Emergency communication method and system based on multi-modal early warning broadcasting terminal

Abstract

The application discloses an emergency communication method and system based on a multi-modal early warning broadcasting terminal, relates to the technical field of emergency communication of early warning broadcasting terminals, and specifically realizes the following scheme: collecting position information of each early warning terminal arranged in a region to be warned, multi-dimensional environment information of the position and self running state information of the early warning terminal, performing environment anomaly identification based on the collected environment information, judging an environment danger coefficient of each early warning terminal, determining a wireless communication wavelength according to the environment interference coefficient and the position distribution of the early warning terminal, constructing a plurality of hop-by-hop relay pre-communication links for each lost early warning terminal, calculating an evaluation coefficient based on the communication state of the early warning terminal, the network center degree, the prediction length and the length of the pre-communication link, selecting a communication chain based on the evaluation coefficient, and providing an adaptive and multi-factor combined optimization emergency communication solution.

Description

Technical Field

[0001] This application relates to the field of emergency communication technology for early warning broadcasting terminals, specifically to an emergency communication method and system based on a multimodal early warning broadcasting terminal. Background Technology

[0002] In areas prone to subsidence, such as key dam areas, several early warning terminals are deployed to collect data such as soil moisture or displacement. Some early warning terminals may lose contact due to public network signal interruption caused by regional subsidence. Therefore, early warning terminals need to maintain signal transmission through wireless signal networking. However, wireless communication signals are also susceptible to interference and transmission weakening due to harsh environments. Therefore, there is an urgent need for an early warning terminal that can maintain communication capabilities to quickly search for lost terminals and reconstruct emergency communication links to ensure that early warning information can be continuously and reliably transmitted and broadcast.

[0003] In the prior art, CN119922524A discloses an emergency communication method and system based on a multimodal early warning broadcasting terminal. This method is designed for flash flood disasters. It triggers multimodal terminal early warning by combining soil moisture and rainfall intensity thresholds, utilizes a radar array to periodically scan surface micro-deformations and calculates displacement rates based on Doppler frequency shift, generates a terrain risk heat map using a dynamic feature library, and constructs a low-power, long-range wireless mesh self-organizing link to perform multi-hop redundant transmission when the public network is interrupted. It also dynamically allocates data packet forwarding priorities according to the early warning level, synchronizing the audible and visual early warning signals with the disaster risk gradient, forming a hierarchical response link linking deformation monitoring and communication. However, this scheme relies solely on the combination of soil moisture and rainfall thresholds for disaster perception, resulting in a single environmental monitoring dimension. It cannot perceive the real-time impact of multimodal factors such as smoke, temperature, and particulate matter concentration on wireless communication quality in fire scenarios. This leads to the need to use fixed wireless parameters when searching for lost terminals, making it difficult to adaptively penetrate complex disaster environments such as dense smoke and high humidity. Furthermore, while the solution prioritizes data packet forwarding based on the warning level and constructs self-organizing links based on preset protocols in terms of communication networking, it fails to comprehensively and quantitatively assess the dynamic status of relay base stations, such as their remaining power, bandwidth usage, topology network centrality, and environmental hazard coefficients. This makes it difficult to identify and avoid vulnerable nodes that are about to lose power, have exhausted bandwidth, or are located in high-risk areas among multiple alternative paths. Consequently, the selected communication links are prone to breakage due to the unexpected failure of individual nodes during actual operation, making it difficult to meet the high requirements for the resilience and survivability of communication networks in emergency scenarios.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] In view of this, the purpose of this application is to provide an emergency communication method and system based on a multimodal early warning broadcasting terminal, which can realize an adaptive, multi-factor joint optimization emergency communication solution.

[0006] According to a first aspect of this application, an emergency communication method based on a multimodal early warning broadcasting terminal is provided, comprising: The system collects location information, multi-dimensional environmental information of the location, and operational status information of each early warning terminal deployed in the area to be warned. Based on the multi-dimensional environmental information of each early warning terminal, it identifies environmental anomalies. By setting the communication adjustment duration, it constructs a dynamic risk distribution map of the area to be warned and assesses the environmental hazard coefficient of each early warning terminal based on the dynamic risk distribution map. The system obtains the public network signal status of all early warning terminals, designates early warning terminals with public network connections as network base stations, and those without public network connections as out-of-connection early warning terminals. It calculates the environmental interference coefficient based on the multi-dimensional environmental information collected by the network base stations, and determines the appropriate wireless communication wavelength by combining the environmental interference coefficient with the location distribution of out-of-connection early warning terminals. It then constructs multiple hop-by-hop relay pre-communication links for each out-of-connection early warning terminal. Based on the operational status information of each relay base station, the remaining predicted working time, communication status score and network centrality of each relay base station are calculated. For each pre-communication link, the forwarding sequence number of each early warning terminal in the pre-communication link is determined, and the communication status score, network centrality and environmental risk coefficient of the network base station and relay base station are coupled. The contribution of each relay base station and network base station in the pre-communication link is obtained by correcting the forwarding sequence number and the exponential decay function of the prediction duration. For each pre-communication link, the stability of the pre-communication link is evaluated based on the contribution of the internal relay base station and the network base station, and a penalty term is constructed with the number of link nodes. The pre-communication link is selected as the communication link based on the evaluation results.

[0007] Furthermore, the multi-dimensional environmental information includes temperature, humidity, and particulate matter concentration; Its own operating status information includes wireless communication wavelength, battery level, power consumption, effective transmission rate, bandwidth limit, and occupied bandwidth; The method for identifying environmental anomalies is as follows: For each type of environmental information, a warning threshold and a warning time threshold are set for each environmental type. When each type of environmental information exceeds the corresponding warning threshold and warning time threshold, it is identified as an environmental anomaly.

[0008] Furthermore, the method for constructing the dynamic risk distribution map is as follows: For each environmental type identified as an environmental anomaly, a warning level is assigned. The warning levels for each environmental type are summed to calculate the risk level of the warning terminal. Based on the risk level and location distribution of each warning terminal, the risk level of any location within the warning area is calculated using an inverse distance weighted interpolation method, forming a continuous risk distribution map. The risk distribution map is updated by adjusting the communication interval as set, thereby generating a dynamic risk distribution map.

[0009] Furthermore, the method for analyzing the environmental hazard coefficient is as follows: Centered on the early warning terminal, a risk analysis neighborhood is defined with a preset spatial radius. The risk level values ​​assigned by the dynamic risk distribution map within this neighborhood are extracted as a sample set. The mean of this sample set is calculated to characterize the baseline risk intensity of the early warning terminal. At the same time, the variance of this sample set is calculated to quantify the volatility of the risk distribution within the neighborhood. The mean and variance are coupled through a weighted fusion method to generate the environmental hazard coefficient of the early warning terminal.

[0010] Furthermore, the step of determining the wireless communication wavelength is as follows: Based on the humidity, temperature, and particulate matter concentration collected by the network base station, normalization was performed on each data point. Differential weights were assigned according to the degree of influence of each environmental factor on wireless signal attenuation, and the environmental interference coefficient was calculated by weighted summation. The straight-line distance is calculated based on the location of the network base station and the location of the loss-of-connection warning terminal. Combined with the environmental interference coefficient, the communication wavelength is adaptively optimized through a preset exponential function mapping model to determine the optimal working wavelength for stable wireless communication under the current distance and comprehensive environmental disturbances. The exponential function mapping model is as follows: in, For wireless communication wavelength, This is the proportionality coefficient. The straight-line distance. As an environmental sensitivity index, Environmental interference coefficient; The strategy for constructing pre-communication links is as follows: Based on a defined wireless communication wavelength, starting from a network base station, the system first searches for and establishes connections with lost-connection warning terminals within its direct wireless coverage area. Then, using the successfully connected lost-connection warning terminals as relay base stations, the system continues to search for and extend links to lost-connection warning terminals further away. This process establishes several pre-communication links consisting of single or multiple hops for each lost-connection warning terminal through hop-by-hop relay.

[0011] Furthermore, the analysis method for the predicted duration is as follows: The prediction duration is based on the remaining battery power and average power consumption in its own operating status information; By setting a communication adjustment duration, the battery level and power consumption in the self-operating status information are discretized to obtain the battery level and power consumption at each time interval. By setting N communication adjustment duration windows, the system slides along the time scale with one communication adjustment duration as the step size to calculate the average battery level and average power consumption within each window. The prediction duration is then determined based on the average battery level and average power consumption.

[0012] Furthermore, the communication status is determined based on the effective transmission rate, bandwidth limit, and occupied bandwidth in the early warning terminal's own operating status information; By setting N communication adjustment duration windows, and sliding along the time scale with one communication adjustment duration as the step size, the average effective transmission rate and average remaining bandwidth are calculated based on the effective transmission rate, bandwidth limit, and occupied bandwidth within each window. The ratio of the calculated average effective transmission rate to the expected rate is used as the rate score; The bandwidth score is obtained by dividing the average remaining bandwidth by the bandwidth limit of the warning terminal. The communication status is obtained by summing the weights.

[0013] Furthermore, the method for determining the network centrality is as follows: Based on the public network signal status and wireless communication update status of each early warning terminal, early warning terminals that have stopped updating both public network signal status and wireless communication update status are determined to be in a disconnected state. A comprehensive network centrality score is obtained by weighted summing of the number of disconnected terminals within the communication range of the early warning terminal with normal communication and the distance of each of them to the early warning terminal with normal communication.

[0014] Furthermore, the method for selecting pre-communication links as communication links is as follows: Calculate the evaluation coefficient for each pre-communication link. When calculating the evaluation coefficient of the entire link, the product of the contributions of all early warning terminals in the link is multiplied by a penalty term for the number of early warning terminals in the pre-communication link. The pre-communication link with the highest evaluation coefficient is selected as the communication link.

