A method and system for processing SMS sending based on scene adaptation

Abstract

This application provides a method and system for processing SMS sending based on scene adaptation. Among them, the method includes: constructing a three-dimensional communication feature map through building information models and infrared thermal imaging data, predicting SMS request distribution by combining signal attenuation field analysis, integrating cross-operator network quality data to form a fused communication quality field, and finally intelligently scheduling SMS sending according to network status and service requirements. This method realizes precise SMS routing optimization based on real-time communication quality and crowd distribution in complex building environments. The technical solution provided by this application can significantly improve the SMS delivery rate and timeliness in complex building environments, especially the communication reliability is significantly improved in crowded areas and signal blind spots.

Description

Technical Field

[0001] This application relates to the technical field of short message distribution processing, and particularly to a short message distribution processing method and system based on scenario adaptation. Background Art

[0002] In the modern building environment, the short message distribution service faces complex communication challenges. Especially in large buildings (such as shopping malls, hospitals, office buildings), due to the influence of physical obstacles such as wall materials, metal structures, and elevator shafts, the signal propagation loss is serious, and communication blind spots are easily formed. In addition, the dynamic change of crowd density (such as the gathering of people during peak hours) further exacerbates the instability of short message transmission. Therefore, there is an urgent need for a short message distribution processing method that can combine building structure characteristics and real-time crowd flow data to optimize the signal propagation path, improve the short message transmission efficiency, and ensure communication reliability in emergency situations.

[0003] Currently, traditional short message distribution methods mainly rely on fixed communication base stations and simple signal strength detection technologies. These methods usually cannot effectively cope with the signal attenuation problem in complex building environments, especially in areas with high metal density or crowded areas, where signal interruption or delay is likely to occur. In addition, the existing technologies lack the comprehensive analysis ability of building structure characteristics and dynamic changes in human flow, resulting in low short message distribution efficiency and inability to meet the communication requirements in high-concurrency scenarios.

[0004] Existing technical solutions usually adopt static communication models and cannot adapt to the changes of the building environment in real time. For example, during peak hours, the gathering of people may cause signal overload in local areas, while the existing systems cannot dynamically adjust the bandwidth allocation or forwarding strategy. In addition, the existing solutions lack accurate analysis of building structures (such as wall materials, elevator shafts), resulting in inaccurate prediction of the signal attenuation field and inability to effectively eliminate communication blind spots. Therefore, there is an urgent need for a short message distribution processing method that can combine building information models, real-time crowd flow data, and dynamic communication optimization. Summary of the Invention

[0005] The embodiments of this application provide a short message distribution processing method and system based on scenario adaptation to solve the problems of low efficiency and poor reliability in short message distribution processing in the prior art.

[0006] In a first aspect, the embodiments of this application provide a short message distribution processing method based on scenario adaptation, including:

[0007] Obtaining spatial topology data by parsing the building information model, and combining with the personnel residence heat map collected by the infrared thermal imaging array to generate a three-dimensional communication feature map including signal propagation loss weights and crowd density gradients;

[0008] Based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map, the difference in the absorption rate of the wall material is detected to construct a three-dimensional signal attenuation field including the reflection interference of the elevator shaft;

[0009] According to the communication blind area boundary data in the three-dimensional signal attenuation field, combined with the position coordinates of the fire channel, the temporal and spatial distribution of the SMS request volume in each partition is predicted;

[0010] Accessing the Internet SMS gateway cluster, collecting base station signaling data and user terminal online status to generate a thermal map of cross-operator network quality, and mapping the thermal map to the three-dimensional communication feature map to form a fused communication quality field;

[0011] Based on the spatiotemporal distribution of the SMS request volume in each partition and the backlog of the gateway cache queue, combined with the network congestion hotspots in the fusion communication quality field, a message forwarding rule sorted by priority is generated; and the SMS sending operation is executed according to the message forwarding rule.

[0012] Optionally, the access to the Internet SMS gateway cluster generates a thermal map of the network quality across operators by collecting base station signaling plane data and user terminal online status, including:

[0013] Through the multi-operator protocol interface, each SMS gateway cluster is synchronously connected to obtain the signaling plane data of each base station in real time. The signaling plane data includes the control signaling interaction record between the base station and the user terminal, the terminal registration status and the signaling response delay;

[0014] At the same time, the terminal measurement data reported by the user terminal is collected, including the reference signal received power, signal interference and noise ratio, and terminal moving speed;

[0015] The signaling plane data collected in the same time window and the terminal measurement data are temporally and spatially matched according to the base station coverage area, abnormal terminal data are eliminated, and the valid data point set is retained. Based on the valid data point set, the signaling response delay distribution and signal strength distribution of each terminal within the coverage area of ​​each base station are extracted;

[0016] Divide the base station coverage area into a number of grid cells according to the terminal location distribution density, each grid cell contains the measurement data of at least one terminal;

[0017] The signaling response delay and signal strength in each grid cell are weightedly calculated to obtain the network quality score of the grid cell. The network quality scores of adjacent grid cells are smoothed using a spatial interpolation method to generate a thermal map of cross-operator network quality.

[0018] Optionally, mapping the thermal layer to the three-dimensional communication feature map to form a fused communication quality field includes: mapping the central coordinates of each grid unit in the network quality thermal layer to the corresponding spatial positions in the three-dimensional communication feature map;

[0019] Adjusting the contribution coefficient of the network quality score according to the population density gradient and signal propagation loss weight at the corresponding spatial positions in the three-dimensional communication feature map to update the network quality score;

[0020] Overlaying the updated network quality score with the communication parameters in the three-dimensional communication feature map to form a fused communication quality field.

[0021] Optionally, generating a message forwarding rule sorted by priority based on the spatio-temporal distribution of the SMS request volume in each partition and the backlog volume in the gateway cache queue, in combination with the network congestion hotspots in the fused communication quality field, includes:

[0022] Fusing the coordinates of the high-density regions and the group movement direction vectors in the spatio-temporal distribution of the SMS request volume in each partition, and calculating the matching degree between the movement direction clustering result and the historical request volume growth rate by associating the signal strength attenuation rate of the network congestion hotspots to generate an emergency channel weight coefficient;

[0023] Based on the peak period characteristics in the gateway queue backlog volume, in combination with the network congestion hotspots and the signal attenuation gradient of adjacent access points, generating a cross-floor diversion ratio by matching the inflection points of the attenuation gradient and the physical topology distance;

[0024] According to the geometric boundary data of the weak signal area, extracting the idle channel coverage range of adjacent access points within the network congestion hotspots, and generating a relay forwarding priority sequence by analyzing the correlation strength between the access point response delay and the signal compensation requirement;

[0025] Performing multi-dimensional fusion on the emergency channel weight coefficient, the cross-floor diversion ratio, and the relay forwarding priority sequence to construct a forwarding rule trigger matrix, and the forwarding rule trigger matrix is dynamically updated by associating the change rate of the group movement direction vector and the growth rate of the gateway cache queue backlog volume;

[0026] Based on the critical parameters of the forwarding rule trigger matrix, simulating message path conflicts in the digital twin environment, and generating a message forwarding rule sorted by priority by reversely correcting the matching degree threshold and the signal compensation parameters.

[0027] Optionally, the performing multi-dimensional fusion on the emergency channel weight coefficient, the cross-floor diversion ratio, and the relay forwarding priority sequence to construct a forwarding rule trigger matrix includes:

[0028] Fuse the emergency channel weight coefficient with the real-time load and historical traffic data of each channel, generate a basic constraint value in combination with the channel length, and perform dynamic compensation by superimposing the obstacle density to obtain a channel quality evaluation index;

[0029] Based on the floor escape node density, couple the capacity threshold of the escape ladder and the channel quality evaluation index to generate a diversion gradient coefficient for the cross-floor diversion ratio, and fuse the building load limit to output a diversion ratio reference value;

[0030] The relay forwarding priority sequence constructs a node coverage map through the multipath effect of wireless signals, fuses the diversion gradient coefficient to generate a three-dimensional permeability surface, extracts the curvature characteristics of the three-dimensional permeability surface, and generates a priority sorting linked list;

[0031] Input the channel quality evaluation index, diversion ratio reference value, and priority sorting linked list into the building digital twin, extract implicit association rules through tensor decomposition and orthogonally reorganize them to construct a forwarding rule trigger matrix.

[0032] Optionally, the step of generating a diversion gradient coefficient by coupling the capacity threshold of the escape ladder and the channel quality evaluation index based on the floor escape node density for the cross-floor diversion ratio includes:

[0033] Generate a node coverage intensity value based on the floor escape node density according to the visible distance and the number of path branches, trigger connectivity compensation through the visible distance, and correct the topological intensity attenuation by the number of path branches;

[0034] Based on the capacity threshold of the escape ladder and the smoke diffusion rate, generate a real-time traffic attenuation coefficient through the asymmetric superposition of a static capacity baseline and a dynamic attenuation factor, and simultaneously introduce the node coverage intensity value;

[0035] The channel quality evaluation index and the real-time traffic attenuation coefficient are fused through a horizontal weight base and a vertical emergency weight, and tensor dot product and topological filtering are performed at the floor connection node to generate a diversion gradient coefficient.

[0036] Optionally, the step of detecting the difference in wall material absorption rate based on the coordinates of the metal-dense area in the three-dimensional communication feature map and constructing a three-dimensional signal attenuation field including elevator shaft reflection interference includes:

[0037] The coordinates of the metal-dense area in the three-dimensional communication feature map dynamically generate a metal interference intensity field through the distribution density of metal components and the spatial correlation of signal reflection;

[0038] Generate an absorption rate reference value through the mapping between the material type and the electromagnetic wave penetration loss, introduce the variation coefficients of the wall moisture content and thickness, and combine the environmental humidity data and the frequency-domain convolution of the reference value to generate an absorption rate dynamic correction parameter;

[0039] The cavity resonance frequency is obtained by analyzing the geometric dimensions of the shaft cavity. The reflection path of the metal shaft is combined with the metal interference intensity field for attenuation compensation, and the compensation result is phase-aligned with the resonance frequency to obtain the reflected interference intensity field.

[0040] By fusing the metal interference intensity field, the dynamic correction parameter of the absorption rate, and the reflected interference intensity field, the metal interference gradient and the absorption rate parameter are polarization-weighted and embedded into the phase constraint of the reflected interference to generate a three-dimensional signal attenuation field.

[0041] Optionally, the method for obtaining the cavity resonance frequency by analyzing the geometric dimensions of the shaft cavity, combining the reflection path of the metal shaft with the metal interference intensity field for attenuation compensation, and phase-aligning the compensation result with the resonance frequency to obtain the reflected interference intensity field includes:

[0042] Combined with the rigid node constraint standing wave mode analysis of the building vertical channel, the geometric dimensions of the shaft cavity are analyzed through the calculation of the ratio of the cross-sectional perimeter to the height to generate the cavity resonance frequency.

[0043] The reflection path of the metal shaft is attenuated and compensated by the multipath reflection characteristics of the metal interference intensity field. Combining the Fresnel reflection coefficient and the geometric projection of the signal incident angle, the harmonic order of the cavity resonance frequency is matched to generate the reflection path attenuation compensation parameter.

[0044] The radial gradient of the reflection path attenuation compensation parameter is helically convolved with the phase of the cavity resonance frequency, and after truncating the standing wave energy threshold, the reflected interference intensity field is generated.

[0045] Optionally, the method for predicting the spatio-temporal distribution of the short message request volume in each zone according to the communication blind zone boundary data in the three-dimensional signal attenuation field and combining the position coordinates of the fire channel includes:

[0046] The communication blind zone boundary data is discretized into three-dimensional grid cells, and the signal attenuation coefficient of the three-dimensional grid cells and their three-dimensional Manhattan distance to the nearest fire channel entrance are calculated to generate an accessibility matrix.

[0047] A dynamic path network is constructed based on the position coordinates of the fire channel, time windows are divided and associated with the historical short message request volume data to generate the request volume change baseline of the three-dimensional grid cells.

[0048] The three-dimensional grid cells and the dynamic path network are spatially superimposed, the shortest connected path is calculated and the wall penetration loss compensation value is fused to construct a path loss correction model.

[0049] The signal attenuation coefficient is decomposed into a periodic component and a residual component in time series, and combined with the path loss correction model to generate the real-time signal coverage quality index of the three-dimensional grid cells.

[0050] Fuse the reachability matrix, real-time signal coverage quality index, and request volume change baseline into a multi-dimensional feature tensor, generate a spatio-temporal coupling feature map through a channel attention mechanism, and output the predicted value of the short message request volume for the future time slice.

[0051] In a second aspect, the present application provides a short message sending processing system based on scenario adaptation, including:

[0052] An acquisition module, which obtains spatial topology data by parsing a building information model, and combines the personnel residence heat map collected by an infrared thermal imaging array to generate a three-dimensional communication feature map including signal propagation loss weights and crowd density gradients;

[0053] A construction module, based on the coordinates of metal-dense areas in the three-dimensional communication feature map, detects the difference in wall material absorption rates, and constructs a three-dimensional signal attenuation field including elevator shaft reflection interference;

[0054] A prediction module, according to the communication blind area boundary data in the three-dimensional signal attenuation field, combines the fire channel position coordinates to predict the spatio-temporal distribution of the short message request volume in each area;

[0055] A generation module, used to access the Internet short message gateway cluster, generate a heat map layer of cross-operator network quality by collecting base station signaling plane data and user terminal online status, and map the heat map layer to the three-dimensional communication feature map to form a fused communication quality field; based on the spatio-temporal distribution of the short message request volume in each area and the backlog volume of the gateway cache queue, combine the network congestion hot spots in the fused communication quality field to generate a message forwarding rule sorted by priority;

[0056] A processing module, used to perform a short message sending operation according to the message forwarding rule.

