Communication self-organizing network method and system applied to infrastructure without signal site

Abstract

The present invention discloses a communication ad hoc network method and system for use in infrastructure construction sites with no signal, belonging to the field of communication technology. The method includes: determining the spatial distribution of node devices to obtain each initial planned path; creating a four-dimensional spectrum map and dynamically adjusting the modulation mode; determining the quality of each initial planned path; establishing service priorities; determining whether energy replenishment is required; establishing a communication-energy consumption optimization algorithm to obtain a network optimization solution; constructing a node device trust model to determine the optimal transmission path; constructing a node fault prediction model to predict node device failures and trigger maintenance instructions. The present invention can quickly build a stable communication network, reduce costs, improve communication reliability and security, optimize energy consumption management, extend the battery life of node devices, ensure the continuity of key services, adapt to dynamic construction environments, and realize intelligent and automated communication management.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a communication self-organizing network method and system applied to infrastructure without signal sites. Background Art

[0002] Infrastructure construction sites with no signal are often located in mountainous areas, tunnels, underground projects, offshore platforms, and other areas far from cities. Traditional communication networks (such as 4G / 5G and fiber optics) in these areas have no coverage or extremely weak signals. These scenarios typically have the following characteristics: A lack of public network coverage prevents carrier base station signals from reaching them, rendering conventional communication devices like mobile phones and walkie-talkies inoperable. The environments are complex, with terrain obstacles (such as tunnels and forests), electromagnetic interference (such as high-voltage equipment), and extreme weather (such as heavy rain and cold weather) further hindering communication. Furthermore, the highly dynamic nature of construction workers, vehicles, and equipment requires a flexible and adaptable communication solution, as their locations frequently change.

[0003] Ad hoc networks are wireless network technologies that autonomously form networks without fixed infrastructure. Their core features include decentralization, self-organization and self-healing, and multi-hop transmission. Typical technologies include Wi-Fi Mesh, ZigBee, and LoRa, as well as interference-resistant protocols designed specifically for industrial applications. However, existing ad hoc network technologies have numerous shortcomings. Some solutions rely on satellite communications, resulting in high costs and significant latency. Furthermore, network security is insufficient, making it difficult to effectively defend against external attacks and data leaks. These shortcomings make existing technologies unable to meet the demand for real-time, low-cost, and highly reliable communications in infrastructure construction sites with limited signal availability. Summary of the Invention

[0004] The present invention provides a communication self-organizing network method and system for application in infrastructure construction sites with no signal, which can quickly build a stable communication network, reduce costs, improve communication reliability and security, optimize energy consumption management, extend the battery life of node equipment, ensure the continuity of key services, adapt to dynamic construction environments, and realize intelligent and automated communication management.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] A communication ad hoc network method applied to infrastructure construction sites without signal comprises the following steps:

[0007] Based on multiple node devices deployed in the construction area, a three-dimensional topological structure of the construction area is constructed. Combined with the lidar scanning carried by the drone, a digital elevation model of the construction area is generated, the spatial distribution of the node devices is determined, and the initial planning paths are obtained;

[0008] Create a four-dimensional spectrum map based on the spatial distribution of node devices and dynamically adjust the modulation method;

[0009] Based on the spatial distribution of node devices and the four-dimensional spectrum map, a path quality prediction algorithm is established to determine the quality of each initially planned path;

[0010] Establish service priorities based on the quality of each initially planned path;

[0011] Obtain the remaining power of each node device and determine whether energy replenishment is needed. If necessary, energy replenishment is performed according to the corresponding service priority;

[0012] Based on the quality of each initially planned path, service priority, and energy replenishment results, combined with the remaining power of each node device, a communication-energy consumption optimization algorithm is established to obtain a network optimization solution and perform network optimization.

[0013] Based on the optimized node devices in the network, a node device trust model is constructed to determine the optimal transmission path;

[0014] Obtain historical fault operation data of each node device, build a node fault prediction model, predict node device failures, and trigger maintenance instructions.

[0015] Preferably, the process of determining the spatial distribution of node devices is as follows:

[0016] Node devices include master node devices, backbone node devices, relay node devices, sensor node devices, mobile node devices and energy node devices;

[0017] Obtain the motion status data of the node device, including acceleration, angular velocity, and magnetic field strength;

[0018] Obtain the relative distance between each node device, combine the motion status data of the node device to obtain the three-dimensional coordinates of each node device, and construct the three-dimensional topological structure of the construction area;

[0019] Obtain the size and shape of the construction area from the data repository, plan the drone's flight path, and use the drone's onboard lidar equipment to emit laser pulses and measure the time difference between reflected lasers to determine the distance to the target object.

[0020] The data collected by the LiDAR equipment is stored in the form of a point cloud. Each point contains three-dimensional coordinates and reflection intensity information. The collected point cloud data is imported into the geographic information system software for preprocessing. The point cloud data is converted into regular grid data through interpolation algorithms to generate a digital elevation model.

[0021] The digital elevation model represents the elevation information of the terrain in the form of a two-dimensional grid, where each grid cell corresponds to an elevation value;

[0022] The generated digital elevation model is preloaded into the storage unit of each node device in the three-dimensional topological structure of the construction area. The node device accesses the digital elevation model to obtain the terrain information of the construction area, performs terrain analysis, and obtains each initial planning path.

[0023] Preferably, a four-dimensional spectrum map is created to dynamically adjust the modulation mode. The process is as follows:

[0024] The wide frequency band is divided into multiple sub-bands. The RF front-end of the node device scans each sub-band in sequence within 5 seconds to obtain the signal strength, interference strength, and center frequency within the sub-band;

[0025] The center frequency, the timestamp of the scan, the signal strength and the interference strength, and the three-dimensional coordinates of the node device are stored in the local storage unit of the node device to obtain a four-dimensional spectrum map;

[0026] Based on the signal strength and interference strength in the four-dimensional spectrum map, the channel quality of the sub-band is evaluated to obtain a channel quality evaluation result of the sub-band;

[0027] Comparing the channel quality assessment result of the sub-frequency band with the channel quality assessment results of each reference sub-frequency band stored in the data repository, the modulation mode corresponding to the channel quality assessment result of the same reference sub-frequency band is the modulation mode corresponding to the channel quality assessment result of the sub-frequency band;

[0028] The channel quality assessment result of the sub-band is obtained as follows:

[0029] Compare the signal strength with the signal evaluation threshold stored in the data repository. If the signal strength is less than the corresponding evaluation threshold, the channel quality level of the corresponding sub-frequency band is level 1, and the channel quality of the corresponding sub-frequency band is poor.