[0015] According to a second aspect of this application, an emergency communication system based on a multimodal early warning broadcasting terminal is provided, for executing the aforementioned emergency communication method based on a multimodal early warning broadcasting terminal, comprising: The data acquisition and analysis module is used to collect the location information, multi-dimensional environmental information of the location, and the operating status information of each early warning terminal deployed in the area to be warned. Based on the multi-dimensional environmental information of each early warning terminal, environmental anomalies are identified. By setting the communication adjustment time, a dynamic risk distribution map of the area to be warned is constructed. The environmental risk coefficient of each early warning terminal is evaluated based on the dynamic risk distribution map. The link construction module is used to obtain the public network signal status of all early warning terminals, use early warning terminals with public network connection as network base stations, and use those without public network connection as lost early warning terminals. The module calculates the environmental interference coefficient based on the multi-dimensional environmental information collected by the network base stations, and determines the appropriate wireless communication wavelength by combining the environmental interference coefficient with the location distribution of the lost early warning terminals. Multiple hop-by-hop relay pre-communication links are constructed for each lost early warning terminal. The terminal analysis module is used to calculate the remaining working prediction time, communication status score and network centrality of each relay base station based on its own operating status information. The terminal coupling module is used to determine the forwarding sequence number of each early warning terminal in each pre-communication link, couple the communication status score, network centrality and environmental hazard coefficient of the network base station and the relay base station, and obtain the contribution of each relay base station and the network base station in the pre-communication link by correcting the forwarding sequence number and the exponential decay function of the prediction duration. The link selection module evaluates the stability of each pre-communication link based on the contribution of the internal relay base station and the network base station, and constructs a penalty term with the number of link nodes. Based on the evaluation results, the pre-communication link is selected as the communication link.

[0016] One embodiment of the above application has the following advantages or beneficial effects: This application collects location information, multi-dimensional environmental information of the location, and operational status information of each early warning terminal deployed in the area to be warned. Based on the collected environmental information, it identifies environmental anomalies and determines the environmental hazard coefficient of each early warning terminal. It determines the wireless communication wavelength based on the environmental interference coefficient and the location distribution of the early warning terminals, and constructs multiple hop-by-hop relay pre-communication links for each lost early warning terminal. Based on the communication status of the early warning terminal, network centrality, prediction duration, and the length of the pre-communication links, it calculates evaluation coefficients and selects communication links based on the evaluation coefficients. This application provides an adaptive, multi-factor joint optimization emergency communication solution. This application utilizes a gridded approach to deploy multimodal early warning terminals in the disaster-prone area. It collects real-time environmental information such as temperature, humidity, and particulate matter concentration, as well as status data including the terminals' wireless parameters, battery level, power consumption, and transmission rate. On one hand, it constructs a dynamic risk distribution map based on environmental anomalies and calculates environmental hazard coefficients. On the other hand, it quantifies the real-time impact of the environment on signals as an environmental interference coefficient, which is then used in an exponential function model to adaptively determine the optimal wireless communication wavelength for searching for lost terminals. This allows for the rapid establishment of multiple alternative pre-communication links for each signal-interrupted early warning terminal in complex and ever-changing disaster environments. This design effectively overcomes the challenge of wireless communication obstruction caused by multimodal interference factors in disaster environments, enabling the rapid location of lost nodes and the establishment of alternative communication connections even in complex scenarios such as fires and mudslides.

[0017] In the link optimization decision-making stage, this application comprehensively considers the communication status, network centrality, environmental hazard coefficient, and available time dynamically predicted based on remaining power and average power consumption of each early warning terminal in each pre-communication link. It corrects the contribution by introducing an exponential decay function of forwarding sequence number and predicted time, and calculates the overall evaluation coefficient of the link using a squared penalty term based on the total number of nodes in the link. Finally, it selects the best link for network deployment. This evaluation system closely links node performance with disaster risk, ensuring that the contribution of terminals in high-risk areas or nodes about to run out of energy is reasonably reduced during link selection. This prioritizes the construction of communication links that avoid danger, have strong node endurance, and prominent central positions, significantly improving the survival time and overall stability of the emergency local area network. Simultaneously, this solution deeply integrates environmental perception and early warning with communication resource scheduling. Based on a discrete sliding window of communication adjustment time, it dynamically predicts terminal usage time and assesses communication status change trends, achieving orderly network deployment during terminal disconnection and reconstruction. This significantly reduces the risk of cascading interruptions caused by local node failures, providing an adaptive, highly robust, and multi-factor jointly optimized emergency communication solution for gridded large-scale early warning systems.

[0018] Other effects of the above-mentioned alternative methods will be described below in conjunction with specific embodiments. Attached Figure Description

[0019] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of this application. Wherein: Figure 1 This is a schematic diagram of an emergency communication method based on a multimodal early warning broadcasting terminal provided in an embodiment of this application; Figure 2 This is a network distribution diagram of the pre-communication link provided in an embodiment of this application; Figure 3 This is a block diagram of an emergency communication system based on a multimodal early warning broadcasting terminal, provided in an embodiment of this application. Detailed Implementation

[0020] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0021] To facilitate understanding of this application, the embodiments of this application will be briefly described below: In areas prone to subsidence, such as key dam areas, numerous early warning terminals are deployed to collect data such as soil moisture or displacement. Some of these terminals may lose connection due to public network signal interruptions caused by regional subsidence. Therefore, it is necessary for these early warning terminals to maintain signal transmission through wireless network networking. However, wireless communication signals are also susceptible to interference and transmission weakening due to harsh environments. Thus, there is an urgent need for an early warning terminal capable of maintaining communication capabilities to quickly search for lost terminals and reconstruct emergency communication links to ensure that early warning information can be continuously and reliably transmitted and broadcast. Furthermore, there are significant differences among terminals in terms of remaining battery power, bandwidth usage, topology location, and the degree of danger of their surrounding environment. Traditional solutions use fixed wireless parameters for searching and select communication links solely based on hop count or signal strength. This approach cannot adaptively penetrate complex environments and is difficult to identify vulnerable nodes. Consequently, restored communication links are easily interrupted again due to the failure of individual nodes, failing to meet the high requirements for resilience and survivability of communication networks in emergency scenarios. This application quantifies multimodal environmental interference into an environmental interference coefficient, substitutes it into an exponential function model to adaptively determine the optimal wireless communication wavelength to accurately search for lost terminals and establish multiple alternative pre-communication links. Then, by comprehensively considering communication status, network centrality, environmental risk coefficient, and dynamically predicted availability, the link evaluation coefficient is calculated through a contribution multiplication and hop count squared penalty mechanism, and the best network is selected. In this way, emergency communication links with high node efficiency, short and robust paths, and the maximum survival period are automatically selected in harsh environments.

[0022] See Figure 1 This is a flowchart illustrating an emergency communication method based on a multimodal early warning broadcasting terminal provided in an embodiment of this application. Figure 1 The execution subject of the method shown can be a combination of software and / or hardware, specifically, it can be one or more of various types of terminals, hardware systems, cloud computing, etc.

[0023] Figure 1 An emergency communication method based on a multimodal early warning broadcasting terminal, as shown, includes steps 1 to 5, as detailed below: Step 1: Collect the location information, multi-dimensional environmental information of the location, and the operating status information of each early warning terminal deployed in the area to be warned. Based on the multi-dimensional environmental information of each early warning terminal, identify environmental anomalies. By setting the communication adjustment time, construct a dynamic risk distribution map of the area to be warned. Based on the dynamic risk distribution map, assess the environmental risk coefficient of each early warning terminal.

[0024] The core purpose of this step is to establish a precise understanding of the environmental situation before establishing communication links, thereby transforming communication strategy formulation from "passive response" to "proactive adaptation." It achieves a deep integration of communication resource scheduling and real-time environmental risk, with the advantage of building a forward-looking decision-making foundation based on multi-dimensional environmental perception. Traditional emergency communications often rely solely on network connectivity to build links after a disaster, neglecting the actual degree of danger at the terminal's location, easily leading to critical information in high-risk areas not being transmitted preferentially. This step, by collecting multi-dimensional environmental information and identifying anomalies, can quantify abstract environmental risks into specific and comparable environmental hazard coefficients for each early warning terminal. This makes subsequent relay selection and link screening no longer a blind search for network optimization, but rather prioritizes ensuring the communication survivability of high-risk areas. The construction of a dynamic risk distribution map further activates static geographical location information into a risk evolution map with a time dimension, capturing the changing trends and spatial spread of risks, allowing the system to anticipate impending threats. This approach of "knowing the environment before communicating" fundamentally improves the targeting and effectiveness of emergency communications, ensuring that valuable communication resources are accurately deployed to where they are most needed.

[0025] In this embodiment, the multi-dimensional environmental information includes temperature, humidity, and particulate matter concentration; Its own operating status information includes wireless communication wavelength, battery level, power consumption, effective transmission rate, bandwidth limit, and occupied bandwidth; The method for identifying environmental anomalies is as follows: For each type of environmental information, a warning threshold and a warning time threshold are set for each environmental type. When each type of environmental information exceeds the corresponding warning threshold and warning time threshold, it is identified as an environmental anomaly.

[0026] This step of performing environmental anomaly analysis and constructing a dynamic risk distribution map for each early warning terminal is because changes in environmental parameters during a disaster are often not isolated, single signals, but rather dynamic processes involving the correlation and coupling of multiple physical quantities within a specific terrain unit. Traditional monitoring systems mostly issue alarms based on a single threshold, such as comparing whether temperature exceeds a standard or particulate matter concentration exceeds a limit. This approach cannot distinguish between short-term fluctuations caused by sunrise warming or routine dust storms and the multi-parameter coordinated anomalies exhibited by true disaster precursors. This application, by setting early warning thresholds and early warning time thresholds for each type of environmental information, can issue early warnings for single environmental factors. This multi-locking mechanism significantly suppresses false alarms caused by noise from a single sensor or transient interference in principle, ensuring that every issued early warning signal is based on a multimodal evidence chain, guaranteeing the accuracy and authority of the early warning broadcast.

