Ultralow-delay data graph transmission method and system based on 5G

Abstract

The invention discloses an ultra-low delay data graph transmission method and system based on 5G, and relates to the technical field of unmanned aerial vehicle data transmission, and the method effectively manages the switching of an unmanned aerial vehicle between different signal towers through a hybrid neural network and a dynamic programming algorithm, so as to ensure the continuity of data transmission and the stability of signals. The method comprises the following steps: firstly, acquiring flight data of an unmanned aerial vehicle and preprocessing; constructing a hybrid neural network including spatial feature extraction, time feature extraction and signal attenuation feature extraction; and finally, based on real-time flight information, predicting and determining an optimal signal tower switching point by using the neural network. In addition, the invention further provides a technology for dynamically adjusting the size of the data window, the frequency of data collection is dynamically adjusted according to the flight speed and signal change of the unmanned aerial vehicle so as to adapt to different environmental conditions, and therefore the data transmission efficiency is improved, and delay is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of unmanned aerial vehicle data transmission, and in particular to an ultra-low latency data image transmission method and system based on 5G. Background Art

[0002] With the rapid expansion of drone applications, especially in urban monitoring, traffic management, and emergency rescue, the demand for real-time data transmission from drones is increasing. Especially in urban environments, buildings are densely packed and signal towers are unevenly distributed, which makes drones face challenges of unstable signal coverage and data transmission delays when performing tasks. Traditional data transmission methods often rely on static signal tower connection strategies and lack the ability to dynamically adjust according to the real-time environment, which limits the application efficiency of drones in complex environments. To solve these problems, the use of 5G communication is a viable solution because 5G technology provides higher data transmission rates and lower latency. However, even 5G technology may encounter signal instability problems in highly dynamic environments, especially when drones frequently shuttle between high-rise buildings and multiple 5G signal base stations.

[0003] For example, the Chinese patent with publication number CN114885304A discloses a 5G networked drone data transmission system and method based on FlexE technology. The system includes: a drone terminal, a 5G base station and a drone cloud platform; the drone terminal is used to label the data packets to be transmitted according to the category, and queue the labeled data packets; the 5G base station uses FlexE as the bearer network to provide a dedicated data path with sufficient bandwidth and deterministic delay for the labeled data packets, and sends the data packets to the core network; the core network then sends the data packets to the drone cloud platform; the drone cloud platform is used to receive the data transmitted by the drone terminal, and process the received data and receive operation instructions for the drone terminal. Based on this system, the invention also proposes a 5G networked drone data transmission method based on FlexE technology. It can reduce delays, save energy, reduce costs, increase system capacity and large-scale device connections.

[0004] The above methods have the problems raised by this background technology. Therefore, developing a method that can respond to environmental changes in real time and intelligently manage signal tower switching is crucial to improving the operating efficiency of drones and the reliability of data transmission. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of this application is to provide an ultra-low latency data image transmission method and system based on 5G, which can dynamically adjust the signal switching point of the drone when it shuttles between various signal towers, so as to improve the transmission quality of the data image transmission task, based on the characteristics of short latency under 5G signals but limited coverage of signal towers and frequent switching, so as to improve the transmission quality of the data image transmission task. The specific scheme is as follows: In the first aspect, the present application provides an ultra-low latency data image transmission method based on 5G, including: Acquire historical flight data of the UAV within the coverage area of ​​multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; Based on the preprocessed historical flight data, a hybrid neural network is constructed; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; The real-time flight information of the UAV is obtained, and the signal switching point of the UAV within the coverage range of the multiple target signal towers is determined by using the hybrid neural network.

[0006] As an optional implementation, the preprocessing includes: Performing random phase shift on the quadrature amplitude modulation data to generate target training samples for the hybrid neural network training; The actual delay distribution is calculated based on the historical flight data, and a random delay distribution is generated. The first time series.

[0007] As an optional implementation, the constructing of a hybrid neural network includes: The spatial feature extraction branch is constructed based on a multi-layer convolutional neural network to extract the spatial position and environmental layout features of the UAV relative to the target signal tower and target environmental obstacles; Among them, each layer of convolutional neural network is connected to a batch normalization layer and a ReLU activation function; A time feature extraction branch is constructed based on a multi-layer long short-term memory network to extract the time dependency and dynamic changes of the historical flight data; wherein the multi-layer long short-term memory network includes: a first time step sliding window method; At least one fully connected layer is used to form a signal attenuation feature extraction branch to analyze the distance and signal strength data from the drone to each signal tower, and predict signal attenuation and potential switching points; each fully connected layer is equipped with a discard layer.

[0008] As an optional implementation, the historical flight data further includes: flight speed, network change information; the first time step sliding window method includes: dynamically adjusting the time window size based on the flight speed of the drone and the network signal change information; The dynamically adjusting the time window size includes: Based on the flight mission of the UAV, a scene mode is determined; the scene mode includes a first scene mode, a second scene mode and a third scene mode; When the drone is in the first scene mode, the flight speed of the drone is detected, and a first initial time window size is set; in response to the flight speed of the drone being greater than or equal to a first speed threshold, the time window size is reduced; in response to the flight speed of the drone being less than a second speed threshold, the time window size is increased; When the drone is in a second scene mode, detecting a signal-to-noise ratio of an environment in which the drone is located, and setting a second initial time window size, in response to the signal-to-noise ratio being greater than or equal to a first signal-to-noise ratio threshold, reducing the time window size, and in response to the signal-to-noise ratio being less than a second signal-to-noise ratio threshold, increasing the time window size; When the drone is in a third scene mode, the signal strength of the drone is detected and a third initial time window size is set. In response to the signal strength being greater than or equal to a first signal threshold, the time window size is reduced. In response to the signal strength being less than a second signal threshold, the time window size is increased.

