CN117320024A - Low-altitude network coverage optimization method based on digital twinning - Google Patents

Abstract

The invention discloses a low-altitude network coverage optimization method based on digital twinning, which comprises the following steps: s1, obtaining a base station antenna pattern through simulation, constructing a base station antenna radiation model, and reconstructing a three-dimensional environment model through unmanned aerial vehicle aerial photography; s2, combining a three-dimensional environment model and a base station antenna radiation model to construct a digital twin body covered by a low-altitude network; s3, base station deployment planning, configuration optimization and low-altitude network coverage scheme are carried out based on the digital twin body. The invention introduces the wave beam optimizing technology for network coverage and optimization thereof, and improves the benefit of network coverage; and the network coverage of the low-altitude airspace and the ground is jointly optimized, so that the development requirement of the low-altitude airspace communication network is met.

Description

A digital twin-based low-altitude network coverage optimization method

Technical field

The present invention relates to low-altitude network coverage optimization, and in particular to a digital twin-based low-altitude network coverage optimization method.

Background technique

Base stations are widely used in communication systems. Most of the existing base station coverage methods use the side lobes of ground base station antennas for coverage. They do not apply beam optimization technology, and are deficient in optimizing the efficiency and effect of coverage. In addition, the existing coverage is only for ground communication networks, and there is no overall optimization for communication network needs in low-altitude airspace.

Contents of the invention

The purpose of the present invention is to overcome the shortcomings of the existing technology and provide a low-altitude network coverage optimization method based on digital twins.

The object of the present invention is achieved through the following technical solution: a low-altitude network coverage optimization method based on digital twins, including the following steps:

S1. Obtain the base station antenna pattern through simulation, use this to construct the base station antenna radiation model, and reconstruct the three-dimensional environment model through drone aerial photography;

S2. Combine the three-dimensional environment model and base station antenna radiation model to build a digital twin for low-altitude network coverage;

S3. Based on the digital twin, base station deployment planning and configuration optimization are carried out to obtain a low-altitude network coverage solution.

The beneficial effects of the present invention are: the present invention introduces beam optimization technology for network coverage and optimization to improve the efficiency of network coverage; it jointly optimizes network coverage in low-altitude airspace and ground to meet the needs of the development of low-altitude airspace communication networks.

Description of drawings

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

Figure 2 is a schematic diagram of a 4×8 dual polarization array antenna;

Figure 3 is a schematic diagram of the constructed Cartesian coordinate system and three-dimensional polar coordinate system;

Figure 4 is a schematic diagram of the polar coordinate model centered on the base station antenna;

Figure 5 shows the polar coordinate model centered on the UAV receiving antenna.

Detailed ways

The technical solution of the present invention will be described in further detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following description.

As shown in Figure 1, a low-altitude network coverage optimization method based on digital twins is characterized by: including the following steps:

S1. Obtain the base station antenna pattern through simulation, use this to construct the base station antenna radiation model, and reconstruct the three-dimensional environment model through drone aerial photography;

In the embodiment of this application, according to the common design of MIMO antennas, we simulate a 4×8 dual-polarized array antenna arranged as shown in Figure 2. In this array antenna, there are two half-wave dipole antennas placed vertically at each position, that is, ±45° polarization.

S101. Establish the Cartesian coordinate system and the three-dimensional polar coordinate system as shown in Figure 3. The angle between the vector of the three-dimensional polar coordinate system and the positive direction of the Z-axis is the polar angle θ, and the angle between the projection of the vector on the XOY plane and the positive direction of the X-axis As the azimuth angle φ; place the MIMO antenna panel on the YOZ plane, at this time φ′=0, simulate the antenna at the current angle and obtain the initial gain V(θ′,φ′) of the array antenna;

_{The MIMO antenna} adopts _{an M z}

According to the MIMO antenna array unit arrangement, the array steering vector corresponding to the (θ,φ) direction is calculated by the following formula:

Among them, v _y (θ,φ) and v _z (θ,φ) represent the antenna unit steering vectors in the Y-axis and Z-axis directions respectively, d _y and d _z are the antenna unit spacing in the horizontal and vertical directions, λ is the wavelength, And a(θ,φ) is the array steering vector of the entire antenna array;

For a dual-polarized array antenna, the array steering vectors of M _z ×M _y antenna units corresponding to each polarization direction are:

a ₁ (θ,φ)＝|f ₁ (θ,φ)|a(θ,φ)

a ₂ (θ,φ)＝|f ₂ (θ,φ)|a(θ,φ)

