Multi-mode 5G Internet of Things card signal intelligent coverage method and system

Abstract

The invention relates to the field of Internet of Things equipment, in particular to a multi-mode 5G Internet of Things card signal intelligent coverage method and system. The method comprises the following steps: acquiring a project progress information set, and based on the project progress information set, carrying out multi-dimensional cooperative channel sensing and space-time correlation analysis to generate a space-time signal atlas set dynamically evolved along with a physical space structure; the method comprises the following steps: acquiring an unmanned aerial vehicle cluster task and a state data set, and on the basis of the unmanned aerial vehicle cluster task and state data set, performing communication role dynamic allocation and multi-modal link optimization by constructing a task-driven distributed collaborative decision to generate an unmanned aerial vehicle cluster dynamic ad hoc network instruction set; and performing network toughness enhancement and adaptive optimization processing based on the unmanned aerial vehicle cluster dynamic ad hoc network instruction set and execution feedback data thereof, and generating and outputting a signal intelligent coverage efficiency report. According to the invention, in the 5G signal intelligent coverage process, the network deployment efficiency, the communication reliability and the overall toughness in scenes such as an intelligent construction site and emergency communication can be significantly improved.

Description

Technical Field

[0001] This application relates to the field of Internet of Things (IoT) devices, and in particular to a method and system for intelligent coverage of multimodal 5G IoT card signals. Background Technology

[0002] In the fields of complex engineering and emergency support, drone swarms, with their advantages of mobility and flexibility, have become a key carrier for achieving wide-area, dynamic intelligent coverage of 5G IoT signals and data transmission. The quality of their communication coverage directly determines the reliability of remote control, real-time monitoring and massive data backhaul, and is an important communication infrastructure for building digital twin systems in smart construction sites, emergency rescue and complex industrial scenarios.

[0003] However, existing drone-assisted signal coverage methods generally rely on static or pre-set network deployment and fixed relay strategies. They lack the ability to conduct real-time and collaborative signal exploration and perception of dynamically changing physical space structures. Furthermore, they cannot intelligently allocate multi-modal communication roles and adaptively optimize network configuration based on task requirements and real-time link status. This results in blind spots in signal coverage, insufficient network resilience, and difficulty in ensuring stable and efficient end-to-end communication services in complex and time-varying environments. Summary of the Invention

[0004] This application provides a method and system for intelligent coverage of multimodal 5G IoT card signals to solve the above-mentioned technical problems.

[0005] In a first aspect, this application provides a method for intelligent coverage of multimodal 5G IoT card signals, the method comprising:

[0006] Obtain a set of project progress information, and based on the set of project progress information, perform multi-dimensional collaborative channel perception and spatiotemporal correlation analysis to generate a set of spatiotemporal signal maps that dynamically evolve with the physical space structure.

[0007] A dataset of drone cluster tasks and statuses is obtained. Based on the dataset and the spatiotemporal signal map set, a task-driven distributed collaborative decision-making system is constructed to dynamically allocate communication roles and optimize multimodal links, thereby generating a dynamic self-organizing network instruction set for the drone cluster.

[0008] Based on the command set of the UAV swarm dynamic self-organizing network and its execution feedback data, network resilience enhancement and adaptive optimization are performed to generate and output a signal intelligent coverage performance report.

[0009] The above technical solutions transform the traditional static and passive signal coverage mode into a dynamic, proactive, and adaptive intelligent system. This system can accurately predict and respond to complex and dynamic environmental changes, achieve optimal matching between network resources and real-time business needs, and significantly improve network deployment efficiency, communication reliability, and overall resilience in scenarios such as smart construction sites and emergency communications.

[0010] Optionally, the generation of the spatiotemporal signal spectrum set that dynamically evolves with the physical spatial structure includes:

[0011] Based on the project progress information set, a 5G signal detection task is dynamically triggered and executed to obtain a real-time communication detection dataset of the target space.

[0012] Based on the real-time communication detection dataset, a voxelized three-dimensional signal map of the target space is constructed and updated, wherein each voxel unit is associated with multi-dimensional signal attributes and physical environment attributes.

[0013] A temporal fusion algorithm is used to perform correlation analysis on the voxelized three-dimensional signal map of each update to generate the spatiotemporal signal map set that characterizes the spatial distribution and evolution of the signal.

[0014] Optionally, the dynamic triggering and execution of the signal detection task includes:

[0015] Based on the project progress information set, the key change nodes and periodic performance degradation warnings of the project's physical spatial structure are analyzed, and active signal detection task instructions are generated.

[0016] The key change nodes include infrastructure construction milestones and completion events for updating the layout of operational scenarios;

[0017] Based on the active signal detection mission command, a detection formation consisting of a lead drone and a follower drone is dispatched into the target space:

[0018] The detection formation is controlled to perform gridded active detection to collect multi-band 5G channel depth information of preset sampling points, and simultaneously performs broad-spectrum passive listening to map the distribution information of spatial electromagnetic interference sources, which together constitute the real-time communication detection dataset.

[0019] Optionally, constructing and updating the voxelized three-dimensional signal map of the target space includes:

[0020] The real-time communication detection dataset is analyzed to extract the operational space coordinates and multi-dimensional signal features of each data point;

[0021] Based on the workspace coordinates, the target space is divided into a set of regular voxel units, and an initial voxelized 3D signal map is established for each voxel unit:

[0022] The physical attribute layer is composed of material and topological connectivity information inferred from the physical environment information of the data points; the signal fingerprint layer is composed of signal characteristics of the corresponding 5G base station access point; and the dynamic evolution layer is composed of signal change trends recorded based on historical signal data.

[0023] The newly acquired real-time communication detection dataset is fused and analyzed with the corresponding voxel units in the voxelized three-dimensional signal map using an incremental data fusion algorithm, so as to update the voxelized three-dimensional signal map.

[0024] Optionally, the method for generating a dynamic self-organizing network instruction set for unmanned aerial vehicle (UAV) swarms includes:

[0025] Based on the aforementioned drone cluster task and status dataset, the task intensity and real-time status information of each drone node are analyzed to perform initial task allocation and generate an initial task information set.

[0026] Based on the spatiotemporal signal map set, a UAV relay network topology characterized by chain connections is constructed to extend the intelligent signal coverage range.

[0027] Based on the initial task information set and the UAV relay network topology, a multimodal communication role is assigned to each UAV node. The multimodal communication role includes direct mode, relay mode and self-organizing network mode.

[0028] Optionally, the initial task allocation based on the analysis of task intensity and real-time status information of each UAV node includes:

[0029] The task and status dataset of the drone cluster is analyzed, and the spatial distribution of tasks, data throughput requirements and latency constraints are extracted as task intensity indicators. The remaining power of each drone node, the data load of its 5G IoT card and health status are extracted as real-time status information.

[0030] Based on the task intensity index and the real-time status information, a dynamic weighted evaluation algorithm is used to calculate the task suitability score for each UAV node.

[0031] Based on the task suitability score, the tasks to be executed are divided into multiple task subsets, and with the goal of maximizing the overall task completion efficiency and network lifetime, the task subsets are assigned to the corresponding drone nodes.

[0032] Optionally, the construction of the UAV relay network topology characterized by chain connections includes:

[0033] Based on the spatiotemporal signal map set, identify signal coverage blind spots and weak coverage areas in the target space;

[0034] The node with the best current signal quality in the UAV cluster is taken as the root node of the chain network. Based on the signal propagation characteristics and spatial obstacle information provided by the spatiotemporal signal map set, a path search algorithm is used to select the next hop relay node hop by hop starting from the root node, forming one or more chain paths extending to the signal coverage blind area and weak coverage area.

[0035] Each relay node on the chain path is responsible for forwarding data from its downstream nodes and dynamically adjusting its transmit power and antenna beam direction to maintain the stability and energy efficiency of the link.

[0036] Optionally, assigning multimodal communication roles to each UAV node includes:

[0037] For drone nodes with a good line-of-sight link with the 5G ground base station and a signal strength higher than the direct communication threshold, they are assigned to the direct mode and instructed to communicate directly with the 5G ground base station.

[0038] For drone nodes located in signal attenuation areas but with reliable links to other drone nodes, they are assigned to the relay mode and instructed to forward 5G data from upstream or downstream nodes in a single-hop or multi-hop manner.