[0027] In traditional solutions, even when a dynamic risk distribution map is generated, it is often only used for visualization by command personnel. This is disconnected from scheduling actions such as time slot allocation and relay base station wake-up in the communication network. When a disaster evolves and causes a sudden increase in the load on local early warning terminals or signal attenuation, communication resources cannot automatically adjust to the risk situation. This application, however, dynamically sets the communication adjustment duration based on the time distribution of the dynamic risk distribution map, thereby quantitatively determining the environmental hazard coefficient of each early warning terminal. The communication adjustment duration here expands and contracts with changes in risk. In normal environments, it can be set to 10 minutes; in dangerous areas during the disaster development phase, a more intensive sampling and communication refresh frequency can be assigned; when environmental anomalies are detected, it can be reduced to 30 seconds; and in low-risk areas, resource consumption can be appropriately reduced, achieving a dynamic balance between sensing accuracy and communication energy consumption.

[0028] This invention sets warning thresholds and warning time thresholds for each environmental type. Its core purpose is to solve the problems of high false alarm rates and difficulty in identifying disaster types caused by single-sensor threshold alarms. In real natural environments, various environmental parameters are constantly fluctuating dynamically. For example, the surface temperature at midday in summer can easily exceed 50 degrees Celsius, and strong winds can cause brief localized dust storms and vibrations. If relying solely on a static threshold for a single parameter, the system will interpret all these normal disturbances as precursors to disasters, leading to frequent and invalid alarms from the warning terminal. This not only wastes communication bandwidth and terminal power but also creates a "crying wolf" effect, making on-site personnel slow to react to subsequent genuine disaster warnings. By introducing a "warning time threshold," a logical leap from perceiving phenomena to identifying disaster types is achieved. The warning time threshold requires that environmental parameters continuously exceed the limit, rather than changing instantaneously, thus filtering out occasional, brief vibrations and other interference.

[0029] Temperature warning thresholds need to be differentiated from daily extreme weather values. They are typically set based on historical temperature data of the area to be warned, combined with the type of surface cover (e.g., exposed rock, asphalt pavement, or vegetation). For example, the surface air temperature threshold can be set between 45°C and 55°C in plains and hilly areas, while it can be appropriately raised to 50°C to 60°C on sunny slopes of steep mountains due to their stronger heat retention. The warning time threshold is usually set between 10 and 30 seconds. Temperature warnings are not intended to capture natural warming caused by slow, intense sunlight, but rather to detect sudden heat sources such as open flames or electric arcs. Setting this short window ensures that the system will only flag an abnormal heat source as a warning when heat accumulates rapidly and persistently.

[0030] The warning threshold for particulate matter concentration is commonly characterized by PM2.5. For initial fire warnings, a PM2.5 threshold of 400 μg / m³ can be set. 3 This distinguishes it from the gradual accumulation characteristic of regular smog. The warning time threshold for particulate matter concentration is typically set between 90 and 180 seconds. Unlike gaseous smoke, the diffusion and settling of fine particulate matter have a certain lag. Setting a relatively long time window is to observe whether the concentration increase follows a continuous and smooth trajectory. This is because the thermal buoyancy in the early stages of a fire drives particulate matter to rise rapidly and continuously, while windblown sand or construction dust often exhibits a violent, jagged fluctuation followed by a rapid decline; a longer window clearly distinguishes these two types of physical processes.

[0031] Humidity is a key factor in debris flow disasters, and its early warning threshold focuses more on the degree of soil moisture saturation. Based on soil mechanical properties, a soil volumetric moisture content threshold of 60%-75% can be set, with the early warning time threshold typically set at the level of several hours, such as 30 minutes to 1 hour. Soil saturation in geotechnical mechanics is a slow physical infiltration process; reaching a high moisture content of over 60% from a dry state is not something that can be achieved in minutes. Setting a long window of several hours clearly indicates that the soil has entered a fully saturated mechanical critical state, rather than merely a false impression of surface runoff from a brief heavy rain.

[0032] In this embodiment, the method for constructing the dynamic risk distribution map is as follows: For each environmental type identified as an environmental anomaly, a warning level is assigned. The warning levels for each environmental type are summed to calculate the risk level of the warning terminal. Based on the risk level and location distribution of each warning terminal, the risk level of any location within the warning area is calculated using an inverse distance weighted interpolation method, forming a continuous risk distribution map. The risk distribution map is updated by adjusting the communication interval as set, thereby generating a dynamic risk distribution map.

[0033] In some embodiments, the generation of a dynamic risk distribution map is a dynamic aggregation process from single-point environmental anomalies to regional risk fields: First, for each type of environmental information, such as temperature, smoke, vibration, humidity, and particulate matter concentration, multi-level warning thresholds are set according to their disaster characteristics, and the warning levels are quantified into discrete values ​​(e.g., level 0 normal, level 1 attention, level 2 warning, level 3 severe). Once the monitored value continuously exceeds a certain level threshold, the corresponding level value is assigned. The risk level of the warning terminal is obtained by simply adding all the level values ​​arithmetically. For example, if a terminal has a temperature level of 2, a smoke level of 1, a particulate matter level of 2, a humidity level of 1, and a vibration level of 0, the sum of its group levels is 6.

[0034] Furthermore, the warning level for each environmental type can be divided according to a fixed multiple of the warning threshold. Using the basic warning threshold set for each environmental information type as a benchmark, a situation where the monitored value does not exceed the benchmark threshold is classified as Level 0 (normal). When the monitored value continuously exceeds the benchmark threshold but does not reach 1.5 times the benchmark threshold, it is classified as Level 1; when the monitored value reaches 1.5 times but not 2 times the benchmark threshold, it is classified as Level 2; and when the monitored value reaches 2 times or more of the benchmark threshold, it is classified as Level 3. For example, if the basic warning threshold for temperature is 45°C, then below 45°C is Level 0, 45°C to 67.5°C is Level 1, 67.5°C to 90°C is Level 2, and above 90°C is Level 3. The advantage of this method lies in using a fixed multiple as the scale for level transitions, which allows environmental parameters with different physical dimensions and numerical ranges to be discretized under a unified order of magnitude logic. This eliminates the incomparability caused by differences in data types and provides a mathematical basis for directly adding multiple environmental levels to calculate the risk level. The multiple division itself embeds a nonlinear representation of the severity of the disaster. Doubling the monitoring value often means that the disaster energy is increasing exponentially. Using a fixed multiple as the level boundary perfectly matches the physical law of disaster evolution, so that the level value can faithfully reflect the urgency of environmental anomalies. When performing exponential upgrades after adding related environmental types, the discrete level values ​​generated by the fixed multiple division make the starting point and magnitude of the exponential upgrade more controllable and have clear physical meaning. Thus, the risk value can increase moderately when there is a single-factor anomaly, while the risk value shows an explosive leap when multiple factors are coupled, accurately distinguishing between normal disturbances and real disaster precursors.

[0035] After obtaining the risk level of each early warning terminal, and combining their known location coordinates, each early warning terminal is regarded as a discrete sampling point carrying a risk value. The inverse distance weighted interpolation method is used to calculate the risk value of any location in the area to be warned. That is, all early warning terminals with obtained risk levels are used as known sample points, and their spatial locations and risk level values ​​constitute a sample point set.

[0036] More specifically, the area to be warned can be divided into a uniform grid, with each grid cell representing the location of a risk value to be estimated. All warning terminals with known risk levels are used as known sample points, and their spatial coordinates and corresponding risk level values ​​constitute the interpolation source dataset. In the interpolation calculation phase, for each grid cell, the spatial distance between the center coordinates of that grid cell and each known sample point is calculated one by one.

[0037] The grid side length is primarily referenced to the average deployment spacing of the early warning terminals. Typically, the grid side length is set to one-tenth to one-half of the average nearest neighbor terminal distance, ensuring at least several grid units between any two adjacent terminals to prevent overly coarse interpolation results. If the area to be warned is large and the terminals are sparsely distributed, the grid size can be appropriately increased to reduce computational load; otherwise, small-scale risk variations will be difficult to accurately capture when extracting neighboring samples due to an overly coarse grid. Conversely, for core monitoring areas with densely deployed terminals, the grid size should be reduced to retain local risk details. However, it should be noted that excessively small grids will significantly increase the interpolation computation time. The refresh cycle requirements of the dynamic risk distribution map must also be considered. Too many grids will require traversing a massive number of grids for each update. If the computation time exceeds the set communication adjustment time, the grid size must be increased or a parallel computing strategy must be adopted to meet real-time requirements. In summary, in actual deployment, one-fifteenth of the average spacing between terminals can be used as the initial grid side length. After small-scale simulation verification, it can be fine-tuned according to the precision of spatial variation of risk level and system computing power until a compromise between accuracy and efficiency is reached.

[0038] For any location to be estimated within the warning area, calculate the spatial distance between that location and each known sample point. Based on the obtained distances, calculate the influence weight of each known sample point on the estimated location using the inverse of a specified power of the distance; sample points that are closer are assigned higher weights, and sample points that are farther away are assigned lower weights. Sum the risk level values ​​of all known sample points with their corresponding weights, and divide by the sum of all weights for normalization. The result is the risk level of the location to be estimated.

[0039] Then, a heatmap is generated by color rendering according to color levels from low to high. This heatmap is then overlaid with a geographic base map, warning terminal icons, warning type labels, and auxiliary information such as the location of the warning terminals, forming a continuous risk distribution map. As timestamped environmental data continuously flows in during each collection cycle, the above calculation and mapping process is executed cyclically. The dynamic risk distribution map is dynamically refreshed over time, not only showing the direction and speed of the spread of dangerous areas but also identifying areas where risks are steadily increasing or rapidly deteriorating. This provides an intuitive and quantitative basis for subsequent time-series analysis of environmental hazard coefficients, setting communication adjustment durations, and optimizing local area network links.

