Scene self-adaption-based short message issuing processing method and system

Abstract

The invention provides a short message issuing processing method and system based on scene self-adaption. The method comprises the steps that a three-dimensional communication characteristic spectrum is constructed through a building information model and infrared thermal imaging data, short message request distribution is analyzed and predicted in combination with a signal attenuation field, cross-operator network quality data is integrated to form a converged communication quality field, and finally short message issuing is intelligently scheduled according to the network state and service requirements. According to the method, accurate short message routing optimization based on real-time communication quality and crowd distribution in a complex building environment is realized. According to the technical scheme provided by the invention, the short message delivery rate and timeliness in a complex building environment can be remarkably improved, and particularly, the communication reliability in a crowded area and a signal blind area is remarkably improved.

Description

Technical Field

[0001] The present application relates to the technical field of SMS delivery processing, and in particular to a scenario-adaptive SMS delivery processing method and system. Background Art

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

[0003] At present, traditional SMS delivery methods mainly rely on fixed communication base stations and simple signal strength detection technology. These methods are usually unable to effectively deal with the signal attenuation problem in complex building environments, especially in metal-intensive areas or crowded areas, where signal interruption or delay is prone to occur. In addition, the existing technology lacks the ability to comprehensively analyze the structural characteristics of buildings and the dynamic changes of human flow, resulting in low efficiency of SMS delivery and failure to meet the communication needs in high-concurrency scenarios.

[0004] Existing technical solutions usually adopt static communication models and cannot adapt to changes in the building environment in real time. For example, during peak hours, crowds may cause local area signal overload, and the existing system cannot dynamically adjust bandwidth allocation or forwarding strategies. In addition, the existing solutions lack accurate analysis of building structures (such as wall materials and elevator shafts), resulting in inaccurate prediction of signal attenuation fields and inability to effectively eliminate communication blind spots. Therefore, there is an urgent need for a text message delivery 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 the present application provide a scenario-adaptive SMS sending and processing method and system to solve the problems of low efficiency and poor reliability of SMS sending and processing in the prior art.

[0006] In a first aspect, an embodiment of the present application provides a scenario-based adaptive SMS delivery processing method, including: 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; and the SMS sending operation is executed according to the message forwarding rule.

[0007] 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: 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.

[0008] Optionally, mapping the thermodynamic layer to the three-dimensional communication feature map to form a fused communication quality field includes: mapping the center coordinates of each grid cell in the network quality thermodynamic layer 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.

[0009] Optionally, the generating of message forwarding rules sorted by priority based on the spatiotemporal distribution of the amount of SMS requests in each partition and the amount of backlog in the gateway cache queue, combined with the network congestion hotspots in the fusion communication quality field, includes: The high-density area coordinates in the spatiotemporal distribution of the SMS request volume in each partition are integrated with the group moving direction vector, and the matching degree between the moving direction clustering result and the historical request volume growth rate is calculated by associating the signal strength attenuation rate of the network congestion hotspot, so as to generate the emergency channel weight coefficient; Based on the peak period characteristics of the gateway queue backlog, combined with the network congestion hotspot and the signal attenuation gradient of adjacent access points, the cross-floor diversion ratio is generated by matching the attenuation gradient inflection point with the physical topology distance; According to the geometric boundary data of the weak signal area, the coverage range of the idle channels of the adjacent access points in the network congestion hotspot is extracted, and the relay forwarding priority sequence is generated by analyzing the correlation strength between the access point response delay and the signal compensation demand; The emergency channel weight coefficient, the cross-floor diversion ratio and the relay forwarding priority sequence are multi-dimensionally integrated 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 cache queue backlog; Based on the critical parameters of the forwarding rule trigger matrix, the message path conflict is simulated in the digital twin environment, and the message forwarding rules sorted by priority are generated by reversely correcting the matching threshold and signal compensation parameters.

[0010] Optionally, the emergency channel weight coefficient, the cross-floor diversion ratio and the relay forwarding priority sequence are multi-dimensionally integrated to construct a forwarding rule trigger matrix, including: The emergency channel weight coefficient is integrated with the real-time load and historical traffic data of each channel, combined with the channel length to generate a basic constraint value, and the obstacle density is superimposed for dynamic compensation to obtain a channel quality evaluation index; The cross-floor diversion ratio is based on the floor escape node density, coupled with the escape ladder capacity threshold and the channel quality evaluation index to generate a diversion gradient coefficient, and the building load-bearing limit is integrated to output a diversion ratio benchmark value; The relay forwarding priority sequence constructs a node coverage map through the wireless signal multipath effect, 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 chain list; The channel quality evaluation index, diversion ratio benchmark value and priority sorting list are input into the building digital twin, and the implicit association rules are extracted through tensor decomposition and orthogonal recombination to construct the forwarding rule trigger matrix.

[0011] Optionally, the step of generating a diversion gradient coefficient based on the cross-floor diversion ratio based on the floor escape node density, coupling the escape ladder capacity threshold and the channel quality evaluation index includes: The density of floor escape nodes is based on the visible distance and the number of path branches to generate the node coverage strength value. The connectivity compensation is triggered by the visible distance, and the number of path branches corrects the topological strength attenuation. Based on the escape ladder capacity threshold and the smoke diffusion rate, a real-time traffic attenuation coefficient is generated by asymmetric superposition of the static capacity baseline and the dynamic attenuation factor and the introduction of the node coverage strength value; 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 the diversion gradient coefficient.

[0012] Optionally, the detecting the difference in absorption rate of wall materials based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map and constructing a three-dimensional signal attenuation field including elevator shaft reflection interference includes: The coordinates of the metal-dense area of ​​the three-dimensional communication characteristic map dynamically generate the metal interference intensity field through the correlation between the distribution density of metal components and the signal reflection space; The absorption rate benchmark value is generated by mapping the material type and the electromagnetic wave penetration loss, and the wall moisture content and thickness variation coefficient are introduced. The ambient humidity data is combined with the frequency domain convolution of the benchmark value to generate the absorption rate dynamic correction parameter. The cavity resonance frequency is obtained according to the analysis of the geometric dimensions of the shaft cavity, and the metal shaft reflection path is combined with the metal interference intensity field for attenuation compensation. The compensation result is aligned with the phase of the resonance frequency to obtain the reflection interference intensity field; By fusing the metal interference intensity field, the dynamic correction parameter of the absorption rate and the reflection interference intensity field, the metal interference gradient and the absorption rate parameter are polarization-weighted and embedded in the reflection interference phase constraint to generate a three-dimensional signal attenuation field.

[0013] Optionally, the cavity resonance frequency is obtained according to the analysis of 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 aligned with the phase of the resonance frequency to obtain the reflection interference intensity field, including: Combined with the standing wave modal analysis of the rigid node constraints of the vertical channel of the building, the geometric dimensions of the shaft cavity are analyzed by calculating the ratio of the cross-sectional perimeter to the height to generate the cavity resonance frequency; The attenuation compensation of the metal well reflection path is performed through the multipath reflection characteristics of the metal interference intensity field, and the harmonic order of the cavity resonance frequency is matched in combination with the geometric projection of the Fresnel reflection coefficient and the signal incident angle 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 the reflection interference intensity field is generated after the standing wave energy threshold is truncated.

[0014] Optionally, the predicting the spatiotemporal distribution of the amount of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and in combination with the position coordinates of the fire channel includes: Discretize the communication blind area boundary data into three-dimensional grid cells, calculate the signal attenuation coefficient of the three-dimensional grid cells and the three-dimensional Manhattan distance to the nearest fire channel entrance, and generate an accessibility matrix; Building a dynamic path network based on the fire channel location coordinates, dividing the time window and associating the historical SMS request volume data, and generating a request volume change baseline for the three-dimensional grid unit; The three-dimensional grid unit is spatially superimposed with the dynamic path network, the shortest connected path is calculated and the wall penetration loss compensation value is integrated 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 the path loss correction model to generate a real-time signal coverage quality index of the three-dimensional grid unit; The reachability matrix, real-time signal coverage quality index and request volume change baseline are fused into a multi-dimensional feature tensor, a spatiotemporal coupling feature map is generated through a channel attention mechanism, and the predicted value of the SMS request volume in the future time slice is output.

[0015] In a second aspect, the present application provides a scenario-adaptive SMS delivery processing system, including: The acquisition module obtains spatial topology data by analyzing the building information model, and generates a three-dimensional communication feature map containing signal propagation loss weights and crowd density gradients by combining the personnel residence heat map collected by the infrared thermal imaging array; A construction module detects the difference in absorption rate of wall materials based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map, and constructs a three-dimensional signal attenuation field including the reflection interference of the elevator shaft; A prediction module, which predicts the spatiotemporal distribution of the amount of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and the position coordinates of the fire channel; A generation module is used to access the Internet SMS gateway cluster, collect base station signaling plane data and user terminal online status to generate a thermal layer of cross-operator network quality, and map the thermal layer to the three-dimensional communication feature map to form a fused communication quality field; 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 fused communication quality field, generate message forwarding rules sorted by priority; The processing module is used to execute the SMS sending operation according to the message forwarding rule.

[0016] In the embodiment of the present application, by analyzing the building information model to obtain spatial topological 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, which can accurately reflect the signal propagation characteristics and crowd concentration degree of different regions, and provide a data basis for subsequent communication optimization. Based on the coordinates of the metal-dense area in the three-dimensional communication feature map, the difference in the absorption rate of the wall material is detected, and a three-dimensional signal attenuation field containing the reflection interference of the elevator shaft is constructed. It can identify key influencing factors such as the metal structure, wall material and elevator shaft in the building, accurately calculate signal attenuation and reflection interference, and form a signal propagation model that is more in line with the real environment, thereby improving the accuracy of communication quality prediction. 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 spatiotemporal distribution of the SMS request volume in each partition is predicted. It can combine the signal blind area and the personnel flow path (such as the fire channel) to dynamically predict the SMS service demand in different areas at different times, and provide an accurate spatiotemporal distribution reference for resource scheduling. Access the Internet SMS gateway cluster, collect base station signaling data and user terminal online status to generate a cross-operator network quality thermal map, and map the thermal map to the three-dimensional communication feature map to form a fused communication quality field, which can integrate multi-operator network status data, evaluate the communication quality of each area in real time, and integrate with the three-dimensional environment model to form a global communication situation awareness, ensuring that the SMS delivery strategy adapts to the actual network conditions. 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 fused communication quality field, generate message forwarding rules sorted by priority; execute SMS delivery operations according to the message forwarding rules, which can comprehensively consider business needs, network load and communication quality, dynamically optimize SMS delivery priorities, ensure efficient delivery of SMS in high-demand areas and areas with good networks, and at the same time alleviate the transmission pressure in congested areas and improve overall communication efficiency.

[0017] The technical solution of this application significantly improves the reliability and timeliness of SMS transmission in complex building environments through multi-dimensional data fusion and dynamic optimization, especially in crowded areas, signal blind spots and network congestion scenarios.

[0018] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 A flowchart of a scenario-adaptive SMS sending processing method provided by the present application is shown; Figure 2 A structural schematic diagram of a scenario-adaptive SMS sending and processing system provided by the present application is shown. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0022] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, 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 of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0023] The researchers took into account that in the terminal SMS delivery service, there are usually problems with communication efficiency in complex building environments. Traditional SMS delivery methods are difficult to cope with multiple challenges such as signal attenuation, crowd gathering and emergency communications. In order to solve these problems, an in-depth analysis of the building structure characteristics, signal propagation laws and dynamic changes in human flow was conducted, and a scenario-based SMS delivery processing method 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 the distribution of SMS requests, and integrates cross-operator network quality data to form a fused communication quality field, and finally intelligently schedules SMS delivery according to network status and business needs. This method realizes accurate SMS routing optimization based on real-time communication quality and crowd distribution in complex building environments.