[0039] For task sub-clusters that perform collaborative computing and data-intensive exchange, they are assigned to the self-organizing network mode, and are instructed to build a local mesh network inside them for 5G data sharing and processing, and elect one or more gateway nodes to connect to the chain network or 5G ground base station.

[0040] Optionally, the generation and output of the intelligent coverage performance report includes:

[0041] Based on the execution feedback data of the UAV swarm dynamic self-organizing network instruction set and the spatiotemporal signal map set, a multi-dimensional performance evaluation algorithm is used to analyze and aggregate quantitative evaluation results that include real-time link performance indicators, signal coverage integrity and network topology robustness.

[0042] The quantitative evaluation results are compared and analyzed spatiotemporally with historical performance data to generate a signal intelligent coverage performance report that includes network resilience rating and adaptive adjustment suggestions. The report is then output to the engineering management terminal in the form of a visual map.

[0043] Secondly, this application provides a multi-mode 5G IoT card signal intelligent coverage system, the system comprising:

[0044] The spatiotemporal signal map module is used to acquire the project progress information set, and based on the project progress information set, to perform multi-dimensional collaborative channel perception and spatiotemporal correlation analysis to generate a spatiotemporal signal map set that dynamically evolves with the physical space structure.

[0045] The dynamic networking instruction module is used to acquire the UAV cluster task and status dataset. Based on the UAV cluster task and status dataset and the spatiotemporal signal spectrum set, it generates a dynamic self-organizing network instruction set for the UAV cluster by constructing task-driven distributed collaborative decision-making, dynamically allocating communication roles and optimizing multimodal links.

[0046] The signal performance report module is used to perform network resilience enhancement and adaptive optimization processing based on the UAV cluster dynamic self-organizing network instruction set and its execution feedback data, and generate and output a signal intelligent coverage performance report. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application;

[0049] Figure 2 A flowchart illustrating a method for intelligent coverage of multimodal 5G IoT card signals, provided in one embodiment of this application;

[0050] Figure 3 This is a schematic diagram of the structure of a multimodal 5G IoT card signal intelligent coverage system provided in an embodiment of this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0052] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0053] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0054] Existing drone-assisted signal coverage methods generally rely on static or pre-defined network deployments and fixed relay strategies. They lack the ability to conduct real-time and collaborative signal exploration and perception of dynamically changing physical spatial structures. Furthermore, they cannot intelligently allocate multi-modal communication roles and adaptively optimize network configuration based on mission requirements and real-time link status. This results in blind spots in signal coverage, insufficient network resilience, and difficulty in ensuring stable and efficient end-to-end communication services in complex and time-varying environments.

[0055] Based on this, this application provides a method and system for intelligent signal coverage of multimodal 5G IoT cards. First, it integrates project progress information and active UAV detection data to construct and continuously update a spatiotemporal signal map reflecting dynamic changes in the physical environment. Then, using this map as an environmental situational awareness base, and combining real-time task requirements and UAV status, it dynamically allocates communication roles and optimizes multimodal links through a distributed collaborative decision-making algorithm, generating UAV swarm self-organizing network commands. Finally, based on command execution feedback, it performs network performance evaluation and adaptive optimization, generating a signal intelligent coverage effectiveness report, which is then output to implementation personnel. This solution revolutionizes the traditional static, passive signal coverage mode into a dynamic, proactive, and adaptive intelligent system, capable of accurately predicting and responding to complex dynamic environmental changes, achieving optimal matching of network resources and real-time service needs, and significantly improving network deployment efficiency, communication reliability, and overall resilience in scenarios such as smart construction sites and emergency communications.

[0056] Figure 1 This is a schematic diagram illustrating an application scenario provided by this application. In the process of intelligent 5G signal coverage, the method provided in this application can significantly improve network deployment efficiency, communication reliability, and overall resilience in scenarios such as smart construction sites and emergency communications.

[0057] Specifically, the method of this application is applied to any server that communicates with an engineering management platform and an unmanned aerial vehicle (UAV) management system. The server obtains engineering progress information from the engineering management platform and UAV swarm task and status datasets from the UAV management system. First, it integrates engineering progress information with UAV active detection data to construct and continuously update a spatiotemporal signal map reflecting dynamic changes in the physical environment. Then, using this map as the environmental situational awareness base, and combining real-time task requirements and UAV status, it dynamically allocates communication roles and optimizes multimodal links through a distributed collaborative decision-making algorithm, generating UAV swarm self-organizing network commands. Finally, based on command execution feedback, it performs network performance evaluation and adaptive optimization, generating a signal intelligent coverage effectiveness report, which is then output to the implementation personnel.

[0058] For specific implementation details, please refer to the following examples.

[0059] Figure 2 This is a flowchart illustrating a method for intelligent coverage of multimodal 5G IoT card signals according to an embodiment of this application. The method of this embodiment can be applied to servers in the above scenarios. Figure 2 As shown, the method includes:

[0060] S201. Obtain the project progress information set, and based on the project progress information set, perform multi-dimensional collaborative channel perception and spatiotemporal correlation analysis to generate a spatiotemporal signal spectrum set that dynamically evolves with the physical spatial structure.

[0061] The project progress information set can be a data collection that dynamically reflects changes in the physical spatial structure of the construction site. Its content includes: construction stages (such as foundation, main structure, and interior decoration), completion times of key milestones, hoisting locations and times of large components (such as steel structures and precast wall panels), movement trajectories of heavy machinery, and plans for the construction and dismantling of temporary facilities. The data originates from the project management platform. Multi-dimensional collaborative channel sensing can be performed by a swarm of drones equipped with 5G IoT cards, integrating active detection and passive listening, and covering multiple frequency bands (such as Sub-6GHz and millimeter waves) for comprehensive 5G channel status information collection. Spatiotemporal correlation analysis involves correlating and comprehensively analyzing 5G signal data collected at different time points with their corresponding spatial locations and project progress status to uncover the patterns of signal propagation characteristics changing over time (project progress) and space. The spatiotemporal signal atlas set refers to the core data product of the analysis output, a dynamically updated, three-dimensional voxel-based database. Each voxel unit contains not only the real-time signal strength, signal-to-noise ratio, and multipath characteristics of the current location (5G signal fingerprint), but also...

[0062] Specifically, in dynamic scenarios such as large-scale engineering construction, emergency rescue, or complex industrial inspection, the physical environment is constantly and drastically changing. This change directly and profoundly alters the propagation path, fading characteristics, and interference patterns of wireless signals, especially high-frequency 5G signals. Traditional signal assessment methods, which rely on static geographic information systems, pre-set propagation models, or periodic manual drive tests, cannot construct an accurate cognitive model reflecting the real-time state of the signal environment due to their inherent lag and low spatial resolution, resulting in a lack of reliable basis for subsequent network optimization decisions. This step aims to fundamentally solve this core contradiction, and its necessity is reflected in three aspects: First, by introducing engineering progress information sets as prior knowledge of environmental evolution and a trigger source for perception tasks, signal detection activities are proactively synchronized with key changes in the physical space, realizing a paradigm shift from passive response to proactive perception and prediction; Second, by performing multi-dimensional collaborative channel sensing and implementing spatiotemporal correlation analysis, discrete and heterogeneous detection data are fused to generate a voxelized spatiotemporal signal atlas with temporal evolution dimensions. This atlas not only represents the fine three-dimensional distribution of the current signal field, but also embeds the law of signal evolution as the project progresses, forming a high-fidelity digital twin of the communication environment. Finally, as the only and unified environmental situation input for all subsequent intelligent decisions, the accuracy and dynamism of this atlas are the cornerstone for the entire method to achieve self-adaptation and intelligence.

[0063] S202. Obtain the UAV cluster task and status dataset. Based on the UAV cluster task and status dataset and spatiotemporal signal map set, construct task-driven distributed collaborative decision-making, dynamically allocate communication roles and optimize multimodal links to generate a dynamic self-organizing network instruction set for the UAV cluster.

