Static Networks Research Articles

Static and dynamic brain functional connectivity patterns in patients with unilateral moderate-to-severe asymptomatic carotid stenosis

Background and purposeAsymptomatic carotid stenosis (ACS) is an independent risk factor for ischemic stroke and vascular cognitive impairment, affecting cognitive function across multiple domains. This study aimed to explore differences in static and dynamic intrinsic functional connectivity and temporal dynamics between patients with ACS and those without carotid stenosis.MethodsWe recruited 30 patients with unilateral moderate-to-severe (stenosis ≥ 50%) ACS and 30 demographically-matched healthy controls. All participants underwent neuropsychological testing and 3.0T brain MRI scans. Resting-state functional MRI (rs-fMRI) was used to calculate both static and dynamic functional connectivity. Dynamic independent component analysis (dICA) was employed to extract independent circuits/networks and to detect time-frequency modulation at the circuit level. Further imaging-behavior associations identified static and dynamic functional connectivity patterns that reflect cognitive decline.ResultsACS patients showed altered functional connectivity in multiple brain regions and networks compared to controls. Increased connectivity was observed in the inferior parietal lobule, frontal lobe, and temporal lobe. dICA further revealed changes in the temporal frequency of connectivity in the salience network. Significant differences in the temporal variability of connectivity were found in the fronto-parietal network, dorsal attention network, sensory-motor network, language network, and visual network. The temporal parameters of these brain networks were also related to overall cognition and memory.ConclusionsThese results suggest that ACS involves not only changes in the static large-scale brain network connectivity but also dynamic temporal variations, which parallel overall cognition and memory recall.

Altered Static and Dynamic Functional Network Connectivity and Combined Machine Learning in Stroke.

Stroke is a condition characterized by damage to the cerebral vasculature from various causes, resulting in focal or widespread brain tissue damage. Prior neuroimaging research has demonstrated that individuals with stroke present structural and functional brain abnormalities, evident through disruptions in motor, cognitive, and other vital functions. Nevertheless, there is a lack of studies on alterations in static and dynamic functional network connectivity in the brains of stroke patients. Fifty stroke patients and 50 healthy controls (HCs) underwent resting-state functional magnetic resonance imaging (rs-fMRI) scanning. Initially, the independent component analysis (ICA) method was utilized to extract the resting-state network (RSN). Subsequently, the disparities in static functional network connectivity both within and between networks among the two groups were computed and juxtaposed. Following this, five consistent and robust dynamic functional network connectivity (dFNC) states were derived by integrating the sliding time window method with k-means cluster analysis, and the distinctions in dFNC between the groups across different states, along with the intergroup variations in three dynamic temporal metrics, were assessed. Finally, a support vector machine (SVM) approach was employed to discriminate stroke patients from HCs using FC and FNC as classification features. Comparing the stroke group to the healthy control (HC) group, the stroke group exhibited reduced intra-network functional connectivity (FC) in the right superior temporal gyrus of the ventral attention network (VAN), the left calcarine of the visual network (VN), and the left precuneus of the default mode network (DMN). Regarding static functional network connectivity (FNC), we identified increased connectivity between the executive control network (ECN) and dorsal attention network (DAN), salience network (SN) and DMN, SN-ECN, and VN-ECN, along with decreased connectivity between DAN-DAN, ECN-SN, SN-SN, and DAN-VN between the two groups. Noteworthy differences in dynamic FNC (dFNC) were observed between the groups in states 3 to 5. Moreover, stroke patients demonstrated a significantly higher proportion of time and longer mean dwell time in state 4, alongside a decreased proportion of time in state 5 compared to HC. Finally, utilizing FC and FNC as features, stroke patients could be distinguished from HC with an accuracy exceeding 70% and an area under the curve ranging from 0.8284 to 0.9364. In conclusion, our study reveals static and dynamic changes in large-scale brain networks in stroke patients, potentially linked to abnormalities in visual, cognitive, and motor functions. This investigation offers valuable insights into the neural mechanisms underpinning the functional deficits observed in stroke, thereby aiding in the diagnosis and development of targeted therapeutic interventions for affected individuals.

