A reliable data transmission method and system based on multi-path collaboration
By employing a multi-path collaborative data reliable backhaul method, and utilizing evaluation and prediction models and reinforcement learning algorithms to optimize the network and dynamically select the optimal backhaul link, the problems of unaccessible paths and low transmission efficiency in aircraft data backhaul are solved, achieving stable and efficient data transmission.
Patent Information
- Application Number
- CN202410090229.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-23
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-01-23
AI Technical Summary
In existing technologies, aircraft data acquisition and transmission cannot automatically adjust the backhaul link according to flight altitude, which may lead to unaccessible situations and make it impossible to select the optimal backhaul link while ensuring transmission efficiency and cost.
A reliable data backhaul method based on multi-path collaboration is adopted. By establishing an evaluation and prediction model, using reinforcement learning algorithms to optimize the cyclic unit network, and combining optimization algorithms to select the optimal backhaul link, the data transmission path is dynamically adjusted.
It enables dynamic adjustment of the data backhaul path based on actual conditions, reducing the risk of data loss, improving transmission stability and efficiency, and ensuring timely data delivery.
Smart Images

Figure CN117914435B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data return processing, and in particular to a reliable data return method and system based on multi-path collaboration. Background Art
[0002] An airplane is a means of aerial transportation that uses the lift generated by its wings to fly in the air. As one of the important means of transportation in modern society, airplanes have an important impact on globalization, economic development, cultural exchanges, etc. With the development of technology, the design and performance of aircraft are also constantly improving.
[0003] Aircraft data collection is a process that involves using aircraft as a platform to collect various types of data. It may include using aircraft to monitor weather conditions, atmospheric composition, geographical and geological changes, or monitoring traffic flow, road conditions, etc., or when natural disasters occur, aircraft can quickly collect information from disaster areas to assist in rescue operations, or monitor crop growth, soil moisture, etc. The advantage of aircraft data collection is its high mobility and wide coverage, and it can collect a large amount of data in a short period of time.
[0004] When using aircraft for data collection, data backhaul is a key process, which involves transmitting the data back to ground base stations or other receiving points after the aircraft collects the data. In existing technologies, data can be backhauled in a variety of ways, including satellite communications, radio frequencies, or through physical media after the aircraft lands. Therefore, the challenges of aircraft data collection and backhaul include ensuring the reliability and efficiency of transmission, processing large amounts of data, and ensuring the security of data transmission.
[0005] However, in the prior art, when the aircraft collects data and transmits it back, it is not possible to automatically adjust the backhaul link according to the aircraft's flight altitude, which may result in the aircraft being too high in altitude and being unable to access the ground base station. In addition, when the aircraft collects data and transmits it back, it cannot select a backhaul link with a shorter transmission delay based on the amount of data being transmitted back, and it cannot select a backhaul link with the longest service life, which may lead to transmission efficiency and cost problems.
[0006] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention
[0007] In order to solve the above problems, the present invention proposes a method and system for reliable data return based on multi-path collaboration to realize multi-path collaborative data return.
[0008] In order to achieve the above object, the present invention adopts the following technical solution:
[0009] In a first aspect, the present invention provides a method for reliable data transmission based on multi-path collaboration, and the method for reliable data transmission based on multi-path collaboration comprises the following steps:
[0010] S1. Determine the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and divide the distance levels according to the difference distance;
[0011] S2. Establish an evaluation prediction model using the distance level and the flight speed as stable efficiency values, and use the evaluation prediction model to predict the return volume of the aircraft collected data;
[0012] S3. Use the return data to simulate the data return process of the aircraft under different distance levels, design the multi-path return link under the corresponding distance level based on the return process, and determine the data transmission delay;
[0013] S4. According to the data transmission delay result and the optimization algorithm, the maximum usage time of the multi-path backhaul link under different backhaul amounts is selected as the final backhaul link.
[0014] Preferably, establishing an evaluation prediction model by taking the distance level and the flight speed as stable efficiency values, and using the evaluation prediction model to predict the return amount of the aircraft collected data includes the following steps:
[0015] S21. Set the distance level as the key factor affecting the evaluation prediction results, use the flight speed as the stable value to determine the variable relationship between it and the distance, and use the variable relationship and the distance level as the model input, and the data feedback amount as the model output result;
[0016] S22, using a reinforcement learning algorithm to optimize and design the recurrent unit network, setting the environmental state and learning state of the recurrent unit network to obtain an initial prediction model;
[0017] S23, coordinate the set model input, variable relationship and the set recurrent unit network to obtain an evaluation prediction model for the data return amount;
[0018] S24, using time series analysis technology to set a number of different numerical values within the range of distance level and flight speed and input them into the evaluation prediction model;
[0019] S25. Determine the amount of data to be sent back by obtaining environmental data of the aircraft's flight position under given distance level and flight speed conditions through an evaluation prediction model.
[0020] Preferably, the optimization and design operation of the recurrent unit network is implemented by using a reinforcement learning algorithm, and setting the environmental state and learning state of the recurrent unit network to obtain an initial prediction model includes the following steps:
[0021] S221, inputting the distance level into the recurrent unit network to generate an initial model structure, and presetting the termination condition of the reinforcement learning optimization to obtain the model structure state under different time step conditions;
[0022] S222, obtaining an optimal search action pair based on an optimal hybrid search strategy of a deep network, and performing a conversion operation on the model structure state to obtain an initial weight of a recurrent unit network;
[0023] S223, using the initial weights to set the training ratio and momentum factor of the recurrent unit network, and constructing reinforcement learning parameters according to the characteristics of the reinforcement learning algorithm;
[0024] S224. The recurrent unit network is trained based on the reinforcement learning parameters, and the environmental state and learning state of the recurrent unit network are set in combination with the additional momentum method to obtain an initial prediction model.