[0024] The technical solution of this application can be mainly applied to the following technical scenarios: Large commercial complexes: In large commercial complexes such as shopping malls and shopping centers, the crowd density is high and unevenly distributed, which can easily lead to local signal overload or communication blind spots. This solution dynamically adjusts the SMS delivery strategy by analyzing the heat map of crowd flow and building structure characteristics in real time to ensure smooth communication.

[0025] 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, accurately predicts communication blind spots, and optimizes the SMS delivery path based on the location of the emergency channel to ensure the reliability of medical communications.

[0026] Office buildings and industrial parks: Office buildings and industrial parks usually contain a lot of metal structures and elevator shafts, which are prone to signal reflection and interference. This solution optimizes the signal propagation path and improves the efficiency of SMS delivery by analyzing the building information model.

[0027] Transportation hubs and airports: In transportation hubs such as airports and train stations, there are dense and mobile traffic, and communication demand fluctuates greatly. This solution predicts the temporal and spatial distribution of SMS requests and dynamically allocates bandwidth resources to ensure communication stability in high-concurrency scenarios.

[0028] Emergency communication scenario: In natural disasters or public safety incidents, the stability of the communication network is crucial. This solution combines the location of the fire channel and the bandwidth allocation strategy of the emergency channel to ensure smooth communication in emergency situations.

[0029] In the process of developing this solution, we analyzed the communication pain points in complex building environments based on the actual needs of terminal SMS delivery services. By combining building information modeling (BIM) and infrared thermal imaging technology, we constructed a three-dimensional communication feature map to accurately depict the spatial distribution of signal propagation loss and crowd density. Furthermore, we introduced digital twin technology to simulate the communication performance in crowd gathering scenarios, and optimized the SMS delivery efficiency by dynamically adjusting the forwarding rules. This process not only solves the communication blind spot problem that traditional methods cannot cope with, but also provides reliable technical support for emergency communication scenarios.

[0030] The technical solution of this application achieves precise optimization of the SMS delivery process by combining building information modeling, infrared thermal imaging technology and digital twin technology. This solution is suitable for large commercial complexes, hospitals, office buildings, transportation hubs and emergency communication scenarios, and can effectively improve the efficiency of SMS transmission and ensure the stability and reliability of communications. By analyzing the structural characteristics of buildings and the dynamic changes of human flow in real time, this solution provides an innovative solution for terminal SMS delivery services.

[0031] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0032] Figure 1 A flowchart of a method for processing SMS delivery based on scenario adaptation is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes: 101. Obtain spatial topological data by analyzing the building information model, and generate a three-dimensional communication feature map containing signal propagation loss weights and crowd density gradients by combining the personnel residence heat map collected by the infrared thermal imaging array; In this step, the building information model is a virtual building model created by digital means to provide accurate spatial structure and material information.

[0033] Spatial topological data are obtained by parsing the building information model, including building structure information, which is used to construct accurate indoor maps and provide a basis for subsequent analysis.

[0034] The heat map of human presence is collected by an infrared thermal imaging array, showing the distribution of crowd density in different areas and helping to identify areas with high crowd density.

[0035] The signal propagation loss weight reflects the impact of building materials on wireless signals and is calculated based on the material properties. The crowd density gradient shows the distribution trend of the crowd in space and is used to evaluate the flow of people in the area.

[0036] The three-dimensional communication feature map integrates the signal propagation loss weight and the crowd density gradient to guide wireless network planning and form a comprehensive understanding of the wireless communication environment within the building.

[0037] In the embodiment of the present application, first, the spatial topology data is extracted from the building information model. Then, the infrared thermal imaging array is used to collect the thermal map of the personnel in the current environment. Then, based on these data, the algorithm is applied to calculate the signal propagation loss weight of each area, and the crowd density gradient is determined according to the distribution of personnel. Finally, all the information is combined to form a detailed three-dimensional communication feature map.

[0038] In a large shopping mall, the research team first analyzed the building information model data of the shopping mall and obtained detailed building structure and material information, including key data such as wall material, metal structure distribution, and elevator shaft location. At the same time, the team deployed a high-precision infrared thermal imaging array to monitor the customer's stay in real time and generate a high-resolution heat map of customer stay.

[0039] Next, the team calculated the signal propagation loss weight of each area based on the electromagnetic properties of building materials (such as the difference in signal absorption rates of concrete, metal, and glass). At the same time, combined with the heat map of personnel residence, the crowd density gradient was analyzed to identify high-density areas (such as shopping mall entrances and dining areas) and low-density areas (such as warehouses and equipment rooms).

[0040] Finally, the team integrated these 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 crowd density. This step provides basic data support for subsequent communication optimization, enabling the team to accurately identify high-density areas and weak-signal areas, and provides a scientific basis for dynamically adjusting SMS delivery strategies.

[0041] 102. Based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map, detect the difference in absorption rate of the wall material and construct a three-dimensional signal attenuation field including the reflection interference of the elevator shaft; In this step, the metal-dense area coordinates refer to the location information of a building containing a large amount of metal materials, and the metal-dense area detector is used to identify the area with concentrated metal components in the building.

[0042] The difference in wall material absorption rate describes the absorption capacity of different wall materials to wireless signals. Use the wall absorption rate difference calculator to evaluate the impact of different materials on signals.

[0043] Elevator shaft reflection interference is a signal reflection problem caused by the metal structure in the elevator shaft, which causes the signal strength to weaken.

[0044] The 3D signal attenuation field is a 3D model that shows the signal strength attenuation at various locations inside a building, taking into account factors such as differences in wall material absorption and interference from elevator shaft reflections.

[0045] In the embodiment of the present application, the metal-dense areas and their coordinates in the building are first identified, and the differences in absorptivity of different wall materials are measured and recorded. Then, these data are used to construct a three-dimensional signal attenuation field containing elevator shaft reflection interference. This process requires a detailed analysis of the distribution of metal structures inside the building and experimental measurement of the differences in absorptivity of different wall materials. The resulting three-dimensional signal attenuation field can intuitively display the signal strength attenuation at various locations inside the building, providing a basis for subsequent communication optimization.

[0046] In the example of the shopping mall mentioned above, the research team further analyzed the metal-intensive areas inside the building, such as elevator shafts, steel structure supports, and metal decorations. Using high-precision measuring equipment, the team experimentally measured the signal absorption rates of different wall materials (such as concrete, glass, and metal plates) and recorded the differences in absorption rates. The team then used this data to construct a three-dimensional signal attenuation field. This model not only takes into account the reflection interference of the elevator shaft, but also combines the signal absorption of different floors and areas to accurately simulate the signal propagation path and attenuation law in the building.

[0047] Through this step, the team was able to accurately identify signal blind spots (such as near elevators and metal-dense areas) and take corresponding measures to enhance signal coverage, such as adjusting antenna positions, adding signal relay equipment, or optimizing SMS delivery paths. This process provided a scientific basis for subsequent communication optimization and significantly improved the communication quality within the shopping mall.

[0048] 103. Predicting the spatiotemporal distribution of the amount of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and the position coordinates of the fire channel; In this step, the communication blind area 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 the area with insufficient signal coverage.

[0049] The fire escape location coordinates are the specific location information of the emergency evacuation route within the building, which is crucial for emergency response.

[0050] The temporal and spatial distribution of SMS request volume describes the changes in SMS sending demand in different time periods and regions, and is used to reasonably allocate resources.

[0051] Predicting the spatiotemporal distribution of SMS request volumes in each zone can help improve emergency response speed and ensure unimpeded communications in key areas.

[0052] In the embodiment of the present application, first, the communication blind area boundary data is extracted from the three-dimensional signal attenuation field and combined with the known fire channel location coordinates. Then, the SMS request patterns in different areas are analyzed using historical data to predict the SMS request volume in each area in the future time period. Then, these prediction results are combined with the communication blind area information to adjust the resource allocation plan. Finally, all the data are integrated to generate a detailed SMS request volume forecast in the time and space dimensions.

[0053] In the aforementioned shopping mall application, the research team used the data of the three-dimensional signal attenuation field and combined it with the location coordinates of the fire escape to predict the temporal and spatial distribution of SMS requests in each partition. Especially during holidays, the team found that the number of SMS requests in areas close to the fire escape (such as the mall entrance and the dining area) increased significantly, while the number of SMS requests in communication blind spots (such as underground parking lots and equipment rooms) was relatively low.

[0054] Based on this data, the team developed optimization strategies, such as adding temporary communication base stations in high-density areas, deploying signal relay equipment in communication blind spots, and dynamically adjusting the priority of SMS delivery. These measures ensure that in emergency situations, communications in key areas (such as near fire passages) will not be interrupted, thereby significantly improving the efficiency of emergency response and the reliability of the communication network.

[0055] 104. Accessing the Internet SMS gateway cluster, collecting base station signaling plane 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; In this step, the Internet SMS gateway cluster refers to a distributed processing system composed of multiple operators' SMS gateways. The base station signaling plane data includes key indicators such as signaling interaction success rate, signaling delay and signaling load. The online status of the user terminal reflects the terminal connection stability and signal reception quality. The cross-operator network quality thermal layer is a network performance visualization distribution map generated by integrating multi-operator data. The converged communication quality field is a three-dimensional communication quality model that integrates building environment characteristics and real-time network status.

[0056] In the embodiment of the present application, the SMS gateway clusters of each operator are first connected through a standardized interface to collect base station signaling data and terminal status information in real time. Then, the multi-operator data is aligned in time and space and correlation analysis is performed to calculate the comprehensive network quality score of each region. Then, the score result is converted into a thermal layer, and coordinate matching and numerical fusion are performed with the pre-built three-dimensional communication feature map. Finally, a fused communication quality field containing building structure characteristics, crowd distribution and real-time network status is output.

[0057] In the application of shopping malls, the system collected real-time signaling data of 42 base stations and the connection status of 3,865 terminals in the mall by connecting to the gateways of the three major operators. After data analysis, it was found that the southeast corner dining area had a significant decline in network quality during peak meal times (signaling delay increased by 35%). The system merged the network quality score of the area (62 points) with the crowd density data in the three-dimensional model (peak value of 1.8 people / ㎡), and finally marked the area as a "high-load communication area", providing a decision-making basis for subsequent SMS scheduling.

[0058] 105. 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 fused communication quality field, generate message forwarding rules sorted by priority.

[0059] Among them, the message forwarding rules include emergency channel bandwidth allocation, load balancing and diversion, and relay forwarding strategies in weak signal areas, and the triggering conditions are dynamically adjusted by analyzing the trend of personnel movement direction.

[0060] In this step, the gateway cache queue backlog refers to the number of messages to be processed in the network device, reflecting the current communication load.

[0061] The network congestion hotspots in the fused communication quality field refer to the areas with poor signal quality and high communication delay identified by the three-dimensional communication feature map.

[0062] Message forwarding rules are a set of guidelines that guide how to prioritize different types of messages, ensuring that high-priority messages are processed in a timely manner.

