CN-119789165-B - Ultralow-delay data graph transmission method and system based on 5G

CN119789165BCN 119789165 BCN119789165 BCN 119789165BCN-119789165-B

Abstract

The application discloses an ultralow-delay data graph transmission method and system based on 5G, and relates to the technical field of unmanned aerial vehicle data transmission. The method comprises the following steps: firstly, acquiring flight data of an unmanned aerial vehicle and preprocessing; constructing a hybrid neural network comprising spatial feature extraction, temporal feature extraction and signal attenuation feature extraction; finally, based on real-time flight information, predicting and determining an optimal signal tower switching point by utilizing the neural network; in addition, the application also provides a technology for dynamically adjusting the size of the data window, and the frequency of data acquisition is dynamically adjusted according to the flying speed and signal change of the unmanned aerial vehicle so as to adapt to different environmental conditions, thereby improving the data transmission efficiency and reducing the delay.

Inventors

YANG YI
JIANG HEJIE
LU JUNHANG
ZENG GANG
BIN LIN
LI YI

Assignees

北方天途航空技术发展（北京）有限公司

Dates

Publication Date: 20260512
Application Date: 20250311

Claims (5)

1. The ultra-low delay data graph transmission method based on 5G is characterized by comprising the following steps: acquiring historical flight data of an unmanned aerial vehicle in coverage areas of a plurality of target signal towers and preprocessing, wherein the historical flight data comprises quadrature amplitude modulation data, access point data, signal attenuation data, flight speed and network change information, and the quadrature amplitude modulation data is an in-phase component and a quadrature component of a wireless link baseband signal; constructing a hybrid neural network based on the preprocessed historical flight data, wherein the hybrid neural network comprises a spatial feature extraction branch, a temporal feature extraction branch and a signal attenuation feature extraction branch; the space feature extraction branch is a multi-layer convolutional neural network, each layer is followed by batch normalization and ReLU activation, and the space feature extraction branch is used for extracting the space position and the environment layout features of the unmanned aerial vehicle relative to the target signal tower and the target environment barrier; The time feature extraction branch is a multi-layer long-short-term memory network and comprises a time step sliding window method, and the size of a time window is dynamically adjusted according to three scene modes corresponding to the unmanned aerial vehicle flight speed, the environment signal-to-noise ratio and the signal strength and a preset threshold rule; The signal attenuation characteristic extraction branch consists of at least one full-connection layer, and a discarding layer is arranged behind each layer and is used for extracting and analyzing the distance from the unmanned aerial vehicle to each signal tower and the signal intensity data, and predicting signal attenuation and potential switching points; Acquiring real-time flight information of the unmanned aerial vehicle, and determining signal switching points of the unmanned aerial vehicle in coverage areas of the target signal towers by utilizing the hybrid neural network; the construction of the hybrid neural network comprises the steps of preprocessing the historical flight data, and inputting the spatial feature extraction branch, the temporal feature extraction branch and the signal attenuation feature extraction branch; the pretreatment comprises the following steps: Performing random phase shift on the quadrature amplitude modulation data to generate a target training sample for training the hybrid neural network; Calculating an actual delay profile based on the historical flight data, generating a signal containing random delays Is a first time sequence of (a); the method comprises the steps of obtaining real-time flight information of the unmanned aerial vehicle, obtaining a prediction result at least comprising signal quality score, signal attenuation prediction and interference level evaluation by utilizing the hybrid neural network, solving according to a first dynamic programming scheme based on state definition, decision variables, transfer equations, objective functions, initialization and boundary conditions based on the prediction result, the distance from the unmanned aerial vehicle to each signal tower, signal intensity and environment layout information, and determining an optimal signal switching point sequence of the unmanned aerial vehicle in the coverage range of a plurality of target signal towers with the aim of maximizing signal quality and minimizing switching cost.
