Method for synchronizing data of relay node configuration in ad hoc network based on dynamic topology

Abstract

The application discloses a method for synchronizing configuration data of a relay node in a dynamic topology-based ad hoc network, comprising the following steps: obtaining target area structure information and node coordinates of the relay node, and classifying disaster information to obtain first disaster information; obtaining a historical environmental parameter sequence, processing the disaster information, and obtaining second disaster information; combining the first disaster information to obtain fused disaster information; iteratively optimizing a communication topology network according to the fused disaster information, and predicting switching information loss of the relay node and disaster blocking of the node coordinates according to the historical environmental parameter sequence, and optimizing; converging and optimizing to obtain an optimal communication topology network, and performing configuration and data synchronization communication. The application solves the technical problem that the existing relay node configuration technology in the ad hoc network cannot simultaneously consider spatial static disaster risks and time dynamic environmental changes for systematic iterative optimization, resulting in insufficient network stability in a disaster environment.

Description

Technical Field

[0001] This invention relates to the field of wireless ad hoc network communication technology, and more specifically to a method for synchronizing configuration data of ad hoc network relay nodes based on dynamic topology. Background Technology

[0002] Wireless ad hoc networks offer advantages such as flexible deployment and strong resilience, making them widely used in scenarios like urban fire emergency communications. The topology of relay nodes directly impacts the coverage and reliability of the communication network. In urban fire emergency communications, relay nodes deployed in densely populated high-rise building areas or hazardous material storage areas are far more likely to fail due to fire spread or building collapses than nodes deployed in open areas. Failure to relay nodes will trigger network reconstruction, resulting in data retransmission and command loss.

[0003] In existing ad hoc network relay node configuration technologies, the configuration process typically only focuses on communication performance parameters such as signal strength and transmission bandwidth; disaster risk assessments often rely on static spatial geographic information for judgment, but the fire environment is constantly changing, and static information is difficult to reflect the ever-changing fire situation; at the same time, existing technologies consider different indicators such as communication coverage, node failure risk, and transmission efficiency independently, without forming a unified optimization scheme. Summary of the Invention

[0004] This invention provides a method for synchronizing configuration data of relay nodes in ad hoc networks based on dynamic topology. This invention solves the technical problem that existing ad hoc network relay node configuration technologies struggle to systematically iteratively optimize both static spatial disaster risks and dynamic temporal environmental changes, leading to insufficient network stability under disaster environments.

[0005] In view of the above problems, the present invention provides a method for synchronizing configuration data of relay nodes in ad hoc networks based on dynamic topology, the method comprising: The system acquires regional structure information within the target area and obtains the coordinates of multiple relay nodes deployed within the target area. It then classifies the amount of disaster information to obtain multiple first disaster information quantities. Among these, the multiple relay nodes are used to construct a communication topology network. Multiple historical environmental parameter sequences obtained from monitoring by multiple relay nodes within a preset time range are acquired, and disaster information is processed to obtain multiple second disaster information quantities. These are then combined with the multiple first disaster information quantities to obtain multiple fused disaster information quantities. Based on the multiple fused disaster information quantities, the communication topology network is iteratively optimized, including the prediction of switching information loss of multiple relay nodes and the prediction of disaster isolation of multiple node coordinates based on multiple historical environmental parameter sequences. The convergence optimization yields the optimal communication topology network, which is then configured and used for data synchronization communication.

[0006] One or more technical solutions provided in this invention have at least the following technical effects or advantages: By acquiring regional structural information and relay node physical coordinates and calculating relative coordinates, and combining this with a coordinate disaster prediction agent to classify disaster information, the spatial accuracy of disaster assessment is improved. By acquiring historical environmental parameter sequences, extracting peak values ​​and variation coefficients, and calculating secondary disaster information, disaster assessment is extended from a spatial static dimension to a temporal dynamic dimension, uncovering disaster precursor information from historical monitoring data. By fusing spatial static risk and temporal dynamic risk to obtain fused disaster information, spatiotemporal integration of disaster assessment is achieved, significantly improving the comprehensiveness, accuracy, and dynamic adaptability of the assessment. By using coverage thresholds as hard constraints to screen candidate topologies, and utilizing multi-intelligent branch voting to predict blocking rates, coverage ratio, blocking rate, and information loss are weighted and fused into a topology score, a comprehensive quantitative assessment of coverage capability, disaster resistance reliability, and information transmission efficiency is achieved. Through multiple iterations, the optimal configuration is selected, enhancing the system's communication service quality and survivability in complex disaster environments. By setting clear convergence conditions and providing a complete connection from optimal solution output to node activation, link configuration, and communication synchronization, a complete closed loop from theoretical optimization to actual network deployment is achieved, reducing the risk of data loss and communication interruption due to network switching.

[0007] In summary, this invention solves the technical problem that existing ad hoc network relay node configuration technologies are unable to systematically iteratively optimize both static spatial disaster risks and dynamic temporal environmental changes, resulting in insufficient network stability under disaster environments. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a flowchart illustrating the method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram illustrating the flow of spatial static and temporal dynamic disaster information fusion in the self-organizing network relay node configuration data synchronization method based on dynamic topology provided in this embodiment of the invention. Detailed Implementation

[0010] This invention provides a method for synchronizing configuration data of self-organizing network relay nodes based on dynamic topology, which specifically solves the technical problem that existing self-organizing network relay node configuration technologies are unable to systematically iteratively optimize both spatial static disaster risks and temporal dynamic environmental changes, resulting in insufficient network stability under disaster environments.

[0011] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0012] Examples, such as Figure 1 and Figure 2 As shown, this invention provides a method for synchronizing configuration data of relay nodes in an ad hoc network based on dynamic topology. The method includes: S100: Obtain regional structure information within the target area, and obtain the coordinates of multiple relay nodes deployed within the target area. Classify the disaster information to obtain multiple first disaster information quantities. Among them, multiple relay nodes are used to construct a communication topology network.

[0013] Current ad hoc network relay node configuration technologies typically focus only on node communication performance parameters and basic network connectivity, neglecting the geographical differences and potential disaster risks associated with each relay node's location. This results in a lack of refined spatial location support for disaster assessment, making it difficult to quantify risks differently for different nodes. Furthermore, traditional disaster information assessment methods often rely on manual experience or complex physical model calculations. When there are numerous relay nodes in the target area, point-by-point assessment is not only computationally intensive and time-consuming, but also yields subjective results lacking consistency and reproducibility, failing to meet the practical needs of rapid response and dynamic optimization in emergency communication scenarios. Therefore, how to combine regional structural information and the spatial location information of each relay node to achieve rapid, accurate, and standardized classification and assessment of disaster information has become an urgent technical problem to be solved.

[0014] Step S100 in the method provided in this embodiment of the invention includes: Obtain the regional structure information within the target area, including the regional type and area. The physical coordinates of multiple relay nodes deployed in the target area are obtained, and the relative coordinates of multiple nodes with respect to the center coordinates of the target area are calculated as multiple node coordinates. The physical coordinates of multiple nodes are obtained through the positioning information of multiple relay nodes. Based on the regional structure information and multiple node coordinates, disaster information is classified to obtain multiple first disaster information quantities.

[0015] In this embodiment of the invention, the target area refers to the geographical area where communication topology network deployment and disaster monitoring are required. The regional structure information includes the region type and the region area. The region type includes low-rise old urban areas, high-rise business districts, large complexes, chemical industrial parks, etc.; the region area is used to characterize the size of the target area and can be calculated by obtaining the boundary coordinates of the target area through a Geographic Information System (GIS).

[0016] In this embodiment of the invention, multiple relay nodes are pre-deployed at different locations within the target area to construct a communication topology network and realize data relay transmission. Each relay node is equipped with a positioning module, such as a Global Positioning System (GPS) module, a BeiDou positioning module, or other satellite navigation positioning modules. Through the positioning information of each relay node, the physical coordinates of each relay node can be obtained, i.e., its absolute position information in a geographic coordinate system, denoted as latitude and longitude coordinates or Gauss-Kruger projection coordinates.