[0009] As an optional implementation, the output of the hybrid neural network includes: signal quality score, signal attenuation prediction, and interference level assessment; determining the signal switching point of the drone within the coverage range of the multiple target signal towers includes: Using the first dynamic planning scheme, determining the signal switching point of the UAV within the coverage area of ​​multiple target signal towers; The first dynamic programming scheme includes: state definition, decision variables, transfer equations, objective function, initialization and boundary conditions, and dynamic programming solution.

[0010] Secondly, this application also provides an ultra-low latency data image transmission system based on 5G, including: An acquisition unit is used to acquire historical flight data of the UAV within the coverage area of ​​multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; A construction unit constructs a hybrid neural network based on the preprocessed historical flight data; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; The processing unit is used to obtain real-time flight information of the UAV and determine the signal switching point of the UAV within the coverage range of the multiple target signal towers using the hybrid neural network.

[0011] In the third aspect, an optional implementation method of the present application also provides an electronic device, including a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is used to execute the machine-readable instructions stored in the memory, and when the machine-readable instructions are executed by the processor, the machine-readable instructions perform the steps of the above-mentioned 5G-based ultra-low latency data image transmission method.

[0012] In a fourth aspect, an optional implementation of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run, the steps of the above-mentioned 5G-based ultra-low latency data image transmission method are executed.

[0013] Compared with the prior art, the beneficial effects of the present invention are: by introducing a hybrid neural network and a dynamic programming algorithm, intelligent signal switching management of drones under 5G networks is realized, the stability and continuity of data transmission are significantly improved, and the signal interruption problem in traditional methods is reduced. At the same time, by dynamically adjusting the time window size, the data collection frequency can be flexibly adjusted according to the changes in the drone's flight speed and signal quality, reducing transmission delays and improving data transmission efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A flowchart of an ultra-low latency data image transmission method based on 5G provided in an embodiment of the present application; Figure 2 A flow chart of a method for dynamically adjusting the time window size provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of an ultra-low latency data image transmission system based on 5G provided in an embodiment of the present application; Figure 4 A schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0016] 5G networks have the characteristics of high bandwidth and low latency, which enables drones to transmit high-quality data such as video streaming and sensor information in real time. However, the coverage range of 5G network signal towers is relatively small, and drones may need to frequently switch between multiple signal towers during flight to maintain signal stability and continuity. This requires a dynamic adjustment strategy that optimizes the signal tower switching process through hybrid neural networks and dynamic programming, reduces latency and data loss, and ensures that drones can reliably perform tasks in complex environments.

[0017] For this purpose, see Figure 1 As shown, a flowchart of an ultra-low latency data image transmission method based on 5G provided in an embodiment of the present application, the method includes steps S101 to S103, wherein: S101: Acquire flight data of a UAV within the coverage of multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; S102: constructing a hybrid neural network based on the preprocessed historical flight data; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; S103: Acquire real-time flight information of the UAV, and use the hybrid neural network to determine a signal switching point of the UAV within the coverage of the multiple target signal towers.

[0018] In the specific implementation, the present application first collects historical flight data involved in the UAV flight process, including but not limited to: In-phase and Quadrature data (IQ) data, Access Point data (AP) data and signal attenuation data, and cleans and formats the collected data, such as removing noise, standardizing processing, and data synchronization to ensure the accuracy and consistency of the data when it is input into the neural network.

[0019] Among them, IQ data consists of basic signals in wireless communication, including in-phase and quadrature. These data reflect the amplitude and phase of the signal and are key information for signal processing and analysis. AP data involves wireless connection information between the drone and the ground or other network infrastructure, including but not limited to the identification, location, signal strength and network quality of the access point. Signal attenuation data describes the loss of signal strength received by the drone from the signal source (such as GPS satellites, Wi-Fi access points) with distance or other obstacles.

[0020] There are many ways to obtain the above IQ data, AP data, and signal attenuation data. UAVs can directly measure and record these data through integrated wireless communication hardware such as Wi-Fi modules and cellular network modules. These sensors can capture the strength and quality of signals in real time, as well as relevant information about signal towers; software tools can also be used to analyze received signal data, such as estimating signal attenuation and network access point performance by processing raw signal data through algorithms; and detailed data about signal towers and network coverage can be obtained from external network services using a ground control station or cloud platform that communicates with the drone. The choice of these methods can be based on the specific application requirements and operating environment of the drone to ensure the accuracy and real-time nature of the data.

[0021] After acquiring and preprocessing the historical flight data of the above-mentioned UAV, the present application constructs a hybrid neural network to extract features from the above-mentioned data from multiple different dimensions, including: using the spatial feature extraction branch to process and analyze the spatial position information of the UAV relative to the signal tower, and extract position-related features; using the temporal feature extraction branch to analyze the flight dynamics of the UAV, such as speed and movement trajectory, and extract time series data features; using the signal attenuation feature extraction branch to analyze the signal strength data and predict the potential attenuation of the signal in different environments.

[0022] In this application, after the above hybrid neural network is constructed, the flight data updated in real time is input into the trained hybrid neural network. The hybrid neural network comprehensively considers the spatial position, flight time characteristics and signal attenuation, and outputs a signal tower switching suggestion, which optimizes the communication efficiency of the drone based on the predicted signal quality and future flight trajectory.