[0063] Emergency channel bandwidth allocation, load balancing and diversion, and relay forwarding strategies in weak signal areas are specific message forwarding measures used to optimize communication efficiency.

[0064] Triggers are criteria for initiating or adjusting message forwarding policies, usually based on trends in the direction of people moving.

[0065] In the embodiment of the present application, firstly, based on the spatiotemporal distribution and the backlog of the gateway cache queue, combined with the network congestion hotspots in the fusion communication quality field, the communication needs of each area are comprehensively evaluated. Then, the message forwarding rules are formulated, including emergency channel bandwidth allocation, load balancing diversion and relay forwarding strategies in weak signal areas. Then, by real-time monitoring of the trend of personnel movement direction and changes in network congestion hotspots, the trigger conditions are dynamically adjusted to ensure the effectiveness of the rules. Finally, all information is integrated to generate a complete set of dynamically adjustable message forwarding rules. In the application of shopping malls, the research team generated message forwarding rules based on the spatiotemporal distribution, the backlog of the gateway cache queue and the network congestion hotspots in the fusion communication quality field. Especially during holidays, some areas not only have serious backlogs in the gateway cache queue, but also have network congestion hotspots, resulting in increased communication delays. To this end, the team has formulated emergency channel bandwidth allocation and load balancing diversion strategies, such as giving priority to bandwidth resources in areas with high traffic density (such as shopping mall entrances and dining areas), and deploying relay forwarding equipment in weak signal areas (such as underground parking lots). At the same time, the team dynamically adjusts the trigger conditions by analyzing the trend of personnel movement direction and changes in network congestion hotspots in real time. For example, when a large number of people are detected moving toward a fire escape and a network congestion hotspot appears in that area, the system automatically activates the emergency channel bandwidth allocation strategy and optimizes the load balancing solution to ensure smooth communication in key areas. These measures significantly improve communication efficiency, especially in the case of network congestion, to maintain good communication quality.

[0066] 106. Execute a text message sending operation according to the message forwarding rule.

[0067] Among them, message forwarding rules refer to a set of SMS processing strategies sorted by priority, including specific implementation plans such as emergency channel allocation, load balancing and relay forwarding; 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; The execution process needs to monitor network status changes in real time and dynamically adjust the delivery strategy; The operation result feedback is used to optimize the generation of subsequent message forwarding rules.

[0068] In the embodiment of the present application, the currently effective message forwarding rules are first parsed to determine the priority order and specific transmission strategy of SMS messages in each region. Then, the SMS gateway cluster distributes the messages to the corresponding base stations according to the rules, and monitors the changes in network quality in real time during the transmission process. When the transfer of network congestion hotspots or new trends of personnel gathering are detected, the sending path and resource allocation are dynamically adjusted. Finally, the SMS delivery rate and latency data of each region are collected and fed back to the rule generation module for continuous optimization.

[0069] In the application of shopping malls, the system executes SMS delivery 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 channel area is marked as high priority, the emergency channel is immediately activated to ensure the transmission of key messages. By collecting the message delivery status of each area in real time, the system continuously optimizes the forwarding strategy to ensure that the delivery success rate can still be maintained at more than 95% under conditions of dense crowds and network fluctuations.

[0070] 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; based on the building structure characteristics, metal-dense areas and special structures (such as elevator shafts) are identified to establish a three-dimensional signal attenuation field that accurately reflects the actual attenuation status; the communication blind spots and personnel flow channels are combined to predict the temporal and spatial distribution of SMS services to achieve accurate demand prediction; the network status data of multiple operators are integrated to form a real-time communication quality heat map, and integrated with the three-dimensional environmental model to build a global communication situation awareness; the optimal forwarding strategy is dynamically generated by comprehensively considering business needs, network load and communication quality, and intelligent resource scheduling is achieved. The reliable transmission of high-priority messages is ensured through a rule-driven execution mechanism, significantly improving the communication guarantee capability in complex environments. This method realizes the precision, intelligence and reliability of SMS services in building scenarios.

[0071] Furthermore, the method also includes: importing the three-dimensional communication feature map and forwarding rules into the digital twin environment to simulate a crowd gathering scene, and correcting the absorption rate parameters of the three-dimensional signal attenuation field by comparing the gateway response delay with a preset threshold.

[0072] In this step, the digital twin environment is a virtual simulation system that can simulate the behavior of the actual physical environment for testing and optimizing the communication system.

[0073] 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 gateway performance and is used to determine whether the system needs to be optimized.

[0074] Modifying the absorptivity parameters of the 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.

[0075] In the embodiment of the present application, first, the three-dimensional communication feature map and message forwarding rules are imported into the digital twin environment to simulate the actual crowd gathering scene. Then, the accuracy of the existing model is evaluated by comparing the gateway response delay in the simulated environment with the preset threshold. Then, based on the comparison results, the absorption rate parameters in the three-dimensional signal attenuation field are corrected to more accurately reflect the actual situation. Finally, through multiple iterative optimizations, the high accuracy of the model is ensured.

[0076] In the aforementioned shopping mall application, the research team used the digital twin environment to simulate the communication conditions during peak holiday periods and found that the gateway response delay in some areas exceeded the preset threshold. For example, in high-density areas such as dining areas and shopping mall entrances, the gateway response delay increased significantly, resulting in a decrease in the efficiency of SMS delivery. Based on these simulation results, the team corrected the absorption rate parameters of the three-dimensional signal attenuation field, such as adjusting the signal absorption rate of metal-dense areas and wall materials to more accurately reflect the signal propagation characteristics in the actual environment.

[0077] In order to solve the problem of inaccurate multi-operator network quality assessment in complex building environments, in some embodiments, the access Internet SMS gateway cluster collects base station signaling plane data and user terminal online status to generate a cross-operator network quality thermal map, and maps the thermal map to the three-dimensional communication feature map to form a fused communication quality field, including: 1041. Synchronously connect to each SMS gateway cluster through a multi-operator protocol interface to obtain the signaling plane data of each base station in real time.

[0078] Among them, multi-operator protocol interface: a unified access module that supports communication protocols of different operators; Signaling plane data: control signaling interaction records between base stations and terminals (such as call establishment / release records), terminal registration status (online / offline), signaling response delay (terminal response time to base station instructions).

[0079] 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.

[0080] 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; 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.

[0081] 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.

[0082] 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; Grid unit: Divide the base station coverage area (e.g., 500-meter radius) into square grids with a side length of 10 meters; Terminal location distribution density: terminal distribution thermal value calculated by terminal GPS coordinates or base station triangulation.

[0083] In the embodiment of the present application, the terminal density in each grid is calculated based on the terminal positioning data (GPS or base station fingerprint positioning). An adaptive grid division algorithm is used: the grid is automatically reduced to 5 meters in length for terminal dense areas (such as density > 5 units / 10㎡), and expanded to 20 meters in length for sparse areas (density < 1 unit / 10㎡), ensuring that each grid contains at least 1 terminal's measurement data.

[0084] 1044. Perform weighted calculation on the signaling response delay and signal strength in each grid unit to obtain the network quality score of the grid unit, and use a spatial interpolation method to smooth the network quality scores of adjacent grid units to generate a thermal map of the cross-operator network quality.

[0085] Weighted calculation: linearly weight the signaling response delay (weight 0.6) and signal strength (RSRP, weight 0.4); Spatial interpolation: The inverse distance weighted (IDW) algorithm is used to fill in grids without terminal data.

[0086] In the embodiment of the present application, the score of the terminal data in each grid unit is calculated according to the formula: Score = 0.6*(1-delay / maximum delay threshold)+0.4*(RSRP / maximum RSRP); IDW interpolation is used for adjacent grids, and the score of the blank grid is calculated with the inverse of the distance (d is the grid spacing) as the weight: score = Σ(score_i / d_i²) / Σ(1 / d_i²); Finally, a continuous thermal layer covering the entire area is generated.

[0087] 1045. Map the center coordinates of each grid unit in the network quality thermal layer to the corresponding spatial position of the three-dimensional communication feature map; 1046. According to the crowd density gradient and signal propagation loss weight of the corresponding spatial position in the three-dimensional communication characteristic map, adjust the contribution coefficient of the network quality score to update the network quality score; 1047. Superimpose the updated network quality score with the communication parameters of the three-dimensional communication feature map to form a fused communication quality field.

[0088] In the above steps 1045-1047, the contribution coefficient is adjusted: the network quality score is dynamically corrected according to the crowd density (weight 0.7) and the signal propagation loss (weight 0.3). Communication parameter superposition: the corrected score is weighted and fused with the wall attenuation value and metal interference value in the three-dimensional map.

[0089] In the embodiment of the present application, the center coordinates of the thermal layer grid (such as longitude 116.3°E, latitude 39.9°N) are mapped to the corresponding coordinate points of the three-dimensional map. Based on the population density gradient (such as 1.5 people / ㎡) and the signal loss weight (such as elevator shaft area loss +15dB) of the coordinate point, the score is corrected according to the formula: Corrected score = original score * (0.7 * (1-crowd density / peak density) + 0.3 * (1-signal loss / maximum loss)) Finally, the corrected score is superimposed with the communication parameters in the three-dimensional map (such as the attenuation coefficient of the wall material) to generate a fused communication quality field.

[0090] Here is a specific example: First, the system accesses the SMS gateway clusters of China Mobile, China Unicom, and China Telecom through the multi-operator protocol interface, and collects real-time signaling data from 42 base stations in the mall (including an average signaling response delay of 82ms) and measurement data from 3,865 terminals (RSRP ranges from -85dBm to -105dBm). During the peak dining hours, the signaling delay of the base station in the southeast corner dining area suddenly increased to 142ms, and the average RSRP reported by the terminal was less than -98dBm.

[0091] Secondly, the system divides the base station coverage of the dining area into grid units with a side length of 5 meters. It is calculated that the network quality score of the area has dropped from 72 points in the off-peak period to 58 points. After filling the corner grids without terminal data through spatial interpolation, a thermal map of the network quality across operators is generated, showing that the dining area is an orange warning area (score 50-60 points).

[0092] Next, the thermal map was mapped to the three-dimensional communication feature map, and the network quality score was corrected to 51 points according to the formula, combining the peak population density (2.1 people / ㎡) and the signal propagation loss caused by the steel frame structure (+12dB). In the three-dimensional fusion communication quality field, the area was marked as a "severely congested area" and associated with the wall metal density (85%) and the elevator shaft interference coefficient (0.8).

[0093] Finally, the system automatically triggers three strategies based on the marking results of the integrated communication quality field: Lower the priority of SMS forwarding in the dining area from Level 2 to Level 4; Enable load balancing of the adjacent leisure area base station and direct 30% of the message traffic to the west base station; Deploy portable repeater equipment in the steel frame area to increase signal strength by 8dB.

[0094] After implementation, the SMS delivery rate during peak hours in the area increased from 78% to 93%, and the average delay dropped to 89ms.

[0095] This solution achieves accurate characterization and dynamic perception of network quality in complex building scenarios through real-time synchronous collection of multi-operator data, dynamic grid quality assessment and fusion correction of three-dimensional environmental parameters. It effectively solves the evaluation deviation problem caused by the one-sidedness of single operator data and the neglect of building structure interference in traditional methods, provides high-precision communication situation maps for SMS delivery strategies, significantly improves the diversion efficiency in congested areas and the accuracy of relay scheduling in weak signal areas, and ultimately ensures communication reliability and timeliness in high-load scenarios.