[0017] Furthermore, after obtaining the physical coordinates of each relay node, the center coordinates of the target area are first determined. The center coordinates of the target area can be calculated based on the area boundary coordinates in the area structure information, for example, by taking the geometric center of the area boundary coordinates. Then, using the center coordinates of the target area as the reference origin, the relative offset of the physical coordinates of each relay node relative to these center coordinates is calculated, resulting in multiple relative node coordinates. These relative node coordinates directly reflect the relative spatial position of each relay node within the target area, facilitating the subsequent classification and processing of disaster information. The calculated multiple relative node coordinates are then used as the node coordinates for subsequent steps.

[0018] In this embodiment of the invention, the amount of disaster information is used to characterize the degree to which each relay node is affected by a disaster. Specifically, the target area is first divided into multiple grid areas according to a preset grid size. The grid size can be flexibly set according to the size of the target area and the density of node deployment. For example, the target area can be evenly divided into several square grid units of 100 meters × 100 meters, and each grid area has a unique grid number.

[0019] Next, the grid affiliation of each relay node is determined. For each relay node, based on its coordinates within the target area, the node is assigned to the grid area where its coordinates fall. Specifically, the offsets of the node's x and y coordinates relative to the starting coordinates of the target area are calculated. The x-coordinate offset is divided by the grid width and rounded down to obtain the horizontal grid number, and the y-coordinate offset is divided by the grid height and rounded down to obtain the vertical grid number. The horizontal and vertical grid numbers together determine the unique grid area to which the node belongs. If multiple relay nodes are located within the same grid area, these nodes share the disaster frequency statistics for that grid area; if no relay nodes are deployed in a certain grid area, that grid area does not participate in the subsequent calculation of the disaster information coefficient.

[0020] Furthermore, for each grid area within the target area, the number of times a disaster occurred in its historical disaster record data is counted, which is taken as the disaster occurrence frequency for that grid area. Simultaneously, the arithmetic mean of the disaster occurrence frequencies for all grid areas containing relay nodes is calculated, which is taken as the average disaster occurrence frequency. Then, the disaster occurrence frequency for each grid area is divided by this average disaster occurrence frequency to obtain the disaster information content coefficient for that grid area. All relay nodes located within the same grid area use the same disaster information content coefficient. A ratio greater than 1 indicates that the disaster occurrence frequency of that grid area is higher than the overall average, while a ratio less than 1 indicates that it is lower than the average.

[0021] The disaster information coefficient of the grid area to which each relay node belongs is used as the first disaster information of the sample for that node. If the ratio calculation result has extreme values, further smoothing can be performed by truncation or logarithmic transformation to avoid individual extreme values ​​from having an excessive impact on subsequent fusion calculations.

[0022] Based on the aforementioned regional structure information and multiple node coordinates, disaster information is classified, including: Awaken the coordinate disaster prediction intelligent agent; The regional structure information is combined with the coordinates of multiple nodes and input into the coordinate disaster prediction agent. Multiple first disaster information quantities are then classified and output. The training and testing data of the coordinate disaster prediction agent includes a set of sample area structure information extracted from historical disaster record data, a set of sample node coordinates, and a set of labeled sample first disaster information. Each sample first disaster information includes a disaster information coefficient.

[0023] In this embodiment of the invention, the coordinate disaster prediction agent is an artificial intelligence model built based on machine learning technology. It is used to output corresponding disaster information classification results based on the input regional structure information and node coordinates. Before performing disaster information classification, the coordinate disaster prediction agent is first awakened, putting it into a ready state to await the input data.

[0024] As an optional implementation, the coordinate disaster prediction agent can be preloaded into memory and kept in a dormant state when the system starts up. When disaster information classification is required, it can be switched to working state by calling a wake-up command to save computing resources.

[0025] In this embodiment of the invention, the acquired regional structure information is combined with the node coordinates of each relay node to form multiple input samples. Specifically, for the i-th relay node, its node coordinates are concatenated with the regional structure information and input into the coordinate disaster prediction agent. After calculation, the agent outputs the first disaster information quantity corresponding to that node. By traversing all relay nodes, multiple first disaster information quantities can be obtained.

[0026] Coordinate-based disaster prediction agents can be implemented using various machine learning models, such as support vector machines, random forests, gradient boosting decision trees, or deep neural networks. The core function of these agents is to learn the nonlinear mapping relationship between regional structural information and node coordinates and disaster information.

[0027] The training and testing data for the coordinate-based disaster prediction agent includes a set of sample region structure information extracted from historical disaster record data, a set of sample node coordinates, and a set of labeled sample first disaster information. First, historical disaster record data is collected for the target area or similar areas. This historical disaster record data may include information such as the occurrence time, location coordinates, disaster type, disaster scale, and losses caused by historical disaster events.

[0028] Secondly, sample area structural information is extracted from historical disaster record data. For each historical disaster event, structural information such as the area type and area of ​​the region where the event occurred is obtained as sample area structural information.

[0029] Next, sample node coordinates are extracted from historical disaster record data. Multiple sampling points can be selected around the location of a historical disaster event as the center, and the relative coordinates of each sampling point with respect to the disaster location are calculated as the sample node coordinates.

[0030] Finally, the disaster information coefficient for each grid region is calculated based on historical disaster record data. Specifically, the target area is first divided into multiple grid regions according to a preset grid size. For each sampling point in the historical disaster record data, its grid region is determined based on its coordinates. Then, the number of fires occurring in each grid region in the historical disaster record data is counted as the disaster occurrence count for that grid region. Simultaneously, the arithmetic mean of the disaster occurrence counts for all grid regions containing sampling points is calculated as the average disaster occurrence count. The disaster occurrence count for each grid region is divided by the average disaster occurrence count to obtain the disaster information coefficient for that grid region. A ratio greater than 1 indicates that the fire occurrence frequency of that grid region is higher than the overall average, while a ratio less than 1 indicates that it is lower than the average. The disaster information coefficient corresponding to each grid region is used as the label value for all sampling points within that grid region, forming the first disaster information set of the sample. If the ratio calculation results show extreme values, further truncation or logarithmic transformation can be used for smoothing to avoid individual extreme values ​​from excessively affecting subsequent model training.

[0031] The sample region structure information set, sample node coordinate set, and sample first disaster information set are used as training datasets to conduct supervised training on the coordinate disaster prediction agent. During training, the sample region structure information and sample node coordinates are used as inputs, and the sample first disaster information is used as the expected output. The network parameters of the agent are continuously adjusted through the backpropagation algorithm until the classification accuracy of the model on the validation set meets the preset requirements.

[0032] Regarding model structure, taking a multilayer perceptron as an example, its structure includes an input layer with the number of input nodes equal to the sum of the region structural information feature dimension and the node coordinate dimension; two hidden layers with 64 neurons each, using the Modified Linear Unit (ReLU) activation function; and an output layer with one output node, representing the disaster information quantity coefficient. The loss function used during training is the mean squared error loss function, which measures the difference between the model's predicted values ​​and the labeled values. The optimization objective is to minimize this loss function. The convergence condition can be set as the validation set loss function value no longer decreasing in 10 consecutive iterations, or reaching a preset maximum number of training epochs. After training, the test set is input into the trained model, and the model's classification accuracy on the test set is calculated to evaluate the model's generalization ability. If the classification accuracy meets the requirements, the model is considered successfully trained and can be used for actual disaster information quantity classification in target areas.

[0033] The classification accuracy rate typically ranges from 85% to 95%, with the specific value flexibly set based on the sensitivity of the disaster prediction task to accuracy. For example, in urban fire emergency communication scenarios, the tolerance for misjudgments is low, as misjudgments may lead to the incorrect deployment of relay nodes in fire-risk areas, jeopardizing communication reliability. In such cases, a passing accuracy rate of 90% or higher can be set. In general monitoring scenarios, a rate of around 85% is suitable. After training, the trained coordinate disaster prediction agent can be used for classifying actual disaster information in target areas.