[0057] In the embodiments of the present application, spatial topology data is obtained by parsing a building information model, and a three-dimensional communication feature map including signal propagation loss weights and crowd density gradients is generated by combining a thermal map of personnel occupancy collected by an infrared thermal imaging array, which can accurately reflect the signal propagation characteristics and crowd aggregation degree in different regions, providing a data basis for subsequent communication optimization. Based on the coordinates of metal-dense regions in the three-dimensional communication feature map, the difference in the absorption rate of wall materials is detected, and a three-dimensional signal attenuation field including reflection interference from elevator shafts is constructed, which can identify key influencing factors such as metal structures, wall materials, and elevator shafts in the building, accurately calculate signal attenuation and reflection interference, form a signal propagation model more in line with the real environment, and improve the accuracy of communication quality prediction. According to the communication blind area boundary data in the three-dimensional signal attenuation field and the position coordinates of fire channels, the spatio-temporal distribution of short message request volumes in each partition is predicted, which can combine signal blind areas and personnel flow paths (such as fire channels) to dynamically predict the short message service requirements in different regions at different times, providing a precise spatio-temporal distribution reference for resource scheduling. Connect to the Internet short message gateway cluster, generate a thermal layer of cross-operator network quality by collecting base station signaling plane data and user terminal online status, and map the thermal layer to the three-dimensional communication feature map to form a fusion communication quality field, which can integrate multi-operator network status data, evaluate the communication quality of each region in real time, and fuse with the three-dimensional environment model to form a global communication situation awareness, ensuring that the short message sending strategy adapts to the actual network conditions. Based on the spatio-temporal distribution of short message request volumes in each partition and the backlog volume in the gateway cache queue, combined with the network congestion hotspots in the fusion communication quality field, a message forwarding rule sorted by priority is generated; the short message sending operation is performed according to the message forwarding rule, which can comprehensively consider service requirements, network load, and communication quality, dynamically optimize the short message sending priority, ensure the efficient delivery of short messages in high-demand regions and regions with good network conditions, and at the same time relieve the transmission pressure in congested regions and improve the overall communication efficiency.

[0058] The technical solution of the present application significantly improves the reliability and timeliness of short message transmission in complex building environments through multi-dimensional data fusion and dynamic optimization, especially performing outstandingly in crowded areas, signal blind areas, and network congestion scenarios.

[0059] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0061] Figure 1 The flowchart of a method for processing SMS sending based on scene adaptation provided by this application is shown;

[0062] Figure 2 The schematic structural diagram of a system for processing SMS sending based on scene adaptation provided by this application is shown. Detailed implementation manners

[0063] In order to enable those skilled in the art to better understand the solutions of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application.

[0064] In some processes described in the specification, claims and above-mentioned drawings of this application, multiple operations that appear in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order in which they appear in this article or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish each different operation, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., do not represent a sequence, and do not limit that "first" and "second" are of different types.

[0065] R & D personnel considered that in the terminal SMS sending service, the problem of communication efficiency in complex building environments is usually faced. Traditional SMS sending methods are difficult to cope with multiple challenges such as signal attenuation, crowd gathering, and emergency communication. To solve these problems, in-depth analysis of building structure characteristics, signal propagation laws, and dynamic changes in human flow was carried out, and a method for processing SMS sending based on scenes was proposed. This solution constructs a three-dimensional communication feature map through building information models and infrared thermal imaging data, combines signal attenuation field analysis to predict SMS request distribution, and integrates cross-operator network quality data to form a fused communication quality field. Finally, it intelligently schedules SMS sending according to network status and service requirements. This method realizes precise SMS routing optimization based on real-time communication quality and crowd distribution in complex building environments.

[0066] The technical solutions of this application can be mainly applied to the following technical scenarios:

[0067] Large commercial complexes: In large commercial complexes such as shopping malls and shopping centers, the crowd density is high and unevenly distributed, which is likely to cause local signal overload or communication blind spots. This solution dynamically adjusts the SMS sending strategy by real-time analyzing the crowd heat map and building structure characteristics to ensure smooth communication.

[0068] Hospitals and medical institutions: The internal structure of hospitals is complex, and metal equipment and wall materials cause serious interference to signal propagation. This solution constructs a three-dimensional signal attenuation field to accurately predict communication blind spots, and optimizes the SMS sending path in combination with the location of emergency channels to ensure the reliability of medical communication.

[0069] Office buildings and industrial parks: Office buildings and industrial parks usually contain a large number of metal structures and elevator shafts, which are prone to signal reflection and interference. This solution analyzes the building information model to optimize the signal propagation path and improve the SMS sending efficiency.

[0070] Transportation hubs and airports: In transportation hubs such as airports and railway stations, the flow of people is intensive and mobile, and the communication demand fluctuates greatly. This solution dynamically allocates bandwidth resources by predicting the spatio-temporal distribution of SMS requests to ensure communication stability in high-concurrency scenarios.

[0071] Emergency communication scenarios: In natural disasters or public safety incidents, the stability of the communication network is crucial. This solution combines the location of fire channels and the bandwidth allocation strategy of emergency channels to ensure smooth communication in emergencies.

[0072] In the process of developing this solution, starting from the actual needs of the terminal SMS sending service, the communication pain points in complex building environments were analyzed. By combining the building information model (BIM) and infrared thermal imaging technology, a three-dimensional communication feature map was constructed to accurately depict the spatial distribution of signal propagation loss and crowd density. Further, digital twin technology was introduced to simulate the communication performance in scenarios of crowd gathering, and the SMS sending efficiency was optimized by dynamically adjusting the forwarding rules. This process not only solves the communication blind spot problem that cannot be handled by traditional methods, but also provides reliable technical support for emergency communication scenarios.

[0073] The technical solution of this application realizes the precise optimization of the SMS sending process by combining the building information model, infrared thermal imaging technology, and digital twin technology. This solution is applicable to large commercial complexes, hospitals, office buildings, transportation hubs, and emergency communication scenarios, and can effectively improve the SMS transmission efficiency and ensure the stability and reliability of communication. By real-time analyzing the building structure characteristics and dynamic changes in the flow of people, this solution provides an innovative solution for the terminal SMS sending service.

[0074] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of this application.

[0075] Figure 1The figure below is a flowchart of a method for processing SMS sending based on scenario adaptation provided by an embodiment of the present application. As Figure 1 shown, the method includes:

[0076] 101. Obtain spatial topology data by parsing a building information model, and combine it with the personnel occupancy heat map collected by an infrared thermal imaging array to generate a three-dimensional communication feature map containing signal propagation loss weights and crowd density gradients;

[0077] In this step, the building information model is a virtual building model created by digital means, used to provide accurate spatial structure and material information.

[0078] The spatial topology data is obtained by parsing the building information model, including building structure information, used to construct an accurate indoor map and provide a basis for subsequent analysis.

[0079] The personnel occupancy heat map is collected by an infrared thermal imaging array, showing the crowd density distribution in different areas and helping to identify areas with high pedestrian flow density.

[0080] The signal propagation loss weight reflects the influence degree of building materials on wireless signals, calculated according to material characteristics. The crowd density gradient shows the changing trend of the crowd distribution in space, used to evaluate the personnel flow situation in the area.

[0081] The three-dimensional communication feature map integrates the signal propagation loss weight and the crowd density gradient, guides the wireless network planning, and forms a comprehensive understanding of the wireless communication environment inside the building.

[0082] In an embodiment of the present application, first, extract spatial topology data from the building information model. Then, use an infrared thermal imaging array to collect the personnel occupancy heat map of the current environment. Then, based on these data, apply an algorithm to calculate the signal propagation loss weight of each area, and determine the crowd density gradient according to the personnel distribution. Finally, integrate all the information to form a detailed three-dimensional communication feature map.

[0083] In a large shopping mall, the research team first parsed the building information model data of the shopping mall to obtain detailed building structure and material information, including key data such as wall materials, metal structure distribution, and elevator shaft positions. At the same time, the team deployed a high-precision infrared thermal imaging array to monitor the occupancy of customers in real time and generate a high-resolution personnel occupancy heat map.

[0084] Next, the team calculated the signal propagation loss weight of each area according to the electromagnetic characteristics of building materials (such as the signal absorption rate differences of concrete, metal, and glass). At the same time, combined with the personnel occupancy heat map, analyzed the crowd density gradient, and identified areas with high pedestrian flow density (such as shopping mall entrances, dining areas) and areas with low pedestrian flow density (such as warehouses, equipment rooms).

[0085] Finally, the team integrated this data into a three-dimensional communication feature map, which not only shows the spatial distribution of signal propagation loss but also marks the dynamic change trend of population density. This step provides basic data support for subsequent communication optimization, enabling the team to accurately identify areas with high pedestrian density and weak signal areas, and providing a scientific basis for dynamically adjusting the SMS sending strategy.

[0086] 102. Based on the coordinates of the metal-dense areas in the three-dimensional communication feature map, detect the difference in the absorption rate of wall materials, and construct a three-dimensional signal attenuation field that includes the reflection interference of elevator shafts;

[0087] In this step, the coordinates of the metal-dense areas refer to the location information of areas in the building that contain a large amount of metal materials, and the metal components concentrated areas in the building are identified through a metal-dense area detector.

[0088] The difference in the absorption rate of wall materials describes the ability of different wall materials to absorb wireless signals, and the impact of different materials on signals is evaluated according to a wall absorption rate difference calculator.

[0089] The reflection interference of elevator shafts is a signal reflection problem caused by the metal structure in the elevator shaft, resulting in a weakening of the signal strength.

[0090] The three-dimensional signal attenuation field is a three-dimensional model that shows the signal strength attenuation at various positions inside the building, considering factors such as the difference in the absorption rate of wall materials and the reflection interference of elevator shafts.

[0091] In the embodiments of the present application, first, the metal-dense areas and their coordinates in the building are identified, and the difference in the absorption rate of different wall materials is measured and recorded. Then, these data are used to construct a three-dimensional signal attenuation field that includes the reflection interference of elevator shafts. This process requires a detailed analysis of the metal structure distribution inside the building and experimental measurement of the difference in the absorption rate of different wall materials. The finally generated three-dimensional signal attenuation field can visually show the signal strength attenuation at various positions inside the building, providing a basis for subsequent communication optimization.

[0092] In the above example of the shopping mall, the research team further analyzed the metal-dense areas inside the building, such as elevator shafts, steel structure supports, and metal decorations. Through high-precision measurement equipment, the team experimentally measured the signal absorption rates of different wall materials (such as concrete, glass, metal plates) and recorded the absorption rate differences. Then, the team used these data to construct a three-dimensional signal attenuation field. This model not only considers the reflection interference of elevator shafts but also combines the signal absorption conditions of different floors and areas, accurately simulating the signal propagation path and attenuation law inside the building.

[0093] Through this step, the team can accurately identify signal blind spots (such as near elevators and areas with high metal density), and take corresponding measures to enhance signal coverage, such as adjusting the antenna position, adding signal relay devices, or optimizing the SMS sending path. This process provides a scientific basis for subsequent communication optimization and significantly improves the communication quality within the shopping mall.

[0094] 103. According to the communication blind spot boundary data in the three-dimensional signal attenuation field, combined with the position coordinates of the fire escape routes, predict the spatio-temporal distribution of SMS request volumes in each zone;

[0095] In this step, the communication blind spot boundary data in the three-dimensional signal attenuation field refers to the boundary of the area where the signal strength is lower than a certain threshold, which is used to identify areas with insufficient signal coverage.

[0096] The position coordinates of the fire escape routes are the specific position information of the emergency evacuation paths within the building, which is crucial for emergency response.

[0097] The spatio-temporal distribution of the SMS request volume describes the changes in the SMS sending requirements in different time periods and areas, and is used to reasonably allocate resources.

[0098] Predicting the spatio-temporal distribution of the SMS request volume in each zone helps to improve the emergency response speed and ensure unobstructed communication in key areas.

[0099] In the embodiments of the present application, first, extract the communication blind spot boundary data from the three-dimensional signal attenuation field and combine it with the known position coordinates of the fire escape routes. Then, use historical data to analyze the SMS request patterns in different areas and predict the SMS request volume in each zone in future time periods. Next, combine these prediction results with the communication blind spot information to adjust the resource allocation plan. Finally, integrate all the data to generate a detailed prediction of the SMS request volume in the time and space dimensions.

[0100] In the application in the aforementioned shopping mall, the research team used the data of the three-dimensional signal attenuation field and combined it with the position coordinates of the fire escape routes to predict the spatio-temporal distribution of the SMS request volume in each zone. Especially during holidays, the team found that the SMS request volume in areas near the fire escape routes (such as the mall entrance and dining areas) increased significantly, while the SMS request volume in communication blind spots (such as the underground parking lot and equipment rooms) was relatively low.

[0101] Based on these data, the team formulated optimization strategies, such as adding temporary communication base stations in areas with high pedestrian flow density, deploying signal relay devices in communication blind spots, and dynamically adjusting the SMS sending priority. These measures ensure that in case of an emergency, communication in key areas (such as near the fire escape routes) will not be interrupted, thus significantly improving the efficiency of emergency response and the reliability of the communication network.

[0102] 104. Connect to the Internet SMS gateway cluster, collect base station signaling plane data and user terminal online status, generate a heat map of cross-operator network quality, and map the heat map to the three-dimensional communication feature map to form a fused communication quality field;

[0103] In this step, the Internet SMS gateway cluster refers to a distributed processing system composed of multiple operator SMS gateways. The base station signaling plane data includes key indicators such as signaling interaction success rate, signaling delay, and signaling load. The user terminal online status reflects the terminal connection stability and signal reception quality. The heat map of cross-operator network quality is a visualized distribution map of network performance generated by integrating data from multiple operators. The fused communication quality field is a three-dimensional communication quality model that integrates building environment characteristics and real-time network status.

[0104] In the embodiments of this application, first connect to each operator's SMS gateway cluster through a standardized interface to collect base station signaling plane data and terminal status information in real time. Then, perform spatio-temporal alignment and correlation analysis on the multi-operator data, and calculate the comprehensive network quality score for each region. Then, convert the scoring results into a heat map, and perform coordinate matching and numerical fusion with the pre-constructed three-dimensional communication feature map. Finally, output a fused communication quality field that includes building structure characteristics, population distribution, and real-time network status.