The effects of neurofeedback training on behavior and brain functional networks in children with autism spectrum disorder.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder with an unclear pathogenesis to date. Neurofeedback (NFB) had shown therapeutic effects in patients with ASD. In this study,we analyzed the brain functional networks of children with ASD and investigated the impact of NFB targeting the beta rhythm training on these networks. The Autism Behavior Checklist (ABC) and Social Response Scale (SRS) were employed to evaluate the effects of NFB training on the behavioral abilities of children with ASD. We compared the differences in static and dynamic brain functional networks between ASD and Typically Developing (TD) children, also explored the changes in these networks in ASD children after 20 sessions of NFB training. The Weighted Phase Lag Index (wPLI) was used to construct static functional networks, and the Fuzzy Entropy (FuzzyEn) algorithm was further employed to measure the complexity of static functional connectivity and construct dynamic functional networks. This allowed the analysis of functional connectivity and fluctuations in the static functional networks of ASD and TD children, as well as the time variability of the dynamic functional networks. Additionally, the study explored the changes in brain functional networks and behavioral scales before and after NFB training. Results from behavioral scales indicated significant improvements in cognitive, communication, language, and social scores in ASD children following NFB intervention. EEG analysis revealed that static functional connectivity was lower, connectivity variability was higher, and temporal variability was greater in ASD children compared to TD children. Following NFB training, increased functional connectivity, reduced connectivity variability in the Delta frequency band, and decreased temporal variability were observed in ASD children. The results revealed abnormalities in both static and dynamic functional networks in children with ASD, with NFB training showed potential to modulate these networks. While our results showed that NFB training can assist participants in regulating connectivity and temporal variability in specific brain regions, robust evidence for its effectiveness in alleviating core symptoms of ASD remained limited.

An Enhanced Q-Learning MAC Protocol for Energy Efficiency and Convergence in Underwater Sensor Networks

Under-Water Sensor Networks (UWSNs) are important for applications like oceanographic data collection, environmental protection, monitoring, and disaster response. These UWSN networks face challenges in energy efficiency and protocol convergence because of dynamic underwater ecosystems limited in number and low channel capacity. This paper covers and finds real-world solutions for these challenges by proposing an adaptive Q-Learning Medium Access Control (MAC) protocol for UWSNs. The methodology used in this paper involves utilizing Q-Learning, a reinforcement learning technique, to make sensor nodes autonomously refine their transmission strategies in real-time, enhancing energy consumption and improving protocol convergence. The protocol was implemented using the NS 3 network simulator, which offers a detailed and real-world environment for analyzing the protocol’s performance. Extensive simulations were conducted, and experiments were used to analyze the performance of the proposed protocol. The results showcase significant improvements over other and traditional MAC protocols, with a 13% to 19% increase in energy efficiency and channel utilization enhancement in static and mobile network scenarios. The adaptive Q-Learning MAC protocol provides a robust solution for the challenges of UWSNs, offering energy efficiency and faster convergence times. This research significantly contributes to the advancement of efficient and adaptive underwater communication protocols, paving the way for future development in the field

SEISMICITY of the ALTAI and SAYAN region in 2020

An analysis of the seismicity of the Altai-Sayan region in 2020 is given. The stationary seismic network in the reporting period consisted of 53 registration points, work was carried out on the modernization and retrofitting of recording equipment. In addition to the stationary network, four local time networks functioned in the region: one investigated the seismic process and accumulated data for constructing the Moho boundary using the method of receiving functions in the Altai Republic, the other three recorded man-made seismic processes in the area of coal mining enterprises in Kuzbass and the Novosibirsk region. In 2020 8438 earthquakes have been registered in the region, 4655 of them occurred in the Chui-Kurai zone of the Altai Mountains, for which a separate overview of seismicity is presented. The total seismic energy released in the earthquake foci of the Altai-Sayan region in 2020 amounted to 2.191012 J, which is a record low value in the entire history of instrumental observa tions. The slope of the linear part of the earthquake recurrence graph for 2020 has not undergone significant changes compared to the previous reporting period. The strongest in 2020 earthquake in the region (ML=5.3) occurred on October 22 at 13h38m on the southern slope of the Eastern Sayan ridge.

Systematic Risks of the Global Lithium Supply Chain Network: From Static Topological Structures to Cascading Failure Dynamics.