[0064] The drone swarm task and status dataset can include two types of information: first, business requirements issued by the task management platform; and second, real-time status information reported by the drones' onboard sensors and 5G IoT cards, such as location, battery level, remaining payload, real-time throughput of the 5G IoT card, signal reception strength, and link connection status. This data originates from the drone management system. Task-driven distributed collaborative decision-making means that each node in the drone swarm does not completely rely on the central controller, but rather, under unified rules and local information exchange, jointly evaluates task requirements, its own status, and the global environment map, competing or negotiating to arrive at the overall optimal networking solution. Dynamic allocation of communication roles can involve real-time assignment of functions within the communication network to each drone node equipped with a 5G IoT card. Core roles include: direct mode (directly connecting to the 5G ground base station as a regular terminal), relay mode (acting as a transparent forwarding node, extending network coverage), and self-organizing network mode (acting as a mesh node, forming a local high-bandwidth network with its peers). Multimodal link optimization can involve dynamically selecting access frequency bands, adjusting transmit power, and beamforming based on roles and real-time channel conditions to optimize overall link performance. The command set for a dynamic self-organizing network of drone swarms can be a series of executable commands issued to specific drone nodes, including: target flight position, hovering waypoint, assigned communication role, and 5G link parameters that need to be established or maintained.

[0065] Specifically, after obtaining accurate environmental situational awareness, the core decision-making challenge lies in how to schedule and control a swarm of drones equipped with 5G IoT cards to form an optimal communication coverage network. In this scenario, communication tasks exhibit diverse quality-of-service requirements and spatiotemporal distribution characteristics. Drone node resources (such as power, payload, and communication module capabilities) are limited and their states change in real time, while also needing to strictly adhere to the complex propagation constraints revealed by the spatiotemporal signal map. Traditional centralized, pre-programmed, or rule-based networking methods, due to their rigid topology and slow response to dynamic factors, struggle to achieve accurate and efficient adaptation between network resources and immediate task requirements. The key necessity of this step lies in introducing and implementing a task-driven distributed collaborative decision-making mechanism. This mechanism integrates the drone swarm's task objectives, individual states, and global environmental map, enabling the swarm to autonomously and collaboratively complete the dynamic allocation of communication roles and joint optimization of multimodal links with limited central intervention through distributed algorithms. This ensures that the network topology is not statically preset, but dynamically generated based on real-time needs (such as a sudden increase in the demand for high-definition video backhaul in a certain area) and real-time conditions (such as a decrease in the power of critical relay nodes), realizing a fundamental shift from "network adapting to fixed tasks" to "network dynamically shaping to serve immediate tasks." The generated dynamic self-organizing network instruction set is the specific output of this intelligent decision-making process, directly guiding the communication behavior and cooperative relationships of each UAV in space.

[0066] S203. Based on the UAV swarm dynamic self-organizing network instruction set and its execution feedback data, perform network resilience enhancement and adaptive optimization processing, and generate and output a signal intelligent coverage performance report.

[0067] Execution feedback data can be link-level performance data (such as actual throughput, latency, packet loss rate, and bit error rate) and command execution status (such as whether the target location was successfully reached and whether the role switch was successful) transmitted in real time by the drone node through the 5G IoT card and airborne system during the execution of self-organizing network commands. The signal intelligent coverage performance report can be a comprehensive evaluation document output by the system. The content includes not only statistics of 5G air interface link KPIs (key performance indicators) (such as average rate, coverage, and latency distribution), but also analysis of network topology robustness, early warning of potential risks (such as critical relay nodes that are about to exit due to low battery), and optimization suggestions based on historical data comparison.

[0068] Specifically, any model- and strategy-based decision-making will inevitably experience performance deviations in actual execution due to model errors, environmental disturbances, or sudden failures. A system lacking closed-loop feedback and self-evaluation capabilities has incomplete intelligence and its performance cannot be consistently guaranteed. In UAV-assisted dynamic coverage scenarios, the actual link performance after command execution, the robustness of the network topology, and potential risks must be continuously monitored, evaluated, and used to guide system optimization. The necessity of this step lies in constructing a complete closed loop from decision-making and execution to evaluation and optimization, which is a core component for achieving enhanced network resilience and long-term adaptive evolution. By collecting real execution data from UAV nodes and comparing it with the expected goals of the command set, it can promptly identify abnormal degradation or local failures in network performance. Based on this, the system can trigger adaptive optimization processes, such as dynamically adjusting transmission parameters, switching relay paths, or reassigning roles, thereby effectively enhancing the network's resilience in the face of uncertainty. Furthermore, through in-depth aggregation and spatiotemporal comparative analysis of feedback data, environmental maps, and historical performance, this step can generate a signal intelligent coverage performance report that surpasses traditional performance statistics. This report not only presents quantitative indicators, but also includes network health diagnostics, resilience ratings, and root cause optimization recommendations, achieving a leap from simple network monitoring to proactive network prognosis and management, and providing high-value insights for operational decisions.

[0069] The method provided in this embodiment first integrates project progress information with UAV active detection data to construct and continuously update a spatiotemporal signal map reflecting dynamic changes in the physical environment. Then, using this map as the environmental situational awareness base, and combining real-time task requirements and UAV status, a distributed collaborative decision-making algorithm dynamically allocates communication roles and optimizes multimodal links, generating UAV swarm self-organizing network commands. Finally, based on command execution feedback, network performance is evaluated and adaptively optimized, generating a signal intelligent coverage effectiveness report, which is then output to implementation personnel. This solution revolutionizes the traditional static, passive signal coverage mode into a dynamic, proactive, and adaptive intelligent system, capable of accurately predicting and responding to complex dynamic environmental changes, achieving optimal matching of network resources and real-time service needs, and significantly improving network deployment efficiency, communication reliability, and overall resilience in scenarios such as smart construction sites and emergency communications.

[0070] In some embodiments, based on the engineering progress information set, a 5G signal detection task is dynamically triggered and executed to obtain a real-time communication detection dataset of the target space; based on the real-time communication detection dataset, a voxelized three-dimensional signal map of the target space is constructed and updated, wherein each voxel unit is associated with multi-dimensional signal attributes and physical environment attributes; a temporal fusion algorithm is used to perform correlation analysis on the voxelized three-dimensional signal map of each update to generate a spatiotemporal signal map set characterizing the spatial distribution and evolution of the signal.

[0071] 5G signal detection tasks can be triggered dynamically based on changes in project progress, aiming to acquire specialized data on 5G signal propagation characteristics and electromagnetic environment status within a target space. Real-time communication detection datasets can be the raw data sets acquired after the 5G signal detection task is executed. Voxelized 3D signal maps can be 3D visualized signal models constructed based on the physical structure of the target space, using regular voxel units as basic units. Each voxel unit is associated with three core attributes: physical environment attributes, multidimensional signal attributes, and dynamic evolution attributes. Multidimensional signal attributes can be multi-dimensional quantitative indicators characterizing the propagation quality and communication capabilities of 5G signals. Physical environment attributes can be physical feature parameters of the target space that affect 5G signal propagation, including the material of the voxel unit (such as concrete, metal, etc., directly affecting signal penetration attenuation), topological connectivity (such as whether the space is enclosed, the presence of obstructions, etc., affecting the signal propagation path), and obstacle distribution density (such as the density of obstacles such as building components and equipment clusters). Temporal fusion algorithms can be algorithms that perform correlation analysis on voxelized 3D signal maps acquired at multiple time points, capable of uncovering signal change patterns between maps at different time points.

[0072] Specifically, traditional 5G signal coverage solutions suffer from core defects such as static nature, single dimension, and spatiotemporal disconnect. They often involve one-time signal detection after project completion, and the resulting two-dimensional maps cannot adapt to the dynamic changes in the physical space during construction. For example, in a 5G coverage project in an industrial park, the initial construction area had no obstructions and signal propagation was smooth. However, as factory walls were built and production equipment was installed, the previously strong coverage area became weak due to new obstructions. Traditional static maps could not capture this change in real time, leading to frequent interruptions in drone communication. At the same time, traditional maps do not associate physical environment attributes with the temporal evolution of signals, making it difficult to support accurate decision-making for drone self-organizing networks. To address the above issues, this step, based on the infrastructure completion and scene layout updates in the project progress information center, triggers an active signal detection task. A detection formation consisting of a lead drone and follower drones is dispatched to fly along a gridded path with a 5m spacing. This process collects multi-band 5G channel depth information (e.g., millimeter-wave band transmission delay of 10ms) at preset sampling points. Simultaneously, broad-spectrum passive listening is performed to map the distribution of spatial electromagnetic interference sources (e.g., interference frequency of 2.4GHz and intensity of 25dBm in a certain area), forming a real-time communication detection dataset. Subsequently, this dataset is analyzed to extract the three-dimensional spatial coordinates (e.g., X=10m, Y=8m, Z=3m) and multi-dimensional signal characteristics of each data point, arranged in 1m increments. The target space is divided into voxel units with a size of ×1m×1m. An initial three-dimensional map is constructed for each voxel, including a physical property layer based on concrete material and closed space topology, a signal fingerprint layer based on the base station signal strength of 45dBm, and a dynamic evolution layer that records historical signal changes. The new detection data is matched and updated with the corresponding voxel through an incremental data fusion algorithm. Finally, a temporal fusion algorithm is used to process the maps on the 7th and 14th days of construction. The maps after the completion node are assigned a weight of 0.7, and abnormal values ​​of signal strength caused by equipment failure (such as instantaneous drop to 15dBm) are removed. The signal attenuation trend is fitted to generate a spatiotemporal signal map set that represents the spatial distribution and evolution law.