[0096] In order to solve the problem of insufficient accuracy in responding to high-density areas and emergencies in the communication system, by integrating the high-density area coordinates in the spatiotemporal distribution and the group movement direction vector, combined with the peak period characteristics in the gateway cache queue backlog and the network congestion hotspots in the fusion communication quality field, the message forwarding rules sorted by priority are generated. This method not only improves the response speed and efficiency of the communication system, but also ensures efficient communication in emergency situations.

[0097] In some embodiments, the message forwarding rules sorted by priority are generated based on the spatiotemporal distribution of the SMS request volume in each partition and the backlog of the gateway cache queue in combination with the network congestion hotspots in the fusion communication quality field in step 104, including: 201. Integrate the high-density area coordinates and group moving direction vectors in the spatiotemporal 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 hotspot, and generate the emergency channel weight coefficient; In step 201, the signal strength attenuation rate of the network congestion hotspot: represents the rate of change of the signal strength in the network congestion area over time or space (such as a decrease of 2dBm per second). The high-density area coordinates in the spatiotemporal distribution refer to the location information of the area where a large number of people gather, which is used to identify key areas. The group movement direction vector describes the direction and speed of personnel flow, which helps to predict future traffic changes. The movement direction clustering result is obtained by clustering analysis of the group movement direction, reflecting the movement trends of different groups. The historical request growth rate is the growth rate calculated based on past data and is used to predict future demand. The emergency channel weight coefficient is a comprehensive indicator used to measure the importance of a specific area.

[0098] In the embodiment of the present application, the high-density area coordinates (such as the coordinate set of area A in the shopping mall) are first extracted, and the group movement direction vector (such as gathering in the northwest direction) is analyzed by the DBSCAN clustering algorithm. Then the historical request growth rate (such as a 120% increase in a 10-minute window) and the cosine similarity of the current movement direction (such as cosθ=0.92) are calculated, combined with the signal attenuation rate (such as -4.2dB / minute), and the weight coefficient is generated according to the formula: Weight coefficient = similarity × (1-decay rate / maximum decay rate) For example, when the attenuation rate reaches 80% of the preset threshold, the weight coefficient of the trigger emergency channel is increased to 0.75.

[0099] 202. Based on the peak period characteristics of the gateway queue backlog, combined with the network congestion hotspot and the signal attenuation gradient of the adjacent access point, the cross-floor diversion ratio is generated by matching the attenuation gradient inflection point with the physical topological distance; In step 202, the peak period feature in the gateway queue backlog refers to the change pattern of the number of pending messages in the network device during the peak period. The signal attenuation gradient of adjacent access points describes the change trend of the signal strength between different access points. The physical topological distance is the actual distance between different access points. The cross-floor diversion ratio is an allocation strategy used to disperse traffic to different floors to reduce the pressure on a single access point.

[0100] In the embodiment of the present application, based on the peak time period of the gateway queue backlog (e.g., the backlog reaches 1200 at 13:00), the signal attenuation gradient of the access points on adjacent floors is extracted (e.g., the attenuation gradient from floor 3 to floor 4 is 12dB). The attenuation inflection point (e.g., the coordinates of the inflection point in the elevator room) is matched with the physical topological distance (e.g., the straight-line distance is 20 meters) through the path planning algorithm (e.g., the A* algorithm), and the diversion ratio is calculated according to the formula: Diversion ratio = (1-attenuation gradient / maximum gradient) × (1-distance / maximum distance) For example, when the distance from the inflection point is 5 meters and the attenuation gradient is 10 dB, the inter-floor diversion ratio is 45%.

[0101] 203. Extract the coverage of idle channels of neighboring access points in the network congestion hotspot 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 demand; In step 203, the weak signal area geometric boundary data refers to the boundary of the area where the signal strength is lower than a certain threshold. The adjacent access point idle channel coverage describes the coverage area of ​​the nearby available channels. 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 based on 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.

[0102] In the embodiment of the present application, the geometric boundary of the weak signal area (such as the polygonal coordinates of the underground parking lot) is extracted, and the idle channels of the adjacent access points are scanned (such as the parking lot entrance base station has 4 idle channels). The correlation strength (R²=0.85) between the response delay (such as an average of 80ms) and the signal compensation requirement (such as +12dB) is analyzed by linear regression, and the priority is sorted according to the formula: Priority = number of idle channels × (1-delay / maximum delay) + compensation requirement weight For example, if an access point has 3 idle channels and a latency of 60ms, the priority score is 8.2 (out of 10).

[0103] 204. Multi-dimensionally integrate the emergency channel weight coefficient, the cross-floor diversion ratio, and the relay forwarding priority sequence to construct a forwarding rule trigger matrix, and dynamically update the forwarding rule trigger matrix by associating the change rate of the group moving direction vector with the growth rate of the gateway cache queue backlog; In step 204, multi-dimensional fusion is an algorithmic framework for integrating data from multiple dimensions into a unified model. The forwarding rule trigger matrix is ​​a comprehensive chart containing all generated parameters for guiding the message forwarding strategy. The rate of change 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, the data of the emergency channel weight coefficient, the cross-floor diversion ratio, and the relay forwarding priority sequence are first collected and integrated. Then, using multi-dimensional data analysis technology, these parameters are fused into a unified forwarding rule trigger matrix. The matrix not only contains static configuration information, but also can be dynamically adjusted according to real-time data. Then, an association model between the rate of change of the group movement direction vector and the growth rate of the gateway cache queue backlog is established to realize a dynamic update mechanism for the forwarding rules. For example, when it is detected that a crowd is gathering rapidly in a certain area, the system automatically adjusts the forwarding rules and prioritizes the allocation of communication resources.

[0104] 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) between the rate of change of the moving direction (e.g., Δθ=10° / minute) and the queue backlog growth rate (e.g., ΔQ=150 items / 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 ​​dynamically updated: New matrix value = original value × (1 + correlation coefficient × update factor) For example, when the correlation coefficient is 0.7, the weight coefficient is updated from 75 to 80.

[0105] 205. Based on the critical parameters of the forwarding rule trigger matrix, simulate the message path conflict in the digital twin environment, and generate the message forwarding rules sorted by priority by reversely correcting the matching threshold and signal compensation parameters.

[0106] In step 205, the forwarding rule trigger matrix is ​​a multidimensional data structure formed by the fusion of the emergency channel weight coefficient, the cross-floor diversion ratio and the relay forwarding priority sequence, which is used to dynamically adjust the message forwarding strategy. The critical parameter refers to the key value in the trigger matrix used to determine whether the forwarding rule needs to be adjusted. The digital twin environment is a virtual simulation platform that can simulate various scenarios in the real world, including the behavior of the communication network. Message path conflict refers to the situation of information congestion or transmission failure that may occur during the communication process. The matching threshold is a standard value for measuring the degree of matching between different forwarding rules. The signal compensation parameter is a technical parameter set to compensate for the weak signal area, which is used to enhance the signal strength or optimize the information transmission path. The message forwarding rules sorted by priority are a set of instructions that arrange the order of information transmission according to urgency and efficiency.

[0107] Specifically, the forwarding rule trigger matrix is ​​loaded in the digital twin model to simulate the SMS delivery path (such as A→B→C). When a path conflict is detected (such as two messages occupying channel X at the same time), a genetic algorithm is used to iteratively optimize the parameters: Randomly generate 100 sets of parameter combinations (matching threshold, compensation parameters); Select the 20 groups with the lowest conflict rate for crossover mutation; After 50 generations of iteration, the optimal parameters (such as threshold 0.63, compensation +13dB) are output.

[0108] Finally, a rule table sorted by priority is generated (such as rule 1: messages with emergency channel weight > 0.8 are transmitted first).

[0109] Here is a specific example: First, the system detected that the transfer hall (coordinates X, Y) was a high-density area (610 requests / minute), the group moved in the direction of the exit (vector θ=30°, v=1.2m / s), and the signal attenuation rate in this area reached -5dB / minute. The emergency channel weight coefficient was calculated to be 0.82.

[0110] Secondly, the gateway queue backlog reached 1,500 at 18:00, and the signal attenuation gradient between adjacent platform layers was 18 dB. After matching the physical distance of 15 meters, the cross-layer diversion ratio was 52%.

[0111] Next, in the weak signal area (bathroom channel), it was detected that the neighboring access point had two idle channels, the response delay was 65ms, and the relay forwarding priority was calculated to be 7.3.

[0112] Then, a three-dimensional trigger matrix (weight 82, diversion 52, priority 7.3) was constructed. When a sudden turn of the crowd was detected (Δθ=25° / minute) and the queue grew by ΔQ=300 items / minute (r=0.72), the matrix weight was updated to 88.

[0113] Finally, the digital twin simulation shows that the conflict rate of channel Y is 23%. After reverse correction, the matching threshold is reduced from 0.7 to 0.62, generating the final rule: Messages with weight > 0.8 are directly connected through the emergency channel; 52% of the backlog messages are diverted to the platform level; Messages with priority > 7 enable relay forwarding.

[0114] After implementation, the SMS delivery rate in the transfer hall increased from 68% to 89%, and the channel conflict rate dropped to 9%.

[0115] In order to further improve the accuracy of the response to high-density areas and emergency situations in the communication system, in some embodiments, the emergency channel weight coefficient, the cross-floor diversion ratio and the relay forwarding priority sequence are multi-dimensionally integrated in step 204 to construct a forwarding rule trigger matrix, including: 301. The emergency channel weight coefficient is integrated with the real-time load and historical traffic data of each channel, and the basic constraint value is generated in combination with the channel length. The obstacle density is superimposed for dynamic compensation to obtain a channel quality evaluation index; In step 301, the emergency channel weight coefficient is a comprehensive indicator 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 based on the traffic data recorded in the past. The channel length refers to the physical distance. The basic constraint value is a benchmark value calculated by combining the real-time load, historical traffic data and channel length. 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 that reflects the overall quality and availability of the channel.

[0116] In an embodiment of the present application, first, the real-time load and historical traffic data of each channel are collected, and the basic constraint value is generated in combination with the channel length. Then, these data are fused with the emergency channel weight coefficient to form a preliminary evaluation framework. Then, the obstacle density data is obtained through on-site surveys or sensors, and it is superimposed on the basic constraint value as a dynamic compensation factor. Specifically, the obstacle density is quantified using a weighted average method or other statistical methods, and the basic constraint value is adjusted to reflect the actual channel conditions. Finally, all information is integrated to generate a comprehensive channel quality evaluation index for subsequent optimization decisions.

[0117] For example, first, the system collects the flow sensor data (current load) and historical traffic database records of each emergency channel in real time, combines the channel length information provided by the GIS map, and generates the basic constraint value through the weighted calculation formula (load weight 40% + historical data 30% + length 30%). Secondly, the obstacle distribution density in the channel is obtained by using lidar scanning or camera image recognition technology, and the logarithmic compensation algorithm is used to dynamically adjust the basic constraint value (for every 10% increase in obstacle density, the quality score decreases by 5%). Then, the emergency channel weight coefficient is introduced into the calculation process as an adjustment factor to ensure that important channels receive a quality score bonus (for every 0.1 increase in the weight coefficient, the score increases by 2%). Finally, the channel quality evaluation index of 0-100 points is output, and a high-quality channel is above 90 points, and an alarm needs to be triggered if the score is below 60 points.