[0105] In the application in a shopping mall, the system connected to the gateways of the three major operators and collected the real-time signaling data of 42 base stations and the connection status of 3,865 terminals in the mall. Through data analysis, it was found that there was a significant decline in network quality (a 35% increase in signaling delay) in the dining area in the southeast corner during the peak dining hours. The system fused the network quality score (62 points) of this area with the population density data (peak value of 1.8 people / ㎡) in the three-dimensional model, and finally marked this area as a "high-load communication area" to provide a decision-making basis for subsequent SMS scheduling.

[0106] 105. Generate a message forwarding rule sorted by priority based on the spatio-temporal distribution of the SMS request volume in each partition and the backlog volume of the gateway cache queue, in combination with the network congestion hotspots in the fused communication quality field.

[0107] Among them, the message forwarding rule includes emergency channel bandwidth allocation, load balancing and shunting, and weak signal area relay forwarding strategies, and the triggering conditions are dynamically adjusted by analyzing the trend of personnel movement direction.

[0108] In this step, the backlog volume of the gateway cache queue refers to the number of messages to be processed in the network device, which reflects the current communication load situation.

[0109] The network congestion hotspots in the fused communication quality field refer to the areas with poor signal quality and high communication delay identified through the three-dimensional communication feature map.

[0110] The message forwarding rule is a set of guidelines that direct how different types of messages are prioritized for processing, ensuring that high-priority messages are processed in a timely manner.

[0111] The emergency channel bandwidth allocation, load balancing shunt, and relay forwarding strategy in weak signal areas are specific message forwarding measures used to optimize communication efficiency.

[0112] The trigger condition is the criterion for initiating or adjusting the message forwarding strategy, usually based on the trend of personnel movement direction.

[0113] In the embodiments of this application, first, based on the spatio-temporal distribution and the backlog of the gateway cache queue, combined with the network congestion hotspots in the converged communication quality field, the communication requirements of each region are comprehensively evaluated. Then, message forwarding rules are formulated, including emergency channel bandwidth allocation, load balancing shunt, and relay forwarding strategy in weak signal areas. Then, by real-time monitoring the trend of personnel movement direction and the change of network congestion hotspots, the trigger condition is dynamically adjusted to ensure the effectiveness of the rules. Finally, all information is integrated to generate a complete set of message forwarding rules that can be dynamically adjusted. In the application in a shopping mall, the research team generated message forwarding rules based on the spatio-temporal distribution, the backlog of the gateway cache queue, and the network congestion hotspots in the converged communication quality field. Especially during holidays, in some areas, not only is the backlog of the gateway cache queue serious, but there are also network congestion hotspots at the same time, resulting in increased communication delays. For this reason, the team formulated emergency channel bandwidth allocation and load balancing shunt strategies. For example, bandwidth resources are preferentially allocated in areas with high pedestrian flow density (such as shopping mall entrances and dining areas), and relay forwarding devices are deployed in weak signal areas (such as underground parking lots). At the same time, the team dynamically adjusts the trigger condition by real-time analyzing the trend of personnel movement direction and the change of network congestion hotspots. For example, when it is detected that a large number of people are moving towards the fire escape direction and there is a network congestion hotspot in this area, the system automatically activates the emergency channel bandwidth allocation strategy and optimizes the load balancing scheme to ensure smooth communication in key areas. These measures have significantly improved communication efficiency, especially maintaining good communication quality even in the case of network congestion.

[0114] 106. Execute the SMS sending operation according to the message forwarding rule.

[0115] Among them, the message forwarding rule refers to a set of SMS processing strategies sorted by priority, including specific implementation plans such as emergency channel allocation, load balancing, and relay forwarding;

[0116] The SMS sending operation refers to the process of transmitting the message to be sent to the target terminal through the optimal path according to the established rules;

[0117] During the execution process, it is necessary to monitor the change of the network status in real time and dynamically adjust the sending strategy;

[0118] The operation result feedback is used to optimize the generation of subsequent message forwarding rules.

[0119] In the embodiments of this application, first, the currently effective message forwarding rules are parsed to determine the priority order of SMS sending in each area and the specific transmission strategy. Then, the messages are distributed to the corresponding base stations according to the rules through the SMS gateway cluster, and the network quality changes are monitored in real time during the transmission process. When it is detected that the network congestion hotspot has shifted or there is a new trend of people gathering, the sending path and resource allocation are dynamically adjusted. Finally, the SMS delivery rate and delay data in each area are collected and fed back to the rule generation module for continuous optimization.

[0120] In the application in a shopping mall, the system executes SMS sending according to the generated forwarding rules. For example, when network congestion is detected in the dining area, the load balancing strategy is automatically enabled to divert some messages to adjacent base stations; when the fire escape area is marked as high priority, the emergency channel is immediately activated to ensure the transmission of critical messages. By collecting the message delivery status in each area in real time, the system continuously optimizes the forwarding strategy to ensure that the delivery success rate remains above 95% in case of dense population and network fluctuations.

[0121] Through the above steps 101 to 106, a three-dimensional communication feature map is constructed by integrating the building information model and real-time personnel distribution data to accurately depict the signal propagation characteristics and crowd gathering characteristics inside the building; metal-dense areas and special structures (such as elevator shafts) are identified based on the building structure characteristics, and a three-dimensional signal attenuation field that accurately reflects the actual attenuation situation is established; the spatio-temporal distribution of SMS services is predicted by combining communication blind spots and personnel flow channels to achieve accurate demand prediction; the network status data of multiple operators are integrated to form a real-time communication quality heat map, which is fused with the three-dimensional environment model to construct a global communication situation awareness; the optimal forwarding strategy is dynamically generated by comprehensively considering service requirements, network load, and communication quality, and resource intelligent scheduling is realized. Through the rule-driven sending execution mechanism, reliable transmission of high-priority messages is ensured, and the communication guarantee ability in complex environments is significantly improved. This method realizes the precision, intelligence, and reliability of SMS services in building scenarios.

[0122] Furthermore, the method further includes: importing the three-dimensional communication feature map and forwarding rules into the digital twin environment to simulate the crowd gathering scenario, and by comparing the gateway response delay with the preset threshold, correcting the absorption rate parameter of the three-dimensional signal attenuation field.

[0123] In this step, the digital twin environment is a virtual simulation system that can simulate the behavior of the actual physical environment and is used to test and optimize the communication system.

[0124] The gateway response delay refers to the time required for a network device to respond to a received data packet, which measures the performance of the communication system. The preset threshold is a standard value for measuring the gateway performance and is used to determine whether the system needs to be optimized.

[0125] The absorption rate parameter of the corrected three-dimensional signal attenuation field is a further optimization of the original model to more accurately reflect the actual situation and ensure the efficient operation of the communication system.

[0126] In the embodiments of the present application, first, the three-dimensional communication feature map and the message forwarding rule are imported into the digital twin environment to simulate the actual crowd gathering scenario. Then, by comparing the gateway response delay in the simulated environment with the preset threshold, the accuracy of the existing model is evaluated. Then, based on the comparison result, the absorption rate parameter in the three-dimensional signal attenuation field is corrected to more accurately reflect the actual situation. Finally, through multiple iterations of optimization, the high precision of the model is ensured.

[0127] In the application in the aforementioned shopping mall, the research team used the digital twin environment to simulate the communication situation during the peak holiday period and found that the gateway response delay in some areas exceeded the preset threshold. For example, in areas with high pedestrian flow density such as the dining area and the mall entrance, the gateway response delay increased significantly, resulting in a decrease in the efficiency of SMS sending. Based on these simulation results, the team corrected the absorption rate parameter of the three-dimensional signal attenuation field, such as adjusting the signal absorption rate of the metal-dense area and the wall material, to more accurately reflect the signal propagation characteristics in the actual environment.

[0128] To solve the problem of inaccurate multi-operator network quality assessment in complex building environments, in some embodiments, the access to the Internet SMS gateway cluster generates a heat map of the cross-operator network quality by collecting base station signaling plane data and the online status of user terminals, and maps the heat map to the three-dimensional communication feature map to form a fused communication quality field, including:

[0129] 1041. Synchronously connect each SMS gateway cluster through a multi-operator protocol interface to obtain the signaling plane data of each base station in real time.

[0130] Among them, the multi-operator protocol interface: a unified access module that supports the unified access of different operator communication protocols;

[0131] Signaling plane data: Control signaling interaction records between the base station and the terminal (such as call establishment / release records), terminal registration status (online / offline), and signaling response delay (response time of the terminal to the base station instruction).

[0132] In the embodiment of the present application, each operator gateway is connected through a standardized protocol conversion module (such as a Diameter protocol adapter) to pull the signaling plane data of the base station in real time. In the specific implementation, an asynchronous message queue (such as Kafka) is used to receive data streams from different operators, and the scattered signaling data is aggregated by base station dimension through timestamp alignment and base station ID matching. For example, for the signaling response delay data of a certain base station, the average value is calculated according to a 5-second time window, and abnormal values ​​(such as extreme values ​​exceeding 500ms) are filtered.

[0133] 1042. At the same time, terminal measurement data reported by the user terminal is collected, and the signaling plane data collected in the same time window is time-space matched with the terminal measurement data according to the base station coverage area, abnormal terminal data is eliminated, and a valid data point set is retained. Based on the valid data point set, the signaling response delay distribution and signal strength distribution of each terminal within the coverage area of ​​each base station are extracted;

[0134] Among them, terminal measurement data refers to the wireless signal parameters actively reported by the terminal, including reference signal received power (RSRP, reflecting signal strength), signal interference and noise ratio (SINR, reflecting signal quality), and terminal movement speed (calculated by GPS or inertial sensors). Space-time matching: associate the signaling plane data with the terminal measurement data in the same time period and within the coverage range of the same base station.

[0135] In the embodiment of the present application, the terminal periodically reports the measurement data (e.g., once every 2 seconds) through the pre-installed SDK. On the server side, the terminal registration status in the signaling plane data is bound to the terminal ID in the measurement data with the base station ID as the index, and abnormal terminal data that is not registered or has a signal strength of 0 is removed. For example, for a terminal within the coverage area of ​​a certain base station, a valid data point set with both signaling delay (step 1041) and RSRP / SINR (this step) is screened out through time-space matching.

[0136] 1043. Divide the base station coverage area into a number of grid cells according to the terminal location distribution density, each grid cell containing measurement data of at least one terminal;

[0137] Grid unit: Divide the base station coverage area (e.g., 500-meter radius) into square grids with a side length of 10 meters;

[0138] Terminal location distribution density: terminal distribution thermal value calculated by terminal GPS coordinates or base station triangulation.

[0139] In the embodiments of the present application, the terminal density within each grid is calculated based on terminal positioning data (GPS or base station fingerprint positioning). An adaptive grid division algorithm is adopted: for terminal-dense areas (such as density > 5 terminals / 10㎡), the grid is automatically shrunk to a side length of 5 meters, and for sparse areas (density < 1 terminal / 10㎡), it is expanded to a side length of 20 meters to ensure that the measurement data of at least 1 terminal is included in each grid.

[0140] 1044. The signaling response delay and signal strength within each grid cell are weighted and calculated to obtain the network quality score of the grid cell, and the spatial interpolation method is used to smooth the network quality scores of adjacent grid cells to generate a heat map of cross-operator network quality.

[0141] Weighted calculation: linearly weight the signaling response delay (weight 0.6) and signal strength (RSRP, weight 0.4);

[0142] Spatial interpolation: The inverse distance weighted (IDW) algorithm is used to fill the grids without terminal data.

[0143] In the embodiments of the present application, for the terminal data within each grid cell, the score is calculated according to the formula:

[0144] Score = 0.6 * (1 - delay / maximum delay threshold) + 0.4 * (RSRP / maximum RSRP value);

[0145] For adjacent grids, IDW interpolation is used to calculate the score of blank grids with the reciprocal of the distance (d is the grid spacing) as the weight:

[0146] Score = Σ(score_i / d_i²) / Σ(1 / d_i²);

[0147] Finally, a continuous heat map covering the entire area is generated.

[0148] 1045. Map the central coordinates of each grid cell in the network quality heat map to the corresponding spatial positions in the three-dimensional communication feature map;

[0149] 1046. According to the population density gradient and signal propagation loss weight at the corresponding spatial positions in the three-dimensional communication feature map, adjust the contribution coefficient of the network quality score to update the network quality score;

[0150] 1047. Superimpose the updated network quality score and the communication parameters in the three-dimensional communication feature map to form a fused communication quality field.

[0151] In the above steps 1045 - 1047, the contribution coefficient is adjusted: the network quality score is dynamically corrected according to the population density (weight 0.7) and signal propagation loss (weight 0.3). The communication parameters are superimposed: the corrected score is weighted and fused with the wall attenuation value and metal interference value in the three-dimensional atlas.

[0152] In the embodiment of the present application, the central coordinates of the heat map grid (such as longitude 116.3°E and latitude 39.9°N) are mapped to the corresponding coordinate points in the three-dimensional atlas. Based on the population density gradient (such as 1.5 people / ㎡) and signal loss weight (such as +15dB loss in the elevator shaft area) at this coordinate point, the score is corrected according to the formula:

[0153] Corrected score = original score * (0.7 * (1 - population density / peak density) + 0.3 * (1 - signal loss / maximum loss))

[0154] Finally, the corrected score is superimposed with the communication parameters (such as the wall material attenuation coefficient) in the three-dimensional atlas to generate a fused communication quality field.

[0155] The following is a specific example:

[0156] First, the system accesses the SMS gateway clusters of China Mobile, China Unicom, and China Telecom through the multi-operator protocol interface, and real-time collects the signaling plane data of 42 base stations in the mall (including the average signaling response delay of 82ms) and the measurement data of 3865 terminals (the RSRP range is from -85dBm to -105dBm). During the peak dining period, the signaling delay of the base station in the southeast corner dining area suddenly increases to 142ms, and the average RSRP reported by the terminal is lower than -98dBm.

[0157] Second, the system divides the coverage area of the base station in the dining area into grid cells with a side length of 5 meters. It is calculated that the network quality score in this area drops from 72 points during non-peak hours to 58 points. After filling the corner grids without terminal data through spatial interpolation, a heat map of cross-operator network quality is generated, showing that the dining area is an orange warning area (score 50 - 60 points).