[0073] The method provided in this embodiment, which dynamically triggers detection based on the engineering progress information set, can respond to changes in physical space in real time. The constructed voxelized 3D map can accurately map the signal and environmental correlation of different spatial points. The spatiotemporal signal map set formed after time-series fusion includes both spatial distribution and temporal evolution, perfectly solving the core pain points of traditional technologies. It is a key prerequisite for realizing intelligent coverage of multimodal 5G signals and supporting dynamic self-organizing networks of drone swarms.

[0074] In some embodiments, based on the engineering progress information set, key change nodes and periodic performance degradation warnings of the engineering physical space structure are analyzed to generate active signal detection mission instructions; key change nodes include infrastructure construction milestones and layout update completion events of the work scenario; based on the active signal detection mission instructions, a detection formation consisting of a lead drone and a follower drone is dispatched into the target space: the detection formation is controlled to perform gridded active detection to collect multi-band 5G channel depth information of preset sampling points, and simultaneously perform broad-spectrum passive listening to map the distribution information of spatial electromagnetic interference sources, which together constitute a real-time communication detection dataset.

[0075] Key change nodes can be landmark events during the construction process that significantly impact the physical structure and signal propagation environment of the target space. Periodic performance degradation warnings can be warning messages issued based on a preset time period (e.g., weekly, monthly) or equipment runtime, combined with historical signal performance data, to predict potential degradation trends in 5G signal coverage quality. Construction milestones can be key time points in the construction of large-scale infrastructure projects, such as tunnel construction reaching the halfway point or the completion of 5G ground base station foundation construction; these nodes directly change the layout of signal transmission sources and the basic propagation environment. Layout update completion events can be events in the target coverage area where the physical environment layout affecting signal propagation is adjusted and completed, including expansion or reduction of construction areas, completion of building wall construction, and installation of large equipment. Detection formations can be collaborative combinations of drones performing 5G signal detection tasks, consisting of one lead drone and several follower drones. Gridded active detection can be the core method for detection formations to perform signal acquisition. This involves dividing the target space into regular grids according to preset grid sizes (e.g., 5m×5m, 10m×10m). The detection formation flies sequentially according to the grid order, actively transmitting detection signals and receiving feedback at preset sampling points in each grid to collect 5G channel-related data. Multi-band 5G channel depth information can be detailed parameters reflecting channel transmission quality collected at preset sampling points for core 5G operating frequency bands (e.g., Sub-6GHz band, millimeter-wave band). Spatial electromagnetic interference source distribution information can be obtained through broad-spectrum passive listening, representing a set of information characterizing the location, intensity, and characteristics of electromagnetic interference sources within the target space.

[0076] Specifically, traditional 5G signal detection technologies suffer from core defects such as trigger lag, incomplete data, and incomplete coverage. They often employ fixed-cycle or one-time detection after completion, failing to respond to dynamic changes in the physical space during construction. For example, after the completion of workshop wall construction in a factory (a work scenario layout update event), areas with previously good signal coverage become weak due to obstruction. Traditional technologies fail to detect this in time, leading to frequent interruptions in subsequent drone communication. Furthermore, traditional detection methods often involve random sampling by a single drone, collecting signals from only a single frequency band and ignoring electromagnetic interference. For instance, detection at a construction site may fail to detect interference from tower cranes, resulting in signal compatibility failure. To address the above issues, this step first analyzes the project progress information set, extracting infrastructure construction milestones (such as the completion of 5G base station installation and commissioning) and work scenario layout update events (such as the completion of construction of Workshop No. 3). Combined with monthly periodic performance attenuation warnings, an early warning is triggered when the predicted signal attenuation reaches 12dBm, generating an active signal detection task instruction containing the detection range (such as longitude XX°XX′-XX°XX′, latitude XX°XX′-XX°XX′) and a 5m×5m grid size. Subsequently, a detection formation consisting of one lead drone and six follower drones is dispatched into the target space. The lead drone plans a gridded path, and the follower drones collect channel depth information (such as 9ms transmission delay and 38dB signal-to-noise ratio for the 700MHz band) at preset sampling points at the center of each grid. Simultaneously, spatial electromagnetic interference sources (such as interference frequency of 2.4GHz and intensity of 30dBm in a certain area) are captured through broad-spectrum passive listening. After deduplication and format standardization by the lead drone, a real-time communication detection dataset is formed.

[0077] The method provided in this embodiment identifies key change nodes and dynamically triggers detection based on the engineering progress information set and periodic performance degradation warnings (such as a monthly warning cycle). It adopts a "leader + follow" UAV formation, combined with gridded active detection and broad-spectrum passive listening, which can achieve full coverage without omissions and simultaneously acquire channel data and interference information. It perfectly solves the pain points of traditional technologies and is the key to ensuring the accuracy of subsequent signal spectrum and supporting UAV self-organizing network decision-making.

[0078] In some embodiments, the real-time communication detection dataset is parsed to extract the operational space coordinates and multi-dimensional signal features of each data point. Based on the operational space coordinates, the target space is divided into a set of regular voxel units, and an initial voxelized three-dimensional signal map is established for each voxel unit: a physical attribute layer composed of material and topological connectivity information inferred from the physical environment information of the data points, a signal fingerprint layer composed of signal features of the corresponding 5G base station access point, and a dynamic evolution layer based on historical signal data recording signal change trends. The newly acquired real-time communication detection dataset is fused and analyzed with the corresponding voxel units in the voxelized three-dimensional signal map through an incremental data fusion algorithm to update the voxelized three-dimensional signal map.

[0079] The physical attribute layer stores physical environment characteristic parameters corresponding to voxel units, including core parameters such as material type, topological connectivity, and obstacle density, providing data support for analyzing the impact of the physical environment on signals. The signal fingerprint layer stores signal characteristics and multi-dimensional signal characteristics of the corresponding 5G base station access point within the voxel unit, forming a unique signal identifier for each voxel unit, enabling rapid identification of the signal source and propagation quality in that area. The dynamic evolution layer stores historical signal data and signal change trends of voxel units, supporting the tracing of past signal states and prediction of future changes, providing support for subsequent time-series fusion to generate a spatiotemporal signal map. The incremental data fusion algorithm is an efficient map update algorithm that updates data only for voxel units corresponding to newly acquired real-time communication detection data, without requiring a full reconstruction of the entire map. Through data matching, conflict handling, and weight allocation, it fuses new data with existing map data, ensuring the accuracy and timeliness of the map.