[0118] 302. The cross-floor diversion ratio is based on the floor escape node density, coupled with the escape ladder capacity threshold and the channel quality evaluation index to generate a diversion gradient coefficient, and the building load-bearing limit is integrated to output a diversion ratio reference value; In step 302, the cross-floor diversion ratio is an allocation strategy used to disperse traffic to different floors. The floor escape node density refers to the number of emergency exits on each floor. The escape ladder capacity threshold is the maximum load capacity of each escape ladder. The diversion gradient coefficient is a comprehensive indicator used to measure the flow distribution between different floors. The building load-bearing limit refers to the maximum weight that the building structure can withstand. The diversion ratio benchmark value is an allocation standard used to guide the actual diversion operation.

[0119] In the embodiment of the present application, first, based on the floor escape node density and the escape ladder capacity threshold, the preliminary cross-floor diversion ratio is calculated. Then, combined with the channel quality evaluation index, a diversion gradient coefficient is generated that reflects the difference in flow distribution between floors. Then, the building load-bearing limit is considered to ensure that the diversion scheme does not cause additional pressure on the building structure. Specifically, through mathematical modeling and optimization algorithms, the optimal diversion ratio that meets all conditions is found. Finally, all parameters are integrated to output a diversion ratio benchmark value that conforms to the actual situation to guide actual operations.

[0120] For example, first, based on the building BIM model, the escape node distribution data of each floor is extracted, and the density of escape nodes per square meter is calculated in combination with the real-time flow of people counted by infrared thermal imaging. Secondly, according to the escape ladder capacity threshold set by the fire protection regulations (such as 0.3 square meters per person), combined with the channel quality evaluation index obtained in step 301, the gradient descent algorithm is used to solve the optimal diversion gradient coefficient. Then, the load-bearing data of each floor is monitored by the building structure sensor, and the constrained optimization model is used to ensure that the diversion plan does not exceed the load-bearing limit (such as load-bearing ≤500kg per square meter). Finally, the diversion ratio benchmark value of each floor is output, and the accuracy is controlled within the range of ±2%.

[0121] 303. The relay forwarding priority sequence constructs a node coverage map through the wireless signal multipath effect, 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 chain table; 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 the signal reaches the receiving end through multiple paths. The node coverage map is a coverage map built based on the multipath effect. The three-dimensional permeability surface is a three-dimensional model that shows the signal penetration capability of different areas. The curvature feature is a key geometric feature extracted from the three-dimensional permeability surface. The priority sorting list is an ordered list used to guide the signal transmission order.

[0122] In the embodiment of the present application, first, a node coverage map is constructed using the multipath effect of wireless signals to depict the coverage of the wireless network. Then, a three-dimensional permeability surface is generated in combination with the diversion gradient coefficient to show the information transmission capabilities of different areas. Then, the curvature characteristics of the three-dimensional permeability surface are extracted to identify the best information transmission path. Specifically, image processing technology and machine learning algorithms are used to analyze the curvature characteristics and find the optimal path. Finally, a priority sorting list is generated based on the analysis results to ensure that key information can be efficiently transmitted in the shortest time.

[0123] 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 a ray tracing algorithm. Secondly, use the diversion gradient coefficient of step 302 as a permeability adjustment factor, and use the Kriging interpolation method to generate a three-dimensional permeability surface (grid accuracy 0.1 meters). Then, use the differential geometry algorithm to extract the surface curvature characteristics and identify high permeability channels with a curvature radius of less than 3 meters. Finally, calculate the optimal transmission path between each node based on the Dijkstra algorithm, generate a priority sorted list in descending order of permeability efficiency, and update it every 5 seconds.

[0124] 304. Input the channel quality evaluation index, diversion ratio benchmark value and priority sorting list into the building digital twin, extract implicit association rules through tensor decomposition and orthogonal recombination, and construct a forwarding rule trigger matrix.

[0125] In step 304, the building digital twin 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 regularities extracted from data. Orthogonal recombination is an optimization technique used to recombine data for efficiency. The forwarding rule trigger matrix is ​​a comprehensive chart containing all generated parameters to guide the message forwarding strategy.

[0126] In the embodiment of the present application, first, the channel quality evaluation index, the diversion ratio benchmark value and the priority sorting list are input into the building digital twin. Then, the tensor decomposition technology is used to extract implicit association rules to reveal the potential connections between the data. Then, these rules are reorganized into a new data structure, namely the forwarding rule trigger matrix, through orthogonal recombination. Specifically, advanced data analysis tools and algorithms are used to process complex multidimensional data to ensure that the generated matrix can accurately reflect the actual situation. Finally, all information is integrated to construct a complete forwarding rule trigger matrix to guide actual operations.

[0127] For example, first, the building digital twin is connected to the real-time data stream, and three sets of data are loaded: channel quality evaluation index, diversion ratio benchmark value, and priority sorting list. Secondly, the CP tensor decomposition method (Canonical Polyadic Decomposition) is used to extract the third-order implicit association rules, and the decomposition rank is set to 5 to balance accuracy and efficiency. Then, the feature vector is reorganized through the Gram-Schmidt orthogonalization process to eliminate redundant associations between rules. Finally, a forwarding rule trigger matrix with the dimension of [channel × floor × priority] is constructed, and the matrix element value is the trigger weight (0-1), and an online update mechanism is deployed (10% of the key parameters are incrementally updated every minute).

[0128] Here is a specific example: In an application scenario of a large shopping mall, the research team integrated the high-density area coordinates in the spatiotemporal distribution with the group movement direction vector to generate the emergency channel weight coefficient. The research team used the multipath effect of wireless signals to construct a node coverage map, and combined with the previously generated diversion gradient coefficient to generate a three-dimensional permeability surface. By analyzing the curvature characteristics, the research team identified the best information transmission path and generated a priority sorting list. The research team input the channel quality evaluation index, diversion ratio benchmark value and priority sorting list into the building digital twin. Through tensor decomposition technology, the implicit association rules between the data were discovered, and the forwarding rule trigger matrix was generated through orthogonal recombination. This matrix not only improves the efficiency of information transmission, but also enhances the flexibility and response speed of the system, significantly improving the emergency communication management capabilities of the entire shopping mall.

[0129] In summary, steps 301 to 304 construct a dynamically updated forwarding rule trigger matrix by multi-dimensionally integrating the emergency channel weight coefficient, cross-floor diversion ratio, and relay forwarding priority sequence. 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.

[0130] In order to solve the problem of insufficient accuracy in responding to high-density areas and emergency situations in the communication system, in some embodiments, the step 302 generates a diversion gradient coefficient by basing the cross-floor diversion ratio on the density of floor escape nodes, coupling the escape ladder capacity threshold and the channel quality evaluation index, including: 401. Generate a node coverage strength value based on the floor escape node density based on the visible distance and the number of path branches, trigger connectivity compensation through the visible distance, and correct the topological strength attenuation through the number of path branches; 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 strength value is a comprehensive score that reflects the importance of the node. Connectivity compensation is the process of adjusting the node coverage strength value based on the visible distance. Topological strength attenuation is the process of correcting the node coverage strength value based on the number of path branches.

[0131] In the embodiment of the present application, the node coverage strength value is first generated based on the floor escape node density, visual distance and number of path branches. Then, the visual distance is calculated using computer vision technology and used as a connectivity compensation factor. Then, the number of path branches is analyzed and the topological strength attenuation is corrected using a graph theory algorithm. Finally, all information is integrated to generate a node coverage strength value that fully reflects the node evacuation capability.

[0132] For example, first, the three-dimensional coordinate data of the escape nodes on each floor is extracted through the building BIM model, and the visible distance between each node is calculated using the ray casting algorithm (accuracy 0.1 meter). Secondly, the 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 used: coverage intensity value = visible distance coefficient × ln (connectivity) × node density, where the visible distance coefficient is attenuated and compensated for distances exceeding 30 meters through an S-type function. Finally, the calculation results are normalized (range 0-1) to generate a coverage intensity matrix that reflects the node evacuation capability.

[0133] 402. Based on the escape ladder capacity threshold and the smoke diffusion rate, a real-time traffic attenuation coefficient is generated by asymmetric superposition of the static capacity baseline and the dynamic attenuation factor and introduction of the node coverage strength value; In step 402, the escape ladder capacity threshold is the maximum capacity of each escape ladder. The smoke diffusion rate is the speed at which smoke spreads within 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 traffic attenuation coefficient is a comprehensive indicator used to measure the traffic capacity at different time points. The node coverage strength value is the data generated from the previous step.

[0134] In the embodiment of the present application, firstly, based on the escape ladder capacity threshold and the smoke diffusion rate, asymmetric superposition is performed in combination with the static capacity baseline and the dynamic attenuation factor. Then, the node coverage strength value generated in step 401 is introduced to further refine the model. Then, simulation technology is used to simulate the evacuation process in different scenarios and calculate the real-time traffic attenuation coefficient. Finally, a real-time traffic attenuation coefficient that accurately reflects the current traffic capacity is generated.

[0135] For example, first, obtain the design capacity data of each escape ladder from the fire protection system and establish a static capacity baseline (such as a maximum load of 20 people per ladder). Secondly, monitor the smoke concentration change rate of each floor in real time through the smoke sensor network, use the fluid mechanics model to predict the diffusion trend in the next 3 minutes, and generate a dynamic attenuation factor (0-1 value). Then, the static baseline and the asymmetric attenuation factor are weighted fused (static weight 60% + dynamic 40%), and the node coverage strength value of step 401 is superimposed as an adjustment parameter. Finally, through Monte Carlo simulation verification, a real-time traffic attenuation coefficient table updated at the minute level is output.

[0136] 403. 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 the diversion gradient coefficient.

[0137] In step 403, the channel quality evaluation index is a comprehensive score that reflects the overall quality and availability of the channel. The real-time traffic 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 traffic attenuation coefficient. The diversion gradient coefficient is a comprehensive indicator 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.

[0138] In the embodiment of the present application, the channel quality evaluation index is firstly integrated with the real-time traffic attenuation coefficient, and tensor dot product and topological filtering are performed at the floor connection nodes. Then, the graph theory algorithm and optimization technology are applied to generate the diversion gradient coefficient by combining the horizontal weight base and the vertical emergency weight. Then, through multiple iterations of optimization, it is ensured that the generated coefficient can accurately reflect the actual situation. Finally, a comprehensive diversion gradient coefficient is generated to guide the actual operation.

[0139] For example, first, a three-dimensional evacuation decision tensor is constructed, where the x-axis is the channel quality index (result of step 301), the y-axis is the real-time traffic attenuation coefficient, and the z-axis is the floor topology relationship. Secondly, the tensor dot product operation is performed at the connection node to fuse the horizontal weight (channel quality accounts for 55%) and the vertical emergency weight (traffic attenuation accounts for 45%). Then, a topological filter based on a graph convolutional network is applied to eliminate spatial noise and enhance critical path features. Finally, the dominant mode is extracted through eigenvalue decomposition to generate a diversion gradient coefficient matrix distributed by floor, which is automatically updated every 30 seconds and synchronized to the emergency command system.

[0140] Here is a specific example: In a large shopping mall scenario, the research team first parsed the building information model data, obtained the density of escape nodes on each floor, and generated node coverage strength values ​​based on the visible distance and the number of path branches. Subsequently, the real-time traffic attenuation coefficient was calculated by combining the escape ladder capacity threshold and the smoke diffusion rate. Next, the research team fused the channel quality evaluation index with the real-time traffic attenuation coefficient, performed tensor dot products and topological filtering at the floor connection nodes, and generated the diversion gradient coefficient. This series of steps not only improves the evacuation efficiency in emergency situations, but also enhances the flexibility and response speed of the system, significantly improving the emergency communication management capabilities of the entire shopping mall.