[0158] Then, the heat map is mapped to the three-dimensional communication feature atlas. Combining the peak population density (2.1 people / ㎡) in this area and the signal propagation loss (+12dB) caused by the steel frame structure, the network quality score is corrected to 51 points according to the formula. In the three-dimensional fused communication quality field, this area is marked as a "severe congestion area" and is associated with the wall metal density (85%) and the elevator shaft interference coefficient (0.8).

[0159] Finally, the system automatically triggers three strategies according to the marking results of the fused communication quality field:

[0160] Lower the SMS forwarding priority in the dining area from Level 2 to Level 4;

[0161] Enable load balancing and traffic diversion for the base stations in the adjacent leisure area, and direct 30% of the message traffic to the west base station;

[0162] Deploy portable relay devices in the steel frame structure area to increase the signal strength by 8 dB.

[0163] After implementation, the SMS delivery rate during the peak period in this area has increased from 78% to 93%, and the average delay has dropped back to 89 ms.

[0164] This solution realizes the accurate characterization and dynamic perception of network quality in complex building scenarios through real-time synchronous acquisition of multi-operator data, dynamic grid quality assessment, and three-dimensional environmental parameter fusion and correction. It effectively solves the evaluation deviation problems caused by the one-sidedness of single-operator data and the neglect of building structure interference in traditional methods, provides a high-precision communication situation map for the SMS sending strategy, significantly improves the traffic diversion efficiency in congested areas and the relay scheduling accuracy in weak signal areas, and ultimately ensures the communication reliability and timeliness in high-load scenarios.

[0165] To solve the problem of insufficient accuracy in responding to high-density areas and emergencies in the communication system, by fusing the high-density area coordinates and the group movement direction vector in the spatio-temporal distribution, combining the peak period characteristics in the gateway cache queue backlog, and the network congestion hotspots in the fused communication quality field, a message forwarding rule sorted by priority is generated. This method not only improves the response speed and efficiency of the communication system but also ensures efficient communication in case of emergencies.

[0166] In some embodiments, the step 104 of generating a message forwarding rule sorted by priority based on the spatio-temporal distribution of the SMS request volume in each partition and the gateway cache queue backlog, combined with the network congestion hotspots in the fused communication quality field, includes:

[0167] 201. Fuse the high-density area coordinates and the group movement direction vector in the spatio-temporal distribution of the SMS request volume in each partition, and calculate the matching degree between the moving direction clustering result and the historical request volume growth rate by associating the signal strength attenuation rate of the network congestion hotspots, to generate an emergency channel weight coefficient;

[0168] In step 201, the signal strength attenuation rate of the network congestion hotspot: It represents the change rate of the signal strength over time or space within the network congestion area (such as a decrease of 2 dBm per second). The coordinates of the high-density area in the spatio-temporal distribution refer to the location information of the area where there is a large crowd gathering, which is used to identify key areas. The group movement direction vector describes the direction and speed of the personnel flow, which helps to predict future traffic changes. The result of the movement direction clustering is the result obtained by clustering the group movement directions, reflecting the movement trends of different groups. The historical request volume growth rate is the growth rate calculated based on past data, which is used to predict future demands. The emergency channel weight coefficient is a comprehensive index used to measure the importance of a specific area.

[0169] In the embodiment of the present application, first, the coordinates of the high-density area (such as the coordinate set of area A in the shopping mall) are extracted, and the group movement direction vector (such as gathering in the northwest direction) is analyzed through the DBSCAN clustering algorithm. Subsequently, the cosine similarity between the historical request volume growth rate (such as a 120% increase within a 10-minute window) and the current movement direction (such as cosθ = 0.92) is calculated, combined with the signal attenuation rate (such as -4.2 dB / minute), and the weight coefficient is generated according to the formula:

[0170] Weight coefficient = similarity × (1 - attenuation rate / maximum attenuation rate)

[0171] For example, when the attenuation rate reaches 80% of the preset threshold, the emergency channel weight coefficient is triggered to increase to 0.75.

[0172] 202. Based on the peak period characteristics in the gateway queue backlog, combined with the signal attenuation gradient between the network congestion hotspot and the adjacent access points, by matching the inflection point of the attenuation gradient with the physical topology distance, generate the cross-floor traffic diversion ratio;

[0173] In step 202, the peak period characteristics in the gateway queue backlog refer to the change rule of the number of messages to be processed in the network device during the peak period. The signal attenuation gradient between adjacent access points describes the change trend of the signal strength between different access points. The physical topology distance is the actual distance between different access points. The cross-floor traffic diversion ratio is a distribution strategy used to disperse the traffic to different floors to relieve the pressure on a single access point.

[0174] In the embodiment of the present application, based on the peak period of the gateway queue backlog (such as the backlog reaching 1200 messages at 13:00), the signal attenuation gradient between the adjacent floor access points is extracted (such as the attenuation gradient from floor 3 to floor 4 is 12 dB). Through the path planning algorithm (such as the A* algorithm), the attenuation inflection point (such as the inflection point coordinates of the elevator shaft) is matched with the physical topology distance (such as the straight-line distance of 20 meters), and the traffic diversion ratio is calculated according to the formula:

[0175] Traffic diversion ratio = (1 - attenuation gradient / maximum gradient) × (1 - distance / maximum distance)

[0176] For example, when the distance from the inflection point is 5 meters and the attenuation gradient is 10 dB, a floor-to-floor shunt ratio of 45% is generated.

[0177] 203. Extract the coverage range of idle channels of neighboring access points within the network congestion hotspots according to the geometric boundary data of the weak signal area, and generate a relay forwarding priority sequence by analyzing the correlation strength between the access point response delay and the signal compensation requirement;

[0178] In step 203, the geometric boundary data of the weak signal area refers to the boundary of the area where the signal strength is lower than a certain threshold. The coverage range of idle channels of neighboring access points describes the coverage area of available channels nearby. The access point response delay refers to the time required for the network device to respond to the received data packet. The signal compensation requirement is the compensation requirement determined according to the current signal quality. The relay forwarding priority sequence is a sorted list used to guide the signal transmission order in the weak signal area.

[0179] In the embodiment of the present application, the geometric boundary of the weak signal area (such as the polygon coordinates of the underground parking lot) is extracted, and the idle channels of neighboring access points are scanned (such as there are 4 idle channels at the base station at the parking lot entrance). The correlation strength (R² = 0.85) between the response delay (such as an average of 80 ms) and the signal compensation requirement (such as +12 dB) is analyzed by linear regression, and the priority is sorted according to the formula:

[0180] Priority = Number of idle channels × (1 - Delay / Maximum delay) + Compensation requirement weight

[0181] For example, if an access point has 3 idle channels and a delay of 60 ms, the priority score is 8.2 (full score is 10 points).

[0182] 204. Multidimensionally fuse the emergency channel weight coefficient, the cross-floor shunt ratio, and the relay forwarding priority sequence to construct a forwarding rule trigger matrix, and the forwarding rule trigger matrix is dynamically updated by correlating the change rate of the group movement direction vector and the growth rate of the backlog in the gateway cache queue;

[0183] In step 204, multi-dimensional fusion is an algorithm framework used to integrate data from multiple dimensions into a unified model. The forwarding rule trigger matrix is a comprehensive chart that contains all the generated parameters and is used to guide the message forwarding strategy. The change rate of the group movement direction vector and the growth rate of the gateway cache queue backlog are important parameters for dynamically updating the forwarding rule trigger matrix. In the embodiment of the present application, first, the data of the emergency channel weight coefficient, the cross-floor diversion ratio, and the relay forwarding priority sequence are collected and integrated. Then, using multi-dimensional data analysis technology, these parameters are fused into a unified forwarding rule trigger matrix. This matrix not only contains static configuration information but also can be dynamically adjusted according to real-time data. Then, an association model is established between the change rate of the group movement direction vector and the growth rate of the gateway cache queue backlog to implement a dynamic update mechanism for the forwarding rules. For example, when it is detected that the crowd is quickly gathering in a certain area, the system automatically adjusts the forwarding rules to preferentially allocate communication resources.

[0184] Specifically, the outputs of steps 201-203 are normalized to matrix elements (e.g., weight coefficient 0.75 → matrix value 75). The Pearson correlation coefficient (e.g., r = 0.68) of the movement direction change rate (e.g., Δθ = 10° / minute) and the queue backlog growth rate (e.g., ΔQ = 150 messages / minute) is calculated through a sliding window (5-minute window). When the correlation coefficient exceeds the threshold (e.g., r > 0.6), the matrix is triggered for dynamic update:

[0185] New matrix value = original value × (1 + correlation coefficient × update factor)

[0186] For example, when the correlation coefficient is 0.7, the weight coefficient is updated from 75 to 80.

[0187] 205. Based on the critical parameters of the forwarding rule trigger matrix, simulate message path conflicts in the digital twin environment, and generate a message forwarding rule sorted by priority by reversely correcting the matching degree threshold and the signal compensation parameter.

[0188] In step 205, the forwarding rule trigger matrix is a multi-dimensional data structure formed by fusing the emergency channel weight coefficient, the cross-floor diversion ratio, and the relay forwarding priority sequence, and is used to dynamically adjust the message forwarding strategy. The critical parameter refers to the key value in the trigger matrix for determining whether to adjust the forwarding rule. The digital twin environment is a virtual simulation platform that can simulate various scenarios in the real world, including the behavior of communication networks. Message path conflict refers to the situation of information congestion or transmission failure that may occur during communication. The matching degree threshold is a standard value for measuring the matching degree between different forwarding rules. The signal compensation parameter is a technical parameter set to make up for the weak signal area, used to enhance the signal strength or optimize the information transmission path. The message forwarding rules sorted by priority are a set of instruction sets that arrange the information transmission order according to the urgency and efficiency.

[0189] Specifically, load the forwarding rule trigger matrix in the digital twin model to simulate the SMS sending path (such as A→B→C). When detecting a path conflict (such as two messages simultaneously preempting channel X), use the genetic algorithm to iteratively optimize the parameters:

[0190] Randomly generate 100 groups of parameter combinations (matching degree threshold, compensation parameter);

[0191] Select 20 groups with the lowest conflict rate for crossover and mutation;

[0192] After 50 generations of iteration, output the optimal parameters (such as threshold 0.63, compensation +13dB).

[0193] Finally, generate a rule table sorted by priority (such as Rule 1: Messages with an emergency channel weight > 0.8 are transmitted first).

[0194] The following is a specific example:

[0195] First, the system detects that the transfer hall (coordinates X,Y) is a high-density area (request volume 610 messages / minute), the group movement direction points to the exit (vector θ = 30°, v = 1.2m / s), and the signal attenuation rate in this area reaches -5dB / minute. The calculated emergency channel weight coefficient is 0.82.

[0196] Secondly, the backlog of the gateway reaches 1500 messages at 18:00, the signal attenuation gradient of the adjacent platform layer is 18dB, and after matching the physical distance of 15 meters, the cross-floor diversion ratio of 52% is generated.

[0197] Then, in the weak signal area (toilet passage), it is detected that there are 2 idle channels of the adjacent access point, and the response delay is 65ms. The calculated relay forwarding priority is 7.3 levels.

[0198] Then, construct a three-dimensional trigger matrix (weight 82, shunt 52, priority 7.3). When it is detected that the crowd suddenly turns (Δθ = 25° / minute) and the queue grows by ΔQ = 300 messages / minute (r = 0.72), update the matrix weight to 88.

[0199] Finally, the digital twin simulation shows that the conflict rate of channel Y is 23%. After reverse correction, the matching degree threshold is reduced from 0.7 to 0.62, and the final rules are generated:

[0200] Messages with a weight > 0.8 are directly connected through the emergency channel;

[0201] 52% of the backlogged messages are shunted to the platform layer;

[0202] Messages with a priority > 7 enable relay forwarding.

[0203] After implementation, the SMS delivery rate in the transfer hall is increased from 68% to 89%, and the channel conflict rate is decreased to 9%.

[0204] To further improve the accuracy of the communication system in high-density areas and emergency response, in some embodiments, in step 204, the multi-dimensional fusion of the emergency channel weight coefficient, the cross-floor shunt ratio, and the relay forwarding priority sequence is performed to construct a forwarding rule trigger matrix, including:

[0205] 301. Integrate the real-time load and historical traffic data of each channel with the emergency channel weight coefficient, generate a basic constraint value in combination with the channel length, and perform dynamic compensation by superimposing the obstacle density to obtain a channel quality evaluation index;

[0206] In step 301, the emergency channel weight coefficient is a comprehensive index used to measure the importance of a specific area. The real-time load of each channel is the traffic information on the current channel. The historical traffic data is the traffic volume data based on past records. The channel length refers to the physical distance. The basic constraint value is a reference value calculated by combining the real-time load, historical traffic data, and channel length. The obstacle density refers to the number of obstacles in the channel, which affects the availability of the channel. The channel quality evaluation index is a comprehensive score reflecting the overall quality and availability of the channel.

[0207] In the embodiments of the present application, first, collect the real-time load and historical traffic data of each channel, and generate a basic constraint value in combination with the channel length. Then, fuse these data with the emergency channel weight coefficient to form a preliminary evaluation framework. Next, obtain the obstacle density data through on-site investigation or sensors, and superimpose it as a dynamic compensation factor on the basic constraint value. Specifically, use the weighted average method or other statistical methods to quantify the obstacle density, and adjust the basic constraint value to reflect the actual channel conditions. Finally, integrate all the information to generate a comprehensive channel quality evaluation index for subsequent optimization decisions.

[0208] For example, first, the system collects the flow sensor data (current load) of each emergency channel and the historical passage database records in real time. Combining with the channel length information provided by the GIS map, a basic constraint value is generated through a weighted calculation formula (load weight 40% + historical data 30% + length 30%). Secondly, the distribution density of obstacles in the channel is obtained by using lidar scanning or camera image recognition technology, and the basic constraint value is dynamically adjusted by using a logarithmic compensation algorithm (for every 10% increase in obstacle density, the quality score is reduced by 5%). Then, the emergency channel weight coefficient is introduced into the calculation process as an adjustment factor to ensure that important channels obtain a quality score bonus (for every 0.1 increase in the weight coefficient, the score increases by 2%). Finally, a channel quality evaluation index ranging from 0 to 100 points is output. A channel with a score above 90 is a high-quality channel, and an alarm needs to be triggered if the score is below 60.