[0040] In this embodiment, the method for analyzing the environmental hazard coefficient is as follows: Centered on the early warning terminal, a risk analysis neighborhood is defined with a preset spatial radius. The risk level values ​​assigned by the dynamic risk distribution map within this neighborhood are extracted as a sample set. The mean of this sample set is calculated to characterize the baseline risk intensity of the early warning terminal. At the same time, the variance of this sample set is calculated to quantify the volatility of the risk distribution within the neighborhood. The mean and variance are coupled through a weighted fusion method to generate the environmental hazard coefficient of the early warning terminal.

[0041] Among them, taking the current early warning terminal as the center, a circular spatial range is defined as the risk analysis neighborhood based on the effective coverage distance of wireless communication when the early warning terminal is communicating normally. The selection of the radius needs to take into account both the integrity of the local risk representation and the computational efficiency. It can usually be set to be an integer multiple of the grid resolution of the dynamic risk distribution map to ensure that the neighborhood covers a sufficient number of grid cells.

[0042] When extracting the risk level values ​​assigned by the dynamic risk distribution map within the neighborhood as a sample set, the dynamic risk distribution map is based on a continuous grid generated by inverse distance weight interpolation. Each grid cell in the dynamic risk distribution map carries the risk level value of its location. When extracting the sample set, a circular neighborhood buffer is defined with the spatial coordinates of the target early warning terminal as the center and a preset spatial radius. All grid cells in the buffer are obtained through spatial query, and the risk level values ​​corresponding to these grid cells are extracted one by one. All the extracted risk level values ​​constitute the sample set used for calculating the environmental hazard coefficient of the early warning terminal.

[0043] The core purpose of this method for calculating environmental hazard coefficients is to transform early warning terminals from isolated risk points into comprehensive sensing units of the surrounding risk field situation. Its advantages include overcoming the limitations of judging hazard solely based on the terminal's own risk level, effectively eliminating the risk of misjudgment caused by instantaneous sensor anomalies or localized sporadic interference, and possessing the ability to perceive the spread trend of risk. By introducing a coupling mechanism of neighborhood mean and variance, this method achieves a dual characterization of the hazard response of early warning terminals. The mean quantitatively describes the baseline risk intensity of the terminal, i.e., the average expected severity of the environment, while the variance captures the dramatic fluctuations in the risk distribution within the neighborhood. If the terminal is located on the edge of the fire line or in a high-risk gradient zone, even if its own reading has not reached an extremely high threshold, its neighborhood variance will still be significantly higher, enabling the system to accurately identify terminals at the forefront of rapidly evolving disaster situations. In terms of weight allocation, the basis for setting the weights stems from the trade-off between the reliability of the indicators and the sensitivity of the scenarios. The mean, as a direct representation of the risk intensity, should occupy the dominant weight, while the variance, as a structural indicator reflecting the potential risk of sudden changes, is given a secondary weight but must have sufficient identification sensitivity. Accordingly, the initial weights of the two can be set to 0.7 and 0.3, respectively. The comprehensive environmental risk coefficient is obtained by weighted summation. The setting of the variance weight significantly amplifies the contribution value of areas with more drastic risk fluctuations, thereby ensuring that terminals in areas with steep risk changes are marked first.

[0044] Step 2: Obtain the public network signal status of all early warning terminals. Use early warning terminals with public network connections as network base stations and those without public network connections as out-of-connection early warning terminals. Calculate the environmental interference coefficient based on the multi-dimensional environmental information collected by the network base stations. Combine the environmental interference coefficient with the location distribution of the out-of-connection early warning terminals to determine the appropriate wireless communication wavelength. Construct multiple hop-by-hop relay pre-communication links for each out-of-connection early warning terminal.

[0045] This step quantifies the real-time impact of multi-dimensional environment on wireless signals into an environmental interference coefficient, and adaptively determines the wireless communication wavelength for searching for lost terminals based on this coefficient. This allows for the rapid establishment of multiple alternative pre-communication links for each signal-interrupted early warning terminal in complex and ever-changing disaster scenarios. This method overcomes the blindness of traditional solutions that rely solely on fixed transmission power or single signal strength for searching. Existing methods suffer from severe and dynamically changing wireless signal attenuation in environments such as high temperature, dense smoke, and high humidity. If conventional parameters are still used for searching, insufficient signal penetration or incorrect coverage estimation can easily lead to the failure to find actually connectable lost nodes. This invention, however, uses real-time environmental information such as smoke concentration, humidity, temperature, and particulate matter concentration collected by the base station. Through normalized weighted calculation, it accurately reflects the current signal propagation resistance and interference level of the environmental interference coefficient. Combined with the straight-line distance of the lost terminal, it reverse-engineers the wireless communication wavelength that can penetrate the current environment and establish a reliable connection. This gives the search process dynamic adaptability to the real physical channel, significantly improving the success rate of rediscovering and reconnecting to lost terminals after communication interruptions caused by disasters.

[0046] In this embodiment, the step of determining the wireless communication wavelength is as follows: Based on the humidity, temperature, and particulate matter concentration collected by the network base station, normalization was performed on each data point. Differential weights were assigned according to the degree of influence of each environmental factor on wireless signal attenuation, and the environmental interference coefficient was calculated by weighted summation. The straight-line distance is calculated based on the location of the network base station and the location of the loss-of-connection warning terminal. Combined with the environmental interference coefficient, the communication wavelength is adaptively optimized through a preset exponential function mapping model to determine the optimal working wavelength for stable wireless communication under the current distance and comprehensive environmental disturbances. The exponential function mapping model is as follows: in, For wireless communication wavelength, This is the proportionality coefficient. The straight-line distance. As an environmental sensitivity index, This represents the environmental interference coefficient.

[0047] This method quantifies the complex impact of disaster sites on wireless signals into a comprehensive environmental interference coefficient. and its communication distance Substitute them together into the constructed exponential function model In this system, the optimal wireless communication wavelength is adaptively determined. This breaks away from the mechanical mode of traditional emergency communication, which relies on fixed preset single wavelengths or simple signal strength tests. It deeply binds wavelength selection to the real-time dynamic changes of the physical environment, achieving intelligent adaptive optimization of communication parameters.

[0048] Furthermore, the proportionality coefficient As the basic scaling factor, the straight-line distance The determined physical scale is linearly mapped to the wavelength order. It inherently includes the compensation relationship between the system's desired free-space path loss reference and the antenna's effective aperture: under the same distance and environmental conditions, Increasing the wavelength directly leads to a proportional increase, meaning the system physically chooses a longer wavelength in exchange for a larger antenna receiving area and lower free-space attenuation; conversely... Reducing the wavelength shortens the wavelength, making it suitable for scenarios that require high speed and a clean environment. Essentially, it determines the baseline level for the entire dynamic wavelength adjustment range.

[0049] proportionality coefficient The value is set to 0.05 because this parameter is essentially a physical scaling baseline that maps the communication distance to the required wavelength, and its setting must simultaneously satisfy two hard constraints: diffraction coverage capability and terminal antenna portability; if If the value is too low, the calculated wavelength will be too short and the frequency too high at a typical emergency communication distance of several hundred meters, failing to generate a sufficient Fresnel zone radius to diffract the rubble or obstacles at the disaster site. If the value is too high, the wavelength required for long-distance communication will exceed ten meters, and portable early warning terminals will not be able to achieve the matching antenna physical size. Based on 0.05, the wavelength is calculated to be 15 meters when facing a communication distance of 300 meters under interference-free conditions. This is in the shortwave band that can provide strong diffraction penetration and is also engineering-feasible. When the distance is shortened to 100 meters, the wavelength is automatically shortened to 5 meters to improve the transmission rate. This dynamic range fully covers the diverse needs of emergency scenarios, from short-distance high-speed relay to long-distance penetrating coverage.

[0050] straight-line distance As a linear multiplicative factor, it reflects the free-space propagation loss that a signal must overcome to cross space. The greater the communication distance, the more severe the diffusion and attenuation of electromagnetic waves in free space; the path loss is proportional to the square of the distance. In this formula... Directly with wavelength The direct proportionality means that when the distance between the base station and the lost terminal doubles, the wavelength also doubles linearly. This is because longer wavelengths have lower free space attenuation during long-distance transmission. Increasing the wavelength can effectively compensate for the decrease in signal power caused by the increased distance, ensuring that the receiving end can still obtain a signal with sufficient strength.

[0051] Environmental Sensitivity Index As the controller for exponential amplification, it determines the environmental disturbance coefficient. The intensity and nonlinearity of the effect on wavelength. The value of reflects the system's sensitivity to harsh environments and the severity of punishment: The larger the value, This index term follows The more dramatic the increase, the more aggressive the wavelength response to environmental degradation, exhibiting strong nonlinear stretching. This ensures that the wavelength is rapidly extended to achieve maximum penetration under extreme hazardous conditions such as high humidity and high dust. The smaller the value, the more moderate the wavelength response to environmental changes, the higher the system's tolerance to environmental interference, and the smoother the wavelength adjustment.

[0052] Environmental Sensitivity Index The core purpose of setting it to 1.5 is to precisely control the acceleration of the system escaping towards longer wavelengths when the environment deteriorates. At this value of 1.5, when the environmental interference coefficient... When the index rises from 0 to 0.6, which represents moderate disaster, the index term... It increases approximately 2.46 times, with the wavelength increasing smoothly to match the combined attenuation caused by smoke and humidity, while when When the temperature continues to deteriorate to 1.0, representing an extreme state of dense smoke and high heat, the exponential term increases to approximately 4.48 times, and the wavelength requirement shows a significant nonlinear increase. This wavelength broadening of approximately 4.5 times is just enough to forcefully penetrate the dense smoke and saturated water vapor barrier through the low-frequency diffraction and anti-absorption characteristics of long waves, while avoiding the channel bandwidth being compressed to an extremely narrow range due to the infinite amplification of wavelength, and the complete loss of the ability to transmit key information such as early warning images and videos via multi-hop relay. Thus, an optimal balance is achieved between signal penetration and communication capacity, which conforms to the physical logic of disaster emergency response.