[0030] If the signal strength is not less than the signal evaluation threshold, the interference strength is compared with the interference evaluation threshold stored in the data repository. If the interference strength is less than the interference evaluation threshold, the channel quality level of the corresponding sub-frequency band is level 1, and the channel quality of the corresponding sub-frequency band is good.

[0031] If the interference intensity is not less than the interference assessment threshold, the channel quality level of the corresponding sub-frequency band is level 2, and the channel quality of the corresponding sub-frequency band is medium.

[0032] Preferably, a path quality prediction algorithm is established to determine the quality of each path. The process is as follows:

[0033] Divide the initial planned path into main path, backup path, and emergency path;

[0034] The master node device obtains the path quality evaluation data of each initially planned path, including signal strength, bit error rate, signal-to-noise ratio, delay time, and packet loss rate, and performs weighted summation on the path quality evaluation data of each initially planned path to obtain a comprehensive quality score for each initially planned path;

[0035] The comprehensive quality score of each initial planned path is compared with the comprehensive quality score threshold of the corresponding reference path stored in the data repository. If the comprehensive quality score of the initial planned path is greater than the comprehensive quality score threshold of the corresponding reference path, the path quality of the initial planned path is high; otherwise, the path quality of the path is low;

[0036] If the quality of the primary path is low, the network switches to the backup path with higher quality. Otherwise, the network does not switch.

[0037] If the quality of the primary path and the backup paths are both low, the emergency path with high quality is switched to.

[0038] If the path quality of the primary path, each backup path, and each emergency path is low, an alarm is issued.

[0039] Preferably, establish business priorities, the process is as follows:

[0040] Divide business priorities into the following categories: lowest priority 4, low priority 3, medium priority 2, high priority 1, and highest priority 0;

[0041] Among them, the lowest priority 4 corresponds to non-real-time data transmission, the low priority 3 corresponds to equipment status monitoring, the medium priority 2 corresponds to video monitoring, the high priority 1 corresponds to hazardous gas alarm, and the highest priority 0 corresponds to emergency braking instructions;

[0042] Dedicated frequency bands, dedicated broadband, and dedicated paths are allocated for emergency braking commands. The comprehensive quality scores corresponding to the high-quality paths are sorted from large to small. The path with the largest comprehensive quality score is selected as the dedicated path for emergency braking commands. When the emergency braking command needs to be transmitted, it is transmitted immediately.

[0043] Preferably, the process of determining whether energy replenishment is needed is as follows:

[0044] The energy consumption values ​​of each node device under different environmental conditions were measured in a controlled laboratory environment. The energy consumption values ​​and historical environmental data were divided into a training set and a validation set. The training set was input into the neural network model to establish an energy consumption model, and the validation set was used to verify the energy consumption model.

[0045] Obtain the current energy consumption value and current environmental data of each node device, input them into the trained energy consumption model, and obtain the predicted energy consumption value of each node device;

[0046] The current remaining power of each node device is obtained and subtracted from the corresponding predicted energy consumption value to obtain the comprehensive power of each node device. The comprehensive power of each node device is compared with the power threshold stored in the data storage library in turn to determine whether the comprehensive power of each node device is less than the power threshold. If it is less than, the node device triggers the energy replenishment process and performs energy replenishment. If it is not less than, it is not triggered.

[0047] Preferably, a communication-energy consumption optimization algorithm is established to obtain a network optimization solution. The process is as follows:

[0048] Obtain the energy transmission efficiency of the node device, divide the energy consumption value of the node device by the energy transmission efficiency to obtain the total energy consumption, obtain communication data, including delay time, reliability, and data transmission rate, and perform weighted summation of the communication data to obtain a communication performance score;

[0049] Define the objective function based on the total energy consumption and communication performance score;

[0050] Determine the constraints, including delay constraints, data rate constraints, reliability constraints, transmission power constraints, battery capacity constraints, and energy transmission efficiency constraints:

[0051] Delay constraint: the delay time of data transmission shall not exceed the maximum allowed delay time;

[0052] Data rate constraint: the data transmission rate must not be less than the minimum data transmission rate;

[0053] Reliability constraint: the reliability of data transmission shall not be less than the minimum reliability;

[0054] Transmission power constraint: the transmission power does not exceed the maximum transmission power;

[0055] Battery capacity constraint: the remaining power before energy replenishment is performed must not be less than the minimum remaining power;

[0056] Energy transmission efficiency constraint: the energy transmission efficiency is not less than the minimum efficiency;

[0057] The objective function to be minimized is determined based on the genetic algorithm. The optimal solution includes the optimal delay time, optimal data rate, optimal reliability, optimal transmission power, optimal remaining power, and optimal energy transmission efficiency, which is recorded as the network adjustment scheme.

[0058] Preferably, a node device trust model is constructed to determine the optimal transmission path. The process is as follows:

[0059] Obtain historical performance data of the node device, including data transmission success rate, historical failure rate, and energy efficiency. Compare the obtained data transmission success rate, historical failure rate, and energy efficiency with the data transmission success rate threshold, historical failure rate threshold, and energy efficiency threshold stored in the data repository, and perform a weighted sum to obtain an initial trust level.

[0060] The initial trust degree is compared with each reference trust degree stored in the data repository, and the trust level corresponding to the closest reference trust degree is the trust level of the initial trust degree;

[0061] Based on the trust levels of the node devices, the initial planned path corresponding to the node device with the highest trust level is selected as the optimal transmission path.