[0025] Preferably, the expression of the additional momentum method is:
[0026] L a =(1-mc)·ε·d a ·n j +mc·t i ;
[0027] Where, L a represents the parameter value of the recurrent unit network during the ath training, ε represents the training ratio, a represents the number of training times, mc represents the momentum factor, and d a represents the initial weight of the ath training, n j represents the optimal search action pair under the j-th search strategy, t i Represents the structural state value of the model under the condition of the i-th time step.
[0028] Preferably, using the return amount to simulate the data return process of the aircraft under different distance levels, designing a multipath return link under the corresponding distance level based on the return process and determining the data transmission delay includes the following steps:
[0029] S31. A simulation dynamic model is established for the return amount and the dynamic equation to describe the process of data return under different distance levels, and the corresponding data return amount under each distance level is estimated according to the process results;
[0030] S32, analyzing the data return characteristics at the corresponding level according to the data return amount, and using the analysis results to identify and design a number of return path links at the same distance level;
[0031] S33, using the corresponding return path link to simulate the data return process, determine the starting and ending points of the data when it is returned along the return path, and solve the data transmission delay.
[0032] Preferably, analyzing the data return characteristics at the corresponding level according to the data return amount, and using the analysis results to identify and design a number of return path links at the same distance level includes the following steps:
[0033] S321, analyzing the fluctuation range of the corresponding level according to the data return amount to identify the trend of the return amount, and determining the correlation between the return amount and the flight speed to obtain the data return characteristics;
[0034] S322, judging the relationship between the aircraft and the ground base station during the backhaul according to the flight altitude of the corresponding level and the data backhaul characteristics, and setting different flight altitude designated base station types based on the judgment result;
[0035] S323, setting to use the ground base station when the distance height is less than a preset value, and to use satellite communication when the distance height is greater than the preset value;
[0036] S324, using radar ranging to monitor the flight altitude of the aircraft, the signal quality of the base station and the network congestion status in real time, and automatically switching the base station connection type according to the real-time data;
[0037] S325, selecting a backhaul link type according to the corresponding connection type, and setting a backhaul path link by matching the characteristics of the backhaul link type with the corresponding flight altitude, the backhaul link types including optical fiber, radio wave and satellite link;
[0038] S326. Implement encryption measures for the return path link at each flight altitude to monitor the signal strength, transmission rate and delay of the data link, and use the monitoring results to set the link switching process to automatically trigger when the link quality decreases.
[0039] Preferably, selecting the maximum usage time of the multi-path backhaul link under different backhaul amounts as the final backhaul link according to the data transmission delay result in combination with the optimization algorithm includes the following steps:
[0040] S41, respectively extracting the starting and ending points of the multi-path return link, and presetting the data return position to be between the starting and ending points, and establishing a coordinate system with the lowest point of the return link as the coordinate origin;
[0041] S42, using the coordinate system to determine the change of the return link within the data transmission delay result to determine the change equation of the return path, and using a multi-objective conversion method to convert the change equation into a modulo function;
[0042] S43, using the global optimization capability of the optimization algorithm to find the optimal solution of the remainder function under different backhaul amounts, and substituting the optimal solution as the initial value of the iteration into the iterative method to find the usage time of the backhaul link;
[0043] S44, traversing to obtain the usage time of the multi-path backhaul links, matching the usage time with the backhaul links using visualization technology and displaying them in descending order;
[0044] S45. When the data backhaul amount is at the maximum value, the backhaul link with the highest ranking in terms of duration and backhaul link is selected as the final backhaul link at the current distance level.
[0045] Preferably, using the global optimization capability of the optimization algorithm to find the optimal solution of the remainder function under different backhaul amounts, and substituting the optimal solution as the initial value of the iteration into the iterative method to find the usage time of the backhaul link includes the following steps:
[0046] S431, setting the number of search points of the optimization algorithm in the backhaul link space, defining a state quantity according to the number of search points, and obtaining a variable when performing global optimization based on the state quantity;
[0047] S432, defining the perception range, search moving step length, crowding factor, maximum number of trials, maximum number of display changes and distance range between search points in the optimization algorithm;
[0048] S433, using the modulus function to determine the fitness values of each search point and compare the values, selecting the minimum fitness value as the optimal solution and assigning it to the presentation location in the optimization algorithm;
[0049] S434, using the presentation location to save the optimal record of different return amounts in the same return link optimization process, when the search point updates the state information of its own position, the presentation location compares the updated fitness value with the fitness value of the current search point, and uses the search point corresponding to the minimum fitness as the real-time search point according to the comparison result;
[0050] S435, when the preset maximum number of display changes is reached, the search point position corresponding to the minimum fitness value in the presentation location will be set as the optimal solution of the remainder function;
[0051] S436, substituting the optimal solution into the iteration method and setting the maximum number of iterations and iteration accuracy, when the result reaches the iteration accuracy and is less than the preset number of iterations, setting the output result as the usage time of the return link.
