Network Behavior Research Articles

Modeling and analysis of cascading failures in multilayer higher-order networks

With the increasing application of network science, understanding the behavior of higher-order networks has become particularly important. Especially, the study of multilayer higher-order networks is significant for revealing the interdependence and higher-order interactions in complex systems. This paper proposes a mathematical framework to explore multilayer higher-order networks composed of multiple fully interdependent hypergraphs. Each hypergraph consists of the same number of nodes, and nodes between layers are connected one-to-one. By analyzing the cascading failures of multilayer hypergraphs under different topologies, we found that although the network sizes are the same in the steady state, different topological structures (such as star-like, tree-like, and chain-like) significantly affect the time required to reach a stable state, with the star-like topology reaching stability the fastest. Furthermore, we explored the impact of network parameters on the robustness of networks. We found that in multilayer homogeneous hypergraphs, the robustness of the network becomes stronger with an increase in the average hyperdegree or average hyperedge cardinality; in multilayer heterogeneous hypergraphs, the robustness of the network becomes more fragile as the power exponent increases. Finally, experimental results indicate that with the rise in the number of network layers, the network becomes more fragile.

6G in medical robotics: development of network allocation strategies for a telerobotic examination system.

Healthcare systems around the world are increasingly facing severe challenges due to problems such as staff shortage, changing demographics and the reliance on an often strongly human-dependent environment. One approach aiming to address these issues is the development of new telemedicine applications. The currently researched network standard 6G promises to deliver many new features which could be beneficial to leverage the full potential of emerging telemedical solutions and overcome the limitations of current network standards. We developed a telerobotic examination system with a distributed robot control infrastructure to investigate the benefits and challenges of distributed computing scenarios, such as fog computing, in medical applications. We investigate different software configurations for which we characterize the network traffic and computational loads and subsequently establish network allocation strategies for different types of modular application functions (MAFs). The results indicate a high variability in the usage profiles of these MAFs, both in terms of computational load and networking behavior, which in turn allows the development of allocation strategies for different types of MAFs according to their requirements. Furthermore, the results provide a strong basis for further exploration of distributed computing scenarios in medical robotics. This work lays the foundation for the development of medical robotic applications using 6G network architectures and distributed computing scenarios, such as fog computing. In the future, we plan to investigate the capability to dynamically shift MAFs within the network based on current situational demand, which could help to further optimize the performance of network-based medical applications and play a role in addressing the increasingly critical challenges in healthcare.

Antagonistic behavior of brain networks mediated by low-frequency oscillations: electrophysiological dynamics during internal–external attention switching

Antagonistic activity of brain networks likely plays a fundamental role in how the brain optimizes its performance by efficient allocation of computational resources. A prominent example involves externally/internally oriented attention tasks, implicating two anticorrelated, intrinsic brain networks: the default mode network (DMN) and the dorsal attention network (DAN). To elucidate electrophysiological underpinnings and causal interplay during attention switching, we recorded intracranial EEG (iEEG) from 25 epilepsy patients with electrode contacts localized in the DMN and DAN. We show antagonistic network dynamics of activation-related changes in high-frequency (> 50 Hz) and low-frequency (< 30 Hz) power. The temporal profile of information flow between the networks estimated by functional connectivity suggests that the activated network inhibits the other one, gating its activity by increasing the amplitude of the low-frequency oscillations. Insights about inter-network communication may have profound implications for various brain disorders in which these dynamics are compromised.