[0023] As an optional implementation, the preprocessing includes: performing a random phase shift on the IQ data to generate a target training sample, and calculating the actual delay distribution based on the flight data to generate a random delay containing The first time series.

[0024] In the specific implementation, random phase shift is performed on IQ data to generate target training samples. IQ data contains the amplitude and phase information of the signal. By performing random phase shift on IQ data, different signal interference and environmental changes can be simulated. This process enhances the diversity of training data, allowing the network model to maintain high robustness in different environments.

[0025] For example, randomly shifting the phase in the original IQ data to a value within a range (such as ±30 degrees) to generate a variety of different signal samples can help improve the generalization ability of the model and effectively avoid the overfitting problem.

[0026] In this application, the actual delay distribution is calculated based on the flight data in order to generate a first time series containing random delays. During the flight of the drone, signal transmission is affected by various factors, resulting in uncertainty in the actual delay distribution. By analyzing the flight data, the actual delay distribution is calculated, and on this basis, a first time series containing random delays is generated. The first time series can accurately simulate the actual signal transmission situation.

[0027] In this way, more representative and diverse training data can be provided for the hybrid neural network, thereby improving the performance of the model in practical applications.

[0028] With respect to the above S102: constructing a hybrid neural network based on the pre-processed historical flight data; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch.

[0029] As an optional implementation, constructing a hybrid neural network includes: constructing the spatial feature extraction branch based on a multi-layer convolutional neural network; wherein each layer of the convolutional network is followed by a batch normalization and a ReLU activation function, for extracting the spatial position and environmental layout features of the drone relative to the signal tower and the target environmental obstacles; A temporal feature extraction branch is constructed based on a multi-layer long short-term memory network to extract the temporal dependency and dynamic changes in the UAV flight data; wherein the multi-layer long short-term memory network includes: a sliding window method for the first time step; Signal attenuation feature extraction branch: It consists of at least one fully connected layer, each of which is followed by dropout, which is used to analyze the distance and signal strength data from the drone to each signal tower, and predict signal attenuation and potential switching points.

[0030] The present application provides a specific hybrid neural network, including a spatial feature extraction branch, a temporal feature extraction branch and a signal attenuation feature extraction branch.

[0031] S1021: Spatial feature extraction branch, using multi-layer convolutional neural network (CNN) to build the spatial feature extraction branch. Each convolution layer is equipped with batch normalization and ReLU activation function to ensure the effectiveness and stability of feature extraction.

[0032] Exemplary, specific parameters are as follows: Convolutional layer: Each layer contains 32 3x3 filters.

[0033] Batch Normalization: Used to stabilize and speed up the training process.

[0034] ReLU activation function: used to introduce nonlinearity to enhance the expressiveness of the model.

[0035] In this way, the spatial position and layout features of the drone relative to signal towers and environmental obstacles can be effectively extracted, improving the accuracy and robustness of the model.

[0036] S1022: Temporal feature extraction branch, which uses a multi-layer long short-term memory network to process and extract the temporal dependencies and dynamic changes in UAV flight data.

[0037] Exemplary, specific parameters are as follows: LSTM layer: Each layer contains 50 hidden units, and two layers of LSTM are stacked to capture complex temporal dependencies.

[0038] S1023: Signal attenuation feature extraction branch, consisting of at least one fully connected layer (Dense Layer), each fully connected layer is equipped with a dropout layer to prevent overfitting and enhance the generalization ability of the model.

[0039] Exemplary, specific parameters are as follows: Fully connected layer: contains 128 neurons.

[0040] Dropout layer: Set the dropout rate to 0.5.

[0041] In this way, through this hybrid neural network, multi-layer convolution and LSTM networks can efficiently process spatial and temporal features to ensure rapid transmission and processing of data; the signal attenuation feature extraction branch enables the drone to accurately predict signal attenuation and switching points, thereby optimizing signal switching decisions.

[0042] As a possible implementation manner, the historical flight data further includes: flight speed, network change information; the first time step sliding window method includes: Based on the flight speed of the drone and the network signal change information, the window size is dynamically adjusted, including: Based on the flight mission of the UAV, a scene mode is determined; the scene mode includes a first scene mode, a second scene mode and a third scene mode; When the drone is in the first scene mode, the flight speed of the drone is detected, and a first initial time window size is set; in response to the flight speed of the drone being greater than or equal to a first speed threshold, the time window size is reduced; in response to the flight speed of the drone being less than a second speed threshold, the time window size is increased; When the drone is in a second scene mode, detecting a signal-to-noise ratio of an environment in which the drone is located, and setting a second initial time window size, reducing the time window size in response to the signal-to-noise ratio being greater than or equal to a first noise threshold, and increasing the time window size in response to the signal-to-noise ratio being less than a second noise threshold; When the drone is in a third scene mode, the signal strength of the drone is detected and a third initial time window size is set. In response to the signal strength being greater than or equal to a first signal threshold, the time window size is reduced. In response to the signal strength being less than a second signal threshold, the time window size is increased.

[0043] In the specific implementation, this application adopts the first time step sliding window method, that is, the dynamic adjustment strategy of the time window, to cope with the real-time data processing needs of UAVs in different flight scenarios.

[0044] The time window refers to the amount of data collected and processed within a certain period of time. The purpose of dynamically adjusting the time window is to optimize the efficiency of data transmission and processing, and to flexibly adjust the time interval for collecting and processing data according to real-time flight data and network conditions.

[0045] In the specific implementation, the flight data not only includes orthogonal amplitude modulation data (IQ data), access point data (AP data) and signal attenuation data, but also includes flight speed and network change information. Based on these data, the time window size can be dynamically adjusted using the first time step sliding window method.