[0052] Preferably, using the modulus function to determine the fitness values of each search point and comparing the sizes, selecting the minimum fitness value as the optimal solution and assigning it to the presentation location in the optimization algorithm includes the following steps:
[0053] S4331, randomly selecting a position setting state for each search point in the backhaul link space based on the modulus function, and comparing the fitness value of the random position with the original position state, if the fitness value of the original state is greater than the fitness value of the random position, the search point moves forward by a preset step length in the direction of the selected position;
[0054] S4332, if the fitness value of the original state is greater than the fitness value of the random position, repeat the random selection of the position; if the fitness value of the original state is still greater than the random position after the number of repetitions reaches the threshold, randomly select the direction and move forward by a preset step length;
[0055] S4333. Obtain the fitness value of the search point after moving forward, and compare the difference with the fitness of the random position, select the one with the smaller difference as the optimal solution and assign it to the presentation location in the optimization algorithm.
[0056] In a second aspect, the present invention further provides a data reliable return system based on multi-path collaboration, the data reliable return system based on multi-path collaboration includes a height distance division module, a return amount prediction module, a transmission delay judgment module and a return path finding module;
[0057] The height distance division module is used to determine the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and divide the distance level according to the difference distance;
[0058] The feedback volume prediction module is used to establish an evaluation prediction model using the distance level and the flight speed as stable efficiency values, and use the evaluation prediction model to predict the feedback volume of the aircraft collected data;
[0059] The transmission delay judgment module is used to use the return data to simulate the data return process of the aircraft under different distance levels, design the multi-path return link under the corresponding distance level based on the return process, and judge the data transmission delay;
[0060] The backhaul path finding module is used to select the maximum usage time of the multi-path backhaul link under different backhaul amounts as the final backhaul link based on the data transmission delay result combined with the optimization algorithm.
[0061] The beneficial effects of the present invention are:
[0062] 1. The present invention designs a multi-path return link according to the altitude distance between the aircraft and the ground base station, so that other paths can be used when one path cannot receive the signal, reducing the risk of data loss and return failure caused by path problems. At the same time, the evaluation and prediction model established according to the distance level and flight speed between the aircraft and the ground base station can be dynamically adjusted according to actual conditions, providing a more accurate prediction of the data return amount, and designing a multi-path return link at the corresponding distance level and judging the data transmission delay, so as to select the optimal path according to the real-time situation, thereby reducing the delay in data transmission.
[0063] 2. The present invention uses distance level and flight speed as inputs of the evaluation and prediction model, so that it can accurately reflect the impact of environmental variables in actual flight on the amount of data feedback, and uses reinforcement learning algorithm to optimize the recurrent unit network, which is helpful to discover data processing and feedback strategies. The model input, variable relationship and recurrent unit network are coordinated to enable the model to better adapt to different flight conditions and environmental changes. By inputting environmental data of the aircraft's flight position into the model, the data feedback amount can be judged in real time, providing immediate decision support for aircraft data feedback.
[0064] 3. The present invention establishes a simulation dynamic model to accurately describe the process of data return at different distance levels, identifies the best data return path based on data return amount analysis and simulation process, thereby improving the stability and efficiency of data transmission, and calculates the data transmission delay by simulating the data return process, which helps to select the return path with the smallest transmission delay, thereby speeding up data transmission and ensuring timely arrival of data. The purpose of dynamically adjusting the return path according to actual conditions is achieved through analysis of different distance levels and data return amounts.
[0065] 4. The present invention analyzes the starting and ending points of the multi-path return link and the data return position, determines the change of the return path according to the coordinate system and the change equation, and uses the global optimization capability of the optimization algorithm to obtain the optimal solution. It can ensure that the most suitable return path is found under different return amounts and optimize the overall data transmission efficiency. At the same time, the use time of the return link is obtained through the iterative method, providing a variety of path choices for different data return amounts, thereby increasing the flexibility of the system to handle different situations. BRIEF DESCRIPTION OF THE DRAWINGS
[0066] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0067] Figure 1 is a flow chart of a reliable data return method based on multi-path collaboration according to an embodiment of the present invention;
[0068] Figure 2 It is a principle block diagram of a data reliable backhaul system based on multi-path collaboration according to an embodiment of the present invention.
[0069] In the figure:
[0070] 1. Height distance division module; 2. Return quantity prediction module; 3. Transmission delay judgment module; 4. Return path finding module. DETAILED DESCRIPTION
[0071] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0072] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.
[0073] It should be noted that the terms used herein are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that the terms "include" and "have" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0074] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.
[0075] See also Figure 1 The present invention provides a reliable data transmission method based on multi-path collaboration, and the reliable data transmission method based on multi-path collaboration includes the following steps:
[0076] S1. Determine the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and divide the distance levels using the difference distance.
[0077] In this embodiment, determining the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and dividing the distance levels by the difference distance includes the following steps:
[0078] Analyze the aircraft's preset flight path according to the flight plan, including the aircraft's take-off point, landing point, and all turning points and key locations during the flight;
[0079] Identify the specific location of the ground base station, calculate the straight-line distance between the aircraft and the nearest ground base station for each point on the flight path, and find the maximum value among all the calculated distances as the maximum difference distance between the aircraft and the ground base station.
[0080] The maximum difference distance is divided into different distance levels, including short-range, medium-range and long-range levels, and the corresponding communication link requirements are evaluated according to different distance levels. For example, satellite communication is used in the long-range level, while conventional radio communication is used in the short-range level.
[0081] S2. Use the distance level and flight speed as stable efficiency values to establish an evaluation prediction model, and use the evaluation prediction model to predict the return volume of aircraft collected data.
[0082] In this embodiment, the distance level and the flight speed are used as stable efficiency values to establish an evaluation prediction model, and the evaluation prediction model is used to predict the return amount of aircraft collected data, including the following steps:
[0083] S21. Set the distance level as the key factor affecting the evaluation prediction results, use the flight speed as the stable value to determine the variable relationship between it and the distance, and use the variable relationship and the distance level as the model input, and the data feedback amount as the model output result.