[0080] Specifically, traditional signal mapping techniques suffer from core flaws such as single-dimensionality, inefficient updates, and lack of correlation with environmental and temporal characteristics. They are mostly two-dimensional planar models that only reflect the planar distribution of signal strength, ignoring crucial environmental factors like material properties and topological connectivity. For example, in an industrial park's metal equipment warehouse, a traditional two-dimensional map might show adequate signal strength, but in reality, severe attenuation could lead to drone communication interruptions. Furthermore, traditional maps employ a full-update model; even if the target space changes only locally (such as the addition of small obstructions), a full reconstruction is required, which is time-consuming and labor-intensive, making it unsuitable for dynamic engineering scenarios. Simultaneously, traditional maps lack a hierarchical design, failing to simultaneously record signal characteristics, environmental information, and historical changes, resulting in a lack of data support for subsequent temporal fusion. To address the above issues, this step first analyzes the real-time communication detection dataset, extracting the three-dimensional operational space coordinates (e.g., X=20m, Y=15m, Z=6m) and multi-dimensional signal characteristics (e.g., Sub-6GHz band signal strength 45dBm, transmission delay 10ms, signal-to-noise ratio 36dB) of each data point, and simultaneously associating it with physical environment information (e.g., the area's material is concrete, and its topological connectivity is moderate). Based on the target spatial boundary coordinates, the dataset is divided into voxel units of 1m×1m×1m size using a spatial meshing algorithm. An initial voxelized three-dimensional signal map is constructed for each voxel, and the physical attribute layer maps material, topological connectivity, and other information. The fingerprint layer is associated with 5G base station access point features (such as base station ID and operating frequency band 2.6GHz), and the dynamic evolution layer records historical signal data from 3 days ago (such as signal strength 42dBm). An incremental data fusion algorithm is started to locate the corresponding voxel based on the new data coordinates. If there is no environmental change, only the signal fingerprint layer is updated (such as updating the signal strength from 42dBm to 45dBm). If there is an environmental change (such as new obstruction), the physical attribute layer and signal features are updated simultaneously. Conflicting data are fused by weighting the detected data with a weight of 0.8 and the historical data with a weight of 0.2. Finally, a consistency check is completed to ensure that the data of each voxel layer is logically consistent, forming an updated voxelized three-dimensional signal map.

[0081] The method provided in this embodiment constructs a voxelized three-dimensional signal map containing a physical attribute layer, a signal fingerprint layer, and a dynamic evolution layer based on the analysis of real-time detection data. The incremental update only processes the changed areas, which not only achieves accurate correlation between signals and the environment, but also greatly improves update efficiency. It perfectly solves the pain points of traditional technologies and is a key foundation for generating spatiotemporal signal map sets and supporting UAV self-organizing network decision-making.

[0082] In some embodiments, based on the UAV swarm task and status dataset, the task intensity and real-time status information of each UAV node are analyzed to perform initial task allocation and generate an initial task information set; based on the spatiotemporal signal map set, a UAV relay network topology characterized by chain connections is constructed to extend the intelligent signal coverage range; based on the initial task information set and the UAV relay network topology, multimodal communication roles are assigned to each UAV node, including direct mode, relay mode and self-organizing network mode.

[0083] The initial task information set can be the process of dividing the overall 5G signal coverage task into several sub-tasks and assigning them to suitable drone nodes. The core objective is to match the task with the node's capabilities, avoiding node overload or resource waste. Chained connections can be the core connection method of the drone relay network topology. This refers to selecting relay nodes hop-by-hop from the root node to form linear or branching links. Each relay node communicates directly only with its adjacent preceding and following nodes, achieving long-distance signal transmission through data forwarding at each level. It features a simple structure and flexible expansion. The drone relay network topology can be a drone communication connection structure built based on the target space signal distribution characteristics to extend the 5G signal coverage. Its core feature is chained connections, using relay nodes to forward signals and supplement coverage in blind spots and weak coverage areas. Multimodal communication roles can be differentiated communication function assignments based on the drone node's location, signal environment, and task requirements. These roles include direct mode, relay mode, and self-organizing network mode, enabling nodes to adapt to different communication scenario requirements.

[0084] Specifically, traditional drone ad hoc networks suffer from core defects such as rudimentary task allocation, fixed topology, and singular communication roles: task allocation is often averaged or randomized, without considering task intensity and drone real-time status. For example, a construction site might assign a high-data-throughput (8GB per hour) and low-latency (≤30ms) core task to a drone with only 35% battery remaining, causing the task to be interrupted midway; the topology design is detached from signal distribution, and the fixed structure cannot cover blind spots, such as a corner of a park that remains in a signal blind spot due to the lack of a targeted relay link; and the communication roles are not differentiated, with drones with sufficient signal still serving as relays, resulting in wasted resources. Furthermore, collaborative task nodes suffer from significantly increased data transmission latency due to the lack of a dedicated ad hoc network mode. To address the above issues, this step analyzes the drone swarm task and status dataset, extracting task intensity indicators (e.g., coverage area XX°XX′-XX°XX′ E, XX°XX′-XX°XX′ N, data throughput requirement 10GB per hour, latency constraint ≤50ms) and drone real-time status (e.g., drone 1 has 88% remaining battery and 25% data load, drone 2 has 90% health). A dynamic weighted evaluation algorithm (data throughput weight 0.4, latency constraint 0.3, remaining battery weight 0.2, health 0.1) is used to calculate the suitability score, dividing the tasks into 5 subsets and assigning them to the corresponding drones, generating an initial task information set; based on spatiotemporal... The signal map set identifies blind spots (such as the west side of Workshop 3) and weak coverage areas. Taking UAV 1 with the best signal quality (signal strength 49dBm) as the root node, a chain path of "UAV 1 → UAV 3 → UAV 5" is planned through a path search algorithm to extend to the blind spot. The relay node can dynamically adjust the transmission power (e.g., from 22dBm to 26dBm). Combining the initial task and topology, UAV 1 with a signal strength of 49dBm (above the threshold of 40dBm) is assigned direct mode, UAVs 3 and 4 with weak coverage are assigned relay mode, and UAVs 5 and 6 performing collaborative tasks are assigned self-organizing network mode. A local mesh network is constructed, and a dynamic self-organizing network instruction set is generated.

[0085] The method provided in this embodiment accurately allocates tasks by analyzing task intensity and UAV status, constructs a chain relay topology based on spatiotemporal signal maps to fill blind spots, and assigns multimodal communication roles to adapt to different scenarios, perfectly solving the pain points of traditional technologies. It is the key to transforming signal data into executable coverage instructions and achieving intelligent and efficient coverage.

[0086] In some embodiments, the drone cluster task and status dataset is parsed to extract the spatial distribution of tasks, data throughput requirements, and latency constraints as task intensity indicators, and the remaining battery power, data load of its 5G IoT card, and health status of each drone node are extracted as real-time status information. Based on the task intensity indicators and real-time status information, a dynamic weighted evaluation algorithm is used to calculate the task suitability score for each drone node. According to the task suitability score, the tasks to be executed are divided into multiple task subsets, and the task subsets are assigned to the corresponding drone nodes with the goal of maximizing the overall task completion efficiency and network lifetime.

[0087] The task intensity index is a set of core parameters that quantifies the difficulty and resource consumption of 5G signal coverage tasks. It includes three key dimensions: spatial distribution, data throughput requirements, and latency constraints, used to determine the task's capability requirements for drone nodes. Real-time status information is a set of operational status parameters for each drone node at the current moment, including remaining battery power, 5G IoT card data load, and health status, directly determining the drone node's task carrying capacity. The dynamic weighted evaluation algorithm is the core algorithm for calculating the drone node's task suitability score. By assigning dynamic weights to the task intensity index and the drone's real-time status information, and combining the quantified values ​​of each parameter, it calculates the suitability of each drone node for each task. The weights can be dynamically adjusted according to the task type and engineering scenario. The task suitability score is a quantified value calculated by the dynamic weighted evaluation algorithm, used to characterize the degree of matching between a drone node and a specific task subset. A higher score indicates that the node's capabilities better meet the task requirements, resulting in higher task execution efficiency and lower failure risk.

[0088] Specifically, traditional drone initial task allocation suffers from core flaws such as coarse allocation logic, lack of quantitative matching, and single optimization objective. It often adopts an average or random allocation mode without considering the task intensity and the drone's real-time status. For example, a park assigned a core detection task with a data throughput requirement of 12GB per hour and a latency constraint of ≤30ms to a drone with only 30% battery remaining and a data load of 75%, resulting in the task failing due to power failure midway. At the same time, traditional solutions only pursue the rapid completion of tasks, ignoring network lifetime, and concentrate high-load tasks on a few nodes, causing these nodes to be overloaded and exit prematurely, resulting in poor overall cluster stability. To address the above issues, this step analyzes the drone swarm task and status dataset, extracting task intensity indicators (e.g., coverage area of ​​80,000 square meters, data throughput requirement of 10GB per node, core area latency ≤40ms) and drone real-time status (e.g., drone 1 has 85% remaining battery, 25% data load, and 95% health status; drone 2 has 55% remaining battery, 60% data load, and 88% health status). A dynamic weighted evaluation algorithm is used, setting data throughput weight at 0.4, latency constraint at 0.3, remaining battery at 0.2, and health status at 0.1, quantifying each parameter into a score of 0-100. The suitability is calculated using "task parameter score × corresponding weight + node status score × corresponding weight" (e.g., drone 1 has a suitability score of 92 for the core task). Based on the score, the task is divided into 5 subsets, assigning core high-load tasks to highly suited nodes and edge lightweight tasks to low-suited nodes to maximize overall task completion efficiency and network lifetime, generating an initial task information set.

