal-level data yields a heightened efficacy in the domain of state estimation. This work marks a significant achievement in the design of advanced learning architectures in PSSE, contributing and fostering the development of more reliable and secure power system operations.

Young adults with vision impairment in India: Loneliness and social networks.

To examine the prevalence of loneliness and associated factors in young adults with vision impairment (VI), including quality of life (QoL) in India. Two hundred and three VI young adults (18-35 years) and 219 age-matched non-VI (controls) adults completed the loneliness scale, WHOQOL-BREF, Social Network Index (SNI) (network diversity, people in network size and number of embedded network subscales) and questions regarding sociodemographic characteristics and independent mobility. Rasch analysis was used to validate the questionnaires and interval-level scores were generated. Generalised linear models were used to estimate independent associations of sociodemographic factors, VI characteristics, social networks and QoL with loneliness. The prevalence of moderate and severe loneliness in the VI group was 10% (95% CI: 6.5, 15.4) and 4.4% (95% CI: 2.0, 8.2), respectively, and higher than that of controls. The VI group had a worse loneliness score than controls (-1.66 ± 2.25 vs. -2.13 ± 1.85 logits; p = 0.03). Those with ≤12 years and >12 years of education had loneliness scores of -1.58 ± 2.45 and -1.82 ± 1.99 logits, respectively (p = 0.01). Compared with controls, the VI group reported fewer extended family members, neighbours and friends leading to significantly smaller networks and network diversity (all p < 0.001). Loneliness scores demonstrated a significant correlation with only two SNI subscales for both groups: people in network size (r = -0.28 for VI; r = -0.30 for non-VI; p < 0.001 for both) and number of embedded networks (r = -0.22 for VI; r = -0.21 for non-VI; p = 0.002 for both). Both education (β = 0.45; p = 0.04) and QoL (β = -0.27, p = 0.02) were predictors of loneliness. Loneliness was commonly experienced by young VI adults and was higher among those with lower levels of education. Loneliness decreased with the presence of a larger number of people in network, suggesting that interventions to increase social activity and participation may be valuable in young VI adults.

A Critical Review of Artificial Intelligence Based Approaches in Intrusion Detection: A Comprehensive Analysis

Intrusion detection (ID) is critical in securing computer networks against various malicious attacks. Recent advancements in machine learning (ML), deep learning (DL), federated learning (FL), and explainable artificial intelligence (XAI) have drawn significant attention as potential approaches for ID. DL-based approaches have shown impressive performance in ID by automatically learning relevant features from data but require significant labelled data and computational resources to train complex models. ML-based approaches require fewer computational resources and labelled data, but their ability to generalize to unseen data is limited. FL is a relatively new approach that enables multiple entities to train a model collectively without exchanging their data, providing privacy and security benefits, making it an attractive option for ID. However, FL-based approaches require more communication resources and additional computation to aggregate models from different entities. XAI is critical for understanding how AI models make decisions, improving interpretability and transparency. While existing literature has explored the strengths and weaknesses of DL, ML, FL, and XAI-based approaches for ID, a significant gap exists in providing a comprehensive analysis of the specific use cases and scenarios where each approach is most suitable. This paper seeks to fill this void by delivering an in-depth review that not only highlights strengths and weaknesses but also offers guidance for selecting the appropriate approach based on the unique ID context and available resources. The selection of an appropriate approach depends on the specific use case, and this work provides insights into which method is best suited for various network sizes, data availability, privacy, and security concerns, thus aiding practitioners in making informed decisions for their ID needs.

An enhanced structural developmental neural network with information saturation for continual unsupervised learning

In this paper, an enhanced structural developmental neural network with information saturation (ESDNNIS) is proposed for continual unsupervised learning. It improves SDNNIS in the following respects: (1) it makes a distinction between the developmental stages of neurons to better regulate the updating of neurons; (2) it adopts the forgetting mechanism to control the network size; (3) it changes the way coverage domain parameter is initialized and updated to divide clusters more accurately; and (4) it adds a parent–child relationship update mechanism to provide better robustness to sample input order. ESDNNIS inherits the unsupervised incremental learning and class segmentation capabilities of SDNNIS, and achieves better class incremental clustering for more complex datasets while using fewer neurons than SDNNIS. The experimental results show that the accuracy of clustering results of ESDNNIS is improved by 10.94% on average compared with SDNNIS, while the number of generated neurons is reduced by 71.93% on average.

Impact of the Russia–Ukraine Conflict on Global Marine Network Based on Massive Vessel Trajectories

Maritime transportation plays a vital role in global trade, and studying the resilience of the global maritime network is crucial for ensuring its sustainable development. Currently, the ongoing conflict between Russia and Ukraine has garnered significant global attention. However, there is a lack of specific research on the impact of the conflict on maritime shipping, particularly the resilience of the global maritime network. This paper proposes a resilience assessment framework under the influence of significant events by combining complex network metrics and network performance indicators from the resilience triangle model. It quantitatively evaluates the resilience changes in the global maritime network before and after the outbreak of the Russia–Ukraine conflict. The experiment utilizes real automatic identification system (AIS) maritime trajectory data to quantify and visualize the changes in global maritime traffic during a 20-day period before and after the conflict, constructing the global maritime network for resilience calculations. The research findings indicate the following changes occurred after the Russia–Ukraine conflict. Firstly, the global maritime industry experienced overall growth, with increased ship transportation between ports. Transportation in certain regions was negatively affected, with a significant decrease in ship activities in the Black Sea and Adriatic Sea areas. The positions of Russia and Ukraine in the world maritime industry noticeably declined. Secondly, the network connectivity, network size, and network density of the global maritime network significantly increased, indicating an enhanced network resilience. According to our quantitative results, from a topological perspective, we observed the following changes: network connectivity increased by 27.2%, network scale increased by 36.6%, network density increased by 32.4%, and network resilience increased by 18.6%. Thirdly, the global maritime network is characterized by a high degree of heterogeneity, and the impact of conflicts on the heterogeneity of the shipping network is not significant. Finally, the network exhibited a slower performance decline under random attacks, while deliberate attacks led to a sharp decline. Due to the adaptive nature of the maritime network, the resilience of the network improves in terms of its topology following the outbreak of conflicts. After conflict incidents, the rate of performance decline during simulated attacks is lower compared to the pre-conflict period.