Among them, f ₁ (θ, φ) is the gain amplitude of the antenna element in the first polarization direction in the (θ, φ) direction, f ₂ (θ, φ) is the gain amplitude of the antenna in the second polarization direction in the (θ, φ) direction. ,φ) direction gain amplitude, a ₁ (θ,φ) and a ₂ (θ,φ) are the array steering vectors corresponding to the two polarization directions;

Based on the channel state matrix H(θ,φ) and the antenna initial gain V(θ′,φ′), calculate the received signal strength in the (θ,φ) direction, that is, the antenna gain S(θ,φ):

S(θ,φ)＝|H(θ,φ) ^T V(θ′,φ′)| ²

Because each set of angles (θ, φ) corresponds to an antenna gain S (θ, φ), then the antenna gain values of all angles (θ, φ) constitute a picture, and this picture is recorded as the antenna pattern. S _gain , use this antenna pattern as the base station antenna radiation model;

In the embodiment of this application, a drone is used to conduct actual measurement and verification of the antenna pattern. The measurement principle of the antenna pattern ((based on Friis Transmission Formula)) is

Where R is the distance between Tx-Rx, P _R is the received power, P _T is the transmit power, the antenna gain to be measured S _T , the receiving antenna gain S _R ; the distance R between Tx and Rx is measured, the receiving antenna gain S _R , The received power P _R and the transmitted power P _T are used to calculate the antenna gain S _T . These four quantities are also sources of errors, including:

First, construct a polar coordinate model with the base station antenna as the center, as shown in Figure 4. The five-pointed star is the base station and the circle is the receiving antenna. The horizontal plane is the direction plane, the north direction is the direction angle of 0 degrees; the vertical plane is the direction angle. The inclination plane points to the ground with an inclination angle of 0 degrees. Any position in space depends on the direction angle. represented by the inclination angle θ ₁ and the relative distance R

Then, a polar coordinate model with the UAV receiving antenna as the center is constructed. As shown in Figure 5, the five-pointed star is the base station and the cylinder is the receiving antenna; the radial surface is the direction surface and the forward direction is the direction angle of 0 degrees; The longitudinal plane perpendicular to the radial plane is the inclination plane. Any position in space according to the direction angle The inclination angle θ ₂ and the relative distance R are expressed; (the radial plane, longitudinal plane, 0-degree inclination angle, and 0-degree direction angle need to be determined according to the actual antenna.

According to the radio transmission model, the power of the signal received by the antenna The direction angle of the UAV receiving antenna relative to the base station/> Inclination angle θ ₁ , the direction angle of the base station relative to the UAV receiving antenna/> Determined by the inclination angle θ ₂ and the relative distance R between the two;

The antenna gain unit in the above formula is dB, and the power unit is dBm, where P _T is the signal power sent by the base station; For the base station antenna/> Gain in direction;/> is the path loss;/> Receive antenna for drone/> The gain in the direction; w (θ ₁ , d) is the noise at the base station tilt angle θ ₁ and distance R.

In order to measure the gain S _T of the base station antenna in each direction, the pattern S _gain of the base station antenna is formed. During measurement, keep unchanged, measure the direction angle of the receiving antenna relative to the base station/> The received signal power at the inclination angle θ ₁ , and try to eliminate the error caused by the noise w (θ ₁ , d), such as the drone hovering at the same position for multiple measurements to reduce the impact of the error, or the unmanned Calibrate the aircraft before taking off to eliminate errors.

S102. Obtain the necessary information of the area where twins need to be established through the sensing device, and establish a three-dimensional map, that is, a three-dimensional model. Through data processing, for example, using the grid method to downsample the three-dimensional model, by removing some grid vertices and Patches are used to reduce the resolution of the model, such as merging adjacent vertices to reduce the number of vertices. Finally, an environmental model is obtained that has the same building outline as the real three-dimensional environment and does not affect the electromagnetic simulation effect.

S2. Combine the three-dimensional environment model and base station antenna radiation model to build a digital twin for low-altitude network coverage;

The step S2 includes:

The base station antenna pattern obtained by S101 and the three-dimensional environment model obtained by S102, in which the base station coordinates and base station parameters are set according to the real data, and the ray tracing method is used to simulate the base station antenna in the real working environment, consisting of buildings, Indoor compartments, or shadowing or multi-path effects caused by terrain, obtain a digital twin of low-altitude network coverage that is close to reality;

Through the simulation of the digital twin, the effective signal strength, environmental noise, and interference of other base stations on the current location signal at each location under the three-dimensional environment model are obtained.