Self-assembly of mixed-linkage glucan hydrogels formed following EG16 digestion

Mixed-linkage glucans are major components of grassy cell-walls and cereal endosperm. Recently identified plant endo-β-glucanase from the EG16 family cleaves MLGs with strong specificity towards regions with at least four sequential β(1,4)-linked glucose residues. This activity yields a low molecular-weight MLG with a repeating structure of β(1,3)-linked cellotriose that gels rapidly at concentrations as low as 1.0 % w/v. To understand the gelation mechanism, we investigated the structure and behavior using rheology, microscopy, X-ray scattering, and molecular dynamics simulations. Upon digestion, the material's rheological behavior changes from typical polymeric material to a fibrillar network behavior seen for e.g. cellulose nanofibrils. Scanning electron microscopy and confocal microscopy verifies these changes in micro- and nanostructure. Small-angle X-ray scattering shows in-solution self-assembly of MLG through ~10 nm elemental structures. Wide-angle X-ray scattering data indicate that the polymer association is similar to cellulose II, with dominant scattering at d-spacing of 0.43 nm. Simulations of two interacting glucan chains show that β(1,3)-linkages prevent the formation of tight helices that form between β(1,4)-linked d-glucan chains, leading to weaker interactions and less ordered inter-chain assembly. Overall, these data indicate that digestion drives gelation not by enhancement of interactions driving self-assembly, but by elimination of unproductive interactions hindering self-assembly.

Performance enhancement of NOMA multi-user networks with aerial reconfigurable intelligent surfaces and aerial relay

In this article, we propose a combination of multiple aerial reconfigurable intelligent surfaces (ARISs) with an aerial relay (AR) for improving the performance of multiple users adopting the non-orthogonal multiple access (NOMA) scheme. In particular, two groups of ARIS are utilized, one for aiding ground-to-air (G2A) communication and another one for aiding air-to-ground (A2G) communication. Mathematical formulas of outage probability (OP), throughput, and achievable capacity (AC) of every user in the proposed ARISs-AR-NOMA network are derived over the Nakagami-m channel. The propagation environment recommended for the fifth generation and beyond (5G/B5G) is applied to enhance the practical behavior of the proposed network. To confirm the performance enhancement, we compare the OP, throughput, and AC of the proposed system with the conventional AR-NOMA network (i.e., without ARISs). Numerical illustrations demonstrate the great benefits of the proposed ARISs-AR-NOMA network, its transmit power is dramatically lower than that of the conventional network. Moreover, the throughput and AC of the proposed network are considerably higher than the conventional network’s. Based on the effect of key parameters and network behaviors, various effective solutions are recommended to enhance the performance of the proposed ARISs-AR-NOMA network.

Pinning synchronization of a complex network: Nodes, edges and higher-order edges

Abstract In recent years, the interdisciplinary study of complex networks has become increasingly important in fields ranging from biology and physics to sociology and mathematics. This paper focuses on pinning control, an approach essential for achieving coordinated behavior in dynamic networks. We explore recent advancements in pinning control strategies, explaining theoretical frameworks and simulation techniques. Additionally, we discuss the significance of certain structures within networks across different orders. Finally, We conclude with a summary of key insights and propose our outlook on future research.

Enhancing explainability of deep learning models for point cloud analysis: a focus on semantic segmentation

ABSTRACT Semantic segmentation of point clouds plays a critical role in various applications, such as urban planning, infrastructure management, environmental analyses and autonomous navigation. Understanding the behaviour of deep neural networks (DNNs) in analysing point cloud data is essential for improving segmentation accuracy and developing effective network architectures and acquisition strategies. In this paper, we investigate the traits of some state-of-the-art neural networks using indoor and urban outdoor point cloud datasets. We compare PointNet, DGCNN, and BAAF-Net on specifically selected datasets, including synthetic and real-world environments. The chosen datasets are S3DIS, SynthCity, Semantic3D, and KITTI. We analyse the impact of different factors such as dataset type (synthetic vs. real), scene type (indoor vs. outdoor), and acquisition system (static vs. mobile sensors). Through detailed analyses and comparisons, we provide insights into the strengths and limitations not only of different network architectures in handling urban point clouds but also of their data structure. This study contributes to going beyond the mere and unconditional use of AI algorithms, trying to explain DNNs behaviour in point cloud analysis and paving the way for future research to enhance segmentation accuracy and develop possible guidelines both for network design and data acquisition in the geomatics field.

