A UAV ISAC channel estimation method based on multimodal fusion

Abstract

The present invention discloses a UAV ISAC channel estimation method based on multimodal fusion, comprising the following steps: first, constructing a UAV ISAC base station system; second, establishing a state equation and an observation equation of the UAV; third, using an unscented Kalman filter to obtain the UAV first t The optimal state vector at each moment; Fourth, establish the ISAC channel model between the UAV and the ground terminal. The invention is reasonably designed, and the optimal state vector of the UAV is obtained based on the unscented Kalman filter under the multi-modal fusion of the inertial measurement unit IMU and the RTK mobile station to establish the channel model between the UAV and the ground terminal, which reduces the complexity of communication, and considers the trajectory and attitude changes of the UAV, thereby improving the accuracy of channel estimation.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle ISAC, and in particular relates to a unmanned aerial vehicle ISAC channel estimation method based on multi-modal fusion. Background Art

[0002] ISAC technology refers to the technology that perception and communication share the same frequency band and hardware. This technology is considered to be one of the important candidate technologies for the next generation of mobile communications. ISAC technology also integrates a variety of sensors and wireless communication systems to achieve real-time monitoring and control of the physical environment, providing strong support for building an interconnected, intelligent and collaborative digital ecosystem. At the same time, as an important part of the low-altitude economy, drones have become an important platform for realizing the next generation of mobile communication ISAC technology. They can cover a wider range and have more application scenarios. They have received more and more attention in the field of communications. Combining communication perception integration with drone communications can improve its communication performance in multiple dimensions.

[0003] However, the communication and perception performance of UAVs are affected by changes in their posture and motion during flight, but most studies have not comprehensively considered the impact of changes in UAV trajectories and postures on their communication quality.

[0004] Therefore, a reasonably designed UAV ISAC channel estimation method based on multimodal fusion is needed. The optimal state vector of the UAV is obtained based on the unscented Kalman filter under the multimodal fusion of the inertial measurement unit IMU and the RTK mobile station to establish a channel model between the UAV and the ground terminal, which reduces the complexity of communication and considers the changes in the trajectory and attitude of the UAV, thereby improving the accuracy of channel estimation. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a UAV ISAC channel estimation method based on multimodal fusion in view of the deficiencies in the above-mentioned prior art. The method has simple steps and reasonable design. The optimal state vector of the UAV is obtained based on unscented Kalman filtering under the multimodal fusion of the inertial measurement unit IMU and the RTK mobile station to establish a channel model between the UAV and the ground terminal, thereby reducing the complexity of communication and considering the trajectory and attitude changes of the UAV to improve the accuracy of channel estimation.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a UAV ISAC channel estimation method based on multimodal fusion, the method comprising the following steps:

[0007] Step 1: Build the drone ISAC base station system:

[0008] Step 101: The drone ISAC base station system includes a drone ISAC base station and a ground terminal. The drone ISAC base station includes a drone and an ISAC base station integrated on the drone. The bottom of the drone is installed The uniform planar array antenna, the ground terminal is installed Uniform planar array antenna; Indicates the number of rows of uniform planar array antennas at the bottom of the drone, Represents the number of columns of uniform planar array antenna at the bottom of the drone, represents the number of rows of the uniform planar array antenna of the ground terminal, Indicates the number of columns of the uniform planar array antenna of the ground terminal; wherein the drone is equipped with an inertial measurement unit IMU and an RTK mobile station;

[0009] Step 102: Establish a Cartesian coordinate system at the ground terminal; establish a carrier coordinate system at the UAV;

[0010] Step 2: Establish the state equation and observation equation of the drone:

[0011] Step 201: Obtain the quaternion of the state equation through the inertial measurement unit IMU , , , and the state transition coefficient matrix ;

[0012] Step 202: Use a computer to establish a drone The state equation at the moment ;in, Indicates that the drone The state vector at the moment, and , In the Cartesian coordinate system, The three-axis coordinates of the drone at the moment, In the Cartesian coordinate system, The three-axis speed of the drone at this moment, Indicates that the drone The state vector at the moment, represents the system noise; is a positive integer, and ; represents transpose;

[0013] Step 203: Use a computer to establish a drone The observation equation at each moment ;in, Indicates that the drone The observed value of the state at time instant, and , They represent the The predicted value of the distance measurement, azimuth measurement and elevation measurement of the UAV relative to the ground terminal at each moment; They represent the first The predicted value of the yaw angle, pitch angle and roll angle of the drone at each moment; represents the measurement noise, represents a nonlinear observation function;