[0034] This invention acquires the regional structure information of the target area and the physical coordinates of each relay node, and calculates their relative coordinates with respect to the coordinates of the regional center. This precisely quantifies the spatial position of each relay node. Then, a pre-trained coordinate disaster prediction agent is used as input, combining the regional structure information with the node coordinates, to output a first disaster information classification result containing disaster information coefficients. This significantly improves the spatial accuracy and processing efficiency of disaster assessment. The coordinate disaster prediction agent is trained based on historical disaster record data, possessing strong generalization ability and objective and unified evaluation criteria. It can adapt to different regional types and node deployment schemes, and the classification results are reproducible and unaffected by human factors. At the same time, its parallel processing capability can quickly complete the calculation of disaster information for a large number of nodes, providing an accurate and reliable disaster risk data foundation for subsequent iterative optimization of the communication topology network. This effectively enhances the communication stability and overall survivability of the self-organizing network relay system in disaster environments.

[0035] S200: Obtain multiple historical environmental parameter sequences monitored by multiple relay nodes within a preset time range, process the disaster information to obtain multiple second disaster information quantities, and combine them with the multiple first disaster information quantities to obtain multiple fused disaster information quantities.

[0036] In existing ad hoc network relay node configuration technologies, disaster risk assessment of relay nodes typically relies solely on static spatial geographic information, such as the region type where the node is located. This falls under the category of static assessment and lacks consideration of the dynamic changes in the node's environment over time. In fact, disasters are often accompanied by abnormal changes in environmental parameters. Relying solely on static spatial information cannot capture these dynamic precursor signals, leading to serious timeliness deficiencies and insufficient dynamic adaptability in disaster assessment results. Furthermore, existing technologies lack systematic processing methods for historical environmental monitoring data from multiple nodes. The large amount of historical environmental parameter data collected by each relay node cannot be effectively utilized for disaster risk quantification, and its data value is not fully realized. Therefore, how to utilize the historical environmental parameter sequences obtained from the time-dimensional monitoring of each relay node, extract key features reflecting extreme environmental conditions and fluctuations, and organically integrate them with static disaster information to form a comprehensive disaster assessment result that considers both spatial static risk and temporal dynamic risk has become an urgent technical problem to be solved.

[0037] Step S200 in the method provided in this embodiment of the invention includes: Acquire multiple historical environmental parameter sequences obtained from monitoring by multiple relay nodes within a preset time range, including environmental parameters such as ambient temperature; Based on the multiple historical environmental parameter sequences, multiple historical environmental parameter peak values ​​and multiple historical environmental parameter change coefficients are extracted, and multiple second disaster information quantities are calculated. Multiple fused disaster information quantities are obtained by integrating multiple secondary disaster information quantities and multiple primary disaster information quantities.

[0038] In this embodiment of the invention, each relay node, in addition to its communication relay function, also integrates an environmental parameter monitoring sensor for real-time collection of environmental parameters at its location. The environmental parameters include at least ambient temperature; as an optional implementation, they may further include physical quantities related to fire occurrence, such as smoke particulate matter concentration, carbon monoxide concentration, and ambient visibility.

[0039] A preset time range refers to a historical time window preceding the current moment, used to collect sufficient historical data to reflect the changing trends of environmental parameters. The preset time range can be flexibly set according to the type of disaster and monitoring needs. In urban fire emergency communication scenarios, fire intensity typically changes rapidly over tens of minutes to several hours; therefore, the preset time range can be set to the past 10 minutes to 1 hour. A shorter time range is used during the rapid fire spread phase to reflect drastic environmental changes, while a longer time range is used during the fire monitoring and early warning phase to capture temperature change patterns and abnormal trends. Those skilled in the art can flexibly determine the preset time range based on the actual application scenario, guided by the above principles, without requiring creative effort.

[0040] Specifically, a data acquisition command is sent to each relay node. In response to the command, each relay node reads historical environmental parameter data recorded at a preset sampling frequency within a preset time range from its local storage module, and then sends the read historical environmental parameter sequences back to the central processing node. Each relay node corresponds to one historical environmental parameter sequence, which consists of multiple environmental parameter sample values ​​arranged in chronological order.

[0041] In this embodiment of the invention, feature extraction and statistical calculations are performed on the historical environmental parameter sequence of each relay node to quantify the extreme environmental conditions and fluctuations at the node's location over a historical period. The peak values ​​of historical environmental parameters characterize the maximum intensity reached by the environmental parameters within the historical period. For example, the peak environmental temperature reflects the highest temperature at that location during the historical period; a high temperature peak may indicate a fire risk or equipment overheating risk.

[0042] Historical environmental parameter variation coefficients are used to characterize the degree of fluctuation of environmental parameters over a historical period. Taking ambient temperature as an example, the ambient temperature variation coefficient reflects the magnitude of temperature changes at that location; drastic temperature changes may indicate the occurrence of a fire.

[0043] Specifically, based on the extracted peak values ​​of historical environmental parameters and the calculated coefficients of change of historical environmental parameters, a second disaster information quantity is calculated for each relay node. The second disaster information quantity is used to characterize the degree of disaster impact or risk level reflected by historical environmental monitoring data. Its numerical form can be consistent with the first disaster information quantity, for example, both being continuous values ​​or discrete levels between 0 and 1, to facilitate subsequent fusion calculations.

[0044] In this embodiment of the invention, for each relay node, the first disaster information quantity and the second disaster information quantity corresponding to it are fused and calculated to obtain the fused disaster information quantity of the node.

[0045] Fusion computation can employ various methods, including weighted summation, product fusion, fuzzy logic fusion, or neural network-based fusion. As a preferred implementation, weighted summation is used for fusion. Specifically, for the i-th relay node, its fused disaster information volume F... i =α×A i +β×B i A i B represents the first disaster information quantity of the i-th relay node. iLet α be the second disaster information quantity of the i-th relay node, and β be the fusion weight coefficients of the first and second disaster information quantities, respectively. Both α and β are positive numbers, where α + β = 1. The specific values ​​of α and β can be set according to the actual disaster characteristics of the target area. For example, in scenarios where the regional structure has a more significant impact on disasters, the value of α can be appropriately increased; in scenarios where environmental parameter fluctuations have a more significant impact on disasters, the value of β can be appropriately increased.

[0046] The fusion weighting coefficients α and β can be determined using a quantitative calculation method based on correlation analysis. First, historical disaster records are collected from the target area or similar areas. For each historical disaster event, the calculated values ​​of the first and second disaster information quantities at the relay node location before the disaster occurred, as well as the quantified value of the actual severity of the disaster, are recorded. All historical disaster event data are then aggregated to form a dataset containing multiple samples. Each sample contains three fields: first disaster information quantity, second disaster information quantity, and actual disaster severity.

[0047] Secondly, the Pearson correlation coefficient formula was used to calculate the correlation coefficients between the first disaster information quantity and the actual severity of the disaster, as well as the correlation coefficients between the second disaster information quantity and the actual severity of the disaster. The correlation coefficient ranges from -1 to 1. The closer the value is to 1, the stronger the positive correlation, that is, the higher the disaster information quantity, the higher the actual severity of the disaster, and the stronger the ability of the information quantity to characterize disaster risk.

[0048] Finally, the absolute value of each of the two correlation coefficients is divided by the sum of the absolute values ​​of the two correlation coefficients. The resulting ratio is the corresponding fusion weight coefficient. The larger the absolute value of a certain correlation coefficient, the closer the correlation between the disaster information and the actual severity of the disaster, and the higher the weight is assigned in the fusion calculation.

[0049] For example, if the fusion weight coefficients of the first disaster information quantity and the second disaster information quantity are 0.6 and 0.4 respectively, and the first disaster information quantity of a certain relay node is 0.85 and the second disaster information quantity is 0.60, then the fused disaster information quantity is 0.6×0.85+0.4×0.60=0.51+0.24=0.75.