[0053] Environmental interference coefficient As a comprehensive input variable, it is a single quantitative indicator obtained by normalizing and weighting multimodal environmental factors such as humidity, temperature, and particulate matter concentration. The larger the value, the stronger the overall attenuation and interference effect of the current environment on electromagnetic waves. In the formula, Located within the exponential term, this design amplifies the impact of environmental degradation on wavelength exponentially: as K increases from 0 to 1, the wavelength λ will expand. This exponential growth rate is several times higher than that of linear growth. This exponential mapping relationship can truly reflect the nonlinear amplification and destructive effect on signal transmission caused by the superposition of multiple factors in a disaster environment. For example, when smoke and particulate matter coexist, the synergistic absorption and scattering of electromagnetic waves will far exceed the simple sum of the effects of a single factor. Therefore, the wavelength must be shifted to the long-wave direction by a greater extent to maintain basic communication.

[0054] In some embodiments, an increase in particulate matter concentration dramatically increases the absorption and scattering of electromagnetic waves by suspended particles in the air, especially Mie scattering, leading to rapid signal energy attenuation and decreased penetration. This forces the system to migrate to longer wavelengths to reduce scattering loss. High humidity alters the dielectric constant of air, exacerbating polarization shift and energy attenuation in signal transmission. Simultaneously, the resonant absorption of water molecules at specific high frequencies further deteriorates the signal-to-noise ratio. Therefore, particulate matter concentration has the most direct and significant impact on electromagnetic wave scattering and absorption, and thus receives the highest weight (0.55). Humidity is second, as it causes continuous dielectric loss by altering the dielectric constant of air and inducing resonant absorption of water molecules at specific frequencies, resulting in a weight of 0.25. Temperature changes directly alter air density and turbulence intensity, triggering multipath effects and random fading in wireless signals. Specifically, it primarily induces signal flicker and multipath effects by altering air turbulence and density gradients, and its initial weight is correspondingly lower at 0.2. This formula represents the comprehensive quantification of these multimodal effects. distance from the line The linear relationship is integrated to ensure that the determined wavelength is sufficient to overcome the total path loss caused by the current distance and environment, without sacrificing the communication rate due to excessive conservatism. This ensures that in extreme disasters, the wavelength that best matches the current physical conditions can quickly penetrate smoke, hot and humid air masses and dust barriers to accurately search for lost terminals. This lays a solid physical layer connection foundation for the subsequent construction of a highly resilient and long-lived pre-communication link pool and the final optimal networking.

[0055] The strategy for constructing pre-communication links is as follows: Based on a defined wireless communication wavelength, starting from a network base station, the system first searches for and establishes connections with lost-connection warning terminals within its direct wireless coverage area. Then, using the successfully connected lost-connection warning terminals as relay base stations, the system continues to search for and extend links to lost-connection warning terminals further away. This process establishes several pre-communication links consisting of single or multiple hops for each lost-connection warning terminal through hop-by-hop relay.

[0056] The core advantage of this strategy lies in transforming static, single-point coverage into dynamic, chain-like extension, exponentially increasing the recovery radius of emergency communications. When a disaster causes a large number of early warning terminals to lose contact, fixed network base stations cannot directly reach all remote nodes. However, through a hop-by-hop relay mechanism, terminals that successfully regain connection immediately become new search and rescue nodes, pushing the search signal deeper in a "snowballing" manner. This creates a redundant network of alternative paths around the base stations, finding a passable route for each lost terminal.

[0057] In some embodiments, within the warning area, the stable communication radius of network base station A, determined based on the current environmental interference coefficient and adaptive wavelength, is 400 meters. Within this 400-meter radius, network base station A directly searches for and successfully establishes connections with lost connection warning terminals B and C. At this point, lost connection warning terminals B and C are immediately assigned relay functions, acting as relay base stations, and continue searching within their respective 400-meter effective communication ranges. Relay base station B detects lost connection warning terminal D, located approximately 550 meters away and not directly covered by network base station A, in the extended direction. B then establishes a connection with D, forming a two-hop pre-communication link A→B→D. Simultaneously, relay base station C detects and successfully connects with lost connection warning terminal E, located approximately 700 meters away, forming a link A→C→E. Furthermore, if E, after successfully connecting, also acts as a relay base station and continues searching further afield, detecting and establishing a connection with lost connection terminal F, located approximately 1000 meters away, then a three-hop pre-communication link A→C→E→F is formed. During this process, if D happens to be at the edge of C's coverage area, the system will simultaneously generate two links: A→B→D and A→C→D. Similarly, if F is simultaneously covered by links extended by B through multi-hop coverage, there will also be multiple alternative paths. Ultimately, through this relay-style extension, network base station A expands from a single-hop limit of only 400 meters to include all lost terminals within a kilometer range into the pre-communication link network, and retains multiple optional communication links for each remote node, providing sufficient redundancy for subsequent node-state-based optimal networking.

[0058] Step 3: Based on the self-operation status information of each relay base station, calculate the remaining working prediction time, communication status score and network centrality of each relay base station.

[0059] This step introduces a discrete sliding window mechanism based on communication adjustment duration to dynamically and continuously quantify and analyze the terminal status, rather than relying on a single instantaneous snapshot. Existing solutions often directly read the remaining battery percentage as the basis for available time, ignoring the fact that power consumption changes dynamically with service load, resulting in serious distortion of remaining time prediction. This step, however, sets multiple sliding windows with communication adjustment durations and dynamically samples the average battery level and average power consumption within the window to calculate a predicted duration that is closer to the actual load conditions. This upgrades the judgment of terminal battery life from static estimation to dynamic tracking. When evaluating communication status, existing technologies often only use signal strength or current transmission rate as a benchmark, failing to consider the constraints of bandwidth occupancy on relay forwarding capabilities. This step uses the ratio of effective transmission rate to expected rate as the rate score and the ratio of remaining bandwidth to bandwidth limit as the bandwidth score, and then combines them into a communication status through weight allocation. This composite index can accurately identify terminals that appear to have full signal strength but whose bandwidth is heavily occupied and whose actual forwarding capability is exhausted, avoiding misselection as relay base stations and causing a sharp drop in the throughput of the entire link. In the overall solution of this patent, this step plays a crucial role in quantitative evaluation and decision support. It connects upwards to the pool of alternative pre-communication links established in step 2, and downwards provides precise numerical inputs of three items—communication status, network centrality, and predicted duration—for the link selection formula in step 5. This ensures that the calculation of evaluation coefficients no longer relies on fuzzy empirical judgments but is based on rigorous time-series data analysis. As a result, the final selected communication link is not only topologically connected but also possesses continuous and stable transmission capabilities composed of nodes with high endurance, high forwarding capacity, and high centrality, fundamentally improving the survival time and resilience of the emergency local area network.

[0060] In this embodiment, the analysis method for the prediction duration is as follows: The prediction duration is based on the remaining battery power and average power consumption in its own operating status information; By setting a communication adjustment duration, the battery level and power consumption in the self-operating status information are discretized to obtain the battery level and power consumption at each time interval. By setting N communication adjustment duration windows, the system slides along the time scale with one communication adjustment duration as the step size to calculate the average battery level and average power consumption within each window. The prediction duration is then determined based on the average battery level and average power consumption.

[0061] The power consumption of early warning terminals in disaster environments is not a constant value, but fluctuates dramatically with the dynamic scheduling of tasks such as data acquisition frequency, relay forwarding load, and early warning signal broadcasting. Estimating power consumption solely based on the remaining power and nominal power consumption at a given moment, or the global historical average, cannot reflect the actual energy consumption rate of the terminal under the current disaster situation, inevitably leading to a significant deviation between the predicted duration and the actual available time. This method introduces a discrete sliding window mechanism based on communication adjustment duration. Its ingenuity lies in dividing the continuously changing power and consumption time-series data into multiple fixed-duration observation windows, and calculating the average power and average consumption within each window by sliding along the time axis. This is equivalent to performing a dynamically weighted moving average smoothing filter on the terminal's energy consumption trend. The advantages of this approach are twofold: First, it eliminates short-term jitter interference caused by sensor sampling noise and instantaneous load spikes. For example, power consumption spikes caused by instantaneous high-power transmissions from the terminal will not be misjudged as a continuous trend, thus making the prediction results more stable and reliable. Second, the sliding window can continuously track the actual drift of terminal power consumption as the disaster evolves and communication tasks change. When the on-site risk intensifies, causing the terminal to frequently initiate high-power long-wave communication or intensively collect environmental data, the average power consumption within the window increases accordingly, and the prediction time is shortened accordingly. This gives the prediction an adaptive ability to follow the rhythm of the disaster and always reflects the latest power consumption status. Third, by setting the number of windows N, the time scale and conservatism of the prediction can be flexibly controlled. During the rapid escalation of risk, the number of windows can be reduced to strengthen the weight of recent data and achieve rapid response. During the stable phase, the number of windows can be increased to lengthen the observation period and further improve smoothness. This dynamic prediction method provides an accurate node survival time input for link selection in step 5 throughout the patent scheme. This allows the exponential decay term in the evaluation coefficient to truly reflect the risk of terminal failure due to energy depletion. It ensures that the system can eliminate nodes with insufficient remaining available time in advance when making network deployment decisions, and prioritize terminals with long battery life and stable status as relays, thereby significantly extending the life cycle and continuous service capability of the entire emergency communication link.

[0062] In this embodiment, the communication status is determined based on the effective transmission rate, bandwidth limit, and occupied bandwidth in the early warning terminal's own operating status information; By setting N communication adjustment duration windows, and sliding along the time scale with one communication adjustment duration as the step size, the average effective transmission rate and average remaining bandwidth are calculated based on the effective transmission rate, bandwidth limit, and occupied bandwidth within each window. The ratio of the calculated average effective transmission rate to the expected rate is used as the rate score; The bandwidth score is obtained by dividing the average remaining bandwidth by the bandwidth limit of the warning terminal. The communication status is obtained by summing the weights.