2. The 5G-based ultra-low latency data graph transmission method according to claim 1, wherein the dynamically adjusting the time window size comprises: Determining a scene mode based on a flight mission of the unmanned aerial vehicle, wherein the scene mode comprises a first scene mode, a second scene mode and a third scene mode; When the unmanned aerial vehicle is in a first scene mode, detecting the flight speed of the unmanned aerial vehicle, and setting a first initial time window size; the size of the time window is reduced in response to the flight speed of the unmanned aerial vehicle being greater than or equal to a first speed threshold; When the unmanned aerial vehicle is in a second scene mode, detecting the signal-to-noise ratio of the environment in which the unmanned aerial vehicle is located, 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 signal-to-noise ratio threshold, and increasing the time window size in response to the signal-to-noise ratio being less than a second signal-to-noise ratio threshold; When the unmanned aerial vehicle is in a third scene mode, detecting the signal intensity of the unmanned aerial vehicle, setting a third initial time window size, reducing the time window size in response to the signal intensity being greater than or equal to a first signal threshold, and increasing the time window size in response to the signal intensity being less than a second signal threshold.
3. A 5G-based ultra-low latency data graph transmission system for implementing the 5G-based ultra-low latency data graph transmission method of claim 1 or 2, comprising: the system comprises an acquisition unit, a preprocessing unit and a preprocessing unit, wherein the acquisition unit is used for acquiring historical flight data of the unmanned aerial vehicle in the coverage range of a plurality of target signal towers and preprocessing the historical flight data, wherein the historical flight data comprises quadrature amplitude modulation data, access point data, signal attenuation data, flight speed and network change information, and the quadrature amplitude modulation data is an in-phase component and a quadrature component of a wireless link baseband signal; The system comprises a construction unit, a mixed neural network, a signal attenuation characteristic extraction unit and a control unit, wherein the mixed neural network is used for constructing the mixed neural network based on the preprocessed historical flight data, and comprises a spatial characteristic extraction branch, a time characteristic extraction branch and a signal attenuation characteristic extraction branch, wherein the spatial characteristic extraction branch is a multi-layer convolutional neural network, each layer is followed by batch normalization and ReLU activation and is used for extracting the spatial position and the environmental layout characteristic of the unmanned aerial vehicle relative to the target signal tower and the target environmental obstacle; The time feature extraction branch is a multi-layer long-short-term memory network and comprises a time step sliding window method, and the size of a time window is dynamically adjusted according to three scene modes corresponding to the unmanned aerial vehicle flight speed, the environment signal-to-noise ratio and the signal strength and a preset threshold rule; The signal attenuation characteristic extraction branch consists of at least one full-connection layer, and a discarding layer is arranged behind each layer and is used for extracting and analyzing the distance from the unmanned aerial vehicle to each signal tower and the signal intensity data, and predicting signal attenuation and potential switching points; the pretreatment comprises the following steps: Performing random phase shift on the quadrature amplitude modulation data to generate a target training sample for training the hybrid neural network; Calculating an actual delay profile based on the historical flight data, generating a signal containing random delays Is a first time sequence of (a); The processing unit is used for acquiring real-time flight information of the unmanned aerial vehicle and determining signal switching points of the unmanned aerial vehicle in the coverage range of the target signal towers by utilizing the hybrid neural network; The processing unit is further configured to obtain real-time flight information of the unmanned aerial vehicle, obtain a prediction result including at least signal quality score, signal attenuation prediction and interference level evaluation by using the hybrid neural network, and determine an optimal signal switching point sequence of the unmanned aerial vehicle in coverage of a plurality of target signal towers based on the prediction result, distances from the unmanned aerial vehicle to each signal tower, signal intensity and environmental layout information, and according to a first dynamic programming scheme, based on state definition, decision variables, transfer equations, objective functions, initialization and boundary conditions, with the aim of maximizing signal quality and minimizing switching cost.
4. An electronic device comprising a processor, a memory storing machine-readable instructions executable by the processor, the processor configured to execute the machine-readable instructions stored in the memory, the machine-readable instructions when executed by the processor, the processor configured to perform the steps of the 5G-based ultra-low latency data graph transfer method of claims 1 or 2.
5. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when being executed by an electronic device, performs the steps of the 5G based ultra low latency data graph transmission method according to claim 1 or 2.