[0141] In summary, steps 401 to 403 generate the diversion gradient coefficient by multi-dimensionally integrating the floor escape node density, escape ladder capacity threshold and 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 emergency situations through dynamic adjustment strategies, providing strong support for intelligent building management and public safety.

[0142] In order to further improve communication efficiency and signal coverage quality, in some embodiments, the step 102 detects the difference in wall material absorption rate based on the coordinates of the metal-dense area in the three-dimensional communication feature map, and constructs a three-dimensional signal attenuation field including elevator shaft reflection interference, including: 501. Coordinates of metal-dense areas of the three-dimensional communication characteristic map, dynamically generating a metal interference intensity field through correlation between metal component distribution density and signal reflection space; In step 501, the coordinates of the metal-dense area of ​​the three-dimensional communication feature map dynamically generate a metal interference intensity field through the metal component distribution density and the signal reflection spatial correlation. The metal component distribution density refers to the number and distribution of metal structures inside the building. The signal reflection spatial correlation is based on the degree of influence of the metal structure on the reflection of the wireless signal. 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 the parameters according to the signal reflection spatial correlation to finally generate the metal interference intensity field.

[0143] In an embodiment of the present application, first, the coordinates of the metal-dense area are extracted from the three-dimensional communication feature map. Next, the distribution density of metal components in these areas is analyzed, and their spatial correlation with the reflection of wireless signals is evaluated. Then, based on these data, the metal interference intensity field is dynamically generated using electromagnetic simulation technology. Specifically, a ray tracing algorithm is used to simulate the behavior of the signal in the metal-dense area, and the interference intensity of each point is calculated. Finally, all the information is integrated to form a spatial model that comprehensively reflects the metal interference intensity.

[0144] For example, first, the three-dimensional coordinate data of metal components (such as steel mesh and elevator tracks) are extracted through the building BIM model, and the metal distribution density of each voxel (0.5m×0.5m×0.5m) is calculated using the kernel density estimation algorithm. Secondly, the reflection path of electromagnetic waves in the 2.4GHz / 5GHz frequency band between metal structures is simulated based on ray tracing technology, and the multipath delay and signal strength attenuation of each path are counted to generate a spatial correlation matrix. Then, the density distribution and spatial correlation are tensor-producted, and the measurement noise is eliminated through adaptive filtering. Finally, a metal interference intensity field with a scale of 0-100 is output, where areas with an intensity >70 are marked as high-risk interference areas.

[0145] 502. Generate a reference value of absorption rate by mapping material type and electromagnetic wave penetration loss, introduce wall moisture content and thickness variation coefficient, combine ambient humidity data with reference value frequency domain convolution, and generate dynamic correction parameters of absorption rate; In step 502, the material type refers to the specific type of building material, such as concrete, bricks, etc., and includes its physical and chemical properties, which are used to analyze the material's ability to penetrate electromagnetic waves. Electromagnetic wave penetration loss is the amount of loss of the material's ability to penetrate wireless signals, which is specifically manifested as the degree of signal attenuation and is used to evaluate the propagation effect of signals in different materials. The absorption rate reference value is an initial absorption rate value 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 signal absorption rate and is used to adjust the absorption rate reference value to reflect the actual conditions. The thickness variation coefficient is the range of variation 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 absorption rate dynamic correction parameter is a final absorption rate value after comprehensively considering multiple factors, which is used to accurately simulate the signal absorption in the actual environment.

[0146] In the embodiment of the present application, first, a reference value of absorption rate is generated according to the mapping relationship between material type and electromagnetic wave penetration loss. Next, the moisture content and thickness coefficient of variation of the wall are introduced and analyzed in combination with the ambient humidity data. Then, the above parameters are combined with the reference value of absorption rate using frequency domain convolution technology to generate dynamic correction parameters of absorption rate. Specifically, a numerical simulation method is used to process complex data sets to ensure that the generated parameters can accurately reflect the actual situation. Finally, all information is integrated to form a comprehensive dynamic correction parameter of absorption rate.

[0147] 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 benchmark values ​​of each material in the 1-6GHz frequency band. Secondly, a millimeter-wave radar is used to detect the actual moisture content of the wall (accuracy ±1.5%), and a laser rangefinder is used to obtain the thickness variation coefficient (0.1-0.3). Then, the real-time environmental humidity data is convolved with the benchmark value in the frequency domain (sampling rate 10Hz, window function length 256), and the characteristic frequency band (2.4-2.4835GHz) is extracted through wavelet transform. Finally, a multivariate linear regression model is used to generate dynamic correction parameters, and the correction range can reach ±35% of the benchmark value.

[0148] 503. The cavity resonance frequency is obtained according to the geometric dimensions of the shaft cavity, and the metal shaft reflection path is combined with the metal interference intensity field for attenuation compensation. The compensation result is aligned with the phase of the resonance frequency to obtain the reflection interference intensity field; In step 503, the shaft cavity geometry refers to the physical dimensions of the elevator shaft or other cavity, including height, width and depth, which are used to calculate the resonant frequency of the cavity. The cavity resonant frequency is a resonance phenomenon at a specific frequency caused by the geometry of the cavity, including the resonant frequency and its corresponding acoustic characteristics, which is used to analyze the propagation characteristics of the signal in the cavity. The metal shaft reflection path describes the reflection path of the signal in the metal shaft, including the number of reflections, angle and path length, which is used to evaluate the attenuation of the signal in the shaft. Attenuation compensation is the process of compensating for the energy loss caused by various factors during the transmission of the signal, including compensation algorithms and parameters for restoring signal strength. The reflection interference intensity field is a spatial model that describes the reflection interference intensity at different positions, including the reflection interference intensity value of each point, which is used to optimize the signal transmission path and reduce reflection interference.

[0149] In an embodiment of the present application, first, the geometric dimensions of the shaft cavity are analyzed and its resonant frequency is calculated. Next, attenuation compensation is performed by combining the metal shaft reflection path with the metal interference intensity field generated in step 501. Then, the compensation result is adjusted to align with the phase of the resonant frequency to generate a reflection interference intensity field. Specifically, acoustic simulation technology and signal processing algorithms are used to optimize the compensation effect to ensure that the signal can maintain a high quality when passing through the metal shaft. Finally, all information is integrated to form a comprehensive reflection interference intensity field.

[0150] For example, first, the cavity resonance frequency is calculated based on the elevator shaft CAD model (formula: f=c / 2L, c is the speed of light, L is the characteristic size), and the typical shaft resonance frequency range is 300-800MHz. Secondly, the FDTD (finite difference time domain) method is used to simulate the propagation of electromagnetic waves in the metal shaft, and the path loss compensation (compensation amount 3-8dB) is performed in combination with the interference intensity field of step 501. Then, the compensation signal is aligned with the resonance frequency phase (phase difference <5°) through the Hilbert transform to generate an intensity field containing reflection gain / loss values, in which the interference intensity of the reflection hot spot area is increased by 15-20dB.

[0151] 504. By fusing the metal interference intensity field, the dynamic correction parameter of the absorption rate and the reflection interference intensity field, the metal interference gradient and the absorption rate parameter are polarization-weighted and embedded into the reflection interference phase constraint to generate a three-dimensional signal attenuation field.

[0152] In step 504, the metal interference gradient describes the trend of metal interference intensity changing with space, including the rate of change of interference intensity at each point, which is used to evaluate the degree of metal interference received by the signal at different positions. The polarization weighting of the absorption rate parameter is a weighted processing of the absorption rate parameter to reflect its influence in different directions, including weighting coefficients and direction vectors, which are used to more accurately simulate the propagation of signals in complex environments. The reflection interference phase constraint is a technical means for correcting the phase of the reflected signal, including phase adjustment algorithms and parameters 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 spatial signal attenuation model formed after comprehensive consideration of multiple factors, including the signal attenuation value at each position, which is used to guide the design and optimization of communication networks and ensure stable transmission of signals in complex environments.

[0153] In the embodiment of the present application, first, the metal interference intensity field, the dynamic correction parameter of the absorption rate and the reflection interference intensity field are integrated. Then, the metal interference gradient is calculated and polarization weighted processing is performed, combined with the absorption rate parameter. Then, these parameters are embedded in the reflection interference phase constraint to generate a three-dimensional signal attenuation field. Specifically, multidimensional data analysis technology and optimization algorithms are used to process complex multi-source data to ensure that the generated model can accurately reflect the actual situation. Finally, all information is integrated to form a comprehensive three-dimensional signal attenuation field.

[0154] For example, first, the metal interference intensity field (step 501) and the reflection interference intensity field (step 503) are non-uniformly grid fused (the grid size is adaptively adjusted to 0.1-1m). Secondly, the dynamic correction parameter of the absorption rate (step 502) is differentially weighted according to the polarization direction (X / Y / Z axis) (weight ratio 4:3:3). Then, the phase constraint optimization algorithm (least squares fitting) is applied to eliminate the phase conflict between the reflection paths. The final output is 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 1Hz, which can support accurate prediction of the millimeter wave frequency band.

[0155] Here is a specific example: In a 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, the metal interference intensity field was dynamically generated based on the spatial correlation between the distribution density of metal components and signal reflection. Subsequently, the research team generated dynamic correction parameters for the absorption rate based on 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 combined the metal shaft reflection path with the metal interference intensity field generated in step 501 for attenuation compensation. Finally, the metal interference gradient was calculated and polarization weighted processing 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 capabilities of the entire shopping mall.

[0156] In summary, steps 501 to 504 generate a three-dimensional signal attenuation field by multi-dimensionally fusing the coordinates of the metal-dense area, the difference in the absorption rate of the wall material, 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.

[0157] In order to further improve the attenuation compensation accuracy of the reflection path in the shaft cavity, in some embodiments, the cavity resonance frequency is obtained according to the analysis of the geometric dimensions of the shaft cavity in step 503, the metal shaft reflection path is combined with the metal interference intensity field for attenuation compensation, and the compensation result is aligned with the phase of the resonance frequency to obtain the reflection interference intensity field, including: 601. Combined with the standing wave modal analysis of the rigid node constraints of the vertical channel of the building, the geometric dimensions of the shaft cavity are analyzed by calculating the ratio of the cross-sectional perimeter to the height to generate the cavity resonance frequency; In step 601, the rigid node constrained standing wave modal analysis of the vertical channel of the building refers to the structural analysis of the vertical channel inside the building (such as the elevator shaft) to determine its inherent vibration mode. This includes evaluating factors such as the location of rigid nodes, material properties and structural design to predict possible standing wave modes. The shaft cavity geometry refers to the specific physical dimensions of the elevator shaft or other cavity, including the cross-sectional shape, circumference and height. These data are used to calculate the resonant frequency of the cavity. The ratio of the cross-sectional circumference to the height is an important parameter to describe the geometric characteristics of the cavity and is used to analyze its acoustic properties. The cavity resonant frequency is a resonance phenomenon at a specific frequency caused by the geometry of the cavity, which includes the resonant frequency and its corresponding acoustic properties, and is used to optimize the signal transmission path.