[0063] A single indicator cannot accurately characterize the true relaying capability of early warning terminals in emergency communications. Existing technologies, if relying solely on signal strength or instantaneous speed as evaluation criteria, are prone to misjudging terminals with full signal strength but heavily utilized bandwidth as high-quality relays, leading to the loss or delay of important early warning information due to congestion during the relay process. This method decomposes the communication status into two core dimensions: speed performance and bandwidth margin. It calculates speed scores and bandwidth scores separately, and then combines them into a single communication status through weighted allocation, thereby achieving a precise quantification of the terminal's relaying competence.

[0064] The application of the sliding window mechanism here greatly improves the time-series robustness of the evaluation. By setting N communication adjustment duration windows and sliding them along the time axis, the system averages the effective transmission rate and remaining bandwidth within each window. This has the advantage of smoothing out data jitter caused by sudden interference or instantaneous traffic spikes at disaster sites, preventing overly negative evaluations of the terminal status due to a sudden drop in rate or temporary bandwidth saturation, and making the evaluation results closer to the stable and continuous operation of the terminal. At the same time, the sliding window continuously tracks the latest data. When the terminal's load increases due to relay tasks and the remaining bandwidth gradually narrows, the window average can promptly reflect this deterioration trend, ensuring that the communication status is dynamically updated with changes in terminal load, providing timely decision-making basis for subsequent link optimization.

[0065] Furthermore, the effective transmission rate refers to the actual effective data throughput successfully transmitted by the terminal within a specific time interval, rather than the theoretical rate negotiated at the physical layer. It directly reflects the terminal's true ability to complete data transmission under current channel conditions and load. A higher rate indicates stronger terminal forwarding efficiency. The expected rate is a minimum guaranteed rate threshold preset according to the emergency warning service requirements, such as a baseline rate to ensure stable transmission of compressed video streams and sensor data on a single-hop link. The rate score is calculated by the ratio of the average effective transmission rate to the expected rate. When the ratio is greater than 1, it indicates that the terminal's actual forwarding capability exceeds the basic service requirements. A higher score indicates better terminal transmission efficiency. The bandwidth limit is the maximum transmission bandwidth allowed by the terminal's wireless interface or network access capability. It depends on the hardware specifications and spectrum resource configuration, representing the terminal's theoretical throughput limit. Occupied bandwidth refers to the total bandwidth consumed by the terminal currently carrying its own services and providing relay forwarding for other nodes. The average remaining bandwidth is obtained by subtracting the occupied bandwidth from the bandwidth limit and then averaging the results through a sliding window, representing the terminal's unused available forwarding capacity. The bandwidth score is obtained by dividing the average remaining bandwidth by the bandwidth limit, with a value between 0 and 1. A value close to 1 indicates that the terminal has ample idle forwarding space, while a value close to 0 indicates that the bandwidth is about to be exhausted and there is no capacity to accept new relay tasks. Finally, by weighted summing the rate score and bandwidth score, the communication status simultaneously considers the terminal's transmission efficiency and capacity margin, ensuring that only nodes with both high-speed processing capabilities and sufficient bandwidth space can obtain high scores in link selection. This naturally eliminates terminals with degraded performance or excessive load during network deployment, guaranteeing the data throughput and transmission reliability of the entire pre-communication link.

[0066] In this embodiment, the method for determining network centrality is as follows: Based on the public network signal status and wireless communication update status of each early warning terminal, it is determined whether the early warning terminal is in a state of being out of contact. The number of all out-of-contact early warning terminals within the communication range of the early warning terminal that is in normal communication is counted, and the center distance between each of them and the early warning terminal that is in normal communication is calculated. The reciprocal of the center distance is summed and multiplied by the number of out-of-contact early warning terminals to obtain a comprehensive network centrality.

[0067] The formula for calculating network centrality is: in, The network centrality score. This refers to the number of all lost early warning terminals within the communication range of a normally communicating early warning terminal. For the first The center distance between the lost-contact early warning terminal and the normally communicating early warning terminal. This is the distance adjustment coefficient.

[0068] This network centrality is not simply a count of how many other early warning terminals a single early warning terminal can communicate with, but rather a precise identification of its irreplaceable role in the communication link by quantifying the important function of a normal early warning terminal in the communication link.

[0069] The advantage of using this method to calculate network centrality lies in its integration of information from two dimensions—connection scale and spatial density—rather than simply counting the number of lost terminals. This allows for the accurate identification of truly irreplaceable key hub nodes within the local network. By using the product structure of the number of lost terminal warnings and the reciprocal of the distance, this method effectively distinguishes scenarios with similar apparent connection numbers but vastly different actual values. This results in warning terminals with closer and more concentrated lost terminals receiving higher centrality scores, preventing the overestimation of weak connections over long distances.

[0070] The explanation is based on the formula, and the parameters are... This represents the total number of lost communication warning terminals that a normally functioning warning terminal can cover within its effective communication range. A larger value indicates a greater potential scale for connection recovery for that warning terminal, and that its failure will create a wider communication blind spot. It is the fundamental metric for hub value. Parameters Indicates the first The smaller the center distance between a lost-connection warning terminal and a normal communication warning terminal, the closer the lost terminal is, the lower the signal attenuation, the higher the reliability of subsequent connection establishment, and the more scarce the available alternative relay nodes around the lost terminal, the higher the dependence on the normal terminal. The formula... Taking the reciprocal and then summing them up achieves a positive mapping from distance to contribution value; the closer the distance, the larger the reciprocal and the higher the contribution. The summation result comprehensively reflects the spatial density of all lost terminals. Parameters The distance adjustment coefficient (Nt) is used to control the magnitude of the overall centrality score. It can be flexibly scaled according to the sensitivity of distance in the actual scenario. For example, under conditions of strong environmental interference, Nt can be appropriately increased to highlight the advantages of nearby nodes. The larger the final calculated network centrality C, the more prominent and irreplaceable the hub position of the early warning terminal in the local network. The system will prioritize the protection and selection of these key nodes as relays based on this value.

[0071] Step 4: For each pre-communication link, determine the forwarding sequence number of each early warning terminal in the pre-communication link, couple the communication status score of the network base station and the relay base station, the network centrality and the environmental hazard coefficient, and obtain the contribution of each relay base station and the network base station in the pre-communication link by correcting the forwarding sequence number with the exponential decay function of the prediction duration.

[0072] The core objective of this step is to transform the static node evaluation into a forward-looking and location-aware "contribution" quantification mechanism through refined dynamic weight design, thereby selecting truly reliable communication nodes in real emergency scenarios. Its most direct benefit is effectively avoiding the short-sightedness of traditional relay selection, which "only considers the current state and ignores future attenuation." Its advantage lies in its precise simulation of the real-world law of relay base station performance deteriorating with time and distance under disaster conditions by introducing a forwarding sequence number and an exponential decay function for the predicted duration. Simply relying on the coupling of communication status scores, network centrality, and environmental hazard coefficients can only reflect the comprehensive value of the early warning terminal at the current moment, but cannot predict its performance throughout the entire link. The contribution calculation combines the predicted duration with the exponential decay function, causing nodes with shorter remaining working time to have an exponentially amplified attenuation penalty. This forces the system to automatically avoid "short-lived" nodes that are about to run out of power or may soon fail, fundamentally reducing the probability of link breakage. Meanwhile, by using forwarding sequence numbers for location weighting, the cost of information transmission over multiple hops is accurately reflected. Forwarding nodes closer to the information source bear greater responsibility, and their failures have a wider impact; therefore, their contribution is given higher location sensitivity. This coupled and then corrected strategy organically integrates environmental hazards, network importance, communication quality, and survivability into a unified contribution score. This ensures that the communication links ultimately selected are not only composed of the most core and secure nodes in space, but also possess the strongest continuous resilience in time, achieving a leap from "optimal connection" to "resilient survival" for emergency communication resources.

[0073] In this embodiment, the specific steps for selecting the communication link based on the evaluation coefficient are as follows: Analyze the contribution of each early warning terminal in each pre-communication link; The contribution is derived by multiplying the communication status, network centrality, and environmental risk coefficient, forming the basic value of the node's fundamental performance. An exponential decay function based on the forwarding sequence number and prediction duration is introduced for correction. The calculation formula is as follows: in, For the first in the pre-communication link The contribution of each early warning terminal For the first The communication status of each early warning terminal. For the first Network centrality of each early warning terminal For the first The environmental hazard coefficient of each early warning terminal For the first The serial number of each early warning terminal. For the first The prediction duration of each early warning terminal, This is the set attenuation control coefficient.

[0074] This method achieves precise quantification of each early warning terminal in the communication link by integrating node real-time performance, disaster risk, and link survival expectation into a dynamic contribution index. This reflects the first The comprehensive contribution capability of an early warning terminal in the current pre-communication link does not simply mean the strength of the signal, but rather the comprehensive capability score of the node as a relay to provide stable and continuous forwarding services for the entire link. The technical effect is that through this quantitative score, the system can accurately identify and prioritize nodes that have high communication quality, are located in key topology positions, are far from high-risk environments, and have sufficient battery life. At the same time, it automatically reduces the weight of high-risk, long-distance, and short-life nodes, giving the link optimization decision a predictive survivability assessment capability.