[0084] It should be explained that different distance levels are defined based on the maximum difference distance between the preset aircraft flight path and the ground base station, and the aircraft's flight speed is determined as the model stability value. The amount of data to be sent back is determined as the output result of the model, and machine learning technology is used to describe the relationship between the flight speed and distance level and the amount of data sent back.
[0085] S22. Use a reinforcement learning algorithm to optimize and design the recurrent unit network, and set the environmental state and learning state of the recurrent unit network to obtain an initial prediction model.
[0086] It needs to be explained that the reinforcement learning algorithm can adopt the artificial bee colony algorithm, which is a heuristic optimization algorithm inspired by the collaborative behavior of bee colonies. It is used to solve search and optimization problems by simulating the collective intelligent behavior of bees in the process of finding food and selecting beehives.
[0087] Specifically, the optimization and design operation of the recurrent unit network is implemented by using a reinforcement learning algorithm, and the environmental state and learning state of the recurrent unit network are set to obtain an initial prediction model, which includes the following steps:
[0088] S221, inputting the distance level into the recurrent unit network to generate an initial model structure, and presetting the termination condition of the reinforcement learning optimization to obtain the model structure state under different time step conditions;
[0089] S222, obtaining an optimal search action pair based on an optimal hybrid search strategy of a deep network, and performing a conversion operation on the model structure state to obtain an initial weight of a recurrent unit network;
[0090] S223, using the initial weights to set the training ratio and momentum factor of the recurrent unit network, and constructing reinforcement learning parameters according to the characteristics of the reinforcement learning algorithm;
[0091] S224. The recurrent unit network is trained based on the reinforcement learning parameters, and the environmental state and learning state of the recurrent unit network are set in combination with the additional momentum method to obtain an initial prediction model.
[0092] Among them, the expression of the additional momentum method is:
[0093] L a =(1-mc)·ε·d a ·n j +mc·t i ;
[0094] Where, L a represents the parameter value of the recurrent unit network during the ath training, ε represents the training ratio, a represents the number of training times, mc represents the momentum factor, and d a represents the initial weight of the ath training, n j represents the optimal search action pair under the j-th search strategy, t i Represents the structural state value of the model under the condition of the i-th time step.
[0095] S23, coordinate the set model input, variable relationship and the set cyclic unit network to obtain an evaluation prediction model for the data return amount.
[0096] Specifically, the set model input, variable relationship and the set recurrent unit network are implemented in coordination to obtain the evaluation prediction model of the data return amount, including the following steps:
[0097] Determine the model input, such as distance level, flight speed and other variables, set the model output, that is, the predicted value of the data return volume is input into the initial prediction model to determine the relationship between different variables.
[0098] Configure the recurrent unit network to process time series data or features with temporal characteristics, and determine the parameters of the network, including the number of layers and hidden units, to obtain an evaluation and prediction model for the amount of data returned.
[0099] S24. Using time series analysis technology, a number of different sets of numerical values are set within the limited range of distance level and flight speed and input into the evaluation prediction model.
[0100] Specifically, the limited range of the distance level and the flight speed can be set according to actual needs to generate different numerical input combinations and input them into the evaluation prediction model.
[0101] S25. Determine the amount of data to be sent back by obtaining environmental data of the aircraft's flight position under given distance level and flight speed conditions through an evaluation prediction model.
[0102] It should be noted that when determining the amount of data to be sent back, it is necessary to define the environmental data that needs to be obtained under given distance levels and flight speed conditions, determine the specific information that needs to be sent back, select applicable sensors and data sources, and determine the frequency of collecting environmental data based on the aircraft's flight speed and distance level.
[0103] Therefore, by taking the distance level and flight speed as the input of the evaluation and prediction model, it is possible to accurately reflect the impact of environmental variables in actual flight on the amount of data feedback, and using the reinforcement learning algorithm to optimize the recurrent unit network, which is helpful to discover data processing and feedback strategies. At the same time, the model input, variable relationship and recurrent unit network are coordinated to enable the model to better adapt to different flight conditions and environmental changes. By inputting the environmental data of the aircraft's flight position into the model, the data feedback amount can be judged in real time, providing immediate decision support for the aircraft data feedback.
[0104] S3. Use the return data to simulate the data return process of the aircraft under different distance levels. Based on the return process, design the multi-path return link under the corresponding distance level and determine the data transmission delay.
[0105] In this embodiment, the data backhaul process of the aircraft under different distance levels is simulated by using the backhaul volume, and the multipath backhaul link under the corresponding distance level is designed based on the backhaul process and the data transmission delay is determined, including the following steps:
[0106] S31. A simulation dynamic model is established for the feedback amount and the dynamic equation to describe the process of data feedback under different distance levels, and the corresponding data feedback amount at each distance level is estimated based on the process results.
[0107] Specifically, a simulation dynamic model is established for the return amount and the dynamic equation to describe the process of data return under different distance levels. The corresponding data return amount at each distance level is estimated based on the process results, including the following steps:
[0108] A simulation dynamic model describing the data feedback process is established using differential equations, and the data feedback process at different flight stages or positions under different distance level conditions is simulated based on the relationship between the data feedback amount and the variables in the dynamic model.
[0109] According to the initial conditions of the given simulation, including the initial state of the aircraft, environmental conditions, etc. This is the starting point for the model to start simulating. The dynamic model is used for simulation to obtain the data return process under different distance level conditions, and the simulation results are recorded, including the data return amount under different distance level conditions.