[0050] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0051] The integrated disaster information combines spatial static factors and temporal dynamic factors, which can more comprehensively and accurately reflect the comprehensive risk level of each relay node in the actual disaster environment, and provide a more reliable basis for the iterative optimization of the subsequent communication topology network.

[0052] Based on the multiple historical environmental parameter sequences, multiple historical environmental parameter peak values ​​and multiple historical environmental parameter variation coefficients are extracted, and multiple second disaster information quantities are calculated, including: The maximum values ​​within the multiple historical environmental parameter sequences are extracted to obtain the peak values ​​of multiple historical environmental parameters; The deviations of the minimum and maximum values ​​within multiple historical environmental parameter sequences are calculated separately to obtain the variation coefficients of multiple historical environmental parameters. The ratios of peak values ​​of multiple historical environmental parameters to the average peak values ​​of multiple historical environmental parameters, and the ratios of change coefficients of multiple historical environmental parameters to the average change coefficients of multiple historical environmental parameters, are calculated separately to obtain multiple peak information coefficients and multiple change information coefficients. These are then fused together to obtain multiple environmental information coefficients, which serve as multiple secondary disaster information quantities.

[0053] In this embodiment of the invention, for each relay node's historical environmental parameter sequence, a sequence maximum value extraction method is used to traverse all sampled values ​​in the sequence and find the maximum value, which is taken as the peak value of the historical environmental parameter corresponding to that node.

[0054] Specifically, for any relay node, its historical environmental parameter sequence is a dataset consisting of multiple environmental parameter values ​​collected at fixed sampling intervals within a preset time range, arranged chronologically. The total number of sampling points in the dataset depends on the length of the preset time range and the sampling frequency. In this sequence, by comparing the magnitudes of each sample value, the sample value with the largest value is selected as the peak value of the relay node's historical environmental parameter. Taking ambient temperature as an example, if a relay node collected 144 temperature values ​​at 10-minute intervals over the past 24 hours, the maximum value among these 144 temperature values ​​is extracted as the peak value of the node's historical environmental parameter. This peak value reflects the highest temperature experienced by the node over the past 24 hours.

[0055] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0056] In this embodiment of the invention, for each relay node's historical environmental parameter sequence, the minimum and maximum values ​​are extracted, and the deviation between them is calculated as the historical environmental parameter variation coefficient for that node. The historical environmental parameter variation coefficient is used to quantitatively characterize the fluctuation range of environmental parameters over a historical period.

[0057] Specifically, for any relay node, the smallest sampled value in its historical environmental parameter sequence is the minimum value of the sequence, and the largest sampled value is the maximum value. Subtracting the minimum value from the maximum value gives the difference, which is the historical environmental parameter variation coefficient for that relay node. A larger difference indicates greater fluctuation and more drastic changes in the environmental parameters at the node's location over a historical period; conversely, a smaller difference indicates relatively stable environmental parameters. For example, if a relay node's highest temperature in the past minute was 35 degrees Celsius and its lowest was 15 degrees Celsius, its temperature variation coefficient is 20 degrees Celsius. This value directly reflects the temperature fluctuation at the node's location.

[0058] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0059] In this embodiment of the invention, the overall average level of the historical environmental parameter peak values ​​of all relay nodes is first calculated, that is, the arithmetic mean of the historical environmental parameter peak values ​​of all relay nodes is obtained; the overall average level of the historical environmental parameter change coefficients of all relay nodes is then calculated, that is, the arithmetic mean of the historical environmental parameter change coefficients of all relay nodes is obtained.

[0060] Specifically, the average historical environmental parameter peak values ​​of all relay nodes are summed and then divided by the total number of relay nodes to obtain the mean historical environmental parameter peak values. The mean historical environmental parameter peak values ​​represent the overall average level of extreme environmental parameter intensity at the locations of all relay nodes within the target area. The average historical environmental parameter variation coefficients are summed and then divided by the total number of relay nodes to obtain the mean historical environmental parameter variation coefficients. The mean historical environmental parameter variation coefficients represent the overall average level of environmental parameter fluctuation at the locations of all relay nodes within the target area.

[0061] Furthermore, for any relay node, the ratio obtained by dividing its historical environmental parameter peak value by the historical environmental parameter average value is the peak information coefficient of that node; the ratio obtained by dividing its historical environmental parameter variation coefficient by the historical environmental parameter variation coefficient average value is the variation information coefficient of that node.

[0062] The peak information coefficient reflects the intensity of an environmental parameter's extreme level relative to the overall average level: a peak information coefficient greater than 1 indicates that the node's extreme environmental conditions are higher than the target area's average level, meaning the extreme environmental conditions at the node's location are more severe; a peak information coefficient less than 1 indicates that the node's extreme environmental conditions are lower than the target area's average level, meaning the extreme environmental conditions at the node's location are relatively mild. The variation information coefficient reflects the degree of fluctuation of an environmental parameter's extreme level relative to the overall average level: a variation information coefficient greater than 1 indicates that the node's environmental fluctuations are higher than the target area's average level, meaning the environmental changes at the node's location are more drastic; a variation information coefficient less than 1 indicates that the node's environmental fluctuations are lower than the target area's average level, meaning the environmental changes at the node's location are relatively mild.

[0063] In this embodiment of the invention, for any relay node, its peak information coefficient and change information coefficient are fused and calculated to obtain the environmental information coefficient of the node, and the environmental information coefficient is used as the second disaster information of the node.

[0064] The fusion calculation can employ a weighted summation method. Specifically, for each relay node, its peak information coefficient is multiplied by a first weighting coefficient, its variable information coefficient is multiplied by a second weighting coefficient, and the products of the two are summed. The sum obtained is the environmental information coefficient of that node. Both the first and second weighting coefficients are positive numbers, and their sum equals 1.

[0065] Taking fire disasters as an example, this section explains the method for setting the first and second weighting coefficients. Historical fire records of the target area or similar areas are collected. For each fire, the peak information coefficient and the variation information coefficient are calculated for the historical environmental parameter sequence within a preset time window. Simultaneously, the actual burned area or equipment damage rate of the fire is statistically analyzed as a quantitative indicator of disaster severity. Then, the Pearson correlation coefficient between the peak information coefficient and disaster severity, and the Pearson correlation coefficient between the variation information coefficient and disaster severity are calculated separately. The ratio of the two correlation coefficients is used as the basis for weight allocation, i.e., the first weighting coefficient = r 峰值 / (r 峰值 +r 变化 ), Second weighting coefficient = r 变化 / (r 峰值 +r 变化 ), where r 峰值 and r 变化 These are the correlation coefficients between the peak information coefficient, the variable information coefficient, and the severity of the disaster, respectively. For example, if r 峰值 =0.7, r 变化=0.3, then the first weight coefficient = 0.7 / (0.7+0.3) = 0.7, and the second weight coefficient = 0.3 / (0.7+0.3) = 0.3.

[0066] The second disaster information of each relay node not only reflects the extreme conditions of the environmental parameters at the node's location, but also reflects the dynamic fluctuation characteristics of the environmental parameters. The combination of the two can more comprehensively depict the degree of environmental impact on the node over a historical period, providing rich time dimension information for the integration with the first disaster information.

[0067] This invention acquires environmental parameter sequences monitored by relay nodes over historical time periods, extracts peak values ​​reflecting extreme environmental intensity and variation coefficients reflecting environmental fluctuations, and calculates the second disaster information quantity for each node. This expands disaster assessment from a spatial static dimension to a temporal dynamic dimension, effectively mining disaster precursor information contained in historical monitoring data. Furthermore, the peak values ​​and variation coefficients of each node are compared with the corresponding global average, eliminating the influence of differences in environmental background values ​​across different regions. This ensures that the second disaster information quantity for each node has a unified comparison scale and comparability, while also considering the combined effect of environmental extreme values ​​and fluctuation levels on disaster risk. Finally, the first disaster information quantity reflecting spatial static risk and the second disaster information quantity reflecting temporal dynamic risk are fused, achieving spatiotemporal integration of disaster assessment. This significantly improves the comprehensiveness, accuracy, and dynamic adaptability of the assessment results, providing a more scientific and reliable decision-making basis for optimizing communication topology networks, and effectively enhancing the network stability and communication reliability of ad hoc network relay systems in complex disaster environments.