[0046] First, the scene mode is determined according to the flight mission of the drone. The scene mode determines how to adjust the time window size.

[0047] The scene mode can be determined manually or automatically to determine the flight mission and scene mode of the drone.

[0048] For example, the user can manually set the scene mode according to the predefined flight mission and environmental conditions. For example, before the drone takes off, the operator can select a suitable scene mode according to the mission requirements, such as "high-speed cruise mode" or "low-speed reconnaissance mode". This manual setting method ensures that the operator has control over the drone and can make the most appropriate choice according to the specific mission requirements.

[0049] For example, the drone can also automatically determine the scene mode by analyzing the flight mission and environmental changes in real time through built-in sensors and intelligent algorithms.

[0050] For example, the drone identifies the current flight mission through the mission management system, determines the nature of the current mission by reading mission parameters such as flight path, target area, flight speed, etc., and obtains real-time environmental data through sensors, including flight speed, signal strength, signal-to-noise ratio, etc. Based on these data, the complexity and changes of the current flight environment are automatically evaluated.

[0051] Automatically select and switch to the most suitable scene mode based on task recognition and environmental perception results.

[0052] In the specific implementation, this application provides three specific scenario modes, and the applicable scenarios of these three modes are as follows: 1. Speed ​​Sensing Mode: Suitable for scenarios where the flight speed of the drone changes significantly, such as tracking fast-moving targets or high-speed cruising. In this mode, the time window is adjusted according to the flight speed, which can respond more quickly to data changes caused by speed.

[0053] 2. Signal Strength Mode: Suitable for environments with large fluctuations in signal strength, such as densely populated high-rise buildings in cities or remote areas. This mode helps optimize data capture and processing in weak signal areas by dynamically adjusting the time window based on signal strength.

[0054] 3. Signal Complexity Mode: Suitable for environments with severe signal interference or multiple signals, such as industrial areas or areas where multiple communication devices are used. In complex signal environments, frequent updates of the data window help reduce errors and signal loss.

[0055] By selecting the appropriate mode, UAV systems can more effectively adapt to different operating environments and challenges, thereby improving the efficiency and success rate of mission execution.

[0056] In a specific implementation, when the drone is in speed sensing mode, the present application adjusts the time window by detecting the flight speed. For example, an initial time window size can be set, that is, the first initial time window size in the present application. When the flight speed is greater than or equal to the first speed threshold, the time window size is reduced on the basis of the initial time window to respond to environmental changes more quickly; when the flight speed is less than the second speed threshold, the time window size is increased to reduce the frequency of data processing.

[0057] In the signal strength mode, the present application adjusts the time window by detecting the signal-to-noise ratio of the environment in which the drone is located. Exemplarily, an initial time window size can be set, i.e., the second initial time window size in the present application. When the signal-to-noise ratio is greater than or equal to the first noise threshold, the time window size is reduced on the basis of the initial time window to ensure that data can be processed quickly in a high-noise environment; when the signal-to-noise ratio is less than the second noise threshold, the time window size is increased to reduce unnecessary data processing burden.

[0058] In the signal complexity mode, the present application adjusts the time window by detecting the signal strength of the drone. Exemplarily, an initial time window size can be set, that is, the second initial time window size in the present application. When the signal strength is greater than or equal to the first signal threshold, the time window size is reduced to cope with the change in signal strength; when the signal strength is less than the second signal threshold, the time window size is increased to ensure that enough data can be collected for processing when the signal is weak. The purpose of dynamically adjusting the time window is to optimize the efficiency of data collection and processing under different flight and network conditions, and to improve the stability and real-time performance of data transmission. This method can help drones maintain efficient communication and data processing capabilities in complex and changing environments, and ensure the smooth completion of tasks.

[0059] See also Figure 2 , Figure 2 A flow chart of a method for dynamically adjusting the time window size is provided for an embodiment of the present application. In the present application, there are three scene modes, namely, a first scene mode for drone speed, a second scene mode for drone signal strength, and a third scene mode for drone signal complexity. In each scene mode, the present application provides two methods for dynamically adjusting the time window size, and these two methods are respectively suitable for reducing and increasing the time window size.

[0060] As a possible implementation, when the drone is in the first scene mode, the time window size is set The formula is: ; in Represents the flight speed of the drone, is the first initial time window size, is the speed adjustment factor, is the first speed threshold.

[0061] As an optional implementation, when the drone is in the first scene mode, the time window size is set The formula is: ; in is the speed adjustment factor, is the second speed threshold.

[0062] In the first scenario mode of the present application, in view of the immediate impact of speed changes on data collection, when flying at low speed, the drone's demand for real-time data is not as urgent as that of flying at high speed, so it can be collected at a speed lower than the second speed threshold. The time window is increased to reduce the frequency of data processing and save energy. Considering that the data does not change much when the speed is low, the square function can be used to smoothly increase the time window to reduce the fluctuation caused by frequent adjustments.

[0063] Conversely, when flying at high speeds, the drone may need to update data more frequently to cope with rapidly changing environments. When the time window is small, the application uses an exponential function to quickly reduce the time window to improve the frequency and accuracy of data updates.

[0064] These two formulas combine the specific characteristics of speed through and The adjustment of the time window can flexibly respond to the data transmission requirements under different flight conditions. When adjusting these parameters, the actual flight environment and the mission requirements of the UAV should be considered to ensure that the adjustment of the time window is effective.

[0065] For example, aerial monitoring in urban and rural areas requires different data update frequencies depending on flight speed.