[0062] Preferably, a node failure prediction model is constructed to predict node equipment failures and trigger maintenance instructions. The process is as follows:

[0063] Perform feature extraction on historical fault operation data to obtain historical fault operation feature data;

[0064] Design the LSTM network architecture, including the input layer, LSTM layer, fully connected layer, and output layer; set network hyperparameters, including the number of LSTM units, number of layers, learning rate, batch size, activation function, and optimizer;

[0065] The historical node equipment failure feature data is divided into a training set, a validation set, and a test set. The training set is used to train the LSTM neural network prediction model to obtain a node failure prediction model.

[0066] Use the validation set to verify the node failure prediction model, and use the test set to test the node failure prediction model;

[0067] The operating data of the node device is obtained and input into the trained node fault prediction model to obtain the fault prediction result of the node device. If the fault prediction result is that the node device has a fault, a maintenance instruction is triggered.

[0068] A communication ad hoc network system applied to infrastructure construction sites without signal, used to implement the above method, includes:

[0069] The initial planning path establishment module is used to construct the three-dimensional topology of the construction area based on multiple node devices deployed in the construction area. It combines the laser radar scanning carried by the drone to generate a digital elevation model of the construction area, determine the spatial distribution of node devices, and obtain each initial planning path;

[0070] The spectrum management module is used to create a four-dimensional spectrum map based on the spatial distribution of node devices and dynamically adjust the modulation method;

[0071] The path quality assessment module is used to establish a path quality prediction algorithm based on the spatial distribution of node devices and the four-dimensional spectrum map to determine the quality of each initially planned path;

[0072] A service priority management module is used to establish service priorities based on the quality of each initially planned path;

[0073] The energy management module is used to obtain the remaining power of each node device and determine whether energy replenishment is needed. If necessary, energy replenishment is performed according to the corresponding service priority;

[0074] The communication-energy consumption optimization module is used to establish a communication-energy consumption optimization algorithm based on the quality of each initially planned path, service priority, and energy replenishment results, combined with the remaining power of each node device, to obtain a network optimization plan and perform network optimization;

[0075] The path selection module is used to build a node device trust model based on each node device after network optimization and determine the best transmission path;

[0076] The fault prediction and maintenance module is used to obtain the historical fault operation data of each node device, build a node fault prediction model, predict node device failures, and trigger maintenance instructions.

[0077] The present invention has the following beneficial effects:

[0078] This invention leverages dynamic ad hoc networking technology to rapidly establish a stable and reliable communications network within construction sites, reducing costs while significantly improving communication reliability and security. It also optimizes energy management, extends node device battery life, and enhances user experience. Furthermore, it adapts to dynamically changing construction environments, ensuring the continuity of critical services and overall network performance. This provides an intelligent, automated communications solution for infrastructure construction sites experiencing signal deprivation, significantly improving the network's economic benefits and user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 is a flow chart of the method of the present invention;

[0080] Figure 2 Schematic diagram of the modules of the system of the present invention. DETAILED DESCRIPTION

[0081] The technical solutions in the embodiments of the present invention are described clearly and completely below with reference to the accompanying drawings.

[0082] Example 1: Figure 1 As shown, the communication self-organizing network method applied to infrastructure without signal site includes the following steps:

[0083] Based on multiple node devices deployed in the construction area, a three-dimensional topological structure of the construction area is constructed. Combined with the lidar scanning carried by the drone, a digital elevation model of the construction area is generated, the spatial distribution of the node devices is determined, and the initial planning paths are obtained;

[0084] Create a four-dimensional spectrum map based on the spatial distribution of node devices and dynamically adjust the modulation method;

[0085] Based on the spatial distribution of node devices and the four-dimensional spectrum map, a path quality prediction algorithm is established to determine the quality of each initially planned path;

[0086] Establish service priorities based on the quality of each initially planned path;

[0087] Obtain the remaining power of each node device and determine whether energy replenishment is needed. If necessary, energy replenishment is performed according to the corresponding service priority;

[0088] Based on the quality of each initially planned path, service priority, and energy replenishment results, combined with the remaining power of each node device, a communication-energy consumption optimization algorithm is established to obtain a network optimization solution and perform network optimization.

[0089] Based on the node devices after network optimization, a node device trust model is constructed to determine the optimal transmission path;

[0090] Obtain historical fault operation data of each node device, build a node fault prediction model, predict node device failures, and trigger maintenance instructions.

[0091] Determine the spatial distribution of node devices. The process is as follows:

[0092] Node devices include master node devices, backbone node devices, relay node devices, sensor node devices, mobile node devices and energy node devices;

[0093] Obtain the motion state data of the node device, including acceleration, angular velocity, and magnetic field strength;

[0094] Obtain the relative distance between each node device, combine the motion status data of the node device to obtain the three-dimensional coordinates of each node device, and construct the three-dimensional topological structure of the construction area;

[0095] Obtain the size and shape of the construction area from the data repository, plan the drone's flight path, and use the drone's onboard lidar equipment to emit laser pulses and measure the time difference between reflected lasers to determine the distance to the target object.

[0096] The data collected by the LiDAR equipment is stored in the form of a point cloud. Each point contains three-dimensional coordinates and reflection intensity information. The collected point cloud data is imported into the geographic information system software for preprocessing. The point cloud data is converted into regular grid data through interpolation algorithms to generate a digital elevation model.

[0097] The digital elevation model represents the elevation information of the terrain in the form of a two-dimensional grid, where each grid cell corresponds to an elevation value;

[0098] The generated digital elevation model is preloaded into the storage unit of each node device in the three-dimensional topological structure of the construction area. The node device accesses the digital elevation model to obtain the terrain information of the construction area, performs terrain analysis, and obtains each initial planning path.

[0099] The acceleration and angular velocity of node devices, as well as the magnetic field strength, are acquired to provide dynamic position information of the node devices in three-dimensional space, including whether the node devices are moving, as well as the direction and speed of movement. The distance between node devices is measured, providing accurate spatial relationship data and constructing a three-dimensional model reflecting the actual location of the node devices. Drones equipped with lidar perform flight scanning. The lidar emits laser pulses and receives reflected pulses, calculating the time difference to determine the distance to the target object. This can quickly cover large areas and generate high-precision three-dimensional point cloud data. The point cloud data is processed by GIS software and converted into regular grid data to form a digital elevation model. The digital elevation model represents terrain elevation information in the form of a two-dimensional grid, with each grid cell corresponding to an elevation value. This provides node devices with terrain information for terrain analysis and communication path planning.