[0158] In the embodiments of the present application, first, the building information model data is parsed 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, the standing wave modal analysis technique is applied to calculate the natural vibration mode of the cavity. Then, the acoustic simulation tool is used to simulate the response at different frequencies and identify the resonant frequency of the cavity. Specifically, the finite element method or the boundary element method is used to process complex multidimensional data sets to ensure that the generated frequencies can accurately reflect the actual situation. Finally, all the information is integrated to generate a comprehensive list of cavity resonant frequencies.

[0159] For example, first, based on the building BIM model, the precise geometric parameters of the shaft cavity are extracted (cross-section perimeter error <1cm, height error <3cm), and the perimeter-to-height ratio (P / H ratio) is calculated as the key characteristic parameter. Secondly, the finite element modal analysis method is used to solve the characteristic frequency equation that satisfies the rigid boundary conditions in the 10-3000MHz frequency band, and the first 10 resonant modes (including TE and TM modes) are identified. Then, the simulation results are verified by experimental measurement (sweep frequency test using a vector network analyzer with an accuracy of ±0.5MHz), and the mapping relationship between the P / H ratio and the fundamental frequency is established: f0=kc / √(P×H), where k is the shape factor (0.49 for a rectangular cross section) and √ represents the square root. Finally, a parameter table containing the resonant frequencies of each order (f0, 2f0, 3f0...) and the corresponding Q values ​​is output for subsequent phase matching.

[0160] 602. Perform attenuation compensation on the multipath reflection characteristics of the metal well reflection path through the metal interference intensity field, combine the Fresnel reflection coefficient with the geometric projection of the signal incident angle, match the harmonic order of the cavity resonance frequency, and generate a reflection path attenuation compensation parameter; 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 length. The multipath reflection feature refers to the phenomenon that the signal is reflected on multiple paths, which affects the quality of the signal. 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 to calculate the behavior of the signal at the interface based on the incident angle, which is used to adjust the attenuation compensation of the reflection path. The reflection path attenuation compensation parameter is a final compensation parameter after comprehensive consideration of multiple factors, which is used to reduce the signal loss on the reflection path.

[0161] In the embodiments of the present application, first, the reflection path in the metal shaft is analyzed based on the multipath reflection characteristics of the metal interference intensity field. Then, attenuation compensation is performed by combining the geometric projection of the Fresnel reflection coefficient and the signal incident angle. 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.

[0162] For example, first, use a ray tracing engine to simulate multipath propagation in a metal shaft (minimum number of rays: 5000), and record the time delay, incident angle, and number of reflections for each path. Secondly, based on the metal interference intensity field of step 501, apply a compensation amount that decays exponentially with distance to each path (compensation coefficient α = 0.3-0.7dB / m). Then, calculate the Fresnel coefficient of each reflection point (considering the polarization mode, with an accuracy of 0.01), and generate the path weight in combination with the geometric projection of the incident angle (angle resolution 1°). Finally, match the compensated path response with the harmonic component (n×f0) of the cavity resonance frequency, and optimize the frequency-related attenuation compensation parameter matrix (dimension: distance×frequency×angle) through the minimum mean square error algorithm.

[0163] 603. Perform spiral 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 the standing wave energy threshold is cut off.

[0164] 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 rate of change of each point, which is used to evaluate the degree of attenuation 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, which is used to correct the signal phase on the reflection path. The spiral 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 a standard value for screening the standing wave energy, retaining only the effective reflection interference components. The reflection interference intensity field is a three-dimensional model that shows the reflection interference intensity in different areas.

[0165] In the embodiment of the present application, first, the radial gradient of the reflection path attenuation compensation parameter is calculated and spirally convolved with the phase of the cavity resonance frequency. Next, the standing wave energy threshold cutoff technology 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 information is integrated to form a comprehensive reflection interference intensity field.

[0166] For example, first, the radial gradient of the compensation parameter (∂C / ∂r, step size 0.1m) is calculated in the cylindrical coordinate system and spirally convolved with the resonant phase distribution (∂ϕ / ∂z) (kernel size 5×5). Secondly, the standing wave energy threshold is set (usually -20dB of the maximum value), and the morphological opening operation is used to eliminate discrete noise. Then, the complete reflection interference intensity field is reconstructed by three-dimensional interpolation (grid accuracy 0.2m), and special areas are marked: ① phase reversal area (red mark) ② strong interference area (yellow mark) ③ safe transmission area (green mark). The final output intensity field contains vector data (E field / H field distribution) and scalar data (power density), and supports parallel display of multiple frequency bands (700MHz / 2.6GHz / 3.5GHz).

[0167] In summary, steps 601 to 603 generate key frequencies that affect wireless signal transmission by analyzing the vertical channel characteristics of the building, and finally generate a reflection interference intensity field that reflects the actual environmental characteristics by accurately calculating the reflection path attenuation compensation parameters, which significantly improves the accuracy and stability of wireless signal transmission in complex building structures, especially in vertical shaft environments with a large amount of metal materials, and can effectively reduce signal loss and improve communication quality.

[0168] In order to solve the problem of insufficient accuracy of wireless signal coverage in complex building environments, in some embodiments, the step 103 predicts the spatiotemporal distribution of the number of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and the fire channel position coordinates, including: 701. Discretize the communication blind area boundary data into three-dimensional grid units, calculate the signal attenuation coefficient of the three-dimensional grid unit and the three-dimensional Manhattan distance to the nearest fire channel entrance, and generate an accessibility matrix; In step 701, the communication blind area boundary data refers to the boundary information of the area where normal communication cannot be achieved due to signal attenuation. The three-dimensional grid cell is the basic unit after the space is discretized, and is used to describe the attributes within a specific area. The signal attenuation coefficient represents the degree of attenuation 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 distance from each grid cell to the nearest fire passage entrance and the signal attenuation coefficient, which is used to guide evacuation route planning.

[0169] In the embodiment of the present application, first, the communication blind area boundary data is discretized into three-dimensional grid cells. Then, the signal attenuation coefficient of each grid cell is calculated using electromagnetic simulation technology. Then, based on the three-dimensional Manhattan distance algorithm, the distance from each grid cell to the nearest fire passage entrance is calculated. Finally, this information is integrated to generate a comprehensive accessibility matrix. Specifically, a graph theory algorithm is used to process complex multidimensional data sets to ensure that the generated matrix can accurately reflect the actual situation.

[0170] For example, first, the building space is divided into three-dimensional grid units of 0.5m×0.5m×0.5m, and the signal attenuation coefficient of each unit is calculated by ray tracing technology (accuracy ±0.5dB). Secondly, the three-dimensional Manhattan distance from each grid to the nearest fire escape is calculated based on the A* algorithm (considering the wall penetration penalty coefficient). Then, an accessibility matrix containing the following elements is constructed: ① grid coordinates ② attenuation coefficient ③ shortest distance ④ path key points, and the data update frequency is 1Hz. Finally, the accessibility matrix is ​​normalized (range 0-1), and the distance value is converted into an accessibility score (1 means optimal).

[0171] 702. Construct a dynamic path network based on the fire channel location coordinates, divide the time window and associate the historical SMS request volume data, and generate the request volume change baseline of the three-dimensional grid unit; In step 702, the fire escape location coordinates refer to the specific location information of the safety exit and evacuation passage in the building. The dynamic path network is a path selection system that is updated in real time according to time and traffic conditions. The time window is to divide a continuous time period into several time periods for more detailed time series analysis. The historical SMS request volume data records the number of SMS requests in each area over a period of time. The request volume change baseline is a benchmark value that reflects the trend of SMS request volume changes in different time periods and is used to predict future request volumes.

[0172] In the embodiment of the present application, first, a dynamic path network is constructed based on the location coordinates of the fire escape. Next, the time window is divided and the historical SMS request volume data is associated. Then, the time series analysis method is used to generate a request volume change baseline for each three-dimensional grid cell. Specifically, a machine learning algorithm is used to process complex time series data to ensure that the generated baseline can accurately reflect actual demand changes. Finally, all information is integrated to generate a comprehensive request volume change baseline.

[0173] For example, first, divide the time window into 15-minute intervals and associate the historical SMS request data (including location, time, and device type) of the past 30 days. Secondly, use the LSTM network to extract spatiotemporal features and generate the hourly request volume pattern of each grid unit. Then, overlay the real-time crowd heat map (from video analysis) for dynamic correction, and establish a baseline model with the following dimensions: ① workday / holiday pattern ② morning and evening peak characteristics ③ emergency impact factors. The final output is a request volume prediction baseline with an accuracy of ±5%.

[0174] 703. Spatially superimpose the three-dimensional grid unit and the dynamic path network, calculate the shortest connected path, integrate the wall penetration loss compensation value, and construct a path loss correction model; In step 703, the dynamic path network is a path selection system that is updated in real time according to time and traffic conditions. 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 that compensates for the energy loss of the signal when passing through the wall. The path loss correction model is a path loss adjustment model formed after comprehensively considering multiple factors, which is used to optimize the signal transmission path and reduce the loss on the path.

[0175] In the embodiment of the present application, first, the three-dimensional grid unit is spatially superimposed with the dynamic path network. Then, the shortest connected path is calculated and the wall penetration loss compensation value is integrated. Then, based on the above processing results, a path loss correction model is constructed. Specifically, graph theory algorithms and ray tracing technology are used to handle complex path problems to ensure that the generated model can accurately reflect the actual situation. Finally, all information is integrated to form a comprehensive path loss correction model.

[0176] For example, first, the dynamic path network (generated by Dijkstra algorithm) is three-dimensionally superimposed with the grid unit (Z-axis accuracy 0.1m). Secondly, the penetration loss compensation value of each path is calculated (concrete wall +8dB, metal door +12dB). Then, the optimal path weight (distance weight 60%, signal quality 40%) is determined through Bayesian optimization, and a correction model containing the following parameters is established: ① path loss reference value ② real-time correction coefficient ③ redundant path score. The model automatically optimizes the parameters every 5 minutes.

[0177] 704. Perform time series decomposition on the signal attenuation coefficient to obtain a periodic component and a residual component, and generate a real-time signal coverage quality index of the three-dimensional grid unit in combination with the path loss correction model; In step 704, the signal attenuation coefficient indicates the degree of attenuation of the wireless signal when passing through different materials or media. Time series decomposition is the process of decomposing time series data into periodic components and residual components, which is used to identify long-term trends and short-term fluctuations in the data. The periodic component is the recurring part of the data, reflecting the regular changes. The residual component is the random part of the data, reflecting the unpredictable changes. The real-time signal coverage quality index is an indicator formed by comprehensively considering multiple factors, which is used to evaluate the quality of the current signal coverage.

[0178] In the embodiment of the present application, first, the signal attenuation coefficient is time-series decomposed to obtain the periodic component and the residual component. Then, combined with the path loss correction model, the real-time signal coverage quality index of each three-dimensional grid unit is calculated. Specifically, the time series analysis method is used to process complex signal data to ensure 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.

[0179] For example, first, the attenuation coefficient is decomposed into STL time series (period = 24 hours) to separate the diurnal period component and the emergency residual. Secondly, the path loss correction result is convolved with the periodic component (kernel size 3×3×3). Then, it is normalized to a quality index of 0-100 through the Sigmoid function, and three thresholds are set: excellent (>80), qualified (60-80), and poor (<60). The index is updated every 30 seconds and visualized.

[0180] 705. The reachability matrix, the real-time signal coverage quality index and the request volume change baseline are integrated into a multi-dimensional feature tensor, a spatiotemporal coupling feature map is generated through a channel attention mechanism, and a predicted value of the SMS request volume in the future time slice is output.