[0068] S300: Based on the multiple fused disaster information quantities, the communication topology network is iteratively optimized, wherein the optimization is performed based on multiple historical environmental parameter sequences to predict the switching information loss of multiple relay nodes and the disaster isolation of multiple node coordinates; In existing ad hoc network relay node configuration optimization techniques, the topology selection typically focuses solely on maximizing communication coverage and minimizing the number of nodes. This lacks comprehensive consideration of node failure risks and task switching information loss under disaster conditions. Consequently, while the optimized topology performs well under normal conditions, it can lead to a cascading effect during disasters due to the obstruction of individual nodes, resulting in a sharp decline or even paralysis of the overall network performance. Furthermore, existing methods often employ a single prediction model for disaster risk assessment. Differences in training data or algorithms between different models can produce inconsistent prediction results, lacking an effective mechanism to ensure the reliability of predictions and making it difficult to guarantee the credibility of risk assessment results.

[0069] Furthermore, existing technologies lack a complete, systematic iterative optimization framework that encompasses coverage constraint verification, multi-dimensional node evaluation, and comprehensive scoring and ranking. This makes it difficult to achieve an effective balance and synergistic optimization among coverage capability, disaster risk, and information loss. Therefore, constructing a systematic network topology iterative optimization method that takes into account communication coverage constraints, disaster risk prediction, and handover information loss assessment to achieve a comprehensive optimization of communication coverage capability, disaster resilience reliability, and information transmission efficiency has become an urgent technical problem to be solved.

[0070] Step S300 in the method provided in this embodiment of the invention includes: Obtain communication coverage constraints, wherein the communication coverage constraints include the ratio of the coverage area of ​​the communication topology network to the area of ​​the target region being greater than a coverage threshold; Within the plurality of relay nodes, several first relay nodes are randomly selected. Based on the communication coverage area of ​​each relay node, a first communication coverage area ratio is generated. If the first communication coverage area ratio satisfies the communication coverage constraint, subsequent steps are performed; otherwise, relay nodes are reselected until the communication coverage constraint is satisfied. Based on multiple historical environmental parameter sequences, disaster isolation predictions are made for several first relay nodes to obtain multiple first node isolation rates. Based on the amount of multiple disaster information, extract the amount of first disaster information from several first relay nodes as the amount of loss of multiple first handover information; The first topology score is calculated based on the first communication coverage area ratio, the blocking rate of multiple first nodes, and the loss of multiple first handover information. Continue selecting the communication topology network and calculating the topology score, and perform iterative optimization.

[0071] In this embodiment of the invention, communication coverage constraints are basic coverage conditions that the communication topology network must meet within the target area, ensuring that the communication network composed of relay nodes can provide sufficient communication service range for the target area. Specifically, the communication coverage constraint includes the ratio of the overall coverage area of ​​the communication topology network to the total area of ​​the target area being greater than a preset coverage threshold. The specific value of the coverage threshold can be set according to the actual application scenario and communication requirements; the higher the level of communication requirements, the larger the coverage threshold value. The coverage threshold can be flexibly selected within the range of 70% to 95%, for example, set above 90% for high reliability requirements and around 80% for general requirements. Those skilled in the art can reasonably determine the threshold within the above range based on actual business needs and service quality requirements without any creative effort.

[0072] The coverage threshold constraint ensures that the optimized communication topology network does not sacrifice basic communication coverage capabilities in pursuit of other performance indicators, thus guaranteeing the basic service availability of the network. The coverage area can be calculated using a planar geometric method, where the communication coverage areas of each relay node are clipped and merged within the target area boundary, and the union area is calculated as the actual coverage area of ​​the communication topology network.

[0073] In this embodiment of the invention, several relay nodes are randomly selected from all relay nodes deployed in the target area as first relay nodes to construct a candidate communication topology network. The number of selected first relay nodes can be set according to actual needs, and its number should be less than or equal to the total number of all relay nodes.

[0074] Specifically, for the selected first relay nodes, the total communication coverage area of ​​these first relay nodes is calculated based on the communication coverage area of ​​each relay node. Then, the proportion of this total communication coverage area to the total area of ​​the target area is calculated to obtain the first communication coverage area percentage. After calculating the first communication coverage area percentage, this percentage is compared with a preset coverage threshold. If the first communication coverage area percentage is greater than the coverage threshold, the currently selected first relay nodes are determined to meet the communication coverage constraints, and subsequent optimization steps can proceed. If the first communication coverage area percentage is less than or equal to the coverage threshold, the current selection is determined to not meet the constraints, and the current selection result needs to be abandoned. Several relay nodes are then randomly selected from all relay nodes, and the communication coverage area percentage is recalculated and verified. This process is repeated until the randomly selected combination of relay nodes can meet the communication coverage constraints.

[0075] In this embodiment of the invention, for each of the selected first relay nodes that meet the communication coverage constraints, it is necessary to predict the risk of future disaster blockage based on its corresponding historical environmental parameter sequence.

[0076] Disaster isolation refers to the phenomenon where communication links between relay nodes are interrupted or the relay nodes themselves malfunction due to a disaster event. The isolation rate is used to quantitatively characterize the probability or degree of risk that each relay node will suffer disaster isolation in the future.

[0077] Specifically, the historical environmental parameter sequence of each first relay node is input into a pre-built disaster isolation prediction model. Based on the changing patterns and characteristics of the historical environmental parameters, the disaster isolation prediction model outputs a predicted probability or risk assessment result of the node's potential disaster isolation in the future, which serves as the node's first-node isolation rate. The isolation rate can be set between 0 and 1, with a higher value indicating a higher risk of the node experiencing disaster isolation in the future.

[0078] In this embodiment of the invention, the first handover information loss amount is used to characterize the degree of information loss or performance degradation caused by switching the relay task of a relay node to another node when the relay node fails due to a disaster. The first handover information loss amount is closely related to the degree of disaster risk of the location of the relay node.

[0079] Specifically, from the obtained multiple fused disaster information quantities, the fused disaster information quantities corresponding to several selected first relay nodes are extracted, and each extracted fused disaster information quantity is directly used as the first handover information loss quantity for each first relay node. The higher the fused disaster information quantity, the greater the comprehensive disaster risk of the node, and the greater the information loss caused by the task handover if the node fails.

[0080] In this embodiment of the invention, the obtained first communication coverage area ratio, multiple first node blocking rates, and multiple first handover information loss quantities are comprehensively calculated to obtain the topology score of the current candidate topology network, which is a communication network composed of several first relay nodes. This topology score is used to quantitatively evaluate the overall performance of the current candidate topology network.

[0081] Specifically, the blocking rates of multiple first nodes can be centered to obtain a first centered blocking rate; the information loss of multiple first handover nodes can be centered to obtain a first centered information loss. The first communication coverage area ratio, the first centered blocking rate, and the first centered information loss are then weighted and summed according to preset weights to obtain a first topology score. The higher the topology score, the better the overall performance of the candidate topology network.

[0082] In this embodiment of the invention, after evaluating the current candidate topology network, several relay nodes are randomly selected from all relay nodes as a new set of candidate nodes. Communication coverage constraints are checked again, and, under the premise of satisfying the constraints, blocking rate prediction, handover information loss extraction, and topology score calculation are performed. Through repeated iterations, each iteration generates a set of candidate topology networks and their corresponding topology scores. The set of relay nodes with the highest topology score is selected from all iterations as the optimal communication topology network configuration.

[0083] The number of iterations can be preset according to actual needs, such as 1000, 5000, or 10000 times. A termination condition can also be set to terminate the iteration when the topology score no longer improves after a certain number of consecutive iterations, thus achieving a balance between computational efficiency and optimization effectiveness. The final optimal topology network configuration will be used for subsequent relay node configuration and data synchronization communication.