[0100] A three-dimensional topology provides the precise location of each node device, helping to optimize communication link selection and reduce signal attenuation. Digital elevation models enable node devices to adjust communication paths based on terrain characteristics, such as bypassing obstacles or selecting the optimal propagation path. Drone scanning can quickly generate high-precision terrain maps, making it suitable for areas with complex or frequently changing terrain, providing real-time data support for communication path planning.

[0101] Create a four-dimensional spectrum map and dynamically adjust the modulation method. The process is as follows:

[0102] The wide frequency band is divided into multiple sub-bands. The RF front-end of the node device scans each sub-band in sequence within 5 seconds to obtain the signal strength, interference strength, and center frequency within the sub-band;

[0103] The center frequency, the timestamp of the scan, the signal strength and the interference strength, and the three-dimensional coordinates of the node device are stored in the local storage unit of the node device to obtain a four-dimensional spectrum map;

[0104] Based on the signal strength and interference strength in the four-dimensional spectrum map, the channel quality of the sub-band is evaluated to obtain a channel quality evaluation result of the sub-band;

[0105] Comparing the channel quality assessment result of the sub-frequency band with the channel quality assessment results of each reference sub-frequency band stored in the data repository, the modulation mode corresponding to the channel quality assessment result of the same reference sub-frequency band is the modulation mode corresponding to the channel quality assessment result of the sub-frequency band;

[0106] The channel quality assessment result of the sub-band is obtained as follows:

[0107] Compare the signal strength with the signal evaluation threshold stored in the data repository. If the signal strength is less than the corresponding evaluation threshold, the channel quality level of the corresponding sub-frequency band is level 1, and the channel quality of the corresponding sub-frequency band is poor.

[0108] If the signal strength is not less than the signal evaluation threshold, the interference strength is compared with the interference evaluation threshold stored in the data repository. If the interference strength is less than the interference evaluation threshold, the channel quality level of the corresponding sub-frequency band is level 1, and the channel quality of the corresponding sub-frequency band is good.

[0109] If the interference intensity is not less than the interference assessment threshold, the channel quality level of the corresponding sub-frequency band is level 2, and the channel quality of the corresponding sub-frequency band is medium.

[0110] The node device's RF front-end sequentially scans multiple predefined sub-bands within 5 seconds, each covering a specific frequency range. Within each sub-band, the node device measures signal strength, interference strength, and center frequency, which is crucial for assessing the band's communication quality. The measured center frequency, timestamp, signal strength, interference strength, and three-dimensional coordinates of the node device are stored in a local storage unit, forming a four-dimensional dataset, where each point represents the signal and interference conditions of the frequency band at a specific time and spatial location. The channel quality of the sub-band is evaluated based on the signal strength and interference strength to obtain a channel quality assessment result. This assessment result is compared with the reference channel quality assessment results in the data repository to determine the optimal modulation method. Based on the channel quality assessment result, the most appropriate modulation method is selected for each sub-band, such as QPSK, 16QAM, or OFDM.

[0111] Accurate frequency band division and quality assessment effectively utilize available spectrum, avoid wasting resources on inefficient frequency bands, ensure optimal communication quality under different channel conditions, reduce bit error rate and increase data transmission rate.

[0112] Establish a path quality prediction algorithm to determine the quality of each path. The process is as follows:

[0113] Divide the initial planned path into main path, backup path, and emergency path;

[0114] The master node device obtains the path quality evaluation data of each initially planned path, including signal strength, bit error rate, signal-to-noise ratio, delay time, and packet loss rate, and performs weighted summation on the path quality evaluation data of each initially planned path to obtain a comprehensive quality score for each initially planned path;

[0115] The comprehensive quality score of each initial planned path is compared with the comprehensive quality score threshold of the corresponding reference path stored in the data repository. If the comprehensive quality score of the initial planned path is greater than the comprehensive quality score threshold of the corresponding reference path, the path quality of the initial planned path is high; otherwise, the path quality of the path is low;

[0116] If the quality of the primary path is low, the network switches to the backup path with higher quality. Otherwise, the network does not switch.

[0117] If the quality of the primary path and the backup paths are both low, the emergency path with high quality is switched to.

[0118] If the path quality of the primary path, each backup path, and each emergency path is low, an alarm is issued.

[0119] The primary path offers the best signal quality and is the most stable, used for transmitting critical services. The backup path serves as an alternative path in the event of a problem with the primary path, and its quality is slightly lower than the primary path. The emergency path is a last resort when both the primary and backup paths are unavailable. While its quality may be lower, it still provides basic communication support.

[0120] Dynamically select the optimal path to reduce communication failures due to path quality issues. Optimal paths are selected based on path quality scores, reducing unnecessary energy consumption and extending node device battery life. Predictive and preventive maintenance reduces maintenance requirements and costs due to path failures. By ensuring stable transmission of critical services, users' service quality experience is improved. The network maintains stable communications even in dynamically changing construction environments.

[0121] Establish business priorities. The process is as follows:

[0122] Divide business priorities into the following categories: lowest priority 4, low priority 3, medium priority 2, high priority 1, and highest priority 0;

[0123] Among them, the lowest priority 4 corresponds to non-real-time data transmission, the low priority 3 corresponds to equipment status monitoring, the medium priority 2 corresponds to video monitoring, the high priority 1 corresponds to hazardous gas alarm, and the highest priority 0 corresponds to emergency braking instructions;

[0124] Dedicated frequency bands, dedicated broadband, and dedicated paths are allocated for emergency braking commands. The comprehensive quality scores corresponding to the high-quality paths are sorted from large to small. The path with the largest comprehensive quality score is selected as the dedicated path for emergency braking commands. When the emergency braking command needs to be transmitted, it is transmitted immediately.