[0066] For example, high speed adjustment factor : 0.1, used in fast-moving dynamic environments (such as urban areas). Low speed speed adjustment coefficient : 0.02, used in stable, less variable environments (such as rural areas or controlled airspace). First speed threshold : 50 km / h, reaching or exceeding this speed requires a fast data update. Second speed threshold : 20 km / h, below which the data update frequency can be reduced. If the drone speed is 70 km / h (exceeding the first speed threshold); This indicates that during high-speed flight, the data update frequency is accelerated. At this time, the current time window value can be set to the minimum time window value. For example, the minimum time window value can be set to 0.1 seconds.

[0067] If the drone speed drops to 10 km / h (below the second speed threshold); When the drone is flying at low speed, the data update frequency slows down and the time window increases to 1.2 seconds.

[0068] By adjusting and The higher the coefficient, the drone can more effectively manage its data collection frequency in response to varying speeds and environmental conditions. A value of allows the time window to be reduced more quickly, ensuring that the data is updated frequently to capture dynamic changes. Conversely, a smaller The value moderates the increase in window size in situations where less frequent updates are required. This flexibility enhances the operational efficiency and responsiveness of the UAV when performing its mission.

[0069] As an optional implementation, when the drone is in the second scene mode, the time window size is set The formula is: ; in is the signal-to-noise ratio, is the second initial time window size, is the signal-to-noise ratio adjustment coefficient, is the first signal-to-noise ratio threshold.

[0070] As an optional implementation, when the drone is in the second scene mode, the time window size is set The formula is: ; in is the second signal-to-noise ratio threshold.

[0071] In a specific implementation, when the signal-to-noise ratio exceeds When the time window size It will decrease as the signal-to-noise ratio increases. When the signal-to-noise ratio is less than When, time window Will increase.

[0072] In this way, the impact of the signal-to-noise ratio on the time window size will be more accurately reflected, ensuring that the data transmission frequency of the drone can be appropriately adjusted when the signal-to-noise ratio changes significantly. This adjustment is based on the position of the actual signal-to-noise ratio relative to the set threshold, and exponential and logarithmic functions are used to strengthen the adjustment of the window size, making it more sensitive and reasonable to reflect environmental changes.

[0073] In this application, the exponential function To enhance the signal-to-noise ratio When the threshold is higher than the time window The reduction effect of Exceeding the threshold When is a positive number, Rapid growth means that The further increase of the time window will be more significant. This processing method adopted by the present application is suitable for scenarios that need to be more sensitive to high signal-to-noise ratios, ensuring that data transmission is more frequent when the signal is very good, taking advantage of the good signal environment.

[0074] In this application, the logarithmic function Used for Below the second threshold Increase the time window The logarithmic function grows slowly, which means that even if The gap gradually increases, and the increase in the time window is also gradual. The window will not increase dramatically due to a slight signal drop, which can maintain a certain stability and data continuity.

[0075] In this way, extreme adjustments caused by sudden signal changes can be avoided, maintaining a smoother data transmission efficiency. and The mathematical properties of the proposed algorithm can balance the response speed and stability of the time window adjustment while ensuring performance, especially in highly dynamic UAV operating environments.

[0076] Exemplarily, a second initial time window may be set : 1.5 seconds.

[0077] Signal-to-noise ratio adjustment factor : 0.05.

[0078] The first signal-to-noise ratio threshold : 20 dB.

[0079] If the current signal-to-noise ratio is 23 dB; ; In this way, when the signal-to-noise ratio slightly exceeds the first signal-to-noise ratio threshold, the time window is reduced, and the calculated result is 0.49575 seconds, which is significantly reduced compared to the second initial window of 1.5 seconds. This shows that when the signal-to-noise ratio increases, in order to make good use of signal conditions, the frequency of data updates is increased to adapt to the dynamically changing environment. Similarly, the method of the first scene mode can also be referred to to set the minimum time window value, which will not be repeated here.

[0080] As an optional implementation, when the drone is in the third scene mode, the time window size is set The formula is: ; in is the signal strength, is the third initial time window size, is the signal strength adjustment factor, is the first signal threshold, is the rate of change of signal strength.

[0081] As an optional implementation, when the drone is in the third scene mode, the time window size is set The formula is: ; in, is the signal strength adjustment factor, is the second signal threshold.

[0082] In a specific implementation, in addition to the signal strength exceeding the first signal threshold, the speed at which the signal strength increases may also be considered. If the signal strength increases rapidly, it means that the drone may be approaching the signal source, and the time window may be more actively reduced to increase the data update frequency.

[0083] Similarly, when the signal strength is lower than the second signal threshold and the signal strength is decreasing, it means that the drone may be leaving the area with a stronger signal. At this time, the time window should be increased to prevent data loss.

[0084] In this way, the introduction of the change rate of signal strength makes the adjustment of the time window more sensitive and timely, and can better adapt to the actual flight status of the drone in a complex environment. By monitoring the rapid changes in signal strength, the data update frequency can be controlled more accurately, ensuring that data updates are accelerated when the signal is good, and slowed down when the signal is poor to reduce data loss.

[0085] With respect to the above S103: real-time flight information of the UAV is obtained, and the signal switching point of the UAV within the coverage range of the multiple target signal towers is determined by using the hybrid neural network.

[0086] As a possible implementation, the output of the hybrid neural network includes: signal quality score, signal attenuation prediction, and interference level assessment; the determination of the signal switching point of the drone within the coverage range of multiple target signal towers includes: Using the first dynamic planning scheme, determining the signal switching point of the UAV within the coverage area of ​​multiple target signal towers; The first dynamic programming scheme includes: state definition, decision variables, transfer equations, objective function, initialization and boundary conditions, and dynamic programming solution.