In the embodiment of this application, UAV actual network coverage is measured

For a test area of length L and width W, an S-shaped trajectory is used to test its coverage with an accuracy of 10m at a height of 60-120m from the base station antenna level. The coverage performance is calculated and evaluated by the coverage performance evaluation function based on the measured data. , compare the measured network coverage with the digital twin established in S201 to check the accuracy of the digital twin.

In the embodiment of this application, a digital twin visualization interface can also be established

Create a user-friendly interface for visualizing low-altitude network coverage digital twins;

The interface allows users to monitor, control or analyze the behavior of the digital twin. After the visualization interface is created, its performance is continuously monitored, real-time data is collected from real base stations to update and improve the model, while the digital twin provides feedback to help Improve relevant performance parameters and decision-making processes.

S3. Based on the digital twin, base station deployment planning and configuration optimization are carried out to obtain a low-altitude network coverage solution.

S301 uses the coverage performance evaluation function to give the measurement results of the current coverage quality:

The coverage performance of three-dimensional space is estimated by the following formula:

Among them, S(x) represents the coverage index at the x position, P _s represents the effective signal strength, N ₀ represents the environmental noise, and P _b represents the interference of other base stations on the signal.

S302 combines the digital twin and user coverage requirements to design an optimization algorithm to obtain the optimal antenna parameter combination.

The coverage requirements are: provide a target track in three-dimensional space, and optimize the coverage performance on the track by adjusting base station parameters. The base station parameters include antenna angle, downtilt angle, beam forming and transmit power;

Base station parameters are adjusted based on the digital twin. The specific adjustment method includes reinforcement learning algorithms. During each interaction process of reinforcement learning, the action is to adjust the parameters of each base station; the status is the current parameters of each base station; the reward is the trajectory. The average increase in coverage performance at several points.

The goals are:

In the process of reinforcement learning, two networks, the policy network and the value network, are used for learning:

Among them, the policy network is used to interact with the environment and learn policies under the guidance of the value function. The value network is responsible for learning a value function using the data set collected by the interaction between the policy network and the environment, helping the policy network to update the policy, and the gradient of the objective function There is a trajectory return, which is used to update the strategy;

The loss function of the value function is defined in the value network, and the value network parameters are updated through the gradient ascent method; in each round of interaction, the current strategy is first sampled, the gradient of the value function is calculated, the value network parameters are updated, and then Under the guidance of the new value function, update the parameters of the policy network;

After multiple rounds of interactions, when the objective function no longer increases, learning is stopped and the setting parameters of each base station are recorded at this time, thereby obtaining the optimal low-altitude network coverage effect on a given trajectory.

The above description shows and describes a preferred embodiment of the present invention, but as mentioned above, it should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various embodiments. other combinations, modifications and environments, and can be modified through the above teachings or technology or knowledge in related fields within the scope of the inventive concept described herein. Any modifications and changes made by those skilled in the art that do not depart from the spirit and scope of the present invention shall be within the protection scope of the appended claims of the present invention.

Claims (4)

1. A low-altitude network coverage optimization method based on digital twinning is characterized in that: the method comprises the following steps:

s1, obtaining a base station antenna pattern through simulation, constructing a base station antenna radiation model, and reconstructing a three-dimensional environment model through unmanned aerial vehicle aerial photography;

s2, combining a three-dimensional environment model and a base station antenna radiation model to construct a digital twin body covered by a low-altitude network;

s3, base station deployment planning, configuration optimization and low-altitude network coverage scheme are carried out based on the digital twin body.

2. The low-altitude network coverage optimization method based on digital twinning according to claim 1, wherein the method comprises the following steps: the step S1 includes:

s101, establishing a Cartesian coordinate system and a three-dimensional polar coordinate system, wherein an included angle between a vector of the three-dimensional polar coordinate system and a positive Z-axis direction is formedAs a polar angle theta, the angle between the projection of the vector on the XOY plane and the positive direction of the X axis is used as an azimuth angle phi; placing the MIMO antenna panel in the YOZ plane at this timePhi ' =0, performing analog simulation on the antenna at the current angle to obtain the initial gain V (theta ', phi ') of the array antenna;

according to the arrangement of the MIMO antenna array units, the array steering vector corresponding to the (theta, phi) direction is calculated by the following formula:

v _y (θ，φ)＝f(M _y ，θ，φ，λ，dy)

v _z (θ，φ)＝f(M _z ，θ，φ，λ，dy)