[0075] Representing the real-time data transmission efficiency and available bandwidth in the communication link, it directly determines the throughput capacity of the early warning terminal as a node in handling forwarding tasks, and is a positive fundamental factor contributing to the network centrality. The topological irreplaceability of the node in the local network is quantified, reflecting the importance of the early warning terminal as a node in the communication link. It is a structural guarantee for link connectivity and an indicator of environmental risk. by The formula incorporates the potential destruction risk of a disaster to a node, transforming it into a positive safety margin. The more dangerous the environment, the smaller this product factor, allowing physical threats to be directly quantified as a discount on contribution. (Forwarding sequence number) The sequence number indicates the number of hops a node has from the base station in the link. A higher sequence number means the data needs to traverse more nodes before being transmitted back, resulting in higher cumulative latency and failure risk. Therefore, it forms a negative correlation with the contribution in the exponential decay term; prediction duration. The expected lifespan of a node under remaining power and average power consumption is placed in the denominator, which significantly penalizes the contribution of short-life nodes while significantly reducing the degradation impact on long-life nodes; the degradation control coefficient. This is a global sensitivity parameter that adjusts the intensity of this punishment. Larger systems have lower tolerance for long connections and short battery life, and therefore adopt more conservative strategies.

[0076] Attenuation control coefficient The value is set to 0.5, which strikes a reasonable balance between link length penalty and node endurance tolerance: when the node's predicted duration... With serial number When they are close, that is, the first The expected survival time of a jump node is roughly equivalent to the number of jumps, and the exponential decay term is approximately equal to... The contribution rate is retained at 60%, reflecting a moderate penalty without direct elimination; if the node number is late and the battery is running low, such as the 3rd hop node which is predicted to have only 1 minute left, the decay rate drops sharply. The contribution is significantly compressed, effectively preventing endangered nodes at the end of long links from being selected as relays; if the node has sufficient endurance, such as the predicted duration of the 5th hop node being as long as 20 minutes, the attenuation term rises back to [a certain value]. The distance penalty is largely diluted by the long battery life. A value of 0.5 is neither too aggressive, causing all multi-hop links to be over-penalized and degenerate into single-hop direct connections, nor too conservative, allowing remote nodes on the verge of power failure to still receive a high score. This aligns with the dual requirements of path robustness and coverage in emergency scenarios.

[0077] Communication status Network centrality With contribution There is a linear positive correlation; an increase in either one will directly and proportionally increase the contribution, environmental safety margin. It is also positively correlated with the contribution level, but its change direction is different from that of the environmental risk coefficient. Conversely, the higher the environmental risk factor, the lower the contribution, thus constituting a reverse inhibition mechanism against high-risk nodes; forwarding sequence number It exhibits an exponential negative correlation with contribution, and the intensity of the penalty is affected by the prediction duration. Modulation, when When the distance penalty is large enough, it is diluted. Even nodes very close to the base station will experience a sharp drop in contribution due to impending power outages in extremely low-power conditions, constituting a combined penalty mechanism of distance and battery life; prediction duration It is exponentially positively correlated with the degree of contribution. The larger the absolute value of the exponential decay term, the smaller its contribution is to the product of the first three basic performance terms. Conversely, the basic performance is exponentially compressed, allowing the system to naturally avoid nodes that are about to be exhausted when networking.

[0078] Step 5: For each pre-communication link, evaluate the stability of the pre-communication link based on the contribution of the internal relay base station and the network base station, and construct a penalty term with the number of link nodes. Select the pre-communication link as the communication link based on the evaluation results.

[0079] This step constructs a link evaluation model that integrates multi-dimensional dynamic weighting and penalty mechanisms, elevating the selection of pre-communication links from traditional connectivity judgment to survivability prediction. This method abandons the routing logic based solely on the fewest hops or the strongest signal strength. Existing solutions often assign fixed nodes when selecting relay links, ignoring the significant differences between different terminals in communication status, prediction duration, communication topology importance, and environmental disaster risks. This results in selected links being highly susceptible to complete failure due to the unexpected failure of a low-efficiency or high-risk node. This step multiplies the communication status, network centrality, and safety margin derived from the environmental hazard coefficient using a contribution formula, closely linking node performance to disaster risk. The contribution of terminals in high-risk areas is directly reduced. Then, an exponential decay function based on forwarding sequence number and prediction duration is embedded in the contribution. As the node is farther from the base station and the remaining available time is shorter, its contribution decreases exponentially. This mathematically and precisely punishes the interruption risk caused by excessively long links or nodes nearing power failure. The link evaluation coefficient is multiplied by the total contribution and then divided by the square of the number of early warning terminals in the pre-communication link as a penalty. This design strongly suppresses the cumulative failure risk and end-to-end latency caused by excessive hop count, forcing the system to prioritize links with efficient nodes and short, robust paths under the same conditions. This step serves as the final decision fusion and network resilience anchoring mechanism. It integrates the environmental hazard coefficients generated in previous steps, the established redundant pre-communication link pool, and the calculated communication status, network centrality score, and predicted duration into a single quantitative decision equation. This transforms all the preliminary perception and assessment results into a comparable selection criterion, ensuring that the entire emergency local area network setup process no longer relies on any empirical rules. Instead, it automatically selects the optimal communication links through rigorous numerical optimization, which can proactively avoid high-risk areas, rely on key nodes with long endurance and strong forwarding capabilities, and have short, robust paths with high redundancy. This fundamentally maximizes the survival time and resilience of the emergency communication network under extreme disaster conditions.

[0080] In this embodiment, the method for selecting pre-communication links as communication links is as follows: Calculate the evaluation coefficient for each pre-communication link. When calculating the evaluation coefficient of the entire link, the product of the contributions of all early warning terminals in the link is multiplied by a penalty term for the number of early warning terminals in the pre-communication link. The specific formula is as follows: in, The evaluation coefficients for each pre-communication link, This represents the total length of the link. The pre-communication link with the highest evaluation coefficient is selected as the communication link.

[0081] This method constructs an evaluation model highly sensitive to the bottleneck effect and hop redundancy by multiplying the individual contributions of all nodes and incorporating a squared penalty term for the communication link length. The fundamental reason for using multiplication instead of summation is that data transmission in emergency communication links is a hop-by-hop serial relay process. The reliability of the entire link depends entirely on the weakest link, not the average level of the nodes. Multiplication ensures that if the contribution of any single node approaches zero, the entire link's evaluation value collapses, thus fully exposing the single-point failure risk in multi-hop relay to the numerical values. This ensures the system does not mistakenly select links where most nodes perform well but conceal a high-risk or low-endurance node. (Total link length) The reason why the penalty term is placed in the denominator and appears in square form is that in a disaster environment, each additional hop not only means a linear increase in end-to-end latency, but also that the probability of link failure accumulates non-linearly with the increase in the number of nodes. The square penalty term strongly suppresses excessively long links. Compared with a link composed of 3 high-contribution nodes, a link composed of 10 medium-contribution nodes has a much higher score due to the cumulative attenuation of up to 10 hops and the denominator penalty of up to 100 times. As a result, the system naturally tends to choose links with fewer hops and more robust paths. When selecting based on evaluation coefficients, the system calculates the evaluation coefficient EPG for each of the candidate pre-communication links generated in step three. The link with the highest score is selected first for actual network deployment. This selection mechanism means that the final communication link is not simply connected, but must achieve the optimal balance between the quality of individual nodes and the overall simplicity of the path. Even if there are multiple paths that can be connected, only the link whose communication status, centrality, safety margin, and endurance of each hop node have been rigorously verified by multiplication, and whose hop count is strictly constrained by the squared penalty effect, can stand out. This ensures from the decision-making source that the emergency local area network has the longest survival period and the strongest resilience under extreme conditions.

[0082] Table 1: Operating Status Information of Each Early Warning Terminal in the Pre-communication Link Based on the three sets of pre-communication links shown in Table 1, combined with Figure 2 It is known that after the disaster, network base stations BS1 and BS2 maintained network connection, and the system generated three alternative pre-communication links for the lost contact early warning terminal TA2 through searching.

[0083] The pre-communication link A consists of three nodes: BS1→TA1→TA2. Among them, the network base station BS1 has a communication status of 0.98, a network centrality of 0.95, an environmental risk coefficient of only 0.02, and a prediction duration of 38 hours. Its contribution, after exponential decay correction, is 0.900. The contribution of the early warning terminal TA1 as a relay base station is 0.466, and the contribution of the disconnection early warning terminal TA2 is 0.360. The evaluation coefficient is 0.0168 after dividing the product of the three nodes, 0.1510, by the square of the total link length of 3.

[0084] The pre-communication link B consists of 4 nodes: BS1→TA1→TA3→TA2. The early warning terminal TA3, as a relay base station, has a communication status of only 0.75, a network centrality of only 0.48, an environmental risk coefficient as high as 0.40, and a prediction duration of only 15 hours. Its basic performance is already low, and its position at number 3 exponentially increases the penalty of exponential decay, resulting in a contribution of only 0.195. The product of the entire link drops significantly to 0.0289. After dividing by the total link length of 4 squared 16, the evaluation coefficient is only 0.0018.

[0085] The pre-communication link C consists of 3 nodes: BS2→TA4→TA2. The network base station BS2 contributes 0.863. The early warning terminal TA4 has the highest communication status (0.92), the highest network centrality (0.72), the lowest environmental risk factor (0.10), and the longest prediction duration (50 hours) among the three links, with a contribution as high as 0.584. The product of these three values ​​(0.1815) divided by the square of the total link length (3) is evaluated as a coefficient of 0.0202.

[0086] The evaluation coefficients of the three links were 0.0202, which is greater than 0.0168 and greater than 0.0018. Pre-communication link C ranked first with a significant advantage, followed by pre-communication link A, and then pre-communication link B. Therefore, the system selected the path from network base station BS2 via relay base station TA4 to the target disconnection warning terminal TA2 as the final network link. The reason for the victory of pre-communication link C is that relay base station TA4 is superior to relay base station TA1 of pre-communication link A in all aspects. Higher communication status ensures data forwarding efficiency, higher network centrality means that the node is more irreplaceable in the network topology, lower environmental risk coefficient indicates that its location is less threatened by disasters, and longer prediction duration ensures that it will not lose connection due to power depletion during the mission cycle. These advantages, amplified by multiplication, resulted in a significant lead in the score.