[0087] In a specific implementation, when the neural network receives real-time flight data, such as flight speed, signal strength, environmental noise, etc., it will immediately process and predict the next signal conditions or other flight-related parameters. Based on these predictions, dynamic decisions are made, such as adjusting the size of the time window. If the predicted signal will deteriorate, the system can increase the time window to reduce the frequency of data updates and avoid excessive data transmission when the signal is unstable; on the contrary, if the predicted signal is stable, the system can reduce the time window and increase the frequency of data transmission to improve the real-time performance of data processing. In addition, this kind of dynamic adjustment mechanism can also be combined with the three scenarios described above. For example, if there is a prediction of speed in the hybrid network, then the predicted speed of the hybrid network can be combined with the time window adjustment method in the speed scenario mentioned above. The same is true for the signal-to-noise ratio and signal strength, which will not be repeated here.

[0088] After the adjustment decision is implemented, the new operating parameters (such as the adjusted time window size) will affect the subsequent data collection and processing. These newly collected data are sent back to the neural network for analysis again. In this way, the network continuously receives new inputs and adjusts its outputs to form a closed-loop feedback system. This feedback mechanism ensures that the UAV system can continue to learn and adapt to environmental changes and optimize its flight and data transmission strategies. For example, if the actual signal quality does not match the prediction, the system can further adjust the time window according to the actual situation to ensure the stability and accuracy of data transmission.

[0089] For example, when the drone is performing a city monitoring mission, the mode can be set to the third scene mode, and the parameters are set as follows: basic time window , signal strength adjustment factor and , signal strength threshold and .

[0090] Drones collect real-time signal strength , according to the formula: ; ; Calculate the new time window size. Substitute the set parameters into the calculation to obtain a new time window size to cope with changes in signal strength.

[0091] The adjusted time window size is used for subsequent data collection cycles. If the signal strength improves in the next cycle, the time window will be recalculated and appropriately reduced to increase the data update frequency and improve the real-time monitoring effect.

[0092] In the specific implementation, the reason why this application adopts the dynamic programming scheme is that although the hybrid neural network is powerful and can predict and evaluate signal quality, signal attenuation and interference level, it does not directly decide on specific operation steps, but provides decision-making support data. The dynamic programming scheme can combine these prediction information and calculate the optimal signal switching strategy through a series of optimization algorithms. Dynamic programming provides a systematic approach to deal with decision-making problems, especially in complex environments involving multi-step decisions and multi-state considerations. It can ensure that throughout the flight mission, the behavior of the drone is always moving towards the optimization goal (such as minimizing signal switching loss and maximizing signal quality). The output of the hybrid neural network serves as the key parameter of the dynamic programming input, helping to define state transitions and evaluate possible decision effects, so that the entire system can not only utilize the powerful predictive ability of the neural network, but also achieve accurate optimization decisions through dynamic programming.

[0093] Among them, the state definition is used to define the connection status between the drone and each signal tower; the decision variable is used to determine whether to switch from one signal tower to another; the transfer equation is used to describe the conversion logic and conditions between states; the objective function is used to set the optimization goal, such as maximizing signal quality or minimizing switching cost; initialization and boundary adjustment are used to set the initial state and calculate boundary adjustment; dynamic programming is used to solve the switching problem of signal towers.

[0094] As an optional implementation, the state definition includes: defining the state ,in, Indicates at a point in time , the drone connects to the signal tower situation; The decision variables include: ,in, For at time point When the drone flies from the signal tower Switch to the signal tower The decision variables are is a binary variable, when switching occurs , when no switching occurs ; The transfer equation S(t+1,j ) is: ; in, and is the index of the signal tower, indicating the signal tower that the drone is currently connected to. From the signal tower Switch to the signal tower The cost includes: signal quality difference information, predicted signal attenuation information and signal interference information; It's at the time When the drone is at the signal tower Signal quality score; is the collection of all signal towers; The optimization goal is to maximize signal quality and minimize switching cost during the entire flight; The objective function is used to maximize the signal stability at the end of the flight and minimize the total switching cost. The objective function is: ; Where T is the final time point when the flight ends, and j is the index of the signal tower; The initialization and boundary conditions include: setting the initial state based on the signal tower when the drone starts During the flight, a signal tower must be connected at each time point, namely: ; Where T is the final time point at the end of the flight, For each time point during the flight, is the index of the signal tower; The dynamic programming solution includes: from Initially, the optimal values ​​of all possible current states are calculated based on the previous state and the current signal quality and transfer cost; Recursive calculation until , determine the signal switching points from the start to the end by backtracking.

[0095] In a specific implementation, a hybrid neural network provides a signal quality score , and predict future signal conditions. These outputs are used as input parameters for dynamic programming. According to the changes in real-time flight and signal status, the input data of the neural network is updated, the signal quality and transfer cost are re-evaluated, and the subsequent decision path is optimized.

[0096] Among them, the switching cost The calculation of can be determined based on multiple factors, including: When the drone switches from one signal tower to another, it may experience a brief interruption in connectivity, and this time delay can be converted into a cost. During the handover process, if the data transmission is not completed, there may be data loss, and this risk can also be quantified as a cost. Signal switching may require additional processing and energy expenditure, especially in the search for a new signal tower in a weak signal area.