[0110] S32. Analyze the data return characteristics at the corresponding level according to the data return amount, and use the analysis results to identify and design several return path links at the same distance level.
[0111] Specifically, analyzing the data return characteristics at the corresponding level according to the data return amount, and using the analysis results to identify and design a number of return path links at the same distance level includes the following steps:
[0112] S321, analyzing the fluctuation range of the corresponding level according to the data return amount to identify the trend of the return amount, and determining the correlation between the return amount and the flight speed to obtain the data return characteristics;
[0113] S322, judging the relationship with the ground base station during the backhaul according to the flight altitude of the corresponding level and the data backhaul characteristics, and setting different flight altitude designated base station types based on the judgment result;
[0114] S323, setting to use the ground base station when the distance height is less than a preset value, and to use satellite communication when the distance height is greater than the preset value;
[0115] S324, using radar ranging to monitor the flight altitude of the aircraft, the signal quality of the base station and the network congestion status in real time, and automatically switching the base station connection type according to the real-time data;
[0116] S325, selecting a backhaul link type according to the corresponding connection type, and setting a backhaul path link by matching the characteristics of the backhaul link type with the corresponding flight altitude, the backhaul link types including optical fiber, radio wave and satellite link;
[0117] S326. Implement encryption measures for the return path link at each flight altitude to monitor the signal strength, transmission rate and delay of the data link, and use the monitoring results to set the link switching process to automatically trigger when the link quality decreases.
[0118] S33, using the corresponding return path link to simulate the data return process, determine the starting and ending points of the data when it is returned along the return path, and solve the data transmission delay.
[0119] Specifically, using the corresponding return path link to simulate the data return process, determining the start and end points of the data when it is returned along the return path to solve the data transmission delay includes the following steps:
[0120] Clearly define the data return path, set the data transmission speed according to the characteristics and performance of the communication link, simulate the data transmission process according to the defined return path, and record the start and end points of data transmission during the simulation.
[0121] The delay of data transmission is calculated by using the difference between the recorded start and end times, and the performance and delay of data transmission are evaluated based on the simulation results of the delay results.
[0122] Therefore, by establishing a simulation dynamic model, it is possible to accurately describe the data return process at different distance levels, identify the optimal data return path based on data return volume analysis and simulation process, thereby improving the stability and efficiency of data transmission, and calculate the data transmission delay by simulating the data return process, which helps to select the return path with the smallest transmission delay, thereby speeding up data transmission and ensuring timely arrival of data. The purpose of dynamically adjusting the return path according to actual conditions is achieved through analysis of different distance levels and data return volumes.
[0123] S4. According to the data transmission delay result and the optimization algorithm, the maximum usage time of the multi-path backhaul link under different backhaul amounts is selected as the final backhaul link.
[0124] It needs to be explained that the optimization algorithm may adopt an artificial fish school algorithm, which is an optimization algorithm based on simulating the behavior of fish schools and is used to solve search and optimization problems. The artificial fish school algorithm can find the global optimal solution in the search space by simulating the group intelligent behavior of fish schools.
[0125] In this embodiment, selecting the maximum usage time of the multi-path backhaul link under different backhaul amounts as the final backhaul link according to the data transmission delay result combined with the optimization algorithm includes the following steps:
[0126] S41, respectively extracting the starting and ending points of the multi-path return link, and presetting the data return position to be between the starting and ending points, and establishing a coordinate system with the lowest point of the return link as the coordinate origin.
[0127] Specifically, respectively extracting the start and end points of the multi-path return link, and presetting the data return position to be between the start and end points, and establishing a coordinate system with the lowest point of the return link as the coordinate origin includes the following steps:
[0128] Determine the starting and ending points of each return path, analyze the multi-path return links to find the lowest point of each link, that is, the lowest position on the coordinate axis, and preset the data return location on the path between the starting and ending points.
[0129] A coordinate system is established with the lowest point of each return path as the coordinate origin, the starting and ending points of each return path are marked in the coordinate system, and the preset data return position is marked.
[0130] S42, using the coordinate system to determine the change of the return link within the data transmission delay result to determine the change equation of the return path, and using a multi-objective conversion method to convert the change equation into a modulo function.
[0131] Specifically, using the coordinate system to determine the change of the return link within the data transmission delay result to determine the change equation of the return path, and using the multi-objective conversion method to convert the change equation into a remainder function includes the following steps:
[0132] Analyze the data transmission delay results to understand the changing trend of the backhaul link, and in the established coordinate system, establish the mathematical equation of the backhaul link according to the starting and ending points of the backhaul link and the preset data backhaul position, and convert the change equation of the backhaul link into the form of multi-objective transformation.
[0133] Determine the target to be converted, integrate the delay changes of multiple return paths into a comprehensive target function, and express the result of the multi-target conversion as a modulo function.
[0134] S43. Utilize the global optimization capability of the optimization algorithm to obtain the optimal solution of the remainder function under different backhaul amounts, and substitute the optimal solution as the initial value of the iteration into the iterative method to obtain the usage time of the backhaul link.
[0135] It should be explained that the iteration method may adopt Newton iteration algorithm, which can quickly approximate the root of the equation by utilizing the first-order derivative information of the function.
[0136] Specifically, using the global optimization capability of the optimization algorithm to find the optimal solution of the remainder function under different backhaul amounts, and substituting the optimal solution as the initial value of the iteration into the iteration method to find the usage time of the backhaul link includes the following steps:
[0137] S431, setting the number of search points of the optimization algorithm in the backhaul link space, defining a state quantity according to the number of search points, and obtaining variables when performing global optimization based on the state quantity.