Recent years have seen increasing concerns on supply chain risks of lithium, a critical material for achieving e-mobility transition and climate ambitions. These risks propagate both along the life cycle and across national boundaries in a multilayer network. However, most previous studies are either only based on static network measures or focused on individual layers, ignoring dynamic cascading risks and interconnected and interdependent relationships along life cycle stages and across economies. Here, we integrated trade-linked material flow and complex network analyses to investigate intricate interconnections, interdependencies, and systematic risks of the global lithium supply chain. Both static and dynamic measures of the global lithium supply chain network exhibit a "robust-yet-fragile" property: robust for random shocks yet fragile for targeted shocks and robust for small or local disruptions yet fragile for large or cascading failures. Portugal, Brazil, Singapore, Canada, Finland, Norway, South Africa, Israel, Hungary, and the United Arab Emirates are most likely to be affected by supply disruptions. A hypothetical USA-China trade decoupling will increase the severity and susceptibility of network-wide failures by around 5%. Our results call for global collaborations and collective efforts to balance efficiency and security and avoid a "zero-sum game" in securing the lithium supply chain.

Evaluation of In Situ FAPAR Measurement Protocols Using 3D Radiative Transfer Simulations

The fraction of absorbed photosynthetically active radiation (FAPAR) is one of the bio-geophysical Essential Climate Variables assessed through remote sensing observations and distributed globally by space and environmental agencies. Any reliable remote sensing product should be benchmarked against a reference, which is normally determined by means of ground-based measurements. They should generally be aggregated spatially to be compared with remote sensing products at different resolutions. In this work, the effectiveness of various in situ sampling methods proposed to assess FAPAR from flux measurements was evaluated using a three-dimensional radiative transfer framework over eight virtual vegetated landscapes, including dense forests (leaf-on and leaf-off models), open canopies, sparse vegetation, and agricultural fields with a nominal extension of 1 hectare. The reference FAPAR value was determined by summing the absorbed PAR-equivalent photons by either all canopy components, both branches and leaves, or by only the leaves. The incoming and upwelling PAR fluxes were simulated in different illumination conditions and at a high spatial resolution (50 cm). They served to replicate in situ virtual FAPAR measurements, which were carried out using either stationary sensor networks or transects. The focus was on examining the inherent advantages and drawbacks of in situ measurement protocols against GCOS requirements. Consequently, the proficiency of each sampling technique in reflecting the distribution of incident and reflected PAR fluxes—essential for calculating FAPAR—was assessed. This study aims to support activities related to the validation of remote sensing FAPAR products by assessing the potential uncertainty associated with in situ determination of the reference values. Among the sampling schemes considered in our work, the cross shaped sampling schemes showed a particular efficiency in properly representing the pixel scale FAPAR over most of the scenario considered.

Cyclic-coverings of Line graph L(Sn) and its Disjoint Union

A graph is a particular representation of a static network and labeling of a graph can be think of as automatic routing of data in a network topology. A visual representation of data in the form of a graph help us to look deep insight. The data science (e.g., Python Package: a computer programming language) uses graph theory concepts to study and analyze the networks. Analyzing these network is equivalent of finding a set of edges E′ for a graph G such that every vertex of G is incident with at least one edge in E′. Then E′ is called an edge-covering of G. A spanning tree of a connected graph is an example of edge-covering. A finite simple graph G is an (ad, d)-H-antimagic if the following three conditions are satisfied: G has an H-covering (H a subgraph of G), there exists a bijection α : V (G) ∪ E(G) → {1, 2, 3, . . . , |V (G) ∪ E(G)|} and the H-weights constitute an arithmetic progression with common difference d. The above said labeling is called super if α(V ) = {1, 2, . . . , |V |}. In this research article, we focused on studying super C3-antimagic labeling of line graph of a sun-let graph for several differences and C3-supermagic labeling of its disjoint union.

Temporal contact patterns and the implications for predicting superspreaders and planning of targeted outbreak control.

Directly transmitted infectious diseases spread through social contacts that change over time, but outbreak models typically make simplifying assumptions about network structure and dynamics. To assess how common assumptions relate to real-world interactions, we analysed 11 networks from five settings and developed metrics, capturing crucial epidemiological features of these networks. We developed a novel metric, the 'retention index', to characterize the distribution of retained contacts over consecutive time steps relative to fully static and dynamic networks. In workplaces and schools, contacts in the same department formed most of the retained contacts. In contrast, no clear contact type dominated the retained contacts in hospitals, thus reducing overall risk of disease introduction would be more effective than control targeted at departments. We estimated the contacts repetition over multiple days and showed that simple resource planning models overestimate the number of unique contacts by 20%-70%. We distinguished the difference between 'superspreader' and infectious individuals driving 'superspreading events' by measuring how often the individual represents the top 80% of contacts in the time steps over the study duration. We showed an inherent difficulty in identifying 'superspreaders' reliably: less than 20% of the individuals in most settings were highly connected for multiple time steps.