[0089] The method provided in this embodiment quantifies the suitability by analyzing the task intensity index and the real-time status of the nodes, using a dynamic weighted evaluation algorithm, splitting task subsets and allocating them with dual objectives, perfectly solving the pain points of traditional technologies. This is the key to achieving accurate matching of tasks and nodes and ensuring the stable and efficient operation of the cluster.

[0090] In some embodiments, based on a spatiotemporal signal map set, signal coverage blind spots and weak coverage areas in the target space are identified; the node with the best current signal quality in the UAV swarm is taken as the root node of the chain network, and according to the signal propagation characteristics and spatial obstacle information provided by the spatiotemporal signal map set, a path search algorithm is used to select the next hop relay node hop by hop starting from the root node, forming one or more chain paths extending to the signal coverage blind spots and weak coverage areas; wherein, each relay node on the chain path is responsible for forwarding the data of its downstream nodes and dynamically adjusting its transmission power and antenna beam pointing to maintain the stability and energy efficiency of the link.

[0091] Signal coverage blind spots can be areas within the target space where 5G signals are completely unreachable or where signal strength is below the minimum communication threshold (e.g., ≤25dBm). Drones in these areas cannot directly communicate with ground base stations or other nodes, making them key areas for signal coverage improvement. Weak coverage areas can be areas within the target space where 5G signal strength is above the minimum communication threshold but below the stable communication threshold (e.g., 25dBm < signal strength ≤ 35dBm). Communication quality is poor in these areas (high latency, prone to interruptions), requiring relaying to enhance signal stability. The root node can be the starting node of a chain relay network topology, located in a strong signal coverage area. It can directly communicate with 5G ground base stations, undertaking the core functions of link initiation and data aggregation and forwarding, forming the foundation for chain path extension. The path search algorithm can be the core algorithm used to plan chain relay paths. Based on signal propagation characteristics and spatial obstacle information, it can start from the root node and hop-by-hop select the optimal relay node to form an unobstructed, low-attenuation chain path, ensuring efficient signal transmission.

[0092] Specifically, drone relay network topologies suffer from core defects such as fixed topologies, blind node selection, and rigid parameters. They often adopt fixed star or mesh structures without considering signal distribution and environmental characteristics. For example, in a factory coverage project, the signal of the central node in a star topology was blocked by a wall, resulting in a blind spot in the corner of the factory, preventing drones from communicating. Traditional relay nodes are often randomly selected without avoiding obstacles, leading to severe link signal attenuation. For instance, at a construction site, a relay node was blocked by a tower crane, causing the signal to drop from 45dBm to below 20dBm, resulting in communication interruption. Furthermore, the node's transmit power and antenna pointing are fixed, making it unable to cope with dynamic signal changes. To address the above issues, this step analyzes the spatiotemporal signal spectrum set, extracts voxel unit signal strength data, and identifies 25dBm as the minimum communication threshold and 35dBm as the stable communication threshold. This identifies the north side of Workshop 3 (signal strength ≤ 25dBm) as a blind zone and the middle section of Channel 2 (25dBm < signal strength ≤ 32dBm) as a weak coverage area. Simultaneously, it extracts signal propagation characteristics and spatial obstacle information, such as the 22dB attenuation coefficient of concrete walls and obstruction by large equipment. Signal strength data of 50dBm or less is also selected from the drone swarm. A node with a signal-to-noise ratio of 38dB and a signal-to-noise ratio of Bm is selected as the root node. The A* path search algorithm is used to select unobstructed and low-attenuation nodes hop by hop from the root node to plan a chain path of "root node → node B → node C" to extend to the blind zone. The relay node on the command path is initially set to transmit power of 22dBm, with the antenna pointing to the target node. The link quality is monitored in real time. When the signal strength drops to 30dBm, the transmit power is automatically increased to 25dBm, and the antenna beam direction is finely adjusted to maintain link stability and energy efficiency.

[0093] The method provided in this embodiment accurately identifies blind spots and weak coverage areas based on spatiotemporal signal map sets. With the optimal signal node as the root node, a chain path is planned through a path search algorithm, and the relay nodes are dynamically adjusted. This not only specifically fills the coverage gaps but also ensures link stability, perfectly solving the pain points of traditional technologies and is the key to achieving full-scene signal coverage.

[0094] In some embodiments, UAV nodes with good line-of-sight links to 5G ground base stations and signal strength higher than the direct communication threshold are assigned to direct mode, instructing them to communicate directly with the 5G ground base station; UAV nodes located in signal attenuation areas but with reliable links to other UAV nodes are assigned to relay mode, instructing them to forward 5G data from upstream or downstream nodes in a single-hop or multi-hop manner; and task sub-clusters performing collaborative computing and data-intensive exchange are assigned to self-organizing network mode, instructing them to build local mesh networks internally for 5G data sharing and processing, and elect one or more gateway nodes to connect to the chain network or 5G ground base station.

[0095] Direct mode allows drone nodes to establish direct communication connections with 5G ground base stations via line-of-sight links, eliminating the need for relay nodes and enabling direct data transmission and command reception. This mode is suitable for areas with strong signal coverage. Relay mode allows drone nodes located in areas with signal attenuation to receive data from upstream nodes (or base stations) via reliable links, amplify and forward it to downstream nodes, achieving signal extension and relay data transmission. Self-organizing network mode allows nodes within a task sub-cluster to autonomously build local mesh networks, enabling high-speed data sharing and collaborative processing, while simultaneously connecting to external chain networks or 5G ground base stations via gateway nodes.

[0096] Specifically, traditional drone communication solutions suffer from core defects such as a single communication mode and poor scenario adaptability. They only support unified direct communication or relay communication and do not allocate resources according to the actual environment of the nodes. For example, in a certain park, 10 drones with good line of sight to 5G ground base stations and signal strength of 45dBm (far exceeding the threshold for direct communication) were forcibly included in the relay network, resulting in wasted communication resources and a 30% increase in data latency. Some nodes located in signal attenuation areas (signal strength of 32dBm) were unable to communicate with the outside world due to the lack of relay mode support. To address the above issues, this step, based on the UAV relay network topology and initial task information set, first uses GPS positioning and spatiotemporal signal map to determine the line-of-sight link between nodes and the 5G ground base station. Node 1, with a signal strength of 48dBm (above the 40dBm direct communication threshold) and no obstructions, is assigned to direct mode and instructed to communicate directly with the base station. Nodes 2 and 3 are located in the signal attenuation area (signal strength 32dBm and 35dBm, respectively), but have reliable links with Node 1 and each other (signal strength 38dBm, latency 45ms, packet loss rate 0.5%), and are assigned relay mode, forming a multi-hop link of "Node 3 → Node 2 → Node 1 → Base Station". Nodes 4, 5, and 6, which perform collaborative modeling tasks, are divided into task sub-clusters, assigned self-organizing network mode, and instructed to build a local mesh network with high-speed internal exchange of 6GB of detection data. Node 4, with a signal strength of 36dBm and 85% remaining battery power, is elected as the gateway node to access the chain network and communicate with the outside world.

[0097] The method provided in this embodiment designs three multimodal communication roles: direct, relay, and self-organizing network. This allows nodes with strong signals to communicate directly, nodes in the attenuation zone to relay and forward signals, and collaborative task nodes to form an internal network. This perfectly solves the pain points of traditional technologies and is the key to achieving accurate communication and ensuring coverage efficiency in different scenarios.

[0098] In some embodiments, based on the execution feedback data of the UAV swarm dynamic self-organizing network instruction set and the spatiotemporal signal map set, a multi-dimensional performance evaluation algorithm is used to analyze and aggregate quantitative evaluation results that include real-time link performance indicators, signal coverage integrity, and network topology robustness. The quantitative evaluation results are then compared and analyzed with historical performance data in a spatiotemporal manner to generate a signal intelligent coverage performance report that includes network resilience rating and adaptive adjustment suggestions, and output to the engineering management terminal in the form of a visual map.