[0014] Step 3: Use unscented Kalman filtering to obtain the UAV The optimal state vector at each moment is:

[0015] Step 301: determine whether the RTK mobile station outputs three-axis coordinates and three-axis speeds. If the RTK mobile station does not output three-axis coordinates and three-axis speeds, the drone is The state equation at the moment is estimated by using the unscented Kalman filter, and the prior estimate of the state vector predicted in the unscented Kalman filter is used as the first The optimal state vector at the moment ;

[0016] If the RTK mobile station outputs three-axis coordinates and three-axis speeds, then the inertial measurement unit IMU and the measurement data monitored by the RTK mobile station are input, based on the established UAV The state equation of the UAV at the moment The observation equation at the moment is estimated by using the unscented Kalman filter, and the posterior estimate of the state vector updated in the unscented Kalman filter is used as the UAV’s The optimal state vector at the moment ;

[0017] Step 302: From the drone The optimal state vector at the moment Get the drone The three-axis coordinate optimization value at the moment And drone Quaternion optimization value at time ;

[0018] Step 303: According to the drone Quaternion optimization value at time , get the The optimal distance between the drone and the ground terminal at a certain moment , azimuth optimization value and elevation optimization value ;

[0019] Step 4: Establish the ISAC channel model between the drone and the ground terminal:

[0020] According to the drone The optimal state vector and distance optimization value at each moment , azimuth optimization value and elevation optimization value , establish the ISAC channel model between the UAV and the ground terminal.

[0021] The above-mentioned UAV ISAC channel estimation method based on multi-modal fusion, step 201, the specific process is as follows:

[0022] Step 2011: Obtain the first The status information of the drone at each moment , , ;in, , , Indicates the output of the inertial measurement unit IMU in the carrier coordinate system. The yaw angle, pitch angle, and roll angle of the drone at each moment;

[0023] Step 2012: According to , get the Quaternion at time , , , ;

[0024] Step 2013: According to , get the upper left coefficient matrix ;

[0025] according to , and get the lower right coefficient matrix ;in, represents the fourth-order identity matrix, represents the intermediate coefficient matrix, , They represent the output of the inertial measurement unit IMU. The three-axis angular velocity of the UAV at the moment; among them, Indicates The moment and The time difference between the moments;

[0026] Step 2014: , and obtain the state transition coefficient matrix ;in, represents a 6×4 zero matrix, Represents a 4×6 zero matrix.

[0027] In the above-mentioned UAV ISAC channel estimation method based on multimodal fusion, step 301 obtains the measurement prediction value in the unscented Kalman filter estimation, which is specifically as follows:

[0028] Step A: Mark the three-axis coordinates of the input observation equation , the quaternion of the input observation equation is recorded as , and get the rotation matrix caused by the jitter effect ;in, , Represents the rotation matrix Middle Line The element value of the column, and and The value range of is a positive integer from 1 to 3;

[0029] Step B: According to , get the The predicted value of the distance measurement of the drone relative to the ground terminal at a certain moment , azimuth angle measurement prediction value and elevation angle measurement prediction ;in, is the distance measurement noise, is the azimuth measurement noise, is the elevation angle measurement noise;

[0030] Step C: According to , get the The predicted value of the UAV's yaw angle measurement at a moment ;

[0031] according to , get the The predicted value of the UAV's pitch angle measurement at each moment ;

[0032] according to , get the The predicted value of the UAV's roll angle measurement at the moment .

[0033] The above-mentioned UAV ISAC channel estimation method based on multimodal fusion, step 4, the specific process is as follows:

[0034] Step 401: According to the drone Quaternion optimization value at time , get the rotation matrix caused by the optimized jitter effect ;

[0035] Step 402: Establish a channel model between the drone and the ground terminal ;in, represents the channel response matrix, represents pi, represents a natural constant, represents the imaginary unit, Indicates the carrier wavelength; represents the first steering vector, and , represents the second steering vector, and , represents the third steering vector, and , represents the first intermediate vector, and , Indicates that the drone and the steering vector between the ground terminal at each moment, and , represents the fourth steering vector, and , represents the second intermediate vector, , represents transpose, represents the Kronecker product operation, represents the conjugate transpose.