[0138] S432, define the perception range of the search point in the optimization algorithm, the search moving step, the crowding factor, the maximum number of trials, the maximum number of display changes and the distance range between each search point.
[0139] S433, using the modulo function to determine the fitness value of each search point and compare the sizes, select the minimum fitness value as the optimal solution and assign it to the presentation location in the optimization algorithm.
[0140] Among them, using the modulus function to determine the fitness value of each search point and compare the size, selecting the minimum fitness value as the optimal solution and assigning it to the presentation in the optimization algorithm includes the following steps:
[0141] S4331, randomly selecting a position setting state for each search point in the backhaul link space based on the modulus function, and comparing the fitness value of the random position with the original position state, if the fitness value of the original state is greater than the fitness value of the random position, the search point moves forward by a preset step length in the direction of the selected position;
[0142] S4332, if the fitness value of the original state is greater than the fitness value of the random position, repeat the random selection of the position; if the fitness value of the original state is still greater than the random position after the number of repetitions reaches the threshold, randomly select the direction and move forward by a preset step length;
[0143] S4333. Obtain the fitness value of the search point after moving forward, and compare the difference with the fitness of the random position, select the one with the smaller difference as the optimal solution and assign it to the presentation location in the optimization algorithm.
[0144] S434. The presentation location is used to save the optimal record of different return amounts in the same return link optimization process. When the search point updates the status information of its own position, the presentation location compares the updated fitness value with the fitness value of the current search point, and uses the search point corresponding to the minimum fitness as the real-time search point based on the comparison result.
[0145] S435. When the preset maximum number of display changes is reached, the search point position corresponding to the minimum fitness value in the presentation location will be set as the optimal solution of the remainder function.
[0146] S436, substituting the optimal solution into the iteration method and setting the maximum number of iterations and iteration accuracy, when the result reaches the iteration accuracy and is less than the preset number of iterations, setting the output result as the usage time of the return link.
[0147] S44, traverse and obtain the usage time of the multi-path return link, use the visualization technology to match the usage time with the return link and display them in descending order.
[0148] It should be noted that sequential display requires matching the calculated usage time with the corresponding return link to establish a corresponding relationship, and using appropriate visualization tools to match the usage time with the return link and display the relevant steps in order from top to bottom.
[0149] S45. When the data backhaul amount is at the maximum value, the backhaul link with the highest ranking in terms of duration and backhaul link is selected as the final backhaul link at the current distance level.
[0150] Therefore, by analyzing the starting and ending points of the multi-path backhaul link and the data backhaul position, judging the change of the backhaul path according to the coordinate system and the change equation, and using the global optimization capability of the optimization algorithm to obtain the optimal solution, it is possible to ensure that the most suitable backhaul path is found under different backhaul conditions and optimize the overall data transmission efficiency. At the same time, the use time of the backhaul link is obtained through the iterative method, providing a variety of path choices for different data backhaul amounts, thereby increasing the flexibility of the system in handling different situations.
[0151] See also Figure 2 The present invention also provides a data reliable return system based on multi-path collaboration, the data reliable return system based on multi-path collaboration includes a height distance division module 1, a return amount prediction module 2, a transmission delay judgment module 3 and a return path finding module 4;
[0152] The height distance division module 1 is used to determine the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and divide the distance level according to the difference distance;
[0153] The feedback amount prediction module 2 is used to establish an evaluation prediction model using the distance level and the flight speed as stable efficiency values, and use the evaluation prediction model to predict the feedback amount of the aircraft collected data;
[0154] The transmission delay judgment module 3 is used to simulate the data return process of the aircraft under different distance levels by using the return amount, design the multi-path return link under the corresponding distance level based on the return process and judge the data transmission delay;
[0155] The backhaul path finding module 4 is used to select the maximum usage time of the multi-path backhaul link under different backhaul amounts as the final backhaul link according to the data transmission delay result combined with the optimization algorithm.
[0156] In summary, with the help of the above technical solution of the present invention, the present invention designs a multi-path return link according to the height distance between the aircraft and the ground base station, so that other paths can be used when one path cannot receive a signal, reducing the risk of data loss and return failure caused by path problems. At the same time, the evaluation and prediction model established according to the distance level and flight speed between the aircraft and the ground base station can be dynamically adjusted according to the actual situation, providing a more accurate prediction of the data return amount, and designing a multi-path return link under the corresponding distance level and judging the data transmission delay, and being able to select the optimal path according to the real-time situation, thereby reducing the delay in data transmission. The present invention uses the distance level and flight speed as the input of the evaluation and prediction model, so that it can accurately reflect the impact of environmental variables in actual flight on the data return amount, and optimizes the recurrent unit network using the reinforcement learning algorithm, which is helpful to discover data processing and return strategies, and coordinates the model input, variable relationship and recurrent unit network, so that the model can better adapt to different flight conditions and environmental changes, and by inputting the environmental data of the aircraft's flight position into the model, the data return amount can be judged in real time, providing instant decision support for aircraft data return.
[0157] The present invention establishes a simulation dynamic model to accurately describe the process of data return under different distance levels, identifies the best data return path based on the data return amount analysis and simulation process, thereby improving the stability and efficiency of data transmission, and calculates the data transmission delay by simulating the data return process, which helps to select the return path with the smallest transmission delay, thereby accelerating the data transmission speed and ensuring that the data arrives in time. The purpose of dynamically adjusting the return path according to the actual situation is achieved by analyzing the starting and ending points and data return positions of the multi-path return link, judging the change of the return path according to the coordinate system and the change equation, and using the global optimization ability of the optimization algorithm to obtain the optimal solution, which can ensure that the most suitable return path is found under different return amounts, optimize the overall data transmission efficiency, and at the same time, obtain the use time of the return link through the iteration method, provide a variety of path selections for different data return amounts, and increase the flexibility of the system to handle different situations.