[0084] Based on multiple historical environmental parameter sequences, disaster isolation predictions are performed for several first relay nodes, resulting in multiple first node isolation rates, including: Based on disaster communication test data of relay nodes over a historical period, a set of sample environmental parameter sequences and a set of sample isolation results are obtained, where the sample isolation results include yes or no. The set of sample environmental parameter sequences and the set of sample isolation results are divided to obtain multiple sets of training and testing data. Multiple sets of training and testing data were used to conduct training and testing, resulting in multiple intelligent branches for blocking prediction that passed the tests. Multiple historical environmental parameter sequences are input into multiple intelligent branches for barrier prediction, and multiple barrier result sets are output. The proportion of the barrier results is calculated for each branch, and multiple first-node barrier rates are obtained.

[0085] In this embodiment of the invention, in order to construct a disaster isolation prediction model capable of predicting whether a relay node will experience disaster isolation based on environmental parameters, it is first necessary to obtain historical data for model training. Disaster communication test data of relay nodes within a historical time period refers to data obtained through disaster simulation tests conducted at the relay node or recorded during actual disaster events over a past period. This data simultaneously records the environmental parameter conditions at the time of the test and the corresponding communication isolation results.

[0086] From the disaster communication test data, environmental parameter sequences corresponding to each test are extracted to construct a sample environmental parameter sequence set. Simultaneously, the blocking results corresponding to each test are extracted to construct a sample blocking result set. The sample blocking results are binary labeled results, including two values: "Yes" and "No." "Yes" indicates that communication blocking occurred at the relay node under the given environmental parameters, while "No" indicates that communication blocking did not occur at the relay node under the given environmental parameters.

[0087] In this embodiment of the invention, the acquired set of sample environmental parameter sequences and the set of sample blocking results are randomly divided multiple times to generate multiple different sets of training and testing data combinations for training multiple independent intelligent branches for blocking prediction.

[0088] Specifically, in each round of partitioning, all samples are randomly divided into a training set and a test set according to a preset ratio. For example, 70% of the samples can be used as the training set and 30% as the test set. The training set is used to train the blocking prediction intelligent branch, and the test set is used to verify the prediction accuracy of the trained intelligent branch. Through multiple different random partitions, multiple sets of training and test data are obtained. Each set of training and test data contains one training set and one test set, and the composition of the data in each set varies depending on the random partition.

[0089] In this embodiment of the invention, a barrier prediction intelligent branch is trained using each set of training and testing data. The barrier prediction intelligent branch is a binary classification model based on a machine learning algorithm, whose input is a sequence of environmental parameters and whose output is the barrier prediction result.

[0090] Specifically, for each set of training and testing data, the training set is input into a pre-defined machine learning model, such as a decision tree, support vector machine, random forest, or neural network. Through supervised learning, the model learns the mapping relationship between the environmental parameter sequence and the blocking results. After training, the test set of that set is input into the trained model, and the prediction accuracy of the model on the test set is calculated. If the prediction accuracy reaches or exceeds a pre-defined passing threshold, the blocking prediction intelligent branch is deemed to have passed the test and is retained; if the prediction accuracy is lower than the passing threshold, the branch is deemed to have failed and is discarded or retrained.

[0091] The specific value of the preset pass threshold is set according to the reliability requirements of the prediction task. Generally, the pass threshold ranges from 80% to 95%. The higher the threshold, the stricter the prediction accuracy requirements for the retained intelligent branches on the test set, and the higher the individual reliability of the branches; however, an excessively high threshold may result in too few intelligent branches passing the test. In practical engineering applications, in scenarios with high disaster risk sensitivity, where the tolerance for misjudgment is low, the pass threshold can be set to 90% or higher to ensure that each intelligent branch has a high prediction accuracy. In general monitoring scenarios, the pass threshold can be set to around 85% to achieve a balance between branch reliability and the number of branches. In scenarios with preliminary verification or limited data, the pass threshold can be set to 80%. As a standard design choice for those skilled in the art, the pass threshold can be reasonably determined within the above range based on the actual amount of training data and the requirements of the prediction task, without requiring any creative effort.

[0092] In this embodiment of the invention, for each first relay node requiring disaster barrier prediction, its historical environmental parameter sequence is input into multiple qualified barrier prediction intelligent branches. Each intelligent branch independently processes the sequence and outputs a barrier prediction result. Thus, for this relay node, multiple intelligent branches output multiple barrier prediction results, which constitute the barrier result set of this relay node.

[0093] Then, count all blocking results in the blocking result set that are predicted as "yes", calculate the percentage of "yes" results out of the total number of blocking results in the blocking result set, and use this percentage as the first node blocking rate of the relay node. For example, if there are 10 qualified blocking prediction intelligent branches, of which 7 branches predict that the relay node will be blocked and 3 branches predict that it will not be blocked, then the first node blocking rate of the relay node is 70%.

[0094] Based on the first communication coverage area ratio, the blocking rates of multiple first nodes, and the loss of multiple first handover information, the first topology score is calculated, including: Based on the multiple first node blocking rates, a centering calculation is performed to obtain the first centering blocking rate; Based on the loss of multiple first switching information values, a centering calculation is performed to obtain the first centering information loss value; The first topology score is calculated based on the first communication coverage area ratio, the first centering obstruction rate, and the first centering information loss.

[0095] In this embodiment of the invention, each of the several first relay nodes has a corresponding first node blocking rate. In order to combine these first node blocking rates into a single indicator that can represent the overall blocking level of the current candidate topology network, it is necessary to perform a centered calculation on the multiple first node blocking rates.

[0096] The centering calculation can be performed using methods such as the arithmetic mean or the median. As a preferred implementation, the arithmetic mean is used for centering calculation. This involves adding the first node isolation rates of all first relay nodes and then dividing by the total number of first relay nodes. The quotient obtained is the first centering isolation rate. The first centering isolation rate represents the overall average level of disaster isolation risk for all selected nodes in the current candidate topology network.

[0097] In this embodiment of the invention, for a plurality of first relay nodes, each node has a corresponding first handover information loss amount. Similarly, in order to integrate these first handover information loss amounts into a single indicator that can represent the overall handover loss level of the current candidate topology network, it is necessary to perform a centered calculation on multiple first handover information loss amounts.

[0098] The centering calculation can be performed using methods such as the arithmetic mean or the median. As a preferred implementation, the arithmetic mean is used for centering calculation. This involves adding the first handover information loss amounts of all first relay nodes and then dividing by the total number of first relay nodes. The resulting quotient is the first centering information loss amount. The first centering information loss amount represents the average degree of information loss that each node in the current candidate topology network may experience when switching tasks due to disaster risks.

[0099] In this embodiment of the invention, the first centering obstruction rate, the first centering information loss, and the first communication coverage area ratio are used as three evaluation dimensions of the current candidate topology network, and the first topology score is calculated by comprehensively calculating them.

[0100] Specifically, the first topology score can be calculated as follows: First topology score = First communication coverage area percentage + (1 / First centering obstruction rate) + (1 / First centering information loss). The basic logic of this calculation method is that the larger the coverage area percentage, the lower the obstruction rate, and the lower the information loss, the higher the topology score.

[0101] Among them, the proportion of the first communication coverage area, the first centering obstruction rate, and the first centering information loss should all be normalized to the range of 0 to 1 before being included in the calculation, so that the three are in the same dimension and avoid affecting the rationality of the score due to different dimensions.

[0102] Within the plurality of relay nodes, several first relay nodes are randomly selected, and a first communication coverage area percentage is generated based on the communication coverage area of ​​each relay node, including: Obtain the communication coverage area of ​​the relay node; Using the aforementioned communication coverage range, several first node communication ranges are divided with several first relay nodes as centers, and the intersection is processed to obtain the first total communication coverage area; Calculate the ratio of the first total communication coverage area to the target area area to obtain the percentage of the first communication coverage area.