[0099] A multidimensional performance evaluation algorithm is a core algorithm that integrates multiple evaluation indicators to comprehensively quantify the effect of intelligent signal coverage. By assigning scientific weights to different indicators, it calculates a comprehensive evaluation result reflecting coverage quality, link performance, and topology stability, avoiding the one-sidedness of single-indicator evaluation. The quantitative evaluation result can be a comprehensive quantitative data set formed by weighting and integrating real-time link performance indicators, signal coverage integrity, and network topology robustness through a multidimensional performance evaluation algorithm.

[0100] Specifically, traditional signal coverage assessments suffer from core flaws such as single indicators, lack of quantification, absence of historical comparisons, vague recommendations, and unintuitive outputs. They often focus solely on signal strength, neglecting link performance and topology robustness. For example, a traditional assessment might state "good signal coverage" for a project, but the actual link latency reaches 200ms, causing frequent interruptions in drone collaborative missions. Traditional assessments are mostly qualitative descriptions, failing to specify the exact location and proportion of weak coverage areas, requiring engineers to re-detect and locate them. Furthermore, without comparison with historical data, they cannot identify the downward trend in coverage integrity from 90% to 80%. Recommendations are limited to "optimizing network configuration," lacking practicality. The output tables are cumbersome, making it difficult to quickly identify core issues. To address the above issues, this step collects dynamic self-organizing network command execution feedback data from the UAV swarm (such as link latency of 28ms, packet loss rate of 0.3%, and coverage integrity of 85%) and spatiotemporal signal maps. A multi-dimensional performance evaluation algorithm is used, with weights of 0.4 for signal coverage integrity, 0.3 for real-time link performance, and 0.3 for network topology robustness. The data is standardized to a score of 0-100, resulting in a comprehensive performance score of 87.4. A spatiotemporal comparison with historical data (comprehensive score of 76.8 three months ago) generates a "good" network resilience rating and an adaptive adjustment suggestion to "add one relay UAV at the end of channel 2." These are output to the engineering management terminal in the form of a signal coverage heatmap and a link performance line graph.

[0101] The method provided in this embodiment accurately presents coverage performance through multi-dimensional quantitative evaluation, spatiotemporal comparative analysis, targeted suggestions, and visualization output, forming a closed loop of "evaluation-optimization-re-evaluation," perfectly solving traditional pain points and being the key to ensuring continuous improvement in coverage quality.

[0102] Figure 3 This is a schematic diagram of the structure of a multimodal 5G IoT card signal intelligent coverage system provided in an embodiment of this application, as shown below. Figure 3 As shown, a multimodal 5G IoT card signal intelligent coverage system 300 in this embodiment includes: a spatiotemporal signal map module 301, a dynamic networking instruction module 302, and a signal performance report module 303.

[0103] The spatiotemporal signal map module 301 is used to acquire the project progress information set, and based on the project progress information set, to perform multi-dimensional collaborative channel perception and spatiotemporal correlation analysis to generate a spatiotemporal signal map set that dynamically evolves with the physical space structure.

[0104] The dynamic networking instruction module 302 is used to acquire the UAV cluster task and status dataset and, based on the UAV cluster task and status dataset and the spatiotemporal signal spectrum set, to construct task-driven distributed collaborative decision-making, dynamically allocate communication roles and optimize multimodal links to generate a UAV cluster dynamic self-organizing network instruction set.

[0105] The signal performance report module 303 is used to perform network resilience enhancement and adaptive optimization processing based on the UAV cluster dynamic self-organizing network instruction set and its execution feedback data, and generate and output a signal intelligent coverage performance report.

[0106] Optionally, when generating the spatiotemporal signal map module 301 based on the spatiotemporal signal map set that dynamically evolves with the physical spatial structure, it is specifically used for:

[0107] Based on the project progress information set, a 5G signal detection task is dynamically triggered and executed to obtain a real-time communication detection dataset of the target space.

[0108] Based on the real-time communication detection dataset, a voxelized three-dimensional signal map of the target space is constructed and updated, wherein each voxel unit is associated with multi-dimensional signal attributes and physical environment attributes.

[0109] A temporal fusion algorithm is used to perform correlation analysis on the voxelized three-dimensional signal map of each update to generate the spatiotemporal signal map set that characterizes the spatial distribution and evolution of the signal.

[0110] Optionally, when the spatiotemporal signal mapping module 301 performs the signal detection task based on the dynamic trigger, it is specifically used for:

[0111] Based on the project progress information set, the key change nodes and periodic performance degradation warnings of the project's physical spatial structure are analyzed, and active signal detection task instructions are generated.

[0112] The key change nodes include infrastructure construction milestones and completion events for updating the layout of operational scenarios;

[0113] Based on the active signal detection mission command, a detection formation consisting of a lead drone and a follower drone is dispatched into the target space:

[0114] The detection formation is controlled to perform gridded active detection to collect multi-band 5G channel depth information of preset sampling points, and simultaneously performs broad-spectrum passive listening to map the distribution information of spatial electromagnetic interference sources, which together constitute the real-time communication detection dataset.

[0115] Optionally, when constructing and updating the voxelized three-dimensional signal map of the target space, the spatiotemporal signal map module 301 is specifically used for:

[0116] The real-time communication detection dataset is analyzed to extract the operational space coordinates and multi-dimensional signal features of each data point;

[0117] Based on the workspace coordinates, the target space is divided into a set of regular voxel units, and an initial voxelized 3D signal map is established for each voxel unit:

[0118] The physical attribute layer is composed of material and topological connectivity information inferred from the physical environment information of the data points; the signal fingerprint layer is composed of signal characteristics of the corresponding 5G base station access point; and the dynamic evolution layer is composed of signal change trends recorded based on historical signal data.

[0119] The newly acquired real-time communication detection dataset is fused and analyzed with the corresponding voxel units in the voxelized three-dimensional signal map using an incremental data fusion algorithm, so as to update the voxelized three-dimensional signal map.

[0120] Optionally, when generating the UAV swarm dynamic self-organizing network instruction set, the dynamic networking instruction module 302 is specifically used for:

[0121] Based on the aforementioned drone cluster task and status dataset, the task intensity and real-time status information of each drone node are analyzed to perform initial task allocation and generate an initial task information set.

[0122] Based on the spatiotemporal signal map set, a UAV relay network topology characterized by chain connections is constructed to extend the intelligent signal coverage range.

[0123] Based on the initial task information set and the UAV relay network topology, a multimodal communication role is assigned to each UAV node. The multimodal communication role includes direct mode, relay mode and self-organizing network mode.

[0124] Optionally, when the dynamic networking instruction module 302 performs initial task allocation based on the analyzed task intensity and the real-time status information of each UAV node, it is specifically used for:

[0125] The task and status dataset of the drone cluster is analyzed, and the spatial distribution of tasks, data throughput requirements and latency constraints are extracted as task intensity indicators. The remaining power of each drone node, the data load of its 5G IoT card and health status are extracted as real-time status information.

[0126] Based on the task intensity index and the real-time status information, a dynamic weighted evaluation algorithm is used to calculate the task suitability score for each UAV node.

[0127] Based on the task suitability score, the tasks to be executed are divided into multiple task subsets, and with the goal of maximizing the overall task completion efficiency and network lifetime, the task subsets are assigned to the corresponding drone nodes.

[0128] Optionally, when constructing a UAV relay network topology characterized by chain connections, the dynamic networking instruction module 302 is specifically used for:

[0129] Based on the spatiotemporal signal map set, identify signal coverage blind spots and weak coverage areas in the target space;

[0130] The node with the best current signal quality in the UAV cluster is taken as the root node of the chain network. Based on the signal propagation characteristics and spatial obstacle information provided by the spatiotemporal signal map set, a path search algorithm is used to select the next hop relay node hop by hop starting from the root node, forming one or more chain paths extending to the signal coverage blind area and weak coverage area.

[0131] Each relay node on the chain path is responsible for forwarding data from its downstream nodes and dynamically adjusting its transmit power and antenna beam direction to maintain the stability and energy efficiency of the link.

[0132] Optionally, when assigning multimodal communication roles to each UAV node, the dynamic networking instruction module 302 is specifically used for:

[0133] For drone nodes with a good line-of-sight link with the 5G ground base station and a signal strength higher than the direct communication threshold, they are assigned to the direct mode and instructed to communicate directly with the 5G ground base station.