[0158] Those skilled in the art will appreciate that the units, i.e., algorithm steps, of the various examples described in conjunction with this embodiment can be implemented in electronic hardware or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0159] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.
Claims
1. A reliable data transmission method based on multi-path collaboration, characterized in that: The reliable data transmission method based on multi-path collaboration includes the following steps: S1. Determine the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and divide the distance levels according to the difference distance; S2. Establish an evaluation prediction model using the distance level and the flight speed as stable efficiency values, and use the evaluation prediction model to predict the return volume of the aircraft collected data; S3. Use the return data to simulate the data return process of the aircraft under different distance levels, design the multi-path return link under the corresponding distance level based on the return process, and determine the data transmission delay; S4. According to the data transmission delay result and the optimization algorithm, the maximum usage time of the multi-path backhaul link under different backhaul amounts is selected as the final backhaul link; The method of establishing an evaluation prediction model by taking the distance level and the flight speed as stable efficiency values, and using the evaluation prediction model to predict the return amount of the aircraft collected data includes the following steps: S21. Set the distance level as the key factor affecting the evaluation prediction results, use the flight speed as the stable value to determine the variable relationship between it and the distance, and use the variable relationship and the distance level as the model input, and the data feedback amount as the model output result; S22, using a reinforcement learning algorithm to optimize and design the recurrent unit network, setting the environmental state and learning state of the recurrent unit network to obtain an initial prediction model; S23, coordinate the set model input, variable relationship and the set recurrent unit network to obtain an evaluation prediction model for the data return amount; S24, using time series analysis technology to set a number of different numerical values within the range of distance level and flight speed and input them into the evaluation prediction model; S25. Determine the amount of data to be sent back by obtaining environmental data of the aircraft's flight position under given distance level and flight speed conditions through an evaluation prediction model.
2. The method for reliable data transmission based on multi-path collaboration according to claim 1, characterized in that: The optimization and design operation of the recurrent unit network using the reinforcement learning algorithm and setting the environmental state and learning state of the recurrent unit network to obtain the initial prediction model include the following steps: S221, inputting the distance level into the recurrent unit network to generate an initial model structure, and presetting the termination condition of the reinforcement learning optimization to obtain the model structure state under different time step conditions; S222, obtaining an optimal search action pair based on an optimal hybrid search strategy of a deep network, and performing a conversion operation on the model structure state to obtain an initial weight of a recurrent unit network; S223, using the initial weights to set the training ratio and momentum factor of the recurrent unit network, and constructing reinforcement learning parameters according to the characteristics of the reinforcement learning algorithm; S224. The recurrent unit network is trained based on the reinforcement learning parameters, and the environmental state and learning state of the recurrent unit network are set in combination with the additional momentum method to obtain an initial prediction model.
3. The method for reliable data transmission based on multi-path collaboration according to claim 2, characterized in that: The expression of the added momentum method is: L a =(1-mc)·ε·d a ·n j +mc·t i ; Where, L a Represents the parameter value of the recurrent unit network during the a-th training; ε represents the training ratio; a represents the number of training times; mc represents the momentum factor; d a represents the initial weight of the a-th training; n j represents the optimal search action pair under the j-th search strategy; t i Represents the structural state value of the model under the condition of the i-th time step.
4. The method for reliable data transmission based on multi-path collaboration according to claim 1, characterized in that: The method of simulating the data return process of the aircraft under different distance levels by using the return amount, designing the multipath return link under the corresponding distance level based on the return process and determining the data transmission delay includes the following steps: S31. A simulation dynamic model is established for the return amount and the dynamic equation to describe the process of data return under different distance levels, and the corresponding data return amount under each distance level is estimated according to the process results; S32, analyzing the data return characteristics at the corresponding level according to the data return amount, and using the analysis results to identify and design a number of return path links at the same distance level; S33, using the corresponding return path link to simulate the data return process, determine the starting and ending points of the data when it is returned along the return path, and solve the data transmission delay.
5. The method for reliable data transmission based on multi-path collaboration according to claim 4, characterized in that: The method of analyzing the data return characteristics at the corresponding level according to the data return amount, and using the analysis results to identify and design a number of return path links at the same distance level includes the following steps: S321, analyzing the fluctuation range of the corresponding level according to the data return amount to identify the trend of the return amount, and determining the correlation between the return amount and the flight speed to obtain the data return characteristics; S322, judging the relationship between the aircraft and the ground base station during the backhaul according to the flight altitude of the corresponding level and the data backhaul characteristics, and setting different flight altitude designated base station types based on the judgment result; S323, setting to use the ground base station when the distance height is less than a preset value, and to use satellite communication when the distance height is greater than the preset value; S324, using radar ranging to monitor the flight altitude of the aircraft, the signal quality of the base station and the network congestion status in real time, and automatically switching the base station connection type according to the real-time data; S325, selecting a backhaul link type according to the corresponding connection type, and setting a backhaul path link by matching the characteristics of the backhaul link type with the corresponding flight altitude, wherein the backhaul link types include optical fiber, radio wave, and satellite link; S326. Implement encryption measures for the return path link at each flight altitude to monitor the signal strength, transmission rate and delay of the data link, and use the monitoring results to set the link switching process to automatically trigger when the link quality decreases.