Dynamic and static functional network connectivity distinguish symptomatic and non‐symptomatic individuals with CADASIL

Dynamic and static functional network connectivity distinguish symptomatic and non‐symptomatic individuals with CADASIL

Brain networks and intelligence: A graph neural network based approach to resting state fMRI data.

Resting-state functional magnetic resonance imaging (rsfMRI) is a powerful tool for investigating the relationship between brain function and cognitive processes as it allows for the functional organization of the brain to be captured without relying on a specific task or stimuli. In this paper, we present a novel modeling architecture called BrainRGIN for predicting intelligence (fluid, crystallized and total intelligence) using graph neural networks on rsfMRI derived static functional network connectivity matrices. Extending from the existing graph convolution networks, our approach incorporates a clustering-based embedding and graph isomorphism network in the graph convolutional layer to reflect the nature of the brain sub-network organization and efficient network expression, in combination with TopK pooling and attention-based readout functions. We evaluated our proposed architecture on a large dataset, specifically the Adolescent Brain Cognitive Development Dataset, and demonstrated its effectiveness in predicting individual differences in intelligence. Our model achieved lower mean squared errors and higher correlation scores than existing relevant graph architectures and other traditional machine learning models for all of the intelligence prediction tasks. The middle frontal gyrus exhibited a significant contribution to both fluid and crystallized intelligence, suggesting their pivotal role in these cognitive processes. Total composite scores identified a diverse set of brain regions to be relevant which underscores the complex nature of total intelligence. Our GitHub implementation is publicly available on https://github.com/bishalth01/BrainRGIN/.

A local and global feature fusion network for Super-Resolution reconstruction of turbulent flows

The resolution of flow fields represents a significant factor influencing the accuracy of turbulent flow analysis. Nevertheless, the acquisition of high-resolution turbulence data remains a challenge due to the limitations imposed by computing resources. The interpolation method, while capable of achieving high-resolution turbulence at low cost, faces challenges in capturing details of turbulent flows. In this study, a local and global feature fusion network (LGFN) is designed for the reconstruction of high-resolution turbulent flows with high quality. First, dual parallel branches made of dense blocks are introduced into the LGFN to extract local features of turbulent flows. Moreover, the features after the first dense block of each branch are summarized into the self-attention block to obtain global features. The extracted local and global features are aggregated through learnable weight parameters to achieve feature fusion. Finally, the fused turbulence features are scaled to the same dimensional size as the high-resolution turbulence through the implementation of multilayer pixel shuffle layers and convolution layers. The effectiveness of the proposed network was evaluated using datasets of forced isotropic turbulence and channel turbulence. The results demonstrate that the reconstructed velocity fields of the LGFN exhibit the highest degree of similarity to the direct numerical simulation results, in comparison with bicubic interpolation, static convolutional neural network, and super-resolution dense connection network results. In addition, compared to alternative methods, the proposed network effectively captures the characteristics of isotropic or anisotropic turbulence even at larger scales.

Revisiting the Trade-Off Between Accuracy and Robustness via Weight Distribution of Filters.

Adversarial attacks have been proven to be potential threats to Deep Neural Networks (DNNs), and many methods are proposed to defend against adversarial attacks. However, while enhancing the robustness, the accuracy for clean examples will decline to a certain extent, implying a trade-off existed between the accuracy and adversarial robustness. In this paper, to meet the trade-off problem, we theoretically explore the underlying reason for the difference of the filters' weight distribution between standard-trained and robust-trained models and then argue that this is an intrinsic property for static neural networks, thus they are difficult to fundamentally improve the accuracy and adversarial robustness at the same time. Based on this analysis, we propose a sample-wise dynamic network architecture named Adversarial Weight-Varied Network (AW-Net), which focuses on dealing with clean and adversarial examples with a "divide and rule" weight strategy. The AW-Net adaptively adjusts the network's weights based on regulation signals generated by an adversarial router, which is directly influenced by the input sample. Benefiting from the dynamic network architecture, clean and adversarial examples can be processed with different network weights, which provides the potential to enhance both accuracy and adversarial robustness. A series of experiments demonstrate that our AW-Net is architecture-friendly to handle both clean and adversarial examples and can achieve better trade-off performance than state-of-the-art robust models.