[0181] In step 705, the accessibility matrix is ​​a data structure containing the distance from each grid cell to the nearest fire escape entrance and the signal attenuation coefficient. The real-time signal coverage quality index is an indicator formed after comprehensive consideration of multiple factors, and is used to evaluate the quality of the current signal coverage. The request volume change baseline is a benchmark value that reflects the trend of changes in the number of SMS requests in different time periods. The multidimensional feature tensor is a multidimensional data structure containing multiple feature information, which is used as the input of the deep learning model. The channel attention mechanism is a technology used to enhance the performance of neural networks by focusing on important features to improve prediction accuracy. The spatiotemporal coupling feature map is a map that combines time and space features and is used to predict future SMS request volumes.

[0182] In the embodiment of the present application, first, the reachability matrix, the real-time signal coverage quality index and the request volume change baseline are fused into a multi-dimensional feature tensor. Then, the channel attention mechanism is applied to generate a spatiotemporal coupling feature map. Then, based on the above processing results, the predicted value of the SMS request volume in the future time slice is output. Specifically, a deep learning algorithm is used to process complex multi-source data to ensure that the generated prediction value can accurately reflect the actual situation. Finally, all information is integrated to form a comprehensive SMS request volume prediction model.

[0183] For example, first, construct a five-dimensional feature tensor (space x, y, z + time + quality index), and use 3D CNN to extract spatiotemporal features. Secondly, strengthen key features through the channel attention mechanism (SE module) (weight increase by 30-50%). Then, use the TCN network to predict the distribution of request volume in the next 15-60 minutes, and output the prediction results containing the following information: ① hot spot area coordinates ② expected request volume ③ confidence interval (90% confidence level). The system achieved 85% prediction accuracy in actual tests in large shopping malls.

[0184] Here is a specific example: In an application scenario of a large shopping mall, the research team first parsed the building information model data, obtained the communication blind spot boundary data, and discretized it into three-dimensional grid cells. Then, a dynamic path network was constructed based on the location coordinates of the fire channel, and the time window was divided to generate the request volume change baseline for each three-dimensional grid cell. Next, the team calculated the shortest connected path and integrated the wall penetration loss compensation value to construct a path loss correction model. The research team fused the accessibility matrix, real-time signal coverage quality index, and request volume change baseline into a multidimensional feature tensor to obtain a spatiotemporal coupling feature map. Then, based on the above processing results, the predicted value of the SMS request volume for the future time slice was output. This not only improves the efficiency of the communication system, but also enhances the speed and accuracy of emergency response, providing customers with a smoother and safer experience.

[0185] In summary, steps 701 to 705 generate an accurate accessibility matrix and request volume change baseline based on the communication blind spot boundary data and the fire channel location coordinates, and significantly improve the accuracy and stability of wireless signal transmission in complex building structures through the path loss correction model and the spatiotemporal coupling feature map, especially in the presence of a large amount of metal materials and complex layouts, which can effectively reduce signal loss and improve communication quality.

[0186] Figure 2 A schematic diagram of a scenario-based adaptive SMS delivery processing system is provided for an embodiment of the present application. Figure 2 As shown, the system includes: The acquisition module 21 acquires spatial topology data by analyzing the building information model, and generates a three-dimensional communication feature map including signal propagation loss weight and crowd density gradient by combining the personnel residence heat map collected by the infrared thermal imaging array; A construction module 22 detects the difference in absorption rate of wall materials based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map, and constructs a three-dimensional signal attenuation field including the reflection interference of the elevator shaft; The prediction module 23 predicts the spatiotemporal distribution of the number of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and the position coordinates of the fire channel; The generation module 24 is used to access the Internet SMS gateway cluster, collect base station signaling plane data and user terminal online status to generate a thermal layer of the cross-operator network quality, and map the thermal layer to the three-dimensional communication feature map to form a fused communication quality field; 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 fused communication quality field, generate message forwarding rules sorted by priority; The processing module 25 is used to execute the SMS sending operation according to the message forwarding rule.

[0187] Figure 2 The scenario-based adaptive SMS delivery processing system can be executed Figure 1 The implementation principle and technical effects of the SMS sending and processing method based on scene adaptation described in the illustrated embodiment will not be repeated. For a SMS sending and processing system based on scene adaptation in the above embodiment, the specific manner in which each module and unit performs operations has been described in detail in the embodiment of the method, and will not be elaborated here. Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present application, not to limit it; although the present application has been described in detail with reference to the aforementioned embodiments, ordinary technicians in this field should understand that it is still possible to modify the technical solutions recorded in the aforementioned embodiments, or to replace some of the technical features therein with equivalents; and these modifications or replacements do not make the essence of the corresponding technical solution deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A scenario-based SMS sending and 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 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, characterized in that 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, characterized in that 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, characterized in that: 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: The high-density area coordinates in the spatiotemporal distribution of the SMS request volume in each partition are integrated with the group moving direction vector, and the matching degree between the moving direction clustering result and the historical request volume growth rate is calculated by associating the signal strength attenuation rate of the network congestion hotspot, so as to generate the emergency channel weight coefficient; Based on the peak period characteristics of the gateway queue backlog, combined with the network congestion hotspot and the signal attenuation gradient of adjacent access points, the cross-floor diversion ratio is generated by matching the attenuation gradient inflection point with the physical topology distance; According to the geometric boundary data of the weak signal area, the coverage range of the idle channels of the adjacent access points in the network congestion hotspot is extracted, and the relay forwarding priority sequence is generated by analyzing the correlation strength between the access point response delay and the signal compensation demand; The emergency channel weight coefficient, the cross-floor diversion ratio and the relay forwarding priority sequence are multi-dimensionally integrated 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 cache queue backlog; Based on the critical parameters of the forwarding rule trigger matrix, the message path conflict is simulated in the digital twin environment, and the message forwarding rules sorted by priority are generated by reversely correcting the matching threshold and signal compensation parameters.

5. The method according to claim 4, characterized in that The multi-dimensional fusion of the emergency channel weight coefficient, the cross-floor diversion ratio and the relay forwarding priority sequence to construct a forwarding rule trigger matrix includes: The emergency channel weight coefficient is integrated with the real-time load and historical traffic data of each channel, combined with the channel length to generate a basic constraint value, and the obstacle density is superimposed for dynamic compensation to obtain a channel quality evaluation index; The cross-floor diversion ratio is based on the floor escape node density, coupled with the escape ladder capacity threshold and the channel quality evaluation index to generate a diversion gradient coefficient, and the building load-bearing limit is integrated to output a diversion ratio benchmark value; The relay forwarding priority sequence constructs a node coverage map through the wireless signal multipath effect, 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 chain table; The channel quality evaluation index, diversion ratio benchmark value and priority sorting list are input into the building digital twin, and the implicit association rules are extracted through tensor decomposition and orthogonal recombination to construct the forwarding rule trigger matrix.

6. The method according to claim 5, characterized in that The step of generating a diversion gradient coefficient by basing the cross-floor diversion ratio on the floor escape node density, coupling the escape ladder capacity threshold and the channel quality evaluation index includes: The density of floor escape nodes is based on the visible distance and the number of path branches to generate the node coverage strength value. The connectivity compensation is triggered by the visible distance, and the number of path branches corrects the topological strength attenuation. Based on the escape ladder capacity threshold and the smoke diffusion rate, a real-time traffic attenuation coefficient is generated by asymmetric superposition of the static capacity baseline and the dynamic attenuation factor and the introduction of the node coverage strength value; 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 the diversion gradient coefficient.

7. The method according to claim 1, characterized in that The method of detecting the difference in absorption rate of wall materials based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map and constructing a three-dimensional signal attenuation field including the reflection interference of the elevator shaft includes: The coordinates of the metal-dense area of ​​the three-dimensional communication characteristic map dynamically generate the metal interference intensity field through the correlation between the distribution density of metal components and the signal reflection space; The absorption rate benchmark value is generated by mapping the material type and the electromagnetic wave penetration loss, and the wall moisture content and thickness variation coefficient are introduced. The ambient humidity data is combined with the frequency domain convolution of the benchmark value to generate the absorption rate dynamic correction parameter. The cavity resonance frequency is obtained according to the analysis of the geometric dimensions of the shaft cavity, and the metal shaft reflection path is combined with the metal interference intensity field for attenuation compensation. The compensation result is aligned with the phase of the resonance frequency to obtain the reflection interference intensity field; By fusing the metal interference intensity field, the dynamic correction parameter of the absorption rate and the reflection interference intensity field, the metal interference gradient and the absorption rate parameter are polarization-weighted and embedded in the reflection interference phase constraint to generate a three-dimensional signal attenuation field.

8. The method according to claim 7, characterized in that The cavity resonance frequency is obtained according to the analysis of 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 aligned with the phase of the resonance frequency to obtain the reflection interference intensity field, including: Combined with the standing wave modal analysis of the rigid node constraints of the vertical channel of the building, the geometric dimensions of the shaft cavity are analyzed by calculating the ratio of the cross-sectional perimeter to the height to generate the cavity resonance frequency; The attenuation compensation of the metal well reflection path is performed through the multipath reflection characteristics of the metal interference intensity field, and the harmonic order of the cavity resonance frequency is matched in combination with the geometric projection of the Fresnel reflection coefficient and the signal incident angle 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 the reflection interference intensity field is generated after the standing wave energy threshold is truncated.

9. The method according to claim 1, characterized in that: The predicting of the spatiotemporal distribution of the amount of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and the fire channel position coordinates includes: Discretize the communication blind area boundary data into three-dimensional grid cells, calculate the signal attenuation coefficient of the three-dimensional grid cells and the three-dimensional Manhattan distance to the nearest fire channel entrance, and generate an accessibility matrix; Building a dynamic path network based on the fire channel location coordinates, dividing the time window and associating the historical SMS request volume data, and generating a request volume change baseline for the three-dimensional grid unit; The three-dimensional grid unit is spatially superimposed with the dynamic path network, the shortest connected path is calculated and the wall penetration loss compensation value is integrated 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 the path loss correction model to generate a real-time signal coverage quality index of the three-dimensional grid unit; The reachability matrix, real-time signal coverage quality index and request volume change baseline are fused into a multi-dimensional feature tensor, a spatiotemporal coupling feature map is generated through a channel attention mechanism, and the predicted value of the SMS request volume in the future time slice is output.

10. A scenario-adaptive SMS delivery processing system, characterized in that: include: The acquisition module obtains spatial topology data by analyzing the building information model, and generates a three-dimensional communication feature map containing signal propagation loss weights and crowd density gradients by combining the personnel residence heat map collected by the infrared thermal imaging array; A construction module detects the difference in absorption rate of wall materials based on the coordinates of the metal-dense area in the three-dimensional communication characteristic map, and constructs a three-dimensional signal attenuation field including the reflection interference of the elevator shaft; A prediction module, which predicts the spatiotemporal distribution of the amount of SMS requests in each partition based on the communication blind area boundary data in the three-dimensional signal attenuation field and the position coordinates of the fire channel; A generation module is used to access the Internet SMS gateway cluster, collect base station signaling plane data and user terminal online status to generate a thermal layer of cross-operator network quality, and map the thermal layer to the three-dimensional communication feature map to form a fused communication quality field; 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 fused communication quality field, generate message forwarding rules sorted by priority; The processing module is used to execute the SMS sending operation according to the message forwarding rule.

Images