[0209] 302. Generate a diversion gradient coefficient based on the cross-floor diversion ratio, the floor evacuation node density, the coupling evacuation ladder capacity threshold, and the channel quality evaluation index, and output a diversion ratio reference value by integrating the building load limit;

[0210] In step 302, the cross-floor diversion ratio is a distribution strategy for dispersing traffic to different floors. The floor evacuation node density refers to the number of emergency exits on each floor. The evacuation ladder capacity threshold is the maximum load capacity of each evacuation ladder. The diversion gradient coefficient is a comprehensive index for measuring the traffic distribution between different floors. The building load limit refers to the maximum weight that the building structure can bear. The diversion ratio reference value is a distribution standard for guiding actual diversion operations.

[0211] In the embodiment of the present application, first, a preliminary cross-floor diversion ratio is calculated based on the floor evacuation node density and the evacuation ladder capacity threshold. Then, in combination with the channel quality evaluation index, a diversion gradient coefficient reflecting the difference in traffic distribution between floors is generated. Then, considering the building load limit, it is ensured that the diversion plan will not cause additional pressure on the building structure. Specifically, through mathematical modeling and optimization algorithms, the best diversion ratio that meets all conditions is found. Finally, all parameters are integrated to output a diversion ratio reference value that conforms to the actual situation for guiding actual operations.

[0212] For example, first, based on the building BIM model, extract the distribution data of the escape nodes on each floor, and combine it with the real-time pedestrian flow statistics from infrared thermal imaging to calculate the density of escape nodes per square meter. Secondly, according to the escape ladder capacity threshold set by the fire protection code (such as 0.3 square meters per person), and combine with the channel quality evaluation index obtained in step 301, use the gradient descent algorithm to solve the optimal diversion gradient coefficient. Then, monitor the load-bearing data of each floor through building structure sensors, and use the constrained optimization model to ensure that the diversion plan does not exceed the load-bearing limit (such as load-bearing per square meter ≤ 500 kg). Finally, output the reference value of the diversion ratio for each floor, with the accuracy controlled within the range of ±2%.

[0213] 303. The relay forwarding priority sequence constructs a node coverage map through the multipath effect of wireless signals, fuses the diversion gradient coefficient to generate a three-dimensional permeability surface, extracts the curvature features of the three-dimensional permeability surface, and generates a priority sorting linked list;

[0214] In step 303, the relay forwarding priority sequence is a sorted list used to guide the signal transmission order in the weak signal area. The multipath effect of wireless signals refers to the phenomenon that signals reach the receiving end through multiple paths. The node coverage map is a coverage map constructed based on the multipath effect. The three-dimensional permeability surface is a three-dimensional model showing the signal penetration ability of different regions. The curvature feature is the key geometric feature extracted from the three-dimensional permeability surface. The priority sorting linked list is an ordered list used to guide the signal transmission order.

[0215] In the embodiment of the present application, first, use the multipath effect of wireless signals to construct a node coverage map to depict the coverage of the wireless network. Then, combine with the diversion gradient coefficient to generate a three-dimensional permeability surface to show the information transmission ability of different regions. Then, extract the curvature features of the three-dimensional permeability surface to identify the best information transmission path. Specifically, apply image processing technology and machine learning algorithms to analyze the curvature features to find the optimal path. Finally, generate a priority sorting linked list according to the analysis results to ensure that key information can be efficiently transmitted in the shortest time.

[0216] For example, first, deploy a wireless probe array to collect multipath effect data, and construct a three-dimensional node coverage map (with a resolution of 0.5 meters) through the ray tracing algorithm. Secondly, use the diversion gradient coefficient in step 302 as the permeability adjustment factor, and use the Kriging interpolation method to generate a three-dimensional permeability surface (with a grid accuracy of 0.1 meters). Then, use the differential geometry algorithm to extract the surface curvature features and identify the high-permeability channels with a curvature radius less than 3 meters. Finally, calculate the optimal transmission path between each node based on the Dijkstra algorithm, and generate a priority sorting linked list sorted in descending order of penetration efficiency, which is updated every 5 seconds.

[0217] 304. Input the channel quality evaluation index, shunt ratio reference value, and priority sorting linked list into the building digital twin body, extract implicit association rules through tensor decomposition and orthogonal recombination, and construct a forwarding rule trigger matrix.

[0218] In step 304, the building digital twin body is a virtual simulation system that can simulate the behavior of the actual physical environment. Tensor decomposition is a mathematical tool used to extract implicit patterns in complex data. Implicit association rules are potential laws extracted from data. Orthogonal recombination is an optimization technique used to recombine data to improve efficiency. The forwarding rule trigger matrix is a comprehensive chart that contains all the generated parameters and is used to guide the message forwarding strategy.

[0219] In the embodiment of this application, first, input the channel quality evaluation index, shunt ratio reference value, and priority sorting linked list into the building digital twin body. Then, use the tensor decomposition technology to extract implicit association rules and reveal the potential connections between data. Next, reorganize these rules into a new data structure, namely the forwarding rule trigger matrix, through orthogonal recombination. Specifically, adopt advanced data analysis tools and algorithms to process complex multi-dimensional data to ensure that the generated matrix can accurately reflect the actual situation. Finally, integrate all the information to construct a complete forwarding rule trigger matrix for guiding actual operations.

[0220] For example, first, connect the building digital twin body with the real-time data stream and load three groups of data: the channel quality evaluation index, shunt ratio reference value, and priority sorting linked list. Second, adopt the CP tensor decomposition method (Canonical Polyadic Decomposition) to extract third-order implicit association rules, and set the decomposition rank to 5 to balance accuracy and efficiency. Then, recombine the eigenvectors through the Gram-Schmidt orthogonalization process to eliminate the redundant associations between the rules. Finally, construct a forwarding rule trigger matrix with dimensions [channel × floor × priority], where the matrix element value is the trigger weight (0 - 1), and deploy an online update mechanism (incrementally update 10% of the key parameters per minute).

[0221] The following is a specific example:

[0222] In an application scenario of a large shopping mall, the research team integrated the coordinates of high-density areas and the group movement direction vectors in the spatio-temporal distribution to generate the weight coefficients of the emergency channels. The research team constructed a node coverage map using the multipath effect of wireless signals and, combined with the previously generated shunt gradient coefficients, generated a three-dimensional permeability surface. Through the analysis of curvature features, the research team identified the optimal information transfer paths and generated a priority sorting linked list. The research team input the channel quality evaluation indicators, shunt ratio benchmark values, and priority sorting linked list into the building digital twin. Through tensor decomposition technology, implicit association rules between data were discovered, and a forwarding rule trigger matrix was generated through orthogonal recombination. This matrix not only improved the efficiency of information transfer but also enhanced the flexibility and response speed of the system, significantly improving the emergency communication management ability of the entire shopping mall.

[0223] In summary, steps 301 to 304 construct a dynamically updated forwarding rule trigger matrix by multi-dimensionally integrating the weight coefficients of the emergency channels, cross-floor shunt ratios, and relay forwarding priority sequences. This solution significantly improves the efficiency and reliability of the indoor wireless communication system. This method not only improves the understanding of signal propagation loss and crowd density in complex environments but also ensures efficient communication in emergency situations through dynamic adjustment strategies, providing strong support for intelligent building management and public safety.

[0224] To address the problem of insufficient accuracy in responding to high-density areas and emergency situations in the communication system, in some embodiments, the process of generating the shunt gradient coefficient by coupling the escape ladder capacity threshold and the channel quality evaluation indicator based on the floor escape node density in step 302 includes:

[0225] 401. Generate a node coverage intensity value based on the floor escape node density, the visible distance, and the number of path branches. Trigger connectivity compensation through the visible distance and correct the topological intensity attenuation through the number of path branches.

[0226] In step 401, the floor escape node density refers to the number of emergency exits on each floor. The visible distance is the distance that can be directly seen between nodes. The number of path branches is the number of nodes connected to other nodes. The node coverage intensity value is a comprehensive score reflecting the importance of the node. Connectivity compensation is the process of adjusting the node coverage intensity value based on the visible distance. Topological intensity attenuation is the process of correcting the node coverage intensity value based on the number of path branches.

[0227] In the embodiments of the present application, first, a node coverage intensity value is generated based on the density of floor evacuation nodes, the visible distance, and the number of path branches. Then, computer vision technology is used to calculate the visible distance, which is used as a connectivity compensation factor. Next, the number of path branches is analyzed, and graph theory algorithms are adopted to correct the attenuation of topological strength. Finally, all the information is integrated to generate a node coverage intensity value that comprehensively reflects the evacuation ability of the nodes.

[0228] For example, first, the three-dimensional coordinate data of each floor evacuation node is extracted through the building BIM model, and the visible distance (accuracy 0.1 meter) between each node is calculated using the ray casting algorithm. Second, an evacuation path network is constructed based on graph theory analysis, and the number of path branches (connectivity) of each node is counted. Then, a composite calculation formula is adopted: coverage intensity value = visible distance coefficient × ln(connectivity) × node density, where the visible distance coefficient compensates for the attenuation of distances exceeding 30 meters through an S-shaped function. Finally, the calculation results are normalized (in the range of 0 - 1) to generate a coverage intensity matrix reflecting the evacuation ability of the nodes.

[0229] 402. Based on the escape ladder capacity threshold and the smoke diffusion rate, through the asymmetric superposition of a static capacity baseline and a dynamic attenuation factor, and by introducing the node coverage intensity value, a real-time passage attenuation coefficient is generated;

[0230] In step 402, the escape ladder capacity threshold is the maximum load capacity of each escape ladder. The smoke diffusion rate is the propagation speed of smoke in the building. The static capacity baseline is a base value set based on the escape ladder capacity threshold. The dynamic attenuation factor is a coefficient adjusted based on the smoke diffusion rate. The real-time passage attenuation coefficient is a comprehensive index used to measure the passage ability at different time points. The node coverage intensity value is the data generated from the previous step.

[0231] In the embodiments of the present application, first, based on the escape ladder capacity threshold and the smoke diffusion rate, the asymmetric superposition is performed by combining the static capacity baseline and the dynamic attenuation factor. Then, the node coverage intensity value generated in step 401 is introduced to further refine the model. Next, simulation technology is used to simulate the evacuation process in different scenarios, and the real-time passage attenuation coefficient is calculated. Finally, a real-time passage attenuation coefficient that accurately reflects the current passage ability is generated.

[0232] For example, first, obtain the design capacity data of each escape ladder from the fire protection system and establish a static capacity baseline (e.g., the maximum load per trip is 20 people). Second, monitor the change rate of smoke concentration on each floor in real time through the smoke sensor network, use the fluid dynamics model to predict the diffusion trend in the next 3 minutes, and generate a dynamic attenuation factor (0 - 1 value). Then, perform weighted fusion of the static baseline and the asymmetric attenuation factor (static weight 60% + dynamic 40%), and superimpose the node coverage intensity value in step 401 as an adjustment parameter. Finally, verify through Monte Carlo simulation and output a real-time passage attenuation coefficient table updated on a minute-by-minute basis.

[0233] 403. The channel quality evaluation index and the real-time passage attenuation coefficient are fused through the horizontal weight base and the vertical emergency weight, and tensor dot product and topological filtering are performed at the floor connection nodes to generate a flow diversion gradient coefficient.

[0234] In step 403, the channel quality evaluation index is a comprehensive score that reflects the overall quality and usability of the channel. The real-time passage attenuation coefficient is the data generated from the previous step. The horizontal weight base is a base value set based on the channel quality evaluation index. The vertical emergency weight is a coefficient set based on the real-time passage attenuation coefficient. The flow diversion gradient coefficient is a comprehensive index used to measure the flow distribution between different floors. Tensor dot product is a mathematical operation used to fuse multi-dimensional data. Topological filtering is an optimization technique used to remove unnecessary data noise.

[0235] In the embodiment of the present application, first, the channel quality evaluation index and the real-time passage attenuation coefficient are fused, and tensor dot product and topological filtering are performed at the floor connection nodes. Then, graph theory algorithms and optimization techniques are applied, combined with the horizontal weight base and the vertical emergency weight, to generate a flow diversion gradient coefficient. Then, through multiple iterations of optimization, ensure that the generated coefficient can accurately reflect the actual situation. Finally, generate a comprehensive flow diversion gradient coefficient for guiding actual operations.

[0236] For example, first, construct a three-dimensional evacuation decision tensor, where the x-axis is the channel quality index (result of step 301), the y-axis is the real-time passage attenuation coefficient, and the z-axis is the floor topological relationship. Second, perform a tensor dot product operation at the connection nodes to fuse the horizontal weight (channel quality accounts for 55%) and the vertical emergency weight (passage attenuation accounts for 45%). Then, apply a topological filter based on a graph convolutional network to eliminate spatial noise and enhance the key path features. Finally, extract the dominant mode through eigenvalue decomposition to generate a flow diversion gradient coefficient matrix distributed by floor, and this matrix is automatically updated every 30 seconds and synchronized to the emergency command system.

[0237] The following is a specific example:

[0238] In the scenario of a large shopping mall, the research team first analyzed the building information model data, obtained the density of evacuation nodes on each floor, and generated the node coverage intensity value based on the visible distance and the number of path branches. Subsequently, in combination with the evacuation ladder capacity threshold and the smoke diffusion rate, the real-time passage attenuation coefficient was calculated. Next, the research team fused the channel quality evaluation index with the real-time passage attenuation coefficient, performed tensor dot product and topological filtering at the floor connection nodes, and generated the diversion gradient coefficient. This series of steps not only improved the evacuation efficiency in case of emergency, but also enhanced the flexibility and response speed of the system, significantly improving the emergency communication management ability of the entire shopping mall.