[0103] In this embodiment of the invention, each relay node has a specific communication coverage area, which is determined by the hardware performance parameters of the relay node and the terrain conditions of the target area. For relay nodes of the same model, their communication coverage area can generally be regarded as a circular area centered on the node, and the coverage radius depends on the node's transmission power and the propagation environment.

[0104] Specifically, the maximum communication distance of a relay node can be obtained from its equipment specifications, or it can be corrected using measured signal propagation loss data within the target area to obtain the actual effective communication radius of each relay node in the target area environment. Using this communication radius as the radius of a circle and the node coordinates of the relay node as the center, the communication coverage area of ​​that relay node can be determined.

[0105] In this embodiment of the invention, for a number of selected first relay nodes that meet the communication coverage constraints, the communication coverage circular area corresponding to each node is divided on the geographical plane of the target area with the node coordinates of each first relay node as the center and its respective communication coverage radius as the radius, thereby obtaining a number of first node communication ranges.

[0106] Furthermore, the communication ranges of these first nodes are superimposed in the planar coordinate system of the target area. The union of all the communication ranges of the first nodes is taken, and the portions located outside the boundary of the target area are cropped to obtain the total geographical area that all selected first relay nodes can actually cover within the target area. The area of ​​the total geographical area is then calculated, which is the first total communication coverage area.

[0107] In this embodiment of the invention, the calculated first total communication coverage area is divided by the total area of ​​the target area, and the resulting ratio is the first communication coverage area percentage. The first communication coverage area percentage is expressed as a percentage or a decimal, reflecting the overall coverage level of all selected relay nodes in the current candidate topology network within the target area.

[0108] For example, if the total area of ​​the target area is 100 square kilometers, and the union area of ​​the communication coverage of all selected relay nodes within the target area is 85 square kilometers, then the first communication coverage area accounts for 85%.

[0109] It should be noted that the above values ​​are for illustrative purposes only and do not constitute a limitation on the present invention.

[0110] This invention first filters all candidate topologies using a communication coverage threshold as a hard constraint to ensure that the optimized results have basic service coverage capabilities. Then, by integrating multiple intelligent branches for barrier prediction, it performs voting-style barrier prediction on the environmental parameter sequences of candidate nodes, using the percentage of nodes with a "yes" barrier result as the barrier rate of each node. This significantly improves the accuracy and robustness of disaster barrier prediction and avoids the prediction bias risk brought by a single model. On this basis, the barrier rate and handover information loss of candidate nodes are centered to reflect the overall level of the topology. These are then weighted and merged with the coverage area ratio to form a unified topology score, achieving a comprehensive quantitative evaluation of coverage capability, disaster resistance reliability, and information transmission efficiency. Finally, through multiple random selections and iterative optimizations, the configuration scheme with the best comprehensive score is selected from a large number of candidate topology schemes. This ensures that the final determined communication topology network not only meets basic coverage requirements but also has strong resistance to barrier and low handover information loss in disaster environments, comprehensively improving the communication service quality and system survivability of the self-organizing network relay system in complex disaster scenarios.

[0111] S400: Convergence optimization to obtain the optimal communication topology network, and to perform configuration and data synchronization communication.

[0112] In existing ad hoc network relay node optimization configuration technologies, after multiple rounds of iterative optimization, there is often a lack of clear convergence criteria and optimal solution selection mechanisms. This can lead to the optimization process consuming significant computational resources due to infinite loops, or missing better solutions due to premature termination. Furthermore, the final output solution lacks a complete connection between theoretical optimization and actual network deployment, making it difficult to directly apply the optimization results to actual relay node configuration and data synchronization communication. Therefore, establishing an effective convergence determination mechanism to scientifically select the optimal topology network from numerous candidate solutions and transform it into a practically executable node configuration and communication synchronization scheme has become a pressing technical problem that needs to be solved.

[0113] Step S400 in the method provided in this embodiment of the invention includes: After the iterative optimization reaches the convergence condition, the optimal communication topology network with the largest topology score is output. According to the optimal communication topology network, based on multiple relay nodes, the communication topology is configured to perform data synchronization communication.

[0114] In this embodiment of the invention, after multiple rounds of random selection and evaluation calculations, the iterative optimization process needs to terminate the iteration at an appropriate time and select the best one from all the evaluated candidate topologies.

[0115] Convergence criteria are the basis for determining whether the iterative optimization process should be terminated. As one optional implementation, the convergence criterion can be a preset maximum number of iterations, for example, a preset iteration count of 10,000. When the iteration count reaches 10,000, the optimization process is considered converged, and further iteration is stopped. Another optional implementation can be a threshold for the improvement in topology score. That is, when, in several consecutive iterations, the improvement in the highest topology score of the newly generated candidate topology network compared to the previously recorded highest topology score is less than a preset threshold, the optimization process is considered converged, and further iteration is unlikely to achieve significant performance improvement. Therefore, the iteration is terminated to save computational resources. In urban fire emergency communication scenarios, where timeliness is critical, the improvement threshold can be set between 0.5% and 1%, meaning that the iteration terminates when the improvement in the highest topology score is less than the improvement threshold. In scenarios with ample computational resources and high optimization accuracy requirements, the threshold can be set between 0.1% and 0.3%. As another alternative implementation, the convergence condition can also be a combination of the two conditions mentioned above. That is, the iteration is terminated when either the maximum number of iterations or the improvement in the topology score is less than the threshold is met, so as to ensure that the optimization process neither stops too early nor continues indefinitely.

[0116] After the iterative optimization reaches convergence and the iteration terminates, the candidate network with the highest topology score is selected from all evaluated candidate topologies. The relay node combination and topology structure corresponding to this candidate network are determined as the optimal communication topology. This optimal communication topology has the best overall performance among all evaluated schemes. That is, under the premise of satisfying communication coverage constraints, it comprehensively balances the risk of disaster isolation and the loss of handover information, achieving global optimum or near-global optimum.

[0117] In this embodiment of the invention, after determining the optimal communication topology network, the optimal communication topology network configuration scheme is distributed to each relevant relay node in the target area for actual network deployment and configuration.

[0118] Specifically, based on the relay node combination determined by the optimal communication topology network, relay nodes that need to be enabled and those that need to be disabled are marked separately. For selected relay nodes, a node activation command is sent to them to put them into working state, and their communication link parameters are configured according to the optimal topology, including but not limited to the connection relationship between adjacent nodes, data transmission routing path, communication frequency allocation, and transmit power settings. For unselected relay nodes, they can be set to standby or low-power state to save energy consumption.

[0119] Furthermore, after completing the network configuration, each relay node establishes a communication link according to the optimal communication topology and begins data synchronization communication. The data synchronization communication includes real-time environmental monitoring data collected by each relay node, the working status information of each node, and relay data that needs to be transmitted between nodes. Through the collaborative work of each relay node, a stable and reliable communication network covering the target area is formed, achieving efficient data transmission from the collection end to the aggregation end.

[0120] At this point, the entire process of the data synchronization method for configuring relay nodes in a dynamic topology-based ad hoc network has been completed. This method, through a complete link from coverage screening and multi-dimensional evaluation to iterative optimization, ultimately achieves optimal network deployment that combines coverage capability, disaster resilience reliability, and communication efficiency in disaster environments.

[0121] In summary, the embodiments of the present invention have at least the following technical effects: By acquiring regional structural information and relay node physical coordinates and calculating relative coordinates, and combining this with a coordinate disaster prediction agent to classify disaster information, the spatial accuracy of disaster assessment is improved. By acquiring historical environmental parameter sequences, extracting peak values ​​and variation coefficients, and calculating secondary disaster information, disaster assessment is extended from a spatial static dimension to a temporal dynamic dimension, uncovering disaster precursor information from historical monitoring data. By fusing spatial static risk and temporal dynamic risk to obtain fused disaster information, spatiotemporal integration of disaster assessment is achieved, significantly improving the comprehensiveness, accuracy, and dynamic adaptability of the assessment. By using coverage thresholds as hard constraints to screen candidate topologies, and utilizing multi-intelligent branch voting to predict blocking rates, coverage ratio, blocking rate, and information loss are weighted and fused into a topology score, a comprehensive quantitative assessment of coverage capability, disaster resistance reliability, and information transmission efficiency is achieved. Through multiple iterations, the optimal configuration is selected, enhancing the quality of communication services and survivability in complex disaster environments. By setting clear convergence conditions and providing a complete connection from optimal solution output to node activation, link configuration, and communication synchronization, a complete closed loop from theoretical optimization to actual network deployment is achieved, reducing the risk of data loss and communication interruption due to network switching.