[0125] Divide the services into five priorities based on their requirements for real-time performance, importance, and quality of service, with each priority corresponding to a different service type. Assign a dedicated channel for emergency braking instructions: Determine the emergency braking instruction as the highest priority 0 because it requires the highest reliability and lowest latency. Assign dedicated frequency bands, dedicated broadband, and dedicated paths to the highest priority level 0 services to ensure that their transmission is not interfered with by other services. Evaluate the quality of each path, including key indicators such as signal strength, bit error rate, signal-to-noise ratio, delay time, and packet loss rate. Convert the evaluation results into a comprehensive quality score and assign a score to each path. Sort the paths from high to low according to the comprehensive quality score, and select the path with the highest score as the dedicated path for the emergency braking instruction. When the emergency braking instruction needs to be transmitted, use the dedicated path immediately for transmission.

[0126] Dedicated channels are allocated for emergency braking commands, ensuring rapid and reliable transmission in all circumstances, avoiding potentially catastrophic consequences. Network resources are dynamically allocated based on service priorities, prioritizing high-priority services while effectively utilizing remaining resources for lower-priority services. Dedicated paths reduce data transmission latency and potential interference, thereby reducing network latency.

[0127] Selecting the highest-quality path as a dedicated path improves data transmission reliability and reduces the likelihood of packet loss and errors. Allocating appropriate network resources to services of different priorities avoids resource waste, thereby optimizing energy consumption across the entire network. For users, the rapid response and high reliability of critical services (such as emergency braking commands) directly correlate to user satisfaction. In a dynamically changing infrastructure environment, service transmission paths can be rapidly adjusted based on real-time path quality assessment results to adapt to environmental changes. Efficient resource allocation and reliable transmission reduce maintenance requirements caused by communication failures, thereby lowering long-term maintenance costs.

[0128] To determine whether energy replenishment is needed, the process is as follows:

[0129] The energy consumption values ​​of each node device under different environmental conditions were measured in a controlled laboratory environment. The energy consumption values ​​and historical environmental data were divided into a training set and a validation set. The training set was input into the neural network model to establish an energy consumption model, and the validation set was used to verify the energy consumption model.

[0130] Obtain the current energy consumption value and current environmental data of each node device, input them into the trained energy consumption model, and obtain the predicted energy consumption value of each node device;

[0131] The current remaining power of each node device is obtained and subtracted from the corresponding predicted energy consumption value to obtain the comprehensive power of each node device. The comprehensive power of each node device is compared with the power threshold stored in the data storage library in turn to determine whether the comprehensive power of each node device is less than the power threshold. If it is less than, the node device triggers the energy replenishment process and performs energy replenishment. If it is not less than, it is not triggered.

[0132] Energy transmission efficiency refers to the amount of energy consumed when transmitting a certain amount of data. Accurately predicting energy consumption and providing timely energy replenishment prevents battery overdischarge, thereby extending battery life. This ensures that node devices do not fail due to battery exhaustion during critical missions, improving overall network reliability. Dynamically adjusting transmission power and modulation methods optimizes energy use based on energy efficiency and reduces waste. This reduces costs associated with battery replacement and maintenance, and lowers long-term operating costs by predicting and optimizing the energy replenishment process.

[0133] The network can dynamically adjust based on real-time energy consumption data and environmental changes, improving its adaptability to dynamic environments. Monitoring and predicting the energy consumption of different node devices can more evenly distribute network load and avoid overloading certain nodes. For users who rely on network communications, such as remote monitoring or automation systems, a more reliable network means fewer service interruptions and a better experience. Automated energy monitoring and replenishment processes reduce the need for manual intervention, making maintenance more intelligent and efficient.

[0134] Establish a communication-energy consumption optimization algorithm and obtain a network optimization solution. The process is as follows:

[0135] Obtain the energy transmission efficiency of the node device, divide the energy consumption value of the node device by the energy transmission efficiency to obtain the total energy consumption, obtain communication data, including delay time, reliability, and data transmission rate, and perform weighted summation of the communication data to obtain a communication performance score;

[0136] Define the objective function based on the total energy consumption and communication performance score;

[0137] Determine the constraints, including delay constraints, data rate constraints, reliability constraints, transmission power constraints, battery capacity constraints, and energy transmission efficiency constraints:

[0138] Delay constraint: the delay time of data transmission shall not exceed the maximum allowed delay time;

[0139] Data rate constraint: the data transmission rate must not be less than the minimum data transmission rate;

[0140] Reliability constraint: the reliability of data transmission shall not be less than the minimum reliability;

[0141] Transmission power constraint: the transmission power does not exceed the maximum transmission power;

[0142] Battery capacity constraint: the remaining power before energy replenishment is performed must not be less than the minimum remaining power;

[0143] Energy transmission efficiency constraint: the energy transmission efficiency is not less than the minimum efficiency;

[0144] The objective function to be minimized is determined based on the genetic algorithm. The optimal solution includes the optimal delay time, optimal data rate, optimal reliability, optimal transmission power, optimal remaining power, and optimal energy transmission efficiency, which is recorded as the network adjustment scheme.

[0145] The genetic algorithm process is as follows: S1. Encode the decision variables, such as mapping delay time, data rate, reliability, transmission power, remaining power, and energy transmission efficiency into genes in the genetic algorithm; S2. Randomly generate the initial population, where each individual represents a set of parameter settings; S3. Construct a fitness function to evaluate the performance of each individual, that is, the objective function value; S4. Select excellent individuals from the current population to enter the next generation population based on the roulette wheel selection method; S5. Perform a crossover operation on the selected next generation population to obtain a crossover population; S6. Perform a mutation operation on the crossover population to obtain a mutated population; S7. Repeat steps S4-S6 until the maximum number of iterations is reached, stop the calculation and output the current optimal solution, which is the network adjustment plan.

[0146] By implementing a communication-energy optimization algorithm, a more intelligent and efficient network management strategy can be achieved. This not only significantly improves network energy efficiency, extending node device battery life by reducing unnecessary energy consumption, but also reduces operating costs while ensuring service quality. The optimized network operates more stably, reducing service interruptions caused by insufficient energy, thereby enhancing the user experience. Furthermore, it enables the network to adapt to dynamically changing environmental conditions, ensuring optimal performance under varying conditions. Intelligent network adjustment solutions can automatically adjust network parameters, reducing manual intervention and making network management more efficient and convenient.