Value-behavior inconsistency is robust to promote cooperative behavior in structured populations.

The evolution of cooperation is a theme commonly studied in biology, psychology, sociology, and economics. Mechanisms that promote cooperative behavior in structured populations have been intensively studied. However, individuals' values, specifically, their opinions have been rarely taken into account so far. Inspired by cognition dissonance theory, we assume that individuals pay the cost of guiltiness if the behavior is defection but the opinion deviates from defection, and pay the cost of regret if the behavior is cooperation but the opinion deviates from cooperation. For all general stochastic evolutionary dynamics on arbitrary static networks with multiple opinions, we prove in the weak selection limit that: (i) value-behavior inconsistency cost promotes cooperative behavior if and only if the average cost of regret is less than that of guiltiness; (ii) individuals with value-behavior consistency are more abundant than that with value-behavior inconsistency. This is in contrast with other mechanisms that are at work for cooperation for one population structure but not others. Furthermore, it is also validated on an empirical network and for non-weak selection intensity. The value-behavior inconsistency is thus a robust mechanism to promote cooperative behavior in structured populations. Our results shed light on the importance of the co-evolutionary dynamics of opinion and behavior, which opens an avenue for cooperation.

Quantifying and predicting evolutionary networks

The process of network evolution is typically characterized by the emergence of topological structures and the intricate interplay between determinism and stochasticity. Our work introduces a novel structural analysis framework to observe fluctuations in the network evolution process, profoundly influenced by the underlying network generation mechanisms. Based on theoretical reasoning and empirical examination, utilizing synthetic, static, and temporal networks, we arrive at two principal conclusions. Firstly, we reveal that the degree and distance distributions of networks exhibit two dynamic phenomena, convergence and divergence on high- and low-frequency curves, depend on the reflections of the network generation mechanism at both topological extreme values and the overall distribution. Secondly, we develop a novel link prediction method based on the convergence of topological fluctuations, which significantly outperforms benchmark algorithms in both real static and temporal networks. These findings are of considerable significance to the study of real network evolution mechanisms, contributing to a more comprehensive understanding of network behavior over time.

Multimodal predictive modeling: Scalable imaging informed approaches to predict future brain health

Background:Predicting future brain health is a complex endeavor that often requires integrating diverse data sources. The neural patterns and interactions identified through neuroimaging serve as the fundamental basis and early indicators that precede the manifestation of observable behaviors or psychological states. New Method:In this work, we introduce a multimodal predictive modeling approach that leverages an imaging-informed methodology to gain insights into future behavioral outcomes. We employed three methodologies for evaluation: an assessment-only approach using support vector regression (SVR), a neuroimaging-only approach using random forest (RF), and an image-assisted method integrating the static functional network connectivity (sFNC) matrix from resting-state functional magnetic resonance imaging (rs-fMRI) alongside assessments. The image-assisted approach utilized a partially conditional variational autoencoder (PCVAE) to predict brain health constructs in future visits from the behavioral data alone. Results:Our performance evaluation indicates that the image-assisted method excels in handling conditional information to predict brain health constructs in subsequent visits and their longitudinal changes. These results suggest that during the training stage, the PCVAE model effectively captures relevant information from neuroimaging data, thereby potentially improving accuracy in making future predictions using only assessment data. Comparison with Existing Methods:The proposed image-assisted method outperforms traditional assessment-only and neuroimaging-only approaches by effectively integrating neuroimaging data with assessment factors. Conclusion:This study underscores the potential of neuroimaging-informed predictive modeling to advance our comprehension of the complex relationships between cognitive performance and neural connectivity.