Description

Ultralow-delay data graph transmission method and system based on 5G Technical Field The application relates to the technical field of unmanned aerial vehicle data transmission, in particular to an ultralow-delay data graph transmission method and system based on 5G. Background Along with the rapid expansion of the application field of unmanned aerial vehicles, especially in the aspects of urban monitoring, traffic management, emergency rescue and the like, the real-time data transmission requirements for unmanned aerial vehicles are increasing. Especially in urban environments, buildings are dense and signal towers are unevenly distributed, which makes unmanned aerial vehicles face challenges of unstable signal coverage and delay of data transmission when performing tasks. Traditional data transmission methods often rely on static signal tower connection strategies, and lack the ability to dynamically adjust according to real-time environments, which limits the efficiency of unmanned aerial vehicle applications in complex environments. To address these problems, the use of 5G communication is a viable solution because 5G technology provides higher data transmission rates and lower delays. However, even 5G technology may suffer from signal instability in high dynamic environments, especially when unmanned aerial vehicles frequently shuttle high-rise buildings and multiple 5G signal base stations. For example, china patent with publication number CN114885304A discloses a 5G network-connected unmanned aerial vehicle data transmission system and method based on FlexE technology, wherein the system comprises an unmanned aerial vehicle terminal, a 5G base station and an unmanned aerial vehicle cloud platform, the unmanned aerial vehicle terminal is used for tagging data packets to be transmitted according to categories and queuing the tagged data packets, the 5G base station takes FlexE as a bearing network to provide a special data path with enough bandwidth and deterministic delay for the tagged data packets, the data packets are sent to a core network, the core network sends the data packets to the unmanned aerial vehicle cloud platform, and the unmanned aerial vehicle cloud platform is used for receiving data transmitted by the unmanned aerial vehicle terminal, processing the received data and receiving an operation instruction for the unmanned aerial vehicle terminal. Based on the system, the invention also provides a data transmission method of the 5G network-connected unmanned aerial vehicle based on FlexE technology. Delay can be reduced, energy can be saved, cost can be reduced, system capacity can be increased, and large-scale equipment can be connected. The method has the problems in the background technology, so that the method capable of responding to the environmental change in real time and intelligently managing the switching of the signal tower is developed, and is important to improving the operation efficiency and the reliability of data transmission of the unmanned aerial vehicle. Disclosure of Invention Aiming at the defects of the prior art, the application aims to provide an ultralow-delay data graph transmission method and system based on 5G, which can dynamically adjust signal switching points of an unmanned aerial vehicle when the unmanned aerial vehicle shuttles to each signal tower according to the characteristics of short delay under 5G signals and limited switching of signal tower coverage, so as to improve the transmission quality of data graph transmission tasks, and specifically adopts the following scheme: In a first aspect, the present application provides a 5G-based ultralow-latency data graph transmission method, including: Acquiring historical flight data of the unmanned aerial vehicle in the coverage range of a plurality of target signal towers and preprocessing, wherein the historical flight data comprises quadrature amplitude modulation data, access point data and signal attenuation data; constructing a hybrid neural network based on the preprocessed historical flight data, wherein the hybrid neural network comprises a spatial feature extraction branch, a temporal feature extraction branch and a signal attenuation feature extraction branch; And acquiring real-time flight information of the unmanned aerial vehicle, and determining signal switching points of the unmanned aerial vehicle in the coverage range of the target signal towers by utilizing the hybrid neural network. As an alternative embodiment, the preprocessing includes: Performing random phase shift on the quadrature amplitude modulation data to generate a target training sample for training the hybrid neural network; Calculating an actual delay profile based on the historical flight data, generating a signal containing random delays Is a time series of the first time series of (a). As an alternative embodiment, the constructing the hybrid neural network includes: Constructing the spatial feature extraction branc