[0134] For drone nodes located in signal attenuation areas but with reliable links to other drone nodes, they are assigned to the relay mode and instructed to forward 5G data from upstream or downstream nodes in a single-hop or multi-hop manner.

[0135] For task sub-clusters that perform collaborative computing and data-intensive exchange, they are assigned to the self-organizing network mode, and are instructed to build a local mesh network inside them for 5G data sharing and processing, and elect one or more gateway nodes to connect to the chain network or 5G ground base station.

[0136] Optionally, the signal performance reporting module 303 is specifically used for:

[0137] Based on the execution feedback data of the UAV swarm dynamic self-organizing network instruction set and the spatiotemporal signal map set, a multi-dimensional performance evaluation algorithm is used to analyze and aggregate quantitative evaluation results that include real-time link performance indicators, signal coverage integrity and network topology robustness.

[0138] The quantitative evaluation results are compared and analyzed spatiotemporally with historical performance data to generate a signal intelligent coverage performance report that includes network resilience rating and adaptive adjustment suggestions. The report is then output to the engineering management terminal in the form of a visual map.

[0139] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A method for intelligent coverage of multimodal 5G IoT card signals, characterized in that, include: Obtain a set of project progress information, and based on the set of project progress information, perform multi-dimensional collaborative channel perception and spatiotemporal correlation analysis to generate a set of spatiotemporal signal maps that dynamically evolve with the physical space structure. A dataset of drone cluster tasks and statuses is obtained. Based on the dataset and the spatiotemporal signal map set, a task-driven distributed collaborative decision-making system is constructed to dynamically allocate communication roles and optimize multimodal links, thereby generating a dynamic self-organizing network instruction set for the drone cluster. Based on the command set of the UAV swarm dynamic self-organizing network and its execution feedback data, network resilience enhancement and adaptive optimization are performed to generate and output a signal intelligent coverage performance report.

2. The method according to claim 1, characterized in that, The generation of the spatiotemporal signal spectrum set that dynamically evolves with the physical spatial structure includes: Based on the project progress information set, a 5G signal detection task is dynamically triggered and executed to obtain a real-time communication detection dataset of the target space. Based on the real-time communication detection dataset, a voxelized three-dimensional signal map of the target space is constructed and updated, wherein each voxel unit is associated with multi-dimensional signal attributes and physical environment attributes. A temporal fusion algorithm is used to perform correlation analysis on the voxelized three-dimensional signal map of each update to generate the spatiotemporal signal map set that characterizes the spatial distribution and evolution of the signal.

3. The method according to claim 2, characterized in that, The dynamic triggering and execution of the signal detection task includes: Based on the project progress information set, the key change nodes and periodic performance degradation warnings of the project's physical spatial structure are analyzed, and active signal detection task instructions are generated. The key change nodes include infrastructure construction milestones and completion events for updating the layout of operational scenarios; Based on the active signal detection mission command, a detection formation consisting of a lead drone and a follower drone is dispatched into the target space: The detection formation is controlled to perform gridded active detection to collect multi-band 5G channel depth information of preset sampling points, and simultaneously performs broad-spectrum passive listening to map the distribution information of spatial electromagnetic interference sources, which together constitute the real-time communication detection dataset.

4. The method according to claim 3, characterized in that, The construction and updating of the voxelized three-dimensional signal map of the target space includes: The real-time communication detection dataset is analyzed to extract the operational space coordinates and multi-dimensional signal features of each data point; Based on the workspace coordinates, the target space is divided into a set of regular voxel units, and an initial voxelized 3D signal map is established for each voxel unit: The physical attribute layer is composed of material and topological connectivity information inferred from the physical environment information of the data points; the signal fingerprint layer is composed of signal characteristics of the corresponding 5G base station access point; and the dynamic evolution layer is composed of signal change trends recorded based on historical signal data. The newly acquired real-time communication detection dataset is fused and analyzed with the corresponding voxel units in the voxelized three-dimensional signal map using an incremental data fusion algorithm, so as to update the voxelized three-dimensional signal map.

5. The method according to claim 4, characterized in that, The instruction set for generating a dynamic self-organizing network of unmanned aerial vehicle (UAV) swarms includes: Based on the aforementioned drone cluster task and status dataset, the task intensity and real-time status information of each drone node are analyzed to perform initial task allocation and generate an initial task information set. Based on the spatiotemporal signal map set, a UAV relay network topology characterized by chain connections is constructed to extend the intelligent signal coverage range. Based on the initial task information set and the UAV relay network topology, a multimodal communication role is assigned to each UAV node. The multimodal communication role includes direct mode, relay mode and self-organizing network mode.

6. The method according to claim 5, characterized in that, The analysis of task intensity and real-time status information of each UAV node is used to perform initial task allocation, including: The task and status dataset of the drone cluster is analyzed, and the spatial distribution of tasks, data throughput requirements and latency constraints are extracted as task intensity indicators. The remaining power of each drone node, the data load of its 5G IoT card and health status are extracted as real-time status information. Based on the task intensity index and the real-time status information, a dynamic weighted evaluation algorithm is used to calculate the task suitability score for each UAV node. Based on the task suitability score, the tasks to be executed are divided into multiple task subsets, and with the goal of maximizing the overall task completion efficiency and network lifetime, the task subsets are assigned to the corresponding drone nodes.

7. The method according to claim 5, characterized in that, The construction of the UAV relay network topology characterized by chain connections includes: Based on the spatiotemporal signal map set, identify signal coverage blind spots and weak coverage areas in the target space; The node with the best current signal quality in the UAV cluster is taken as the root node of the chain network. Based on the signal propagation characteristics and spatial obstacle information provided by the spatiotemporal signal map set, a path search algorithm is used to select the next hop relay node hop by hop starting from the root node, forming one or more chain paths extending to the signal coverage blind area and weak coverage area. Each relay node on the chain path is responsible for forwarding data from its downstream nodes and dynamically adjusting its transmit power and antenna beam direction to maintain the stability and energy efficiency of the link.

8. The method according to claim 7, characterized in that, The assignment of multimodal communication roles to each UAV node includes: For drone nodes with a good line-of-sight link with the 5G ground base station and a signal strength higher than the direct communication threshold, they are assigned to the direct mode and instructed to communicate directly with the 5G ground base station. For drone nodes located in signal attenuation areas but with reliable links to other drone nodes, they are assigned to the relay mode and instructed to forward 5G data from upstream or downstream nodes in a single-hop or multi-hop manner. For task sub-clusters that perform collaborative computing and data-intensive exchange, they are assigned to the self-organizing network mode, and are instructed to build a local mesh network inside them for 5G data sharing and processing, and elect one or more gateway nodes to connect to the chain network or 5G ground base station.

9. The method according to claim 8, characterized in that, The generated and output signal intelligent coverage performance report includes: Based on the execution feedback data of the UAV swarm dynamic self-organizing network instruction set and the spatiotemporal signal map set, a multi-dimensional performance evaluation algorithm is used to analyze and aggregate quantitative evaluation results that include real-time link performance indicators, signal coverage integrity and network topology robustness. The quantitative evaluation results are compared and analyzed spatiotemporally with historical performance data to generate a signal intelligent coverage performance report that includes network resilience rating and adaptive adjustment suggestions. The report is then output to the engineering management terminal in the form of a visual map.

10. A multimodal 5G IoT card signal intelligent coverage system, characterized in that, The method applied to any one of claims 1-9 includes: The spatiotemporal signal map module is used to acquire the project progress information set, and based on the project progress information set, to perform multi-dimensional collaborative channel perception and spatiotemporal correlation analysis to generate a spatiotemporal signal map set that dynamically evolves with the physical space structure. The dynamic networking instruction module is used to acquire the UAV cluster task and status dataset and, based on the UAV cluster task and status dataset and the spatiotemporal signal spectrum set, to generate a dynamic self-organizing network instruction set for the UAV cluster by constructing task-driven distributed collaborative decision-making, dynamically allocating communication roles and optimizing multimodal links. The signal performance report module is used to perform network resilience enhancement and adaptive optimization processing based on the UAV cluster dynamic self-organizing network instruction set and its execution feedback data, and generate and output a signal intelligent coverage performance report.

Images