[0122] In summary, this invention solves the technical problem that existing ad hoc network relay node configuration technologies are unable to systematically iteratively optimize both static spatial disaster risks and dynamic temporal environmental changes, resulting in insufficient network stability under disaster environments.

[0123] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0124] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0125] This specification and accompanying drawings are merely illustrative examples of the invention and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its scope. Therefore, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is intended to include these modifications and modifications.

Claims

1. A method for configuring data synchronization of relay nodes in a self-organizing network based on dynamic topology, characterized in that, The method includes: The system acquires regional structure information within the target area and obtains the coordinates of multiple relay nodes deployed within the target area. It then classifies the amount of disaster information to obtain multiple first disaster information quantities. Among these, the multiple relay nodes are used to construct a communication topology network. Multiple historical environmental parameter sequences obtained from monitoring by multiple relay nodes within a preset time range are acquired, and disaster information is processed to obtain multiple second disaster information quantities. These are then combined with the multiple first disaster information quantities to obtain multiple fused disaster information quantities. Based on the multiple fused disaster information quantities, the communication topology network is iteratively optimized, including the prediction of switching information loss of multiple relay nodes and the prediction of disaster isolation of multiple node coordinates based on multiple historical environmental parameter sequences. The convergence optimization yields the optimal communication topology network, which is then configured and used for data synchronization communication.

2. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 1, characterized in that, Obtain regional structure information within the target area, and acquire the coordinates of multiple relay nodes deployed within the target area. Classify the disaster information to obtain multiple primary disaster information quantities, including: Obtain the regional structure information within the target area, including the regional type and area. The physical coordinates of multiple relay nodes deployed in the target area are obtained, and the relative coordinates of multiple nodes with respect to the center coordinates of the target area are calculated as multiple node coordinates. The physical coordinates of multiple nodes are obtained through the positioning information of multiple relay nodes. Based on the regional structure information and multiple node coordinates, disaster information is classified to obtain multiple first disaster information quantities.

3. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 2, characterized in that, Based on the aforementioned regional structure information and multiple node coordinates, disaster information is classified, including: Awaken the coordinate disaster prediction intelligent agent; The regional structure information is combined with the coordinates of multiple nodes and input into the coordinate disaster prediction agent. Multiple first disaster information quantities are then classified and output. The training and testing data of the coordinate disaster prediction agent includes a set of sample area structure information extracted from historical disaster record data, a set of sample node coordinates, and a set of labeled sample first disaster information. Each sample first disaster information includes a disaster information coefficient.

4. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 1, characterized in that, Multiple historical environmental parameter sequences monitored by multiple relay nodes within a preset time range are acquired, and disaster information is processed to obtain multiple second disaster information quantities. These are then combined with the multiple first disaster information quantities to obtain multiple fused disaster information quantities, including: Acquire multiple historical environmental parameter sequences obtained from monitoring by multiple relay nodes within a preset time range, including environmental parameters such as ambient temperature; Based on the multiple historical environmental parameter sequences, multiple historical environmental parameter peak values ​​and multiple historical environmental parameter change coefficients are extracted, and multiple second disaster information quantities are calculated. Multiple fused disaster information quantities are obtained by integrating multiple secondary disaster information quantities and multiple primary disaster information quantities.

5. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 4, characterized in that, Based on the multiple historical environmental parameter sequences, multiple historical environmental parameter peak values ​​and multiple historical environmental parameter variation coefficients are extracted, and multiple second disaster information quantities are calculated, including: The maximum values ​​within the multiple historical environmental parameter sequences are extracted to obtain the peak values ​​of multiple historical environmental parameters; The deviations of the minimum and maximum values ​​within multiple historical environmental parameter sequences are calculated separately to obtain the variation coefficients of multiple historical environmental parameters. The ratios of peak values ​​of multiple historical environmental parameters to the average peak values ​​of multiple historical environmental parameters, and the ratios of change coefficients of multiple historical environmental parameters to the average change coefficients of multiple historical environmental parameters, are calculated separately to obtain multiple peak information coefficients and multiple change information coefficients. These are then fused together to obtain multiple environmental information coefficients, which serve as multiple secondary disaster information quantities.

6. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 1, characterized in that, Based on the multiple fused disaster information quantities, the communication topology network is iteratively optimized, including: Obtain communication coverage constraints, wherein the communication coverage constraints include the ratio of the coverage area of ​​the communication topology network to the area of ​​the target region being greater than a coverage threshold; Within the plurality of relay nodes, several first relay nodes are randomly selected. Based on the communication coverage area of ​​each relay node, a first communication coverage area ratio is generated. If the first communication coverage area ratio satisfies the communication coverage constraint, subsequent steps are performed; otherwise, relay nodes are reselected until the communication coverage constraint is satisfied. Based on multiple historical environmental parameter sequences, disaster isolation predictions are made for several first relay nodes to obtain multiple first node isolation rates. Based on the amount of multiple disaster information, extract the amount of first disaster information from several first relay nodes as the amount of loss of multiple first handover information; The first topology score is calculated based on the first communication coverage area ratio, the blocking rate of multiple first nodes, and the loss of multiple first handover information. Continue selecting the communication topology network and calculating the topology score, and perform iterative optimization.

7. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 6, characterized in that, Based on multiple historical environmental parameter sequences, disaster isolation predictions are performed for several first relay nodes, resulting in multiple first node isolation rates, including: Based on disaster communication test data of relay nodes over a historical period, a set of sample environmental parameter sequences and a set of sample isolation results are obtained, where the sample isolation results include yes or no. The set of sample environmental parameter sequences and the set of sample isolation results are divided to obtain multiple sets of training and testing data. Multiple sets of training and testing data were used to conduct training and testing, resulting in multiple intelligent branches for blocking prediction that passed the tests. Multiple historical environmental parameter sequences are input into multiple intelligent branches for barrier prediction, and multiple barrier result sets are output. The proportion of the barrier results is calculated for each branch, and multiple first-node barrier rates are obtained.

8. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 6, characterized in that, Based on the first communication coverage area ratio, the blocking rates of multiple first nodes, and the loss of multiple first handover information, the first topology score is calculated, including: Based on the multiple first node blocking rates, a centering calculation is performed to obtain the first centering blocking rate; Based on the loss of multiple first switching information values, a centering calculation is performed to obtain the first centering information loss value; The first topology score is calculated based on the first communication coverage area ratio, the first centering obstruction rate, and the first centering information loss.

9. The method for synchronizing configuration data of relay nodes in a self-organizing network based on dynamic topology according to claim 6, characterized in that, Within the plurality of relay nodes, several first relay nodes are randomly selected, and a first communication coverage area percentage is generated based on the communication coverage area of ​​each relay node, including: Obtain the communication coverage area of ​​the relay node; Using the aforementioned communication coverage range, several first node communication ranges are divided with several first relay nodes as centers, and the union of these ranges is used to obtain the first total communication coverage area. Calculate the ratio of the first total communication coverage area to the target area area to obtain the percentage of the first communication coverage area.

10. The method for synchronizing relay node configuration data in a self-organizing network based on dynamic topology according to claim 1, characterized in that, Convergence optimization yields the optimal communication topology network, followed by configuration and data synchronization communication, including: After the iterative optimization reaches the convergence condition, the optimal communication topology network with the largest topology score is output. According to the optimal communication topology network, based on multiple relay nodes, the communication topology is configured to perform data synchronization communication.

Images