[0239] In summary, steps 401 to 403 generate the diversion gradient coefficient by performing multi-dimensional fusion of the floor evacuation node density, the evacuation ladder capacity threshold, and the channel quality evaluation index. This solution significantly improves the efficiency and reliability of the indoor wireless communication system. This method not only improves the understanding of signal propagation loss and crowd density in complex environments, but also ensures efficient communication in case of emergency through dynamic adjustment strategies, providing strong support for intelligent building management and public safety.

[0240] To further improve the communication efficiency and signal coverage quality, in some embodiments, step 102 of detecting the difference in wall material absorption rate based on the coordinates of the metal-dense area in the three-dimensional communication feature map and constructing a three-dimensional signal attenuation field including elevator shaft reflection interference includes:

[0241] 501. The coordinates of the metal-dense area in the three-dimensional communication feature map are used to dynamically generate a metal interference intensity field through the correlation between the metal component distribution density and the signal reflection space;

[0242] In step 501, the coordinates of the metal-dense area in the three-dimensional communication feature map are used to dynamically generate a metal interference intensity field through the correlation between the metal component distribution density and the signal reflection space. The metal component distribution density refers to the quantity and distribution of metal structures inside the building. The signal reflection space correlation is based on the degree of influence of metal structures on wireless signal reflection. The metal interference intensity field is a three-dimensional model that shows the metal interference intensity in different areas. This process involves extracting the coordinates of the metal-dense area from the three-dimensional communication feature map, calculating the metal component distribution density, and adjusting parameters according to the signal reflection space correlation to finally generate the metal interference intensity field.

[0243] In the embodiments of the present application, first, the coordinates of metal-dense regions are extracted from the three-dimensional communication feature map. Then, the distribution density of metal components within these regions is analyzed, and the spatial correlation of their reflection of wireless signals is evaluated. Next, based on this data, an electromagnetic simulation technology is used to dynamically generate a metal interference intensity field. Specifically, a ray-tracing algorithm is employed to simulate the behavior of signals in metal-dense regions, and the interference intensity at each point is calculated. Finally, all the information is integrated to form a spatial model that comprehensively reflects the metal interference intensity.

[0244] For example, first, the three-dimensional coordinate data of metal components (such as steel bar meshes and elevator tracks) are extracted through a building BIM model, and the metal distribution density of each voxel (0.5m×0.5m×0.5m) is calculated using a kernel density estimation algorithm. Second, based on ray-tracing technology, the reflection paths of 2.4GHz / 5GHz band electromagnetic waves between metal structures are simulated, the multipath delay and signal strength attenuation of each path are statistically analyzed, and a spatial correlation matrix is generated. Then, a tensor product operation is performed on the density distribution and the spatial correlation, and measurement noise is eliminated through adaptive filtering. Finally, a metal interference intensity field with a scale of 0-100 is output, and the regions with an intensity >70 are marked as high-risk interference regions.

[0245] 502. Generate an absorption rate reference value through the mapping between the material type and the electromagnetic wave penetration loss, introduce the variation coefficients of the wall moisture content and thickness, and combine the environmental humidity data with the frequency-domain convolution of the reference value to generate a dynamic correction parameter for the absorption rate;

[0246] In step 502, the material type refers to the specific types of building materials, such as concrete, bricks, etc., and includes their physical and chemical properties, which are used to analyze the penetration ability of the materials to electromagnetic waves. The electromagnetic wave penetration loss is the loss amount of the penetration ability of the material to wireless signals, specifically manifested as the degree of signal attenuation, which is used to evaluate the propagation effect of signals in different materials. The absorption rate reference value is an initial absorption rate value obtained based on the relationship between the material type and the electromagnetic wave penetration loss, including the standard absorption rate of each material, which is used to preliminarily estimate the signal loss. The wall moisture content refers to the moisture content in the wall material, which affects the absorption rate of the signal and is used to adjust the absorption rate reference value to reflect the actual conditions. The thickness variation coefficient is the variation range of the wall thickness, which is used to correct the absorption rate reference value to adapt to the influence of different wall thicknesses. The environmental humidity data describes the humidity level in the surrounding air and is used to further adjust the absorption rate reference value. The dynamic correction parameter for the absorption rate is a final absorption rate value that comprehensively considers various factors and is used to accurately simulate the signal absorption situation in the actual environment.

[0247] In the embodiments of the present application, first, an absorption rate reference value is generated according to the mapping relationship between the material type and the electromagnetic wave penetration loss. Next, the wall moisture content and the thickness variation coefficient are introduced and analyzed in combination with the ambient humidity data. Then, the frequency-domain convolution technique is used to combine the above parameters with the absorption rate reference value to generate a dynamic correction parameter for the absorption rate. Specifically, a numerical simulation method is adopted to process complex data sets to ensure that the generated parameters can accurately reflect the actual situation. Finally, all the information is integrated to form a comprehensive dynamic correction parameter for the absorption rate.

[0248] For example, first, a material database (including 12 types of building materials such as concrete and brick walls) is established to store the penetration loss reference values of each material in the 1-6 GHz frequency band. Second, a millimeter-wave radar is used to detect the actual moisture content of the wall (accuracy ±1.5%), and the thickness variation coefficient (0.1-0.3) is obtained in combination with a laser rangefinder. Then, the real-time ambient humidity data is subjected to frequency-domain convolution with the reference value (sampling rate 10 Hz, window function length 256), and the characteristic frequency band (2.4-2.4835 GHz) is extracted through wavelet transform. Finally, a multiple linear regression model is applied to generate a dynamic correction parameter, and the correction amplitude can reach ±35% of the reference value.

[0249] 503. The cavity resonance frequency is analyzed based on the geometric dimensions of the shaft cavity, and the reflection path of the metal shaft is combined with the metal interference intensity field for attenuation compensation. The compensation result is phase-aligned with the resonance frequency to obtain the reflected interference intensity field;

[0250] In step 503, the geometric dimensions of the shaft cavity refer to the physical dimensions of the elevator shaft or other cavities, including height, width, and depth, which are used to calculate the resonance frequency of the cavity. The cavity resonance frequency is a resonance phenomenon at a specific frequency caused by the cavity geometry, including the resonance frequency and its corresponding acoustic characteristics, which are used to analyze the propagation characteristics of signals in the cavity. The reflection path of the metal shaft describes the reflection path of the signal in the metal shaft, including the number of reflections, angles, and path lengths, which are used to evaluate the attenuation of the signal in the shaft. Attenuation compensation is a process of compensating for the energy loss of the signal caused by various factors during transmission, including compensation algorithms and parameters, which are used to restore the signal strength. The reflected interference intensity field is a spatial model that describes the reflected interference intensity at different positions, including the reflected interference intensity values at each point, which are used to optimize the signal transmission path and reduce the reflected interference.

[0251] In the embodiment of the present application, first, analyze the geometric dimensions of the shaft cavity and calculate its resonant frequency. Next, combine the metal duct reflection path with the metal interference intensity field generated in step 501 for attenuation compensation. Then, adjust the compensation result to align with the phase of the resonant frequency to generate a reflected interference intensity field. Specifically, use acoustic simulation technology and signal processing algorithms to optimize the compensation effect to ensure that the signal can maintain a high quality when passing through the metal duct. Finally, integrate all the information to form a comprehensive reflected interference intensity field.

[0252] For example, first, calculate the cavity resonant frequency based on the elevator shaft CAD model (formula: f = c / 2L, where c is the speed of light and L is the characteristic dimension). The typical shaft resonant frequency range is 300 - 800 MHz. Second, use the FDTD (finite-difference time-domain) method to simulate the electromagnetic wave propagation in the metal duct, and perform path loss compensation (compensation amount 3 - 8 dB) in combination with the interference intensity field in step 501. Then, align the compensated signal with the phase of the resonant frequency through Hilbert transform (phase difference <5°) to generate an intensity field containing reflected gain / loss values, where the interference intensity in the reflected hot spot area is increased by 15 - 20 dB.

[0253] 504. By fusing the metal interference intensity field, the absorption rate dynamic correction parameter, and the reflected interference intensity field, polarize and weight the metal interference gradient and the absorption rate parameter and embed them into the reflected interference phase constraint to generate a three-dimensional signal attenuation field.

[0254] In step 504, the metal interference gradient describes the trend of the metal interference intensity changing with space, including the interference intensity change rate at each point, and is used to evaluate the degree of metal interference on the signal at different positions. The polarization weighting of the absorption rate parameter is to perform a weighting process on the absorption rate parameter to reflect its influence in different directions, including a weighting coefficient and a direction vector, and is used to more accurately simulate the signal propagation in a complex environment. The reflected interference phase constraint is a technical means for correcting the phase of the reflected signal, including a phase adjustment algorithm and parameters, and is used to ensure that the signal maintains the correct phase when passing through the reflection path. The three-dimensional signal attenuation field is a three-dimensional space signal attenuation model formed by comprehensively considering various factors, including the signal attenuation value at each position, and is used to guide the design and optimization of the communication network to ensure stable signal transmission in a complex environment.

[0255] In the embodiment of the present application, first, fuse the metal interference intensity field, the absorption rate dynamic correction parameter, and the reflected interference intensity field. Next, calculate the metal interference gradient and perform polarization weighting processing, in combination with the absorption rate parameter. Then, embed these parameters into the reflected interference phase constraint to generate a three-dimensional signal attenuation field. Specifically, use multi-dimensional data analysis technology and optimization algorithms to process complex multi-source data to ensure that the generated model can accurately reflect the actual situation. Finally, integrate all the information to form a comprehensive three-dimensional signal attenuation field.

[0256] For example, first, perform non-uniform grid fusion (with the grid size adaptively adjusted from 0.1 to 1 m) on the metal interference intensity field (step 501) and the reflection interference intensity field (step 503). Secondly, perform differential weighting (with a weight ratio of 4:3:3) on the absorption rate dynamic correction parameter (step 502) according to the polarization direction (X / Y / Z axes). Then, apply the phase constraint optimization algorithm (least squares fitting) to eliminate the phase conflict between the reflection paths. Finally, output a three-dimensional attenuation field containing the following elements: ① basic attenuation value (dB / m), ② multipath interference index, ③ polarization sensitivity matrix, with a data update frequency of 1 Hz, which can support accurate prediction in the millimeter wave band.

[0257] The following is a specific example:

[0258] In the scenario of a large shopping mall, the research team first parsed the building information model data to obtain the coordinates of the metal-dense areas and the distribution density of metal components. Then, using electromagnetic simulation technology, based on the distribution density of metal components and the spatial correlation of signal reflection, a metal interference intensity field was dynamically generated. Subsequently, the research team generated the absorption rate dynamic correction parameter according to the relationship between the wall material type and the electromagnetic wave penetration loss. Next, the research team analyzed the geometric dimensions of the elevator shaft, calculated its resonant frequency, and performed attenuation compensation by combining the metal shaft reflection path and the metal interference intensity field generated in step 501. Finally, the metal interference gradient was calculated and polarization weighting was performed. These parameters were embedded in the reflection interference phase constraint to generate a detailed three-dimensional signal attenuation field. This not only improves the stability and reliability of the communication system, but also enhances the flexibility and response speed of the system, significantly improving the emergency communication management ability of the entire shopping mall.

[0259] In summary, steps 501 to 504 generate a three-dimensional signal attenuation field by performing multi-dimensional fusion of the coordinates of the metal-dense areas, the difference in the absorption rate of the wall materials, and the reflection interference of the elevator shaft. This solution significantly improves the efficiency and reliability of the indoor wireless communication system. This method not only improves the understanding of signal propagation loss and crowd density in complex environments, but also ensures efficient communication in emergency situations through dynamic adjustment strategies, providing strong support for intelligent building management and public safety.

[0260] To further improve the accuracy of the attenuation compensation for the reflection path in the shaft cavity, in some embodiments, the cavity resonant frequency obtained by analyzing the geometric dimensions of the shaft cavity in step 503, and the metal shaft reflection path is combined with the metal interference intensity field for attenuation compensation, and the compensation result is phase-aligned with the resonant frequency to obtain the reflection interference intensity field, including:

[0261] 601. Combine the rigid node constraint standing wave mode analysis of the building vertical channel, calculate and analyze the geometric dimensions of the shaft cavity through the ratio of the cross-sectional perimeter to the height, and generate the cavity resonance frequency;

[0262] In step 601, the rigid node constraint standing wave mode analysis of the building vertical channel refers to the structural analysis of the internal vertical channels (such as elevator shafts) of the building to determine their inherent vibration modes. This includes evaluating factors such as the positions of rigid nodes, material properties, and structural designs for predicting possible standing wave modes. The geometric dimensions of the shaft cavity refer to the specific physical dimensions of the elevator shaft or other cavities, including the cross-sectional shape, perimeter, and height. These data are used to calculate the resonance frequency of the cavity. The ratio of the cross-sectional perimeter to the height is an important parameter describing the geometric characteristics of the cavity and is used to analyze its acoustic properties. The cavity resonance frequency is the resonance phenomenon at a specific frequency caused by the cavity geometry, including the resonance frequency and its corresponding acoustic properties, which are used to optimize the signal transmission path.

[0263] In the embodiment of the present application, first, parse the building information model data to obtain the geometric dimensions of the shaft cavity, especially the ratio of the cross-sectional perimeter to the height. Then, based on these data, apply the standing wave mode analysis technology to calculate the inherent vibration modes of the cavity. Next, use the acoustic simulation tool to simulate the responses at different frequencies and identify the resonance frequency of the cavity. Specifically, adopt the finite element method or the boundary element method to process complex multi-dimensional data sets to ensure that the generated frequencies can accurately reflect the actual situation. Finally, integrate all the information to generate a comprehensive list of cavity resonance frequencies.