Influence mechanism of paper mechanical properties: numerical simulation and experimental verification based on a fiber network

Abstract Paper is a kind of renewable material that exists widely and has important application prospects. However, previous studies have mostly focused on the macromechanical properties of paper but lack micro theory based on paper fiber networks. We present a comprehensive experimental and computational study on the mechanical properties of fibers and fiber networks under the influence of microstructure. A beam-spring model was established based on a beam-fiber network to simulate the behavior of fiber networks. Simulations were performed to demonstrate the influence of fiber microstructural parameters such as fiber bond strength, stiffness, failure strength, size, and network density on mechanical features. Mechanical experiments verified that the fiber bond strength had a greater influence on the paper properties than did the fiber strength. This result is highly consistent with that of the model. All the simulations were validated by experimental measurements. Finally, we provided computational insights into the interfiber bond damage pattern with respect to different fiber microlevels and demonstrated that the proposed beam-spring model can be used to predict the response of fiber networks of paper materials. The above research can be used to optimize the formulation, process, and treatment of paper to meet specific application needs.

The incorporation of stacked long short-term memory into intrusion detection systems for botnet attack classification

&lt;p&gt;&lt;span lang="EN-US"&gt;Botnets are a common cyber-attack method on the internet, causing infrastructure damage, data theft, and malware distribution. The continuous evolution and adaptation to enhanced defense tactics make botnets a strong and difficult threat to combat. In light of this, the study's main objective was to find out how well techniques like principal component analysis (PCA), synthetic minority oversampling technique (SMOTE), and long short-term memory (LSTM) can help find botnet attacks. PCA shows the ability to reduce the feature dimensions in network data, allowing for a more efficient and effective representation of the patterns contained. The SMOTE addresses class imbalances in the dataset, enhancing the model's ability to recognize suspicious activity. Furthermore, LSTM classifies sequential data, understanding complex network patterns and behaviors often used by botnets. The combination of these three methods provided a substantial improvement in detecting suspicious botnet activities. We also evaluated the effectiveness using performance metrics such as accuracy, precision, recall, and F1-score. The results showed an accuracy of 96.77%, precision of 88.95%, recall of 88.58%, and F1-score of 88.64%, indicating that the proposed model was reliable in detecting botnet traffic compared to other deep learning models. Furthermore, LSTM can classify sequential data, understanding complex network patterns and behaviors often used by botnets.&lt;/span&gt;&lt;/p&gt;

Challenges in Understanding the Biological and Pharmacological Responses Based on Emergent Complexity in Biological Systems: From Bone Metabolism to General Physiology

Biological systems are complex, and although researchers strive to understand them, the accumulated knowledge often complicates integrative comprehension. Consolidating this knowledge can provide insights into the landscape of specific biological events. Our study on bone metabolism, focusing on the behavior of the receptor activator of nuclear factor kappa B (RANK) and its ligand (RANKL) highlighted the challenges in understanding its role across different cell types. At the same time, the study underscores the importance of exploring interactions between various players (cell types and genes/proteins) in complex systems, which is a core focus of systems biology. Analysis by mathematical models is a potentially powerful tool for describing the dynamic behavior of components in the interaction networks. However, such model-based analyses are limited by parameter availability and reliability. To address this, we proposed two approaches, i.e., sequential simulation and system-wide behavior constraints. Sequential simulation of small dynamic models offers potential in reproducing behavior in larger networks, as seen in toxicity analysis of sunitinib-related adverse effects. System-wide constraints derived from "homeostasis" help reduce the parameter search space in large-scale models, as demonstrated in model-based analysis of the effects of non-steroidal anti-inflammatory drugs (NSAIDs) on the arachidonic acid pathway. These analytical approaches offer insights into biological system dynamics and can enhance our understanding of pharmacological effects that result from perturbations in complexities of biological systems.