[0097] The specific cost calculation method can be the weighted sum of these factors, and the weight distribution depends on the specific needs of the task and the operating environment. For example, if the task is extremely time-sensitive, the weight of time delay can be higher; if the task has extremely high requirements for data integrity, the cost of data loss risk can be given more weight. In this way, the switching cost can be reasonably evaluated according to the actual situation, providing quantitative decision support for dynamic programming.

[0098] In dynamic programming, the recursive calculation is performed until , determining the signal switching point from the start to the end by backtracking is a core step. Here, T represents the planned end time, and the whole process includes two main stages: 1. Recursive calculation: Starting from the initial time point, the optimal state at each time point is calculated step by step using the transfer equation. and every possible tower , according to the previous time point The optimal value for each state is calculated based on the state and signal conditions and costs at the current time point. This process continues until the end time T of the task is reached.

[0099] 2. Backtracking process: Once time T is reached, Start tracing back the signal switching points. This step is performed by looking at the optimal decision chosen at each time point, that is, checking which signal tower to switch from and which signal tower provides the best result. By tracing back, it is clear how the drone should switch from one signal tower to another during the entire mission to ensure the maximization of the signal and the minimization of the cost throughout the flight mission.

[0100] For example, a drone performs an emergency rescue monitoring mission in an urban environment.

[0101] First, the drone’s mission is to monitor urban areas with tall buildings, which requires real-time video data transmission. Since the distribution of urban signal towers is uneven and the signal strength is affected by building obstructions, it is necessary to dynamically adjust the signal connection to ensure the quality of data transmission.

[0102] At the beginning of the mission, the estimated time of the monitoring mission is set to 120 minutes, which is defined as T. When the drone takes off, it selects the signal tower with the strongest signal near the starting point as the initial connection tower. Every minute, the system collects the signal quality of the current signal tower. and from the current tower Cost of switching to other towers .

[0103] Based on the output of the hybrid neural network, the dynamic programming scheme is carried out through the following steps: First, define the state Indicates at time Drone connected to a cell tower and set the decision variables To decide whether to Switch to Tower Then, the state is updated according to the above transfer equation, and the optimal switching strategy is selected to maximize the signal quality and minimize the switching cost.

[0104] During the entire flight, the goal of the objective function optimization is to maximize the signal quality and minimize the switching cost, that is, During this process, real-time adjustments are made based on feedback from signal conditions and neural network predictions, ensuring the drone is always connected to the best signal tower.

[0105] In this way, drones can efficiently complete monitoring tasks in complex urban environments, maintain optimal signal connections, and ensure the stability and real-time nature of data transmission. After the mission is completed, the collected data and the records of dynamic adjustments during execution are used to analyze and optimize the dynamic programming model for future missions, improving the overall performance and reliability of the system.

[0106] For example, if a drone is used for urban traffic monitoring, assume there are three signal towers A, B, and C.

[0107] Signal Quality Rating :A=70,B=50,C=80.

[0108] Switching costs : from A to B=10, from A to C=20, from B to A=5, from B to C=15, from C to A=10, from C to B=10.

[0109] Time : Consider the transition from time 1 to time 2.

[0110] Assuming that the drone is connected to signal tower A at the beginning, the state is defined as (a base value), the rest .

[0111] Calculate the optimal strategy for signal switching from A to B and C: , calculated using concrete values: ; The same calculation : ; At time point 2, connecting to tower C provides the highest overall score Therefore, the optimal strategy is to switch from tower A to tower C.

[0112] In this way, the dynamic planning scheme not only takes into account signal quality and switching costs, but also adjusts the drone's connection strategy in real time to ensure optimal communication efficiency and stability in a changing urban environment. This approach allows drones to make the most appropriate signal tower switching decisions based on real-time conditions when performing monitoring tasks.

[0113] Based on the same inventive concept, the embodiment of the present application also provides a 5G-based ultra-low latency data image transmission system corresponding to the 5G-based ultra-low latency data image transmission method. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned 5G-based ultra-low latency data image transmission method, the implementation of the system can refer to the implementation of the method, and the repeated parts will not be repeated.

[0114] Reference Figure 3 As shown, it is a schematic diagram of the structure of an ultra-low latency data image transmission system based on 5G provided in an embodiment of the present application, and the system includes: The acquisition unit 10 is used to acquire historical flight data of the UAV within the coverage area of ​​multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; A construction unit 20 constructs a hybrid neural network based on the preprocessed historical flight data; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; The processing unit 30 is used to obtain real-time flight information of the UAV and determine the signal switching point of the UAV within the coverage range of the multiple target signal towers using the hybrid neural network.

[0115] The present application also provides an electronic device, such as Figure 4 FIG. 1 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, including: A processor 41 and a memory 42; the memory 42 stores machine-readable instructions executable by the processor 41, and the processor 41 is used to execute the machine-readable instructions stored in the memory 42. When the machine-readable instructions are executed by the processor 41, the processor 41 performs the following steps: Acquire historical flight data of the UAV within the coverage area of ​​multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; Based on the preprocessed historical flight data, a hybrid neural network is constructed; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; The real-time flight information of the UAV is obtained, and the signal switching point of the UAV within the coverage range of the multiple target signal towers is determined by using the hybrid neural network.

[0116] The above-mentioned memory 42 includes internal memory 421 and external memory 422; the memory 421 here is also called internal memory, which is used to temporarily store the calculation data in the processor 41 and the data exchanged with the external memory 422 such as the hard disk. The processor 41 exchanges data with the external memory 422 through the internal memory 421.

[0117] The specific execution process of the above instructions can refer to the steps of the 5G-based ultra-low latency data image transmission method described in the embodiment of the present application, which will not be repeated here.