6. The method for reliable data transmission based on multi-path collaboration according to claim 1, characterized in that: The method of selecting the maximum usage time of the multi-path backhaul link under different backhaul amounts as the final backhaul link according to the data transmission delay result in combination with the optimization algorithm comprises the following steps: S41, respectively extracting the starting and ending points of the multi-path return link, and presetting the data return position to be between the starting and ending points, and establishing a coordinate system with the lowest point of the return link as the coordinate origin; S42, using the coordinate system to determine the change of the return link within the data transmission delay result to determine the change equation of the return path, and using a multi-objective conversion method to convert the change equation into a modulo function; S43, using the global optimization capability of the optimization algorithm to find the optimal solution of the remainder function under different backhaul amounts, and substituting the optimal solution as the initial value of the iteration into the iterative method to find the usage time of the backhaul link; S44, traversing to obtain the usage time of the multi-path backhaul links, matching the usage time with the backhaul links using visualization technology and displaying them in descending order; S45. When the data backhaul amount is at the maximum value, the backhaul link with the highest ranking in terms of duration and backhaul link is selected as the final backhaul link at the current distance level.
7. The method for reliable data transmission based on multi-path collaboration according to claim 6, characterized in that: The method of using the global optimization capability of the optimization algorithm to obtain the optimal solution of the remainder function under different backhaul amounts, and substituting the optimal solution as the initial value of the iteration into the iteration method to obtain the usage time of the backhaul link includes the following steps: S431, setting the number of search points of the optimization algorithm in the backhaul link space, defining a state quantity according to the number of search points, and obtaining a variable when performing global optimization based on the state quantity; S432, defining the perception range, search moving step length, crowding factor, maximum number of trials, maximum number of display changes and distance range between search points in the optimization algorithm; S433, using the modulus function to determine the fitness values of each search point and compare the values, selecting the minimum fitness value as the optimal solution and assigning it to the presentation location in the optimization algorithm; S434, using the presentation location to save the optimal record of different return amounts in the same return link optimization process, when the search point updates the state information of its own position, the presentation location compares the updated fitness value with the fitness value of the current search point, and uses the search point corresponding to the minimum fitness as the real-time search point according to the comparison result; S435, when the preset maximum number of display changes is reached, the search point position corresponding to the minimum fitness value in the presentation location will be set as the optimal solution of the remainder function; S436, substituting the optimal solution into the iteration method and setting the maximum number of iterations and iteration accuracy, when the result reaches the iteration accuracy and is less than the preset number of iterations, setting the output result as the usage time of the return link.
8. The method for reliable data transmission based on multi-path collaboration according to claim 7, characterized in that: The method of using the modulus function to determine the fitness value of each search point and compare the values, selecting the minimum fitness value as the optimal solution and assigning it to the presentation location in the optimization algorithm includes the following steps: S4331, randomly selecting a position setting state for each search point in the backhaul link space based on the modulus function, and comparing the fitness value of the random position with the original position state, if the fitness value of the original state is greater than the fitness value of the random position, the search point moves forward by a preset step length in the direction of the selected position; S4332, if the fitness value of the original state is greater than the fitness value of the random position, repeat the random selection of the position; if the fitness value of the original state is still greater than the random position after the number of repetitions reaches the threshold, randomly select the direction and move forward by a preset step length; S4333. Obtain the fitness value of the search point after moving forward, and compare the difference with the fitness of the random position, select the one with the smaller difference as the optimal solution and assign it to the presentation location in the optimization algorithm.
9. A data reliable return system based on multi-path collaboration, used to implement the data reliable return method based on multi-path collaboration according to any one of claims 1 to 8, characterized in that: The reliable data return system based on multi-path collaboration includes a height distance division module, a return amount prediction module, a transmission delay judgment module and a return path finding module; The height distance division module is used to determine the maximum difference distance between the aircraft and the ground base station according to the preset aircraft flight path, and divide the distance levels according to the difference distance; The feedback amount prediction module is used to establish an evaluation prediction model using the distance level and the flight speed as stable efficiency values, and use the evaluation prediction model to predict the feedback amount of the aircraft collected data; The transmission delay judgment module is used to simulate the data return process of the aircraft under different distance levels by using the return amount, design the multi-path return link under the corresponding distance level based on the return process, and judge the data transmission delay; The backhaul path finding module is used to select the maximum usage time of the multi-path backhaul link under different backhaul amounts as the final backhaul link according to the data transmission delay result combined with the optimization algorithm; The method of establishing an evaluation prediction model by taking the distance level and the flight speed as stable efficiency values, and using the evaluation prediction model to predict the return amount of the aircraft collected data includes: The distance level is set as the key factor affecting the evaluation prediction results, and the flight speed is used as the stable value to determine the variable relationship between it and the distance. The variable relationship and the distance level are used as the model input, and the data feedback amount is used as the model output result. The reinforcement learning algorithm is used to optimize and design the recurrent unit network, and the environmental state and learning state of the recurrent unit network are set to obtain an initial prediction model; The set model input, variable relationship and the set recurrent unit network are implemented in coordination to obtain an evaluation prediction model for the data return amount; Using time series analysis technology, several sets of different numerical values are set within the limited range of distance level and flight speed and input into the evaluation prediction model; The amount of data returned is determined by evaluating the prediction model to obtain environmental data of the aircraft's flight position under given distance levels and flight speed conditions.
Citation Information
Patent Citations
Task unloading method and system adopting unmanned aerial vehicle in edge network and storage medium
CN112188441A
Multi-unmanned aerial vehicle data return routing method oriented to energy balance and high-reliability transmission
CN116456419A