Multi-source biological knowledge-guided hypergraph spatiotemporal subnetwork embedding for protein complex identification

Abstract Identifying biologically significant protein complexes from protein–protein interaction (PPI) networks and understanding their roles are essential for elucidating protein functions, life processes, and disease mechanisms. Current methods typically rely on static PPI networks and model PPI data as pairwise relationships, which presents several limitations. Firstly, static PPI networks do not adequately represent the scopes and temporal dynamics of protein interactions. Secondly, a large amount of available biological resources have not been fully integrated. Moreover, PPIs in biological systems are not merely one-to-one relationships but involve higher order non-pairwise interactions. To alleviate these issues, we propose HGST, a multi-source biological knowledge-guided hypergraph spatiotemporal subnetwork (subnet) embedding method for identifying biologically significant protein complexes from PPI networks. HGST initially constructs spatiotemporal PPI subnets using the scopes and temporal dynamics of proteins derived from multi-source biological knowledge, treating them as dynamic networks through fine-grained spatiotemporal partitioning. The spatiotemporal subnets are then transformed into hypergraphs, which model higher order non-pairwise relationships via hypergraph embedding. Simultaneously, fine-grained amino acid sequence features and coarse-grained gene ontology attributes are introduced for multi-dimensional feature fusion. Finally, protein complexes are identified from the reweighted subnets based on fused feature representations using the core-attachment strategy. Evaluations on four real PPI datasets demonstrate that HGST achieves competitive performance. Furthermore, a series of biological analyses confirm the high biological significance of the complexes identified by HGST. The source code is available at https://github.com/qifen37/HGST.

Measuring the connectedness of the Nigerian banking network and its implications for systemic risk

This study examines fifteen major banks’ network connectedness in the Nigerian banking system via its stock returns. The paper studies both the static and dynamic network connectedness of banks built on the generalized forecast error variance decomposition, using daily data from January 4, 2005, to June 28, 2019, of publicly traded banks. This study finds a substantial total connectedness, with a high pairwise connectedness among the system’s large banks. The dynamic evolution of connectedness in the network reveals that banks’ connectivity increases in response to certain economic episodes. The evolution of the global network's topological properties reveals that it is mainly susceptible to shocks threatening its stability. Additionally, the study computes a composite index of systemic importance for the Nigerian banking system by combining several network centrality metrics using the principal component analysis. The outcome shows that large banks are more centralized in the network, and the larger the scale of assets a bank has, the more systemically relevant the bank is in the network. Since systemic risk emanates from connectedness, frequent assessment of the banking system's connectedness and systemic importance will aid policy decisions. The proposed measure of systemic importance can be incorporated into the CBN’s stress testing mechanism for fast-tracking risk potential banks.

Maximum entropy in dynamic complex networks.

The field of complex networks studies a wide variety of interacting systems by representing them as networks. To understand their properties and mutual relations, the randomization of network connections is a commonly used tool. However, information-theoretic randomization methods with well-established foundations mostly provide a stationary description of these systems, while stochastic randomization methods that account for their dynamic nature lack such general foundations and require extensive repetition of the stochastic process to measure statistical properties. In this work, we extend the applicability of information-theoretic methods beyond stationary network models. By using the information-theoretic principle of maximum caliber we construct dynamic network ensemble distributions based on constraints representing statistical properties with known values throughout the evolution. We focus on the particular cases of dynamics constrained by the average number of connections of the whole network and each node, comparing each evolution to simulations of stochastic randomization that obey the same constraints. We find that ensemble distributions estimated from simulations match those calculated with maximum caliber and that the equilibrium distributions to which they converge agree with known results of maximum entropy given the same constraints. Finally, we discuss further the connections to other maximum entropy approaches to network dynamics and conclude by proposing some possible avenues of future research.

A 3D Coverage Method Involving Dynamic Underwater Wireless Sensor Networks for Marine Ranching Monitoring

In view of the poor adaptability and uneven coverage of static underwater wireless sensor networks (UWSNs) to environmental changes and the need for dynamic monitoring, a three-dimensional coverage method involving a dynamic UWSNs for marine ranching, based on an improved sparrow search algorithm (ISSA), is proposed. Firstly, the reverse learning strategy was introduced to generate the reverse sparrow individuals and fuse with the initial population, and the individual sparrows with high fitness were selected to improve the search range. Secondly, Levy flight was introduced to optimize the location update of the producer, which effectively expanded the local search capability of the algorithm. Finally, the Cauchy mutation perturbation mechanism was introduced into the scrounger location to update the optimal solution, which enhanced the ability of the algorithm to obtain the global optimal solution. When deploying UWSNs nodes, an autonomous underwater vehicle (AUV) was used as a mobile node to assist the deployment. In the case of underwater obstacles, the coverage hole in the UWSNs was covered by an AUV at specific times. The experimental results show that compared with other algorithms, the ISSA has a shorter mobile path and achieves a higher coverage rate, with lower node energy consumption.