A Framework of Pinning Control for Nonperiodical Stable Behaviors of Large-Scale Asynchronous Boolean Networks

A Framework of Pinning Control for Nonperiodical Stable Behaviors of Large-Scale Asynchronous Boolean Networks

Scholarly Communication and Information Behavior in Chinese Social Networks: A Sentiment Analysis of WeChat Academic Communities

WeChat has penetrated into academic communication circles; however, there is still a significant gap in understanding the way sentiment and information behaviors shape scholarly discourse on this platform. The present research aimedto explore the scholarly communication and information behavior in the Chinese social networks. For this purpose, a sentiment analysis of WeChat academic communities was performed. This study adopted a comprehensive methodology that involved collection of 800 WeChat articles on the basis of engagement metrics, followed by the data pre-processing. It involved cleaning, Chinese words segmentation using Jieba library and TF-IDF vectorization for text analysis. Results of SVM model demonstrated robust performance in sentiment analysis with an overall accuracy of 89% and consistent precision and recall rates across the sentiment categories. Comparison with existing studies also highlighted effectiveness of this model in classifying sentiments on WeChat. Utilization of SVM in sentiment analysis advances the theoretical understanding of text classification techniques in social media environments. The findings also provide valuable insights for researchers and practitioners so that they can leverage SVM for effective classification of SVM. Limitations and future research indications have also been explained in the study

Construction and combustion behavior of horizontal two-dimension combustion networks of boron-metal oxides

In order to break through the top-down combustion mode brought by the traditional pillars, it is explored to explore the exploration of delay composition array construction in two-dimensional dimensions. In this study, B-CuO, B-Bi2O3, B-Fe2O3 sticks and combustion networks with good forming properties were prepared with the help of a micro-pen direct ink writing device by dispersing the above materials in DMF with boron and metal oxides as the main body of the charge and F2602 as the binder. The sticks were thermally ignited using a nichrome wire, and the flame propagation behaviors of the sticks with different formulations, spacings and heights were tracked with a high-speed camera, and a series of combustion networks were designed on the premise of not leaping into flames. Results show that the B-CuO stick has the fastest ignition speed level (19.71–29.02 mm ·s−1) at equivalence ratios of 1.0–4.0, followed by B-Bi2O3 (5.99–16.01 mm ·s−1) and B-Fe2O3 is the slowest (1.91–4.94 mm ·s−1). The sticks burned best at an equivalence ratio of 1.0–1.5. A variety of combustion networks were constructed on 50 × 50 mm glass slides by selecting B-CuO, B-Bi2O3, and B-Fe2O3 at the equivalence ratios of 1.0, 1.5, and 1.5, respectively, among which B-CuO had the shortest combustion time (5.17 s), the shortest total combustion network length (252 mm), and 400 mm network could be realized for B-Bi2O3. Construction and 19.85 s, and B-Fe2O3 can realize 608 mm network length and 130.7 s combustion time. Through these studies, the two-dimensional combustion network construction of boron-metal oxides was realized, which provides a new idea for the delay action in small size.

Dynamical manifold dimensionality as characterization measure of chimera states in bursting neuronal networks

Methods that distinguish dynamical regimes in networks of active elements make it possible to design the dynamics of models of realistic networks. A particularly salient example of such dynamics is partial synchronization, which may play a pivotal role in emergent behaviors of biological neural networks. Such emergent partial synchronization in structurally homogeneous networks is commonly denoted as chimera states. While several methods for detecting chimeras in networks of spiking neurons have been proposed, these are less effective when applied to networks of bursting neurons. In this study, we propose the correlation dimension as a novel approach that can be employed to identify dynamic network states. To assess the viability of this new method, we study networks of intrinsically bursting Hindmarsh–Rose neurons with non-local connections. In comparison to other measures of chimera states, the correlation dimension effectively characterizes chimeras in burst neurons, whether the incoherence arises in spikes or bursts. The generality of dimensionality measures inherent in the correlation dimension renders this approach applicable to a wide range of dynamic systems, thereby facilitating the comparison of simulated and experimental data. This methodology enhances our ability to tune and simulate intricate network processes, ultimately contributing to a deeper understanding of neural dynamics.