[0264] For example, first, extract the accurate geometric parameters of the shaft cavity from the building BIM model (the cross-sectional perimeter error < 1 cm, the height error < 3 cm), and calculate the perimeter-height ratio (P / H ratio) as the key characteristic parameter. Second, adopt the finite element modal analysis method to solve the characteristic frequency equation satisfying the rigid boundary conditions in the frequency band of 10 - 3000 MHz, and identify the first 10 resonance modes (including TE and TM modes). Then, verify the simulation results through experimental measurements (perform sweep frequency tests using a vector network analyzer with an accuracy of ±0.5 MHz), and establish the mapping relationship between the P / H ratio and the fundamental frequency: f0 = kc / √(P×H), where k is the shape factor (take 0.49 for a rectangular cross-section), and √ represents the square root. Finally, output a parameter table containing the resonance frequencies of each order (f0, 2f0, 3f0...) and the corresponding Q values for subsequent phase matching.

[0265] 602. Attenuate and compensate the reflection path of the metal shaft through the multipath reflection characteristics of the metal interference intensity field, and combine the geometric projection of the Fresnel reflection coefficient and the signal incident angle to match the harmonic order of the cavity resonance frequency to generate the reflection path attenuation compensation parameter;

[0266] In step 602, the metal shaft reflection path describes the reflection path of the signal in the metal shaft, including the number of reflections, angles, and path lengths. The multipath reflection characteristic refers to the phenomenon of signal reflection on multiple paths, which affects the signal quality. The Fresnel reflection coefficient is a physical quantity that describes the reflection ratio of the signal at the interface, taking into account the signal incident angle and material properties. The geometric projection of the signal incident angle is used to calculate the behavior of the signal at the interface according to the incident angle, and is used to adjust the attenuation compensation of the reflection path. The reflection path attenuation compensation parameter is a final compensation parameter that comprehensively considers various factors and is used to reduce the signal loss on the reflection path.

[0267] In the embodiments of the present application, first, based on the multipath reflection characteristics of the metal interference intensity field, the reflection path in the metal shaft is analyzed. Then, in combination with the Fresnel reflection coefficient and the geometric projection of the signal incident angle, attenuation compensation is performed. Then, the harmonic order of the cavity resonance frequency is matched to further optimize the compensation effect. Specifically, ray tracing algorithms and electromagnetic simulation techniques are used to process complex reflection paths to ensure that the generated compensation parameters can accurately reflect the actual situation. Finally, all information is integrated to generate a comprehensive reflection path attenuation compensation parameter.

[0268] For example, first, use a ray tracing engine to simulate the multipath propagation in the metal shaft (minimum number of rays: 5000), and record the time delay, incident angle, and number of reflections of each path. Secondly, based on the metal interference intensity field in step 501, a compensation amount that decays exponentially with distance is applied to each path (compensation coefficient α = 0.3 - 0.7 dB / m). Then, calculate the Fresnel coefficients of each reflection point (considering the polarization mode, accuracy 0.01), and generate path weights in combination with the geometric projection of the incident angle (angle resolution 1°). Finally, match the compensated path response with the harmonic components (n×f0) of the cavity resonance frequency, and optimize through the least mean square error algorithm to obtain a frequency-related attenuation compensation parameter matrix (dimension: distance × frequency × angle).

[0269] 603. Perform a helical convolution on the radial gradient of the reflection path attenuation compensation parameter and the phase of the cavity resonance frequency, and generate a reflection interference intensity field after truncating with the standing wave energy threshold.

[0270] In step 603, the radial gradient of the reflection path attenuation compensation parameter describes the trend of the reflection path attenuation compensation parameter changing with space, including the change rate at each point, and is used to evaluate the attenuation degree of the signal at different positions. The phase of the cavity resonance frequency is the phase distribution of the signal in the cavity at a specific frequency and is used to correct the signal phase on the reflection path. The helical convolution operation refers to the process of combining the radial gradient of the compensation parameter with the resonance frequency phase information. The standing wave energy threshold is the standard value for screening the standing wave energy, and only the effective reflection interference components are retained. The reflection interference intensity field is a three-dimensional model that shows the reflection interference intensity in different regions.

[0271] In the embodiments of the present application, first, the radial gradient of the reflection path attenuation compensation parameter is calculated and helically convolved with the phase of the cavity resonance frequency. Then, the standing wave energy threshold truncation technique is applied to remove the low-energy part and optimize the signal clarity. Then, based on the above processing results, a reflection interference intensity field is generated. Specifically, numerical simulation methods and advanced data analysis tools are used to process complex multi-source data to ensure that the generated model can accurately reflect the actual situation. Finally, all the information is integrated to form a comprehensive reflection interference intensity field.

[0272] For example, first, the radial gradient of the compensation parameter (∂C / ∂r, step size 0.1 m) is calculated in the cylindrical coordinate system and helically convolved with the resonance phase distribution (∂ϕ / ∂z) (kernel size 5×5). Second, the standing wave energy threshold (usually -20 dB of the maximum value) is set, and morphological opening operation is used to eliminate discrete noise. Then, the complete reflection interference intensity field (mesh accuracy 0.2 m) is reconstructed by three-dimensional interpolation, and special regions are marked: ① phase inversion region (red mark) ② strong interference region (yellow mark) ③ safe transmission region (green mark). The finally output intensity field contains vector data (E-field / H-field distribution) and scalar data (power density) and supports parallel display of multiple frequency bands (700 MHz / 2.6 GHz / 3.5 GHz).

[0273] To sum up, steps 601 to 603 generate the key frequencies affecting wireless signal transmission through the analysis of the characteristics of the building vertical channel, and finally generate a reflection interference intensity field reflecting the characteristics of the actual environment by accurately calculating the reflection path attenuation compensation parameter, which significantly improves the accuracy and stability of wireless signal transmission in complex building structures. Especially in the shaft environment with a large amount of metal materials, it can effectively reduce signal loss and improve communication quality.

[0274] To solve the problem of insufficient accuracy of wireless signal coverage in complex building environments, in some embodiments, in step 103, predicting the spatio-temporal distribution of the short message request volume in each zone according to the communication blind zone boundary data in the three-dimensional signal attenuation field and combining the position coordinates of the fire channel includes:

[0275] 701. Discretize the communication dead zone boundary data into three - dimensional grid cells, calculate the signal attenuation coefficient of the three - dimensional grid cells and their three - dimensional Manhattan distance to the nearest fire - fighting access entrance, and generate an accessibility matrix.

[0276] In step 701, the communication dead zone boundary data refers to the boundary information of the area where normal communication cannot be carried out due to signal attenuation. The three - dimensional grid cell is the basic unit after discretizing space and is used to describe the attributes within a specific area. The signal attenuation coefficient represents the attenuation degree of the wireless signal when passing through different materials or media. The three - dimensional Manhattan distance is the shortest path distance between two points in three - dimensional space and is usually used to evaluate the accessibility between two points. The accessibility matrix is a data structure containing the distances from each grid cell to the nearest fire - fighting access entrance and the signal attenuation coefficient, which is used to guide the evacuation path planning.

[0277] In the embodiment of the present application, first, discretize the communication dead zone boundary data into three - dimensional grid cells. Then, use electromagnetic simulation technology to calculate the signal attenuation coefficient of each grid cell. Next, based on the three - dimensional Manhattan distance algorithm, calculate the distance from each grid cell to the nearest fire - fighting access entrance. Finally, integrate this information to generate a comprehensive accessibility matrix. Specifically, use graph - theory algorithms to process complex multi - dimensional data sets to ensure that the generated matrix can accurately reflect the actual situation.

[0278] For example, first, divide the building space into three - dimensional grid cells of 0.5m×0.5m×0.5m, and calculate the signal attenuation coefficient of each cell (accuracy ±0.5dB) through ray - tracing technology. Second, calculate the three - dimensional Manhattan distance from each grid to the nearest fire - fighting access based on the A* algorithm (considering the wall - penetration penalty coefficient). Then, construct an accessibility matrix containing the following elements: ① grid coordinates ② attenuation coefficient ③ shortest distance ④ path key points, with a data update frequency of 1Hz. Finally, normalize the accessibility matrix (in the range of 0 - 1), and convert the distance value into an accessibility score (1 represents the best).

[0279] 702. Construct a dynamic path network based on the fire - fighting access location coordinates, divide time windows and associate historical SMS request volume data, and generate a request volume change baseline for the three - dimensional grid cells.

[0280] In step 702, the fire escape location coordinates refer to the specific location information of the safety exits and evacuation routes within a building. The dynamic path network is a path selection system that is updated in real time according to time and the flow of people. The time window divides a continuous time period into several time segments for more refined time series analysis. The historical SMS request volume data records the number of SMS requests in each region over a past period. The request volume change baseline is a baseline value that reflects the trend of SMS request volume changes in different time periods and is used to predict future request volumes.

[0281] In the embodiments of the present application, first, a dynamic path network is constructed based on the fire escape location coordinates. Then, the time window is divided and associated with the historical SMS request volume data. Next, using the time series analysis method, the request volume change baseline for each three-dimensional grid cell is generated. Specifically, machine learning algorithms are used to process complex time series data to ensure that the generated baseline can accurately reflect the actual demand changes. Finally, all the information is integrated to generate a comprehensive request volume change baseline.

[0282] For example, first, the time window is divided at 15-minute intervals and associated with the historical SMS request data (including location, time, and device type) for the past 30 days. Second, the LSTM network is used to extract spatio-temporal features to generate the hourly request volume patterns for each grid cell. Then, the real-time people flow heat map (from video analysis) is superimposed for dynamic correction to establish a baseline model including the following dimensions: ① weekday / holiday pattern ② morning and evening peak characteristics ③ impact factor of emergencies. Finally, a request volume prediction baseline with an accuracy of ±5% is output.

[0283] 703. Perform spatial superposition of the three-dimensional grid cells and the dynamic path network, calculate the shortest connected path and fuse the wall penetration loss compensation value to construct a path loss correction model;

[0284] In step 703, the dynamic path network is a path selection system that is updated in real time according to time and the flow of people. The shortest connected path refers to the shortest path from one point to another, taking into account various obstacles and path conditions. The wall penetration loss compensation value is a parameter used to compensate for the energy loss of the signal when passing through the wall. The path loss correction model is a path loss adjustment model formed by comprehensively considering various factors and is used to optimize the signal transmission path and reduce the loss on the path.

[0285] In the embodiments of the present application, first, a three-dimensional grid unit is spatially superimposed with a dynamic path network. Then, the shortest connected path is calculated, and the wall penetration loss compensation value is fused. Next, based on the above processing results, a path loss correction model is constructed. Specifically, graph theory algorithms and ray tracing techniques are used to handle complex path problems, ensuring that the generated model can accurately reflect the actual situation. Finally, all information is integrated to form a comprehensive path loss correction model.

[0286] For example, first, a three-dimensional overlay (with a Z-axis accuracy of 0.1 m) is performed on the dynamic path network (generated by the Dijkstra algorithm) and the grid unit. Second, the penetration loss compensation value of each path is calculated ( +8 dB for concrete walls, +12 dB for metal doors). Then, the optimal path weight is determined through Bayesian optimization (distance weight 60%, signal quality 40%), and a correction model including the following parameters is established: ① the path loss reference value, ② the real-time correction coefficient, and ③ the redundant path score. The parameters of the model are automatically optimized every 5 minutes.

[0287] 704. Perform a time series decomposition on the signal attenuation coefficient to obtain a periodic component and a residual component, and combine the path loss correction model to generate the real-time signal coverage quality index of the three-dimensional grid unit;

[0288] In step 704, the signal attenuation coefficient represents the attenuation degree of the wireless signal when passing through different materials or media. Time series decomposition is a process of decomposing time series data into a periodic component and a residual component, used to identify the long-term trend and short-term fluctuations in the data. The periodic component is the part that repeats in the data, reflecting regular changes. The residual component is the random part of the data, reflecting unpredictable changes. The real-time signal coverage quality index is an index formed by comprehensively considering various factors and is used to evaluate the quality of the current signal coverage.

[0289] In the embodiments of the present application, first, a time series decomposition is performed on the signal attenuation coefficient to obtain a periodic component and a residual component. Then, in combination with the path loss correction model, the real-time signal coverage quality index of each three-dimensional grid unit is calculated. Specifically, time series analysis methods are used to handle complex signal data, ensuring that the generated index can accurately reflect the actual situation. Finally, all information is integrated to form a comprehensive real-time signal coverage quality index.

[0290] For example, first, perform STL time series decomposition on the attenuation coefficient (period = 24 hours) to separate the diurnal cycle component and the emergency event residuals. Second, perform a convolution operation (kernel size 3×3×3) on the path loss correction result and the cycle component. Then, normalize it to a quality index of 0 - 100 through the Sigmoid function, and set three-level thresholds: excellent (>80), qualified (60 - 80), poor (<60). The index is updated every 30 seconds and visually presented.

[0291] 705. Integrate the reachability matrix, the real-time signal coverage quality index, and the baseline of the request volume change into a multi-dimensional feature tensor, generate a spatio-temporal coupling feature map through a channel attention mechanism, and output the predicted value of the SMS request volume for the future time slice.

[0292] In step 705, the reachability matrix is a data structure containing the distance from each grid cell to the nearest fire channel entrance and the signal attenuation coefficient. The real-time signal coverage quality index is an indicator formed by comprehensively considering various factors and is used to evaluate the quality of the current signal coverage. The baseline of the request volume change is a reference value reflecting the change trend of the SMS request volume in different time periods. The multi-dimensional feature tensor is a multi-dimensional data structure containing various feature information and is used as the input for the deep learning model. The channel attention mechanism is a technology used to enhance the performance of the neural network and improves the prediction accuracy by focusing on important features. The spatio-temporal coupling feature map is a map that integrates time and space features and is used to predict the future SMS request volume.

[0293] In the embodiments of the present application, first, integrate the reachability matrix, the real-time signal coverage quality index, and the baseline of the request volume change into a multi-dimensional feature tensor. Then, apply the channel attention mechanism to generate a spatio-temporal coupling feature map. Then, based on the above processing results, output the predicted value of the SMS request volume for the future time slice. Specifically, use a deep learning algorithm to process complex multi-source data to ensure that the generated predicted value can accurately reflect the actual situation. Finally, integrate all information to form a comprehensive SMS request volume prediction model.