[0118] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the ultra-low latency data image transmission method based on 5G described in the above method embodiment are executed. The storage medium can be a volatile or non-volatile computer-readable storage medium.

[0119] An embodiment of the present application also provides a computer program product, which carries a program code. The instructions included in the program code can be used to execute the steps of the 5G-based ultra-low latency data image transmission method described in the above method embodiment. Please refer to the above method embodiment for details, which will not be repeated here.

[0120] The computer program product may be implemented in hardware, software or a combination thereof. In one optional embodiment, the computer program product is implemented as a computer storage medium. In another optional embodiment, the computer program product is implemented as a software product, such as a software development kit (SDK).

[0121] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0122] It should be understood that determining B based on A does not mean determining B only based on A. B can also be determined based on A and / or other information.

[0123] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or 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 the present invention.

[0124] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0125] The preferred embodiments of the present invention disclosed above are only used to help explain the present application. The preferred embodiments do not describe all the details in detail, nor do they limit the present application to specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can understand and use the present application well. The present application is limited only by the claims and their full scope and equivalents.

Claims

1. The ultra-low latency data image transmission method based on 5G is characterized by: include: Acquire historical flight data of the UAV within the coverage area of ​​multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; Based on the preprocessed historical flight data, a hybrid neural network is constructed; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; The real-time flight information of the UAV is obtained, and the signal switching point of the UAV within the coverage range of the multiple target signal towers is determined by using the hybrid neural network.

2. The ultra-low latency data image transmission method based on 5G according to claim 1, characterized in that: The pre-processing comprises: Performing random phase shift on the quadrature amplitude modulation data to generate target training samples for the hybrid neural network training; The actual delay distribution is calculated based on the historical flight data, and a random delay distribution is generated. The first time series.

3. The ultra-low latency data image transmission method based on 5G according to claim 2, characterized in that: The construction of the hybrid neural network comprises: The spatial feature extraction branch is constructed based on a multi-layer convolutional neural network to extract the spatial position and environmental layout features of the UAV relative to the target signal tower and target environmental obstacles; Among them, each layer of convolutional neural network is connected to a batch normalization layer and a ReLU activation function; A time feature extraction branch is constructed based on a multi-layer long short-term memory network to extract the time dependency and dynamic changes of the historical flight data; wherein the multi-layer long short-term memory network includes: a first time step sliding window method; At least one fully connected layer is used to form a signal attenuation feature extraction branch to analyze the distance and signal strength data from the drone to each signal tower, and predict signal attenuation and potential switching points; each fully connected layer is equipped with a discard layer.

4. The ultra-low latency data image transmission method based on 5G according to claim 3 is characterized in that: The historical flight data also includes: flight speed, network change information; the first time step sliding window method includes: dynamically adjusting the time window size based on the flight speed of the UAV and the network signal change information; The dynamically adjusting the time window size includes: Based on the flight mission of the UAV, a scene mode is determined; the scene mode includes a first scene mode, a second scene mode and a third scene mode; When the drone is in the first scene mode, the flight speed of the drone is detected, and a first initial time window size is set; in response to the flight speed of the drone being greater than or equal to a first speed threshold, the time window size is reduced; in response to the flight speed of the drone being less than a second speed threshold, the time window size is increased; When the drone is in a second scene mode, detecting a signal-to-noise ratio of an environment in which the drone is located, and setting a second initial time window size, in response to the signal-to-noise ratio being greater than or equal to a first signal-to-noise ratio threshold, reducing the time window size, and in response to the signal-to-noise ratio being less than a second signal-to-noise ratio threshold, increasing the time window size; When the drone is in a third scene mode, the signal strength of the drone is detected and a third initial time window size is set. In response to the signal strength being greater than or equal to a first signal threshold, the time window size is reduced. In response to the signal strength being less than a second signal threshold, the time window size is increased.

5. The ultra-low latency data image transmission method based on 5G according to claim 4, characterized in that: The output of the hybrid neural network includes: signal quality score, signal attenuation prediction, and interference level assessment; determining the signal switching point of the drone within the coverage range of the multiple target signal towers includes: Using the first dynamic planning scheme, determining the signal switching point of the UAV within the coverage area of ​​multiple target signal towers; The first dynamic programming scheme includes: state definition, decision variables, transfer equations, objective function, initialization and boundary conditions, and dynamic programming solution.

6. A 5G-based ultra-low latency data image transmission system, which is used to implement the 5G-based ultra-low latency data image transmission method according to any one of claims 1 to 5, characterized in that: include: An acquisition unit is used to acquire historical flight data of the UAV within the coverage area of ​​multiple target signal towers and perform preprocessing; wherein the historical flight data includes: orthogonal amplitude modulation data, access point data, and signal attenuation data; A construction unit constructs a hybrid neural network based on the preprocessed historical flight data; wherein the hybrid neural network includes: a spatial feature extraction branch, a temporal feature extraction branch, and a signal attenuation feature extraction branch; The processing unit is used to obtain real-time flight information of the UAV and determine the signal switching point of the UAV within the coverage range of the multiple target signal towers using the hybrid neural network.

7. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the processor is used to execute the machine-readable instructions stored in the memory. When the machine-readable instructions are executed by the processor, the processor executes the steps of the 5G-based ultra-low latency data image transmission method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: A computer program is stored on the computer-readable storage medium. When the computer program is executed by an electronic device, the electronic device performs the steps of the 5G-based ultra-low latency data image transmission method as described in any one of claims 1 to 5.

Images