[0087] Please see Figure 3 Figure 1 is a block diagram of an emergency communication system based on a multimodal early warning broadcasting terminal provided in this application embodiment. As shown in the figure, an emergency communication system based on a multimodal early warning broadcasting terminal includes...

[0088] The data acquisition and analysis module is used to collect the location information, multi-dimensional environmental information of the location, and the operating status information of each early warning terminal deployed in the area to be warned. Based on the multi-dimensional environmental information of each early warning terminal, environmental anomalies are identified. By setting the communication adjustment time, a dynamic risk distribution map of the area to be warned is constructed. The environmental risk coefficient of each early warning terminal is evaluated based on the dynamic risk distribution map. The link construction module is used to obtain the public network signal status of all early warning terminals, use early warning terminals with public network connection as network base stations, and use those without public network connection as lost early warning terminals. The module calculates the environmental interference coefficient based on the multi-dimensional environmental information collected by the network base stations, and determines the appropriate wireless communication wavelength by combining the environmental interference coefficient with the location distribution of the lost early warning terminals. Multiple hop-by-hop relay pre-communication links are constructed for each lost early warning terminal. The terminal analysis module is used to calculate the remaining working prediction time, communication status score and network centrality of each relay base station based on its own operating status information. The terminal coupling module is used to determine the forwarding sequence number of each early warning terminal in each pre-communication link, couple the communication status score, network centrality and environmental hazard coefficient of the network base station and the relay base station, and obtain the contribution of each relay base station and the network base station in the pre-communication link by correcting the forwarding sequence number and the exponential decay function of the prediction duration. The link selection module evaluates the stability of each pre-communication link based on the contribution of the internal relay base station and the network base station, and constructs a penalty term with the number of link nodes. Based on the evaluation results, the pre-communication link is selected as the communication link.

[0089] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0090] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0091] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. An emergency communication method based on a multimodal early warning broadcasting terminal, characterized in that, include: The system collects location information, multi-dimensional environmental information of the location, and operational status information of each early warning terminal deployed in the area to be warned. Based on the multi-dimensional environmental information of each early warning terminal, it identifies environmental anomalies. By setting the communication adjustment duration, it constructs a dynamic risk distribution map of the area to be warned and assesses the environmental hazard coefficient of each early warning terminal based on the dynamic risk distribution map. The system obtains the public network signal status of all early warning terminals, designates early warning terminals with public network connections as network base stations, and those without public network connections as out-of-connection early warning terminals. It calculates the environmental interference coefficient based on the multi-dimensional environmental information collected by the network base stations, and determines the appropriate wireless communication wavelength by combining the environmental interference coefficient with the location distribution of out-of-connection early warning terminals. It then constructs multiple hop-by-hop relay pre-communication links for each out-of-connection early warning terminal. Based on the operational status information of each relay base station, the remaining predicted working time, communication status score and network centrality of each relay base station are calculated. For each pre-communication link, the forwarding sequence number of each early warning terminal in the pre-communication link is determined, and the communication status score, network centrality and environmental risk coefficient of the network base station and relay base station are coupled. The contribution of each relay base station and network base station in the pre-communication link is obtained by correcting the forwarding sequence number and the exponential decay function of the prediction duration. For each pre-communication link, the stability of the pre-communication link is evaluated based on the contribution of the internal relay base station and the network base station, and a penalty term is constructed with the number of link nodes. The pre-communication link is selected as the communication link based on the evaluation results.

2. The method according to claim 1, characterized in that, The multi-dimensional environmental information includes temperature, humidity, and particulate matter concentration; Its own operating status information includes wireless communication wavelength, battery level, power consumption, effective transmission rate, bandwidth limit, and occupied bandwidth; The method for identifying environmental anomalies is as follows: For each type of environmental information, a warning threshold and a warning time threshold are set for each environmental type. When each type of environmental information exceeds the corresponding warning threshold and warning time threshold, it is identified as an environmental anomaly.

3. The method according to claim 2, characterized in that, The method for constructing the dynamic risk distribution map is as follows: For each environmental type identified as an environmental anomaly, a warning level is assigned. The warning levels for each environmental type are summed to calculate the risk level of the warning terminal. Based on the risk level and location distribution of each warning terminal, the risk level of any location within the warning area is calculated using an inverse distance weighted interpolation method, forming a continuous risk distribution map. The risk distribution map is updated by adjusting the communication interval as set, thereby generating a dynamic risk distribution map.

4. The method according to claim 1, characterized in that, The method for analyzing the environmental hazard coefficient is as follows: Centered on the early warning terminal, a risk analysis neighborhood is defined with a preset spatial radius. The risk level values ​​assigned by the dynamic risk distribution map within this neighborhood are extracted as a sample set. The mean of this sample set is calculated to characterize the baseline risk intensity of the early warning terminal. At the same time, the variance of this sample set is calculated to quantify the volatility of the risk distribution within the neighborhood. The mean and variance are coupled through a weighted fusion method to generate the environmental hazard coefficient of the early warning terminal.

5. The method according to claim 1, characterized in that, The steps for determining the wireless communication wavelength are as follows: Based on the humidity, temperature, and particulate matter concentration collected by the network base station, normalization was performed on each data point. Differential weights were assigned according to the degree of influence of each environmental factor on wireless signal attenuation, and the environmental interference coefficient was calculated by weighted summation. The straight-line distance is calculated based on the location of the network base station and the location of the loss-of-connection warning terminal. Combined with the environmental interference coefficient, the communication wavelength is adaptively optimized through a preset exponential function mapping model to determine the optimal working wavelength for stable wireless communication under the current distance and comprehensive environmental disturbances. The exponential function mapping model is as follows: in, For wireless communication wavelength, This is the proportionality coefficient. The straight-line distance. As an environmental sensitivity index, Environmental interference coefficient; The strategy for constructing pre-communication links is as follows: Based on a defined wireless communication wavelength, starting from a network base station, the system first searches for and establishes connections with lost-connection warning terminals within its direct wireless coverage area. Then, using the successfully connected lost-connection warning terminals as relay base stations, the system continues to search for and extend links to lost-connection warning terminals further away. This process establishes several pre-communication links consisting of single or multiple hops for each lost-connection warning terminal through hop-by-hop relay.

6. The method according to claim 2, characterized in that, The analysis method for the predicted duration is as follows: The prediction duration is based on the remaining battery power and average power consumption in its own operating status information; By setting a communication adjustment duration, the battery level and power consumption in the self-operating status information are discretized to obtain the battery level and power consumption at each time interval. By setting N communication adjustment duration windows, the system slides along the time scale with one communication adjustment duration as the step size to calculate the average battery level and average power consumption within each window. The prediction duration is then determined based on the average battery level and average power consumption.

7. The method according to claim 2, characterized in that, The communication status is determined based on the effective transmission rate, bandwidth limit, and occupied bandwidth in the early warning terminal's own operating status information. By setting N communication adjustment duration windows, and sliding along the time scale with one communication adjustment duration as the step size, the average effective transmission rate and average remaining bandwidth are calculated based on the effective transmission rate, bandwidth limit, and occupied bandwidth within each window. The ratio of the calculated average effective transmission rate to the expected rate is used as the rate score; The bandwidth score is obtained by dividing the average remaining bandwidth by the bandwidth limit of the warning terminal. The communication status is obtained by summing the weights.

8. The method according to claim 1, characterized in that, The method for determining network centrality is as follows: Based on the public network signal status and wireless communication update status of each early warning terminal, early warning terminals that have stopped updating both public network signal status and wireless communication update status are determined to be in a disconnected state. A comprehensive network centrality score is obtained by weighted summing of the number of disconnected terminals within the communication range of the early warning terminal with normal communication and the distance of each of them to the early warning terminal with normal communication.

9. The method according to claim 1, characterized in that, The method for selecting pre-communication links as communication links is as follows: Calculate the evaluation coefficient for each pre-communication link. When calculating the evaluation coefficient of the entire link, the product of the contributions of all early warning terminals in the link is multiplied by a penalty term for the number of early warning terminals in the pre-communication link. The pre-communication link with the highest evaluation coefficient is selected as the communication link.

10. An emergency communication system based on a multimodal early warning broadcasting terminal, characterized in that, The system is used to execute an emergency communication method based on a multimodal early warning broadcasting terminal as described in any one of claims 1 to 9, comprising: The data acquisition and analysis module is used to collect the location information, multi-dimensional environmental information of the location, and the operating status information of each early warning terminal deployed in the area to be warned. Based on the multi-dimensional environmental information of each early warning terminal, environmental anomalies are identified. By setting the communication adjustment time, a dynamic risk distribution map of the area to be warned is constructed. The environmental risk coefficient of each early warning terminal is evaluated based on the dynamic risk distribution map. The link construction module is used to obtain the public network signal status of all early warning terminals, use early warning terminals with public network connection as network base stations, and use those without public network connection as lost early warning terminals. The module calculates the environmental interference coefficient based on the multi-dimensional environmental information collected by the network base stations, and determines the appropriate wireless communication wavelength by combining the environmental interference coefficient with the location distribution of the lost early warning terminals. Multiple hop-by-hop relay pre-communication links are constructed for each lost early warning terminal. The terminal analysis module is used to calculate the remaining working prediction time, communication status score and network centrality of each relay base station based on its own operating status information. The terminal coupling module is used to determine the forwarding sequence number of each early warning terminal in each pre-communication link, couple the communication status score, network centrality and environmental hazard coefficient of the network base station and the relay base station, and obtain the contribution of each relay base station and the network base station in the pre-communication link by correcting the forwarding sequence number and the exponential decay function of the prediction duration. The link selection module evaluates the stability of each pre-communication link based on the contribution of the internal relay base station and the network base station, and constructs a penalty term with the number of link nodes. Based on the evaluation results, the pre-communication link is selected as the communication link.

Images