[0147] Build a node device trust model and determine the optimal transmission path. The process is as follows:

[0148] Obtain historical performance data of the node device, including data transmission success rate, historical failure rate, and energy efficiency. Compare the obtained data transmission success rate, historical failure rate, and energy efficiency with the data transmission success rate threshold, historical failure rate threshold, and energy efficiency threshold stored in the data repository, and perform a weighted sum to obtain an initial trust level.

[0149] The initial trust degree is compared with each reference trust degree stored in the data repository, and the trust level corresponding to the closest reference trust degree is the trust level of the initial trust degree;

[0150] Based on the trust levels of the node devices, the initial planned path corresponding to the node device with the highest trust level is selected as the optimal transmission path.

[0151] By analyzing the historical performance data of node devices and calculating the trust of each node device, the model is provided with an empirical basis for the behavior of node devices, making the calculation of trust more objective and accurate.

[0152] The node trust model collects data on each node's past operations, including its data transmission success rate, historical failure rate, and energy efficiency. It then compares these metrics to pre-set thresholds, which represent ideal performance for network nodes. By weighting the deviations from these thresholds, the model generates a composite initial trust score, reflecting the node's overall reliability.

[0153] By implementing a node device trust model, the overall performance and reliability of ad hoc networks are significantly improved. Accurately evaluating the historical and real-time behavior of node devices provides an intelligent path selection mechanism for the network, ensuring that critical data is transmitted over the most reliable and secure paths. This not only optimizes resource allocation by prioritizing the most reliable nodes and paths, thereby improving network efficiency, but also reduces maintenance requirements due to node device failures, effectively lowering operating and maintenance costs.

[0154] Furthermore, the node device trust model enhances network security by identifying and avoiding untrusted node devices, improving defenses against potential security threats. For end users, more reliable network connections mean fewer service interruptions and higher-quality data transmission, significantly improving user satisfaction. The node device trust model's dynamic adjustment capabilities enable the network to adapt to changing environments and conditions, ensuring long-term stability and adaptability.

[0155] Build a node failure prediction model to predict node equipment failures and trigger maintenance instructions. The process is as follows:

[0156] Perform feature extraction on historical fault operation data to obtain historical fault operation feature data;

[0157] Design the LSTM network architecture, including the input layer, LSTM layer, fully connected layer, and output layer; set network hyperparameters, including the number of LSTM units, number of layers, learning rate, batch size, activation function, and optimizer;

[0158] The historical node equipment failure feature data is divided into a training set, a validation set, and a test set. The training set is used to train the LSTM neural network prediction model to obtain a node failure prediction model.

[0159] Use the validation set to verify the node failure prediction model, and use the test set to test the node failure prediction model;

[0160] The operating data of the node device is obtained and input into the trained node fault prediction model to obtain the fault prediction result of the node device. If the fault prediction result is that the node device has a fault, a maintenance instruction is triggered.

[0161] Analyzing historical fault operation data can identify patterns and precursors to failures, enabling prediction of future failures. Real-time monitoring of node device status and fault prediction allows proactive action to avoid potentially serious problems. Once a failure is predicted, maintenance processes are immediately triggered, minimizing the impact on services. This reduces unplanned downtime, ensures the continuity of critical services, and effectively reduces maintenance costs and resource consumption.

[0162] Furthermore, predictive maintenance strategies optimize network resource allocation, prioritizing nodes most likely to fail, thereby enhancing overall network reliability and stability. Intelligent fault management and maintenance processes optimize user experience, reduce service interruptions, improve service quality, and significantly enhance overall network performance and economic benefits.

[0163] Example 2: Figure 2 As shown, a communication ad hoc network system applied to a signal-free infrastructure site, used to implement the method in Example 1, includes:

[0164] The initial planning path establishment module is used to construct the three-dimensional topology of the construction area based on multiple node devices deployed in the construction area. It combines the laser radar scanning carried by the drone to generate a digital elevation model of the construction area, determine the spatial distribution of node devices, and obtain each initial planning path;

[0165] The spectrum management module is used to create a four-dimensional spectrum map based on the spatial distribution of node devices and dynamically adjust the modulation method;

[0166] The path quality assessment module is used to establish a path quality prediction algorithm based on the spatial distribution of node devices and the four-dimensional spectrum map to determine the quality of each initially planned path;

[0167] A service priority management module is used to establish service priorities based on the quality of each initially planned path;

[0168] The energy management module is used to obtain the remaining power of each node device and determine whether energy replenishment is needed. If necessary, energy replenishment is performed according to the corresponding service priority;

[0169] The communication-energy consumption optimization module is used to establish a communication-energy consumption optimization algorithm based on the quality of each initially planned path, service priority, and energy replenishment results, combined with the remaining power of each node device, to obtain a network optimization plan and perform network optimization;

[0170] The path selection module is used to build a node device trust model based on each node device after network optimization and determine the best transmission path;

[0171] The fault prediction and maintenance module is used to obtain the historical fault operation data of each node device, build a node fault prediction model, predict node device failures, and trigger maintenance instructions.

[0172] Example 3: Based on Example 1, the initial planning path establishment module is improved to optimize the process of determining the spatial distribution of node devices:

[0173] When constructing the three-dimensional topological structure of the construction area, an environmental perception algorithm is introduced to enable node devices to perceive environmental changes in real time, including slight changes in terrain, the appearance of temporary obstacles, and changes in electromagnetic interference intensity, and dynamically feed these environmental perception data into the three-dimensional topological structure; at the same time, a reinforcement learning algorithm is used to intelligently optimize the flight path planning of drones, making the flight path of drones in complex environments more efficient and safer, thereby improving the generation efficiency and accuracy of digital elevation models, further optimizing the spatial distribution of node devices, and improving the adaptability and stability of the communication network.

[0174] Example 4: A communication ad hoc network device applied to infrastructure without signal sites includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The method in Example 1 or Example 3 is implemented by executing the computer program by the processor.