wherein v is _y (θ, φ) and v _z (θ, φ) represents the antenna element steering vectors in the Y-axis and Z-axis directions, M _y And M _z The number of antennas in the Y-axis and Z-axis directions, d _y And d _z The space between antenna units in the horizontal and vertical directions is lambda is wavelength, and a (theta, phi) is the array guide vector of the whole antenna array; f () is a generic steering vector calculation function;

if the MIMO antenna is a dual polarized array antenna, M corresponds to each polarization direction _z ×M _y The array steering vectors for the individual antenna elements are:

a ₁ (θ，φ)＝|f ₁ (θ，φ)|a(θ，φ)

a ₂ (θ，φ)＝|f ₂ (θ，φ)|a(θ，φ)

wherein f ₁ (theta, phi) is the gain amplitude of the antenna element in the (theta, phi) direction of the first polarization direction, f ₂ (theta, phi) is the gain amplitude of the antenna element in the (theta, phi) direction of the second polarization direction, a ₁ (θ, φ) and a ₂ (θ, φ) is an array steering vector corresponding to two polarization directions;

based on the channel state matrix H (θ, Φ) and the antenna initial gain V (θ', Φ), the received signal strength in the (θ, Φ) direction, i.e., the antenna gain S (θ, Φ), is calculated:

S(θ，φ)＝|H(θ，φ) ^T V(θ′，φ′)| ²

because each set of angles (θ, φ) corresponds to an antenna gain S (θ, φ), the antenna gain values for all angles (θ, φ) form a picture, which is referred to as an antenna pattern S _gain Taking the antenna pattern as a base station antenna radiation model;

s102, acquiring necessary information of an area needing to be built with a twin body through sensing equipment, building a three-dimensional map, namely a three-dimensional model, and finally obtaining an environment model with the same building outline as a real three-dimensional environment and without affecting electromagnetic simulation effect through data processing.

3. The low-altitude network coverage optimization method based on digital twinning according to claim 1, wherein the method comprises the following steps: the step S2 includes:

the base station antenna pattern obtained in the step S101 and the three-dimensional environment model obtained in the step S102 are used for setting base station coordinates and base station parameters thereof according to real data, and simulation is carried out through a ray tracing method, so that shielding or multipath effects caused by buildings, indoor compartments or topography of the base station antenna in a real working environment are simulated, and a low-altitude network coverage digital twin body close to reality is obtained;

and acquiring the effective signal intensity, the environmental noise and the interference of other base stations on the current position signal of each position under the three-dimensional environment model through the simulation of the digital twin body.

4. The low-altitude network coverage optimization method based on digital twinning according to claim 1, wherein the method comprises the following steps: the step S3 includes:

s301, using a coverage performance evaluation function to give a measurement result of current coverage quality:

the coverage performance of the three-dimensional space is measured by the following formula:

wherein S (x) represents an overlay index of the x position, P _s Representing the effective signal strength, N ₀ Representing ambient noise, including background noise and interference from other environments, P _b Indicating interference of other base stations to the signal;

s302, combining the digital twin body and the coverage requirement of a user, and designing an optimization algorithm to obtain an optimal base station parameter combination;

the coverage requirements are: providing a target track in a three-dimensional space, and enabling the track to reach the best coverage performance by adjusting base station parameters, wherein the base station parameters comprise an antenna angle, a downward inclination angle, an initial gain of an antenna and a transmitting power;

the base station parameters are adjusted based on the digital twin body, the specific adjustment method comprises a reinforcement learning algorithm, and each base station parameter is adjusted by acting in each interaction process of reinforcement learning; the state is the current parameters of each base station; rewarding as trackThe average value of the coverage performance of a plurality of points is increased.

The targets are as follows:

in the reinforcement learning process, two networks, namely a strategy network and a value network, are adopted for learning:

the system comprises a strategy network, a value network, a target function and a data set, wherein the strategy network is used for interacting with an environment, learning strategies under the guidance of the value function, the value network is responsible for learning a value function by using a data set collected by interaction of the strategy network and the environment, helping the strategy network to update the strategy, and a track report exists in the gradient of the target function and is used for updating the strategy;

defining a loss function of a cost function in a value network, and updating the value network parameters by a gradient rising method; in each round of interaction, sampling the current strategy, calculating the gradient of the bid value function, updating the value network parameters, and updating the parameters of the strategy network under the guidance of the new value function;

after the multi-round interaction is carried out, when the objective function is not increased any more, the learning is stopped, and the setting parameters of each base station at the moment are recorded, so that the optimized low-altitude network coverage effect on the given track is obtained.