[0294] For example, first, construct a five-dimensional feature tensor (space x, y, z + time + quality index), and use 3D CNN to extract spatio-temporal features. Second, strengthen the key features (weight increase of 30 - 50%) through the channel attention mechanism (SE module). Then, use the TCN network to predict the request volume distribution for the next 15 - 60 minutes, and output a prediction result containing the following information: ① coordinates of the hot spot area ② expected request volume ③ confidence interval (90% confidence level). The system achieves a prediction accuracy of 85% in the actual measurement of large shopping malls.

[0295] The following is a specific example:

[0296] In an application scenario of a large shopping mall, the research team first analyzed the building information model data, obtained the communication blind zone boundary data, and discretized it into three-dimensional grid cells. Then, based on the position coordinates of the fire escape routes, a dynamic path network was constructed, time windows were divided, and a baseline for the change in request volume for each three-dimensional grid cell was generated. Next, the team calculated the shortest connected path and incorporated the wall penetration loss compensation value to construct a path loss correction model. The research team fused the reachability matrix, real-time signal coverage quality index, and the baseline of request volume change into a multi-dimensional feature tensor to obtain a spatio-temporal coupling feature map. Subsequently, based on the above processing results, the predicted value of the SMS request volume for future time slices was output. This not only improved the efficiency of the communication system but also enhanced the speed and accuracy of emergency response, providing a smoother and safer experience for customers.

[0297] In summary, steps 701 to 705, based on the communication blind zone boundary data and the position coordinates of the fire escape routes, generate an accurate reachability matrix and a baseline for the change in request volume, and through the path loss correction model and the spatio-temporal coupling feature map, significantly improve the accuracy and stability of wireless signal transmission in complex building structures. Especially in the case of a large amount of metal materials and complex layouts, it can effectively reduce signal loss and improve communication quality.

[0298] Figure 2 The following is a schematic structural diagram of a SMS sending and processing system based on scenario adaptation provided by an embodiment of the present application. As Figure 2 shown, the system includes:

[0299] An acquisition module 21, which generates a three-dimensional communication feature map containing signal propagation loss weights and crowd density gradients by analyzing the spatial topology data obtained from the building information model and combining the heat map of personnel residence collected by the infrared thermal imaging array;

[0300] A construction module 22, which detects the difference in the absorption rate of wall materials based on the coordinates of the metal-dense areas in the three-dimensional communication feature map and constructs a three-dimensional signal attenuation field including elevator shaft reflection interference;

[0301] A prediction module 23, which predicts the spatio-temporal distribution of SMS request volumes in each zone according to the communication blind zone boundary data in the three-dimensional signal attenuation field and in combination with the position coordinates of the fire escape routes;

[0302] A generation module 24, which is used to access the Internet SMS gateway cluster, generate a heat map layer of cross-operator network quality by collecting base station signaling plane data and the online status of user terminals, and map the heat map layer to the three-dimensional communication feature map to form a fused communication quality field; based on the spatio-temporal distribution of the SMS request volumes in each zone and the backlog volume in the gateway cache queue, and in combination with the network congestion hotspots in the fused communication quality field, generate a message forwarding rule sorted by priority;

[0303] A processing module 25, configured to perform an SMS sending operation according to the message forwarding rule.

[0304] Figure 2 The described SMS sending processing system based on scenario adaptation can execute Figure 1 The described SMS sending processing method based on scenario adaptation in the illustrated embodiment, the implementation principle and technical effects will not be elaborated further. For the SMS sending processing system based on scenario adaptation in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here in detail. Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A scenario-based SMS sending processing method, characterized in that include: By analyzing the building information model to obtain spatial topology data, combined with the heat map of personnel residence collected by the infrared thermal imaging array, a three-dimensional communication feature map containing signal propagation loss weights and crowd density gradients is generated; Based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map, the difference in the absorption rate of the wall material is detected to construct a three-dimensional signal attenuation field including the reflection interference of the elevator shaft; According to the communication blind area boundary data in the three-dimensional signal attenuation field, combined with the position coordinates of the fire channel, the temporal and spatial distribution of the SMS request volume in each partition is predicted; Accessing the Internet SMS gateway cluster, collecting base station signaling surface data and user terminal online status to generate a thermal map of cross-operator network quality, and mapping the thermal map to the three-dimensional communication feature map to form a fused communication quality field; Based on the spatiotemporal distribution of the SMS request volume in each partition and the backlog of the gateway cache queue, combined with the network congestion hotspots in the fusion communication quality field, a message forwarding rule sorted by priority is generated; The SMS sending operation is executed according to the message forwarding rule.

2. The method according to claim 1, wherein The access to the Internet SMS gateway cluster generates a thermal map of the cross-operator network quality by collecting base station signaling plane data and user terminal online status, including: Through the multi-operator protocol interface, each SMS gateway cluster is synchronously connected to obtain the signaling plane data of each base station in real time. The signaling plane data includes the control signaling interaction record between the base station and the user terminal, the terminal registration status and the signaling response delay; At the same time, the terminal measurement data reported by the user terminal is collected, including the reference signal received power, signal interference and noise ratio, and terminal moving speed; The signaling plane data collected in the same time window and the terminal measurement data are temporally and spatially matched according to the base station coverage area, abnormal terminal data are eliminated, and the valid data point set is retained. Based on the valid data point set, the signaling response delay distribution and signal strength distribution of each terminal within the coverage area of ​​each base station are extracted; Divide the base station coverage area into a number of grid cells according to the terminal location distribution density, each grid cell contains the measurement data of at least one terminal; The signaling response delay and signal strength in each grid cell are weightedly calculated to obtain the network quality score of the grid cell. The network quality scores of adjacent grid cells are smoothed using a spatial interpolation method to generate a thermal map of cross-operator network quality.

3. The method according to claim 1, wherein Mapping the thermal map to the three-dimensional communication feature map to form a fused communication quality field, including: Mapping the center coordinates of each grid cell in the network quality thermal map to a corresponding spatial position of the three-dimensional communication feature map; According to the crowd density gradient and signal propagation loss weight of the corresponding spatial position in the three-dimensional communication feature map, the contribution coefficient of the network quality score is adjusted to update the network quality score; The updated network quality score is superimposed on the communication parameters of the three-dimensional communication characteristic map to form a fused communication quality field.

4. The method according to claim 1, wherein The message forwarding rules sorted by priority are generated based on the spatiotemporal distribution of the SMS request volume of each partition and the backlog of the gateway cache queue, combined with the network congestion hotspots in the fusion communication quality field, including: Fuse the coordinates of high-density regions in the spatio-temporal distribution of the SMS request volumes of each partition with the group movement direction vectors, and calculate the matching degree between the movement direction clustering results and the historical request volume growth rate by associating the signal strength attenuation rate of network congestion hotspots, to generate the emergency channel weight coefficients; Based on the peak period characteristics in the gateway queue backlog, combine the network congestion hotspots with the signal attenuation gradient between adjacent access points, and generate the cross-floor traffic diversion ratio by matching the inflection points of the attenuation gradient with the physical topology distance; According to the geometric boundary data of the weak signal area, extract the coverage range of idle channels of neighboring access points within the network congestion hotspots, and generate the relay forwarding priority sequence by analyzing the correlation strength between the access point response delay and the signal compensation requirements; Perform multi-dimensional fusion of the emergency channel weight coefficients, cross-floor traffic diversion ratio, and relay forwarding priority sequence to construct a forwarding rule trigger matrix, and the forwarding rule trigger matrix is dynamically updated by associating the change rate of the group movement direction vector with the growth rate of the gateway buffer queue backlog; Based on the critical parameters of the forwarding rule trigger matrix, simulate message path conflicts in the digital twin environment, and generate the message forwarding rules sorted by priority by reversely correcting the matching degree threshold and signal compensation parameters; 5. The method according to claim 4, wherein The multi-dimensional fusion of the emergency channel weight coefficients, cross-floor traffic diversion ratio, and relay forwarding priority sequence to construct a forwarding rule trigger matrix includes: Fuse the emergency channel weight coefficients with the real-time load and historical traffic data of each channel, generate a basic constraint value in combination with the channel length, and perform dynamic compensation by superimposing the obstacle density to obtain the channel quality evaluation index; Based on the floor evacuation node density, couple the cross-floor traffic diversion ratio with the evacuation ladder capacity threshold and the channel quality evaluation index to generate a traffic diversion gradient coefficient, and fuse the building load limit to output the traffic diversion ratio reference value; The relay forwarding priority sequence constructs a node coverage map through the multipath effect of wireless signals, fuses the traffic diversion gradient coefficient to generate a three-dimensional permeability surface, and extracts the curvature characteristics of the three-dimensional permeability surface to generate a priority sorting linked list; Input the channel quality evaluation index, traffic diversion ratio reference value, and priority sorting linked list into the building digital twin, extract implicit association rules through tensor decomposition and orthogonal recombination, and construct a forwarding rule trigger matrix.

6. The method according to claim 5, characterized in that The cross-floor traffic diversion ratio based on the floor evacuation node density, coupling the evacuation ladder capacity threshold and the channel quality evaluation index to generate a traffic diversion gradient coefficient includes: Generate a node coverage intensity value based on the visible distance and the number of path branches of the floor evacuation node density, trigger connectivity compensation through the visible distance, and correct the topological strength attenuation by the number of path branches; Based on the evacuation ladder capacity threshold and the smoke diffusion rate, generate a real-time traffic attenuation coefficient by asymmetrically superimposing the static capacity baseline and the dynamic attenuation factor, and introducing the node coverage intensity value at the same time; The channel quality evaluation index and the real-time traffic attenuation coefficient are fused through the horizontal weight base and the vertical emergency weight, and tensor dot product and topological filtering are performed at the floor connection nodes to generate a traffic diversion gradient coefficient.

7. The method according to claim 1, wherein Detecting the absorption rate difference of wall materials based on the coordinates of the metal-dense areas in the three-dimensional communication feature map, and constructing a three-dimensional signal attenuation field including elevator shaft reflection interference, comprising: The coordinates of the metal-dense areas in the three-dimensional communication feature map are used to dynamically generate a metal interference intensity field through the correlation between the metal component distribution density and the signal reflection space; The absorption rate reference value is generated by mapping the material type and the electromagnetic wave penetration loss, and the variation coefficients of the wall moisture content and thickness are introduced. Combining the environmental humidity data and the frequency-domain convolution of the reference value, a dynamically corrected absorption rate parameter is generated; The cavity resonance frequency is obtained by analyzing the geometric dimensions of the shaft cavity. The metal shaft reflection path is combined with the metal interference intensity field for attenuation compensation, and the compensation result is phase-aligned with the resonance frequency to obtain the reflection interference intensity field; By fusing the metal interference intensity field, the dynamically corrected absorption rate parameter, and the reflection interference intensity field, the metal interference gradient and the absorption rate parameter are polarization-weighted and then embedded into the reflection interference phase constraint to generate a three-dimensional signal attenuation field.

8. The method according to claim 7, wherein The method for obtaining the cavity resonance frequency by analyzing the geometric dimensions of the shaft cavity, and combining the metal shaft reflection path with the metal interference intensity field for attenuation compensation, and phase-aligning the compensation result with the resonance frequency to obtain the reflection interference intensity field, comprises: Combining the rigid node constraint standing wave mode analysis of the building vertical channel, and analyzing the geometric dimensions of the shaft cavity through the calculation of the ratio of the cross-sectional perimeter to the height to generate the cavity resonance frequency; The metal shaft reflection path is attenuated and compensated through the multipath reflection characteristics of the metal interference intensity field, and combined with the geometric projection of the Fresnel reflection coefficient and the signal incident angle to match the harmonic order of the cavity resonance frequency, so as to generate the reflection path attenuation compensation parameter; The radial gradient of the reflection path attenuation compensation parameter is spirally convolved with the phase of the cavity resonance frequency, and after truncating by the standing wave energy threshold, the reflection interference intensity field is generated.

9. The method according to claim 1, wherein Predicting the spatio-temporal distribution of the short message request volume in each zone according to the communication blind area boundary data in the three-dimensional signal attenuation field and combining the fire channel position coordinates, comprises: Discretizing the communication blind area boundary data into three-dimensional grid cells, calculating the signal attenuation coefficient of the three-dimensional grid cells and their three-dimensional Manhattan distance to the nearest fire channel entrance to generate an accessibility matrix; Constructing a dynamic path network based on the fire channel position coordinates, dividing time windows and associating historical short message request volume data to generate a baseline of the request volume change of the three-dimensional grid cells; Performing spatial superposition of the three-dimensional grid cells and the dynamic path network, calculating the shortest connected path and fusing the wall penetration loss compensation value to construct a path loss correction model; Performing time series decomposition on the signal attenuation coefficient to obtain a periodic component and a residual component, and combining with the path loss correction model to generate a real-time signal coverage quality index of the three-dimensional grid cells; Fusing the accessibility matrix, the real-time signal coverage quality index, and the request volume change baseline into a multi-dimensional feature tensor, generating a spatio-temporal coupling feature map through a channel attention mechanism, and outputting the predicted value of the short message request volume in the future time slice.

10. A text message sending processing system based on scenario adaptation, characterized in that, Comprising: An acquisition module, which obtains spatial topology data by parsing a building information model, and combines a personnel occupancy thermal map collected by an infrared thermal imaging array to generate a three-dimensional communication feature map including signal propagation loss weights and crowd density gradients; A construction module, which detects the difference in the absorption rate of wall materials based on the coordinates of metal-dense areas in the three-dimensional communication feature map, and constructs a three-dimensional signal attenuation field including elevator shaft reflection interference; A prediction module, which predicts the spatio-temporal distribution of SMS request volumes in each zone according to the communication blind zone boundary data in the three-dimensional signal attenuation field and combines the position coordinates of fire corridors; A generation module, which is used to access an Internet SMS gateway cluster, generates a thermal layer of cross-operator network quality by collecting base station signaling plane data and user terminal online status, and maps the thermal layer to the three-dimensional communication feature map to form a fused communication quality field; based on the spatio-temporal distribution of SMS request volumes in each zone and the backlog volume of the gateway cache queue, and combines the network congestion hotspots in the fused communication quality field to generate a message forwarding rule sorted by priority; A processing module, which is used to perform SMS sending operations according to the message forwarding rule.

Images