Analysis of the dynamical behavior of discrete memristor-coupled scale-free neural networks

The synchronization of neural networks is crucial for neural information processing and represents a key feature of various functional brain diseases. Memristors are ideal electronic components for mimicking biological synapses, among which discrete memristors have the advantage of fast computing speed and are often used in memristor-based neural networks. For these reasons, this paper proposes a novel discrete memristor-coupled Scale-Free neural network (DMSNN). Phase diagrams and time series of membrane potential are employed to analyze the firing pattern coexistence of individual neurons in the network. Furthermore, Spatiotemporal patterns, heat maps of the Spearman correlation coefficient matrix and the values of neuron membrane potential at a particular time point are adopted to declare the spatio-temporal dynamics of the complex neural network, encompassing asynchronization, chimeric state, synchronization and synchronization transition. The study also identifies the phenomenon of topology-induced coexistence and elucidates the underlying reasons for the emergence of chimeric states in the DMSNN as the coupling strength increases. Finally, a hardware implementation platform is constructed using a highly integrated SSD202 processor to validate the accuracy of the DMSNN. The results are consistent with the numerical simulations.

Advancing anomaly detection in cloud environments with cutting‐edge generative AI for expert systems

AbstractAs artificial intelligence (AI) continues to advance, Generative AI emerges as a transformative force, capable of generating novel content and revolutionizing anomaly detection methodologies. This paper presents CloudGEN, a pioneering approach to anomaly detection in cloud environments by leveraging the potential of Generative Adversarial Networks (GANs) and Convolutional Neural Network (CNN). Our research focuses on developing a state‐of‐the‐art Generative AI‐based anomaly detection system, integrating GANs, deep learning techniques, and adversarial training. We explore unsupervised generative modelling, multi‐modal architectures, and transfer learning to enhance expert systems' anomaly detection systems. We illustrate our approach by dissecting anomalies regarding job performance, network behaviour, and resource utilization in cloud computing environments. The experimental results underscore a notable surge in anomaly detection accuracy with significant development of approximately 11%.

Anomalous Network Behaviour Detection in Interoperable Health Systems using Machine Learning in Resource Limited Areas

The connection of devices in distributed environments produces and shares a vast amount of data useful for different organisational decision-making. In healthcare service organisations, for example, multiple e-health systems from different departments or facilities connect and share health data and information. During sharing, proper management is important to ensure the information is secure against intruders. Machine learning as a non-conventional security technique can be used along conventional techniques like firewalls, antivirus and intrusion detection systems to predict future network threats and other anomalies using historical backgrounds and other features. However, some machine learning algorithms have complex computation thus requiring resourceful systems in terms of network bandwidth, CPU power, memory, and storage capacity. In resource-constrained environments, therefore, special consideration is needed to ensure that the analysis of the big data is successful and that the benefits associated with them are effectively obtained. In this paper, a Machine Learning algorithm was selected among four algorithms whose performance was compared through various performance metrics. Classification accuracy, Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Relative Absolute Error (RAE) among other performance metrics were used to compare the ANN, Random forest, Decision trees, and Naïve byes classification algorithms using an extract from CICDDOS2019 dataset. Using the Weka version 3.8.6, the algorithms were compared to choose the best one to classify the data. By using three computers with different resources, the experiments were carried out to determine the performance of those machine learning algorithms. The result revealed that the random forest produced a good average classification performance in resource-limited systems since it surpassed other algorithms in classifying the data at an average of 99 per cent with a low average mean absolute error of 0.0001. Furthermore, as an ensemble that classifies with multiple decision trees algorithm, it likewise uses reasonable time to build and test the model therefore recommended for resource-limited systems.