Claims

1. A communication ad hoc network method applied to infrastructure construction sites without signal, characterized in that: The following steps are involved: Based on multiple node devices deployed in the construction area, a three-dimensional topological structure of the construction area is constructed. Combined with the lidar scanning carried by the drone, a digital elevation model of the construction area is generated, the spatial distribution of the node devices is determined, and the initial planning paths are obtained; Create a four-dimensional spectrum map based on the spatial distribution of node devices and dynamically adjust the modulation method; Based on the spatial distribution of node devices and the four-dimensional spectrum map, a path quality prediction algorithm is established to determine the quality of each initially planned path; Establish service priorities based on the quality of each initially planned path; Obtain the remaining power of each node device and determine whether energy replenishment is needed. If necessary, energy replenishment is performed according to the corresponding service priority; Based on the quality of each initially planned path, service priority, and energy replenishment results, combined with the remaining power of each node device, a communication-energy consumption optimization algorithm is established to obtain a network optimization solution and perform network optimization. Based on the node devices after network optimization, a node device trust model is constructed to determine the optimal transmission path; Obtain historical fault operation data of each node device, build a node fault prediction model, predict node device failures, and trigger maintenance instructions; Determine the spatial distribution of node devices. The process is as follows: Obtain the motion state data of the node device, including acceleration, angular velocity, and magnetic field strength; Obtain the relative distance between each node device, combine the motion status data of the node device to obtain the three-dimensional coordinates of each node device, and construct the three-dimensional topological structure of the construction area; The introduction of environmental perception algorithms enables node devices to perceive environmental changes in real time and dynamically feed these environmental perception data into the three-dimensional topology structure; Obtain the size and shape of the construction area from the data repository, plan the drone's flight path, and use the drone's onboard lidar equipment to emit laser pulses and measure the time difference between reflected lasers to determine the distance to the target object. The data collected by the LiDAR equipment is stored in the form of a point cloud. Each point contains three-dimensional coordinates and reflection intensity information. The collected point cloud data is imported into the geographic information system software for preprocessing. The point cloud data is converted into regular grid data through interpolation algorithms to generate a digital elevation model. The digital elevation model represents the elevation information of the terrain in the form of a two-dimensional grid, where each grid cell corresponds to an elevation value; The generated digital elevation model is preloaded into the storage unit of each node device in the three-dimensional topological structure of the construction area. The node device accesses the digital elevation model to obtain the terrain information of the construction area, performs terrain analysis, and obtains each initial planning path.

2. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: Create a four-dimensional spectrum map and dynamically adjust the modulation method. The process is as follows: The wide frequency band is divided into multiple sub-bands. The RF front-end of the node device scans each sub-band in sequence within 5 seconds to obtain the signal strength, interference strength, and center frequency within the sub-band; The center frequency, the timestamp of the scan, the signal strength and the interference strength, and the three-dimensional coordinates of the node device are stored in the local storage unit of the node device to obtain a four-dimensional spectrum map; Based on the signal strength and interference strength in the four-dimensional spectrum map, the channel quality of the sub-band is evaluated to obtain a channel quality evaluation result of the sub-band; Comparing the channel quality assessment result of the sub-frequency band with the channel quality assessment results of each reference sub-frequency band stored in the data repository, the modulation mode corresponding to the channel quality assessment result of the same reference sub-frequency band is the modulation mode corresponding to the channel quality assessment result of the sub-frequency band; The channel quality assessment result of the sub-band is obtained as follows: Compare the signal strength with the signal evaluation threshold stored in the data repository. If the signal strength is less than the corresponding evaluation threshold, the channel quality level of the corresponding sub-frequency band is level 1, and the channel quality of the corresponding sub-frequency band is poor. If the signal strength is not less than the signal evaluation threshold, the interference strength is compared with the interference evaluation threshold stored in the data repository. If the interference strength is less than the interference evaluation threshold, the channel quality level of the corresponding sub-frequency band is level 1, and the channel quality of the corresponding sub-frequency band is good. If the interference intensity is not less than the interference assessment threshold, the channel quality level of the corresponding sub-frequency band is level 2, and the channel quality of the corresponding sub-frequency band is medium.

3. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: Establish a path quality prediction algorithm to determine the quality of each path. The process is as follows: Divide the initial planned path into main path, backup path, and emergency path; The node device obtains the path quality evaluation data of each initially planned path, including signal strength, bit error rate, signal-to-noise ratio, delay time, and packet loss rate, and performs weighted summation on the path quality evaluation data of each initially planned path to obtain a comprehensive quality score for each initially planned path; The comprehensive quality score of each initial planned path is compared with the comprehensive quality score threshold of the corresponding reference path stored in the data repository. If the comprehensive quality score of the initial planned path is greater than the comprehensive quality score threshold of the corresponding reference path, the path quality of the initial planned path is high; otherwise, the path quality of the path is low; If the quality of the primary path is low, the network switches to the backup path with higher quality. Otherwise, the network does not switch. If the quality of the primary path and the backup paths are both low, the emergency path with high quality is switched to. If the path quality of the primary path, each backup path, and each emergency path is low, an alarm is issued.

4. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: Establish business priorities. The process is as follows: Divide business priorities into the following categories: lowest priority 4, low priority 3, medium priority 2, high priority 1, and highest priority 0; Among them, the lowest priority 4 corresponds to non-real-time data transmission, the low priority 3 corresponds to equipment status monitoring, the medium priority 2 corresponds to video monitoring, the high priority 1 corresponds to hazardous gas alarm, and the highest priority 0 corresponds to emergency braking instructions; Dedicated frequency bands, dedicated broadband, and dedicated paths are allocated for emergency braking commands. The comprehensive quality scores corresponding to the high-quality paths are sorted from large to small. The path with the largest comprehensive quality score is selected as the dedicated path for emergency braking commands. When the emergency braking command needs to be transmitted, it is transmitted immediately.

5. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: To determine whether energy replenishment is needed, the process is as follows: The energy consumption values ​​of each node device under different environmental conditions were measured in a controlled laboratory environment. The energy consumption values ​​and historical environmental data were divided into a training set and a validation set. The training set was input into the neural network model to establish an energy consumption model, and the validation set was used to verify the energy consumption model. Obtain the current energy consumption value and current environmental data of each node device, input them into the trained energy consumption model, and obtain the predicted energy consumption value of each node device; The current remaining power of each node device is obtained and subtracted from the corresponding predicted energy consumption value to obtain the comprehensive power of each node device. The comprehensive power of each node device is compared with the power threshold stored in the data storage library in turn to determine whether the comprehensive power of each node device is less than the power threshold. If it is less than, the node device triggers the energy replenishment process and performs energy replenishment. If it is not less than, it is not triggered.

6. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: Establish a communication-energy consumption optimization algorithm and obtain a network optimization solution. The process is as follows: Obtain the energy transmission efficiency of the node device, divide the energy consumption value of the node device by the energy transmission efficiency to obtain the total energy consumption, obtain communication data, including delay time, reliability, and data transmission rate, and perform weighted summation of the communication data to obtain a communication performance score; Define the objective function based on the total energy consumption and communication performance score; Determine the constraints, including delay constraints, data rate constraints, reliability constraints, transmission power constraints, battery capacity constraints, and energy transmission efficiency constraints: Delay constraint: the delay time of data transmission shall not exceed the maximum allowed delay time; Data rate constraint: the data transmission rate must not be less than the minimum data transmission rate; Reliability constraint: the reliability of data transmission shall not be less than the minimum reliability; Transmission power constraint: the transmission power does not exceed the maximum transmission power; Battery capacity constraint: the remaining power before energy replenishment is performed must not be less than the minimum remaining power; Energy transmission efficiency constraint: the energy transmission efficiency is not less than the minimum efficiency; The objective function to be minimized is determined based on the genetic algorithm. The optimal solution includes the optimal delay time, optimal data rate, optimal reliability, optimal transmission power, optimal remaining power, and optimal energy transmission efficiency, which is recorded as the network adjustment scheme.

7. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: Build a node device trust model and determine the optimal transmission path. The process is as follows: Obtain historical performance data of the node device, including data transmission success rate, historical failure rate, and energy efficiency. Compare the obtained data transmission success rate, historical failure rate, and energy efficiency with the data transmission success rate threshold, historical failure rate threshold, and energy efficiency threshold stored in the data repository, and perform a weighted sum to obtain an initial trust level. The initial trust degree is compared with each reference trust degree stored in the data repository, and the trust level corresponding to the closest reference trust degree is the trust level of the initial trust degree; Based on the trust levels of the node devices, the initial planned path corresponding to the node device with the highest trust level is selected as the optimal transmission path.

8. The communication ad hoc network method applied to infrastructure without signal sites according to claim 1, characterized in that: Build a node failure prediction model to predict node equipment failures and trigger maintenance instructions. The process is as follows: Perform feature extraction on historical fault operation data to obtain historical fault operation feature data; Design the LSTM network architecture, including the input layer, LSTM layer, fully connected layer, and output layer; set network hyperparameters, including the number of LSTM units, number of layers, learning rate, batch size, activation function, and optimizer; The historical node equipment failure feature data is divided into a training set, a validation set, and a test set. The training set is used to train the LSTM neural network prediction model to obtain a node failure prediction model. Use the validation set to verify the node failure prediction model, and use the test set to test the node failure prediction model; The operating data of the node device is obtained and input into the trained node fault prediction model to obtain the fault prediction result of the node device. If the fault prediction result is that the node device has a fault, a maintenance instruction is triggered.

9. A communication ad hoc network system applied to infrastructure construction sites without signals, for implementing the method according to any one of claims 1 to 8, characterized in that: include: The initial planning path establishment module is used to construct the three-dimensional topology of the construction area based on multiple node devices deployed in the construction area. It combines the laser radar scanning carried by the drone to generate a digital elevation model of the construction area, determine the spatial distribution of node devices, and obtain each initial planning path; Determine the spatial distribution of node devices. The process is as follows: Obtain the motion state data of the node device, including acceleration, angular velocity, and magnetic field strength; Obtain the relative distance between each node device, combine the motion status data of the node device to obtain the three-dimensional coordinates of each node device, and construct the three-dimensional topological structure of the construction area; The introduction of environmental perception algorithms enables node devices to perceive environmental changes in real time and dynamically feed these environmental perception data into the three-dimensional topology structure; Obtain the size and shape of the construction area from the data repository, plan the drone's flight path, and use the drone's onboard lidar equipment to emit laser pulses and measure the time difference between reflected lasers to determine the distance to the target object. The data collected by the LiDAR equipment is stored in the form of a point cloud. Each point contains three-dimensional coordinates and reflection intensity information. The collected point cloud data is imported into the geographic information system software for preprocessing. The point cloud data is converted into regular grid data through interpolation algorithms to generate a digital elevation model. The digital elevation model represents the elevation information of the terrain in the form of a two-dimensional grid, where each grid cell corresponds to an elevation value; The generated digital elevation model is preloaded into the storage unit of each node device in the three-dimensional topological structure of the construction area. The node device accesses the digital elevation model to obtain the terrain information of the construction area, performs terrain analysis, and obtains each initial planning path; The spectrum management module is used to create a four-dimensional spectrum map based on the spatial distribution of node devices and dynamically adjust the modulation method; The path quality assessment module is used to establish a path quality prediction algorithm based on the spatial distribution of node devices and the four-dimensional spectrum map to determine the quality of each initially planned path; A service priority management module is used to establish service priorities based on the quality of each initially planned path; The energy management module is used to obtain the remaining power of each node device and determine whether energy replenishment is needed. If necessary, energy replenishment is performed according to the corresponding service priority; The communication-energy consumption optimization module is used to establish a communication-energy consumption optimization algorithm based on the quality of each initially planned path, service priority, and energy replenishment results, combined with the remaining power of each node device, to obtain a network optimization plan and perform network optimization; The path selection module is used to build a node device trust model based on each node device after network optimization and determine the best transmission path; The fault prediction and maintenance module is used to obtain the historical fault operation data of each node device, build a node fault prediction model, predict node device failures, and trigger maintenance instructions.

Images