Stimulation Behavior of Fracture Networks in the Second Hydrate Trial Production Area of China Considering the Presence of Multiple Layers

The fracture network’s stimulation of China’s second hydrate trial production area was investigated. First, the stimulation potential of the fracture network and the influence of well arrangement on hydrate development were explored. Second, the fracture distributions’ influence on development behavior was investigated. Results showed that the fracture network could cause the trial production reservoir to reach the commercial production rate. The average CH4 production rate of unit horizontal well length using the depressurization method and depressurization combined with thermal stimulation (combined method) were 61.3 and 151.5 m3/d with the fracture network and 23.7 and 14.3 m3/d without the fracture network. In addition, without the fracture network, the development behavior of wells arranged in the mixed layer was better than that of wells arranged in the hydrate layer. However, with the fracture network, the result was reversed. With the depressurization method, the best production behavior was obtained by fracturing in the hydrate layer; however, for the combined method, the best production behavior was obtained by fracturing in the hydrate and mixed layer, while fracturing in the free gas layer was useless. This study provides a valuable reference for the hydrate development of China’s trial production reservoir.

Fuzzy Ranking for Enhanced Security: Optimizing Feature Selection in VANET Intrusion Detection

In Vehicular Ad-Hoc Networks (VANETs), ensuring efficient intrusion detection while minimizing false alarms is paramount for maintaining network security and reliability. Traditional methods often struggle to strike this balance effectively. This study proposes a novel approach that employs fuzzy ranking to enhance intrusion detection efficiency in VANETs while mitigating false positive and negative factors. At the core of our approach lies the development of an innovative fuzzy ranking mechanism to identify optimal features extracted from VANET nodes. These features serve as critical indicators of potential security threats within the network. By systematically collecting and analyzing data from various sources including vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, our approach aims to capture nuanced patterns and anomalies indicative of malicious behavior. The fuzzy ranking technique enables us to prioritize features based on their relevance to intrusion detection, thereby facilitating more effective feature selection. This process ensures that only the most informative features are utilized, enhancing the overall efficiency of the intrusion detection system. Features such as packet transmission rates, signal strength variations, route deviation patterns, and network topology characteristics are carefully evaluated and ranked to optimize detection accuracy. To train our intrusion detection model, we leverage machine learning algorithms such as deep neural networks, support vector machines, and decision trees. By utilizing labeled datasets containing examples of both benign and malicious network activities, our model learns to distinguish between normal and anomalous behavior, thus enabling timely threat detection. Furthermore, we integrate fuzzy logic-based anomaly detection techniques to identify previously unseen or zero-day attacks, enhancing the system's robustness. In addition to improving detection accuracy, our approach focuses on preventing harmful warnings caused by false alarms. We achieve this by implementing a comprehensive filtering mechanism that distinguishes between transient anomalies and genuine security threats. Contextual factors such as traffic conditions, environmental variables, and historical network behavior are considered to reduce false positives while ensuring reliable threat detection.Through extensive experimentation and evaluation, we demonstrate the effectiveness of our approach in enhancing intrusion detection efficiency and reducing false positive/negative factors in VANETs. The proposed methodology offers a promising solution to the security challenges faced by VANETs, paving the way for safer and more secure vehicular communication systems in the future.

Elucidating Nonlinear Stress Relaxation in Transient Networks through Two-Dimensional Rheo-Optics.

This study aims to elucidate the origin of nonlinear stress relaxation behaviors in transient networks using a systematically controlled model system consisting of the tetra-armed polyethylene glycols (Tetra-PEG slime) in conjunction with two-dimensional rheo-optics observations. Transient networks, characterized by their temporary cross-links, are extensively utilized in self-healing and robust materials. However, the molecular mechanisms governing their viscoelastic responses to large deformations have remained elusive. This is primarily due to the heterogeneous structures inherent in conventional transient networks and a scarcity of detailed experimental evaluations. By employing Tetra-PEG slime, which is distinguished by its regular structure with uniform strand lengths and functionalities, and the polarization imaging method, we overcome these obstacles. Our results reveal that the damping phenomena observed under large step strains arise from spatially heterogeneous relaxation, predominantly driven by network strand pullout. These insights lay a solid foundation for understanding the intricate rheological properties of transient networks.