[0036] In the above-mentioned UAV ISAC channel estimation method based on multimodal fusion, when the unscented Kalman filter estimation is adopted in step 301, the initialization setting is as follows:

[0037] according to , get the scaling parameter in the unscented Kalman filter ;in, represents the first parameter, and its value is 0.01; Represents the second parameter, and its value is 0. represents the dimension of the state vector, and The value is 10;

[0038] Set the initial state of the state equation and the initial covariance matrix ;in, , is the three-axis coordinate of the drone at the initial moment, is the three-axis speed of the drone at the initial moment, and the initial value of the quaternion is 0. is a 10×10 diagonal matrix, and .

[0039] Compared with the prior art, the present invention has the following advantages:

[0040] 1. The method of the present invention is simple in steps and easy to implement. First, the quaternion and state conversion coefficient matrix of the state equation are obtained through the inertial measurement unit IMU, and then the UAV is established. The state equation of the UAV at the moment; then, based on the quaternion of the state equation, the The observation equation at each moment is expressed by quaternions, so that each quaternion corresponds to a unique rotation, and there are no two different quaternions representing the same rotation, which makes quaternions more reliable when processing rotations and can significantly improve the calculation speed.

[0041] 2. The present invention is based on quaternion and unscented Kalman filter estimation to obtain the optimal state vector of the UAV at each moment, realize the multi-mode fusion of RTK and IMU data, and by judging whether the RTK mobile station outputs three-axis coordinates and three-axis speeds, use the one-step prediction of the unscented Kalman filter, or use the unscented Kalman filter prediction-update mechanism to improve the state optimization of the UAV, and avoid the changes in the attitude of the UAV during flight, the vibration of its own motor, and the jitter caused by airflow, which affect the communication quality.

[0042] 3. Based on the UAV attitude quaternion and coordinate trajectory, the present invention establishes the ISAC channel model between the UAV and the ground terminal, which not only improves the accuracy of channel estimation, but also enhances the adaptability and stability of the UAV in complex environments. The key parameters in the channel are used to carry out the channel reconstruction process. This strategy greatly simplifies the communication process and reduces the complexity of communication.

[0043] 4. The present invention models the ISAC channel model between the UAV and the ground terminal as a physical parameter model, which is determined by the quaternion, distance, azimuth and elevation corresponding to the three-dimensional attitude angle. Compared with the traditional channel modeling, which does not consider the impact of the three-dimensional attitude change of the UAV on the channel modeling, by integrating the quaternion corresponding to the three-dimensional attitude angle change into the ISAC channel modeling, the accuracy of attitude representation can be effectively improved and the channel estimation error can be reduced.

[0044] In summary, the method of the present invention has simple steps and reasonable design. The optimal state vector of the UAV is obtained based on the unscented Kalman filter under the multimodal fusion of the inertial measurement unit IMU and the RTK mobile station to establish a channel model between the UAV and the ground terminal, which reduces the complexity of communication and considers the trajectory and attitude changes of the UAV to improve the accuracy of channel estimation.

[0045] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 The figure is a flowchart of the method of the present invention.

[0047] Figure 2 It is the estimated error diagram of the trajectory of the present invention.

[0048] Figure 3 It is the estimated error diagram of the attitude angle of the present invention. DETAILED DESCRIPTION

[0049] like Figure 1 The method for estimating the ISAC channel of a UAV based on multimodal fusion includes the following steps:

[0050] Step 1: Build the drone ISAC base station system:

[0051] Step 101: The drone ISAC base station system includes a drone ISAC base station and a ground terminal. The drone ISAC base station includes a drone and an ISAC base station integrated on the drone. The bottom of the drone is installed The uniform planar array antenna, the ground terminal is installed Uniform planar array antenna; Indicates the number of rows of uniform planar array antennas at the bottom of the drone, Represents the number of columns of uniform planar array antenna at the bottom of the drone, represents the number of rows of the uniform planar array antenna of the ground terminal, Indicates the number of columns of the uniform planar array antenna of the ground terminal; wherein the drone is equipped with an inertial measurement unit IMU and an RTK mobile station;

[0052] Step 102: Establish a Cartesian coordinate system at the ground terminal; establish a carrier coordinate system at the UAV;

[0053] Step 2: Establish the state equation and observation equation of the drone:

[0054] Step 201: Obtain the quaternion of the state equation through the inertial measurement unit IMU , , , and the state transition coefficient matrix ;

[0055] Step 202: Use a computer to establish a drone The state equation at the moment ;in, Indicates that the drone The state vector at the moment, and , In the Cartesian coordinate system, The three-axis coordinates of the drone at the moment, In the Cartesian coordinate system, The three-axis speed of the drone at this moment, Indicates that the drone The state vector at the moment, represents the system noise; is a positive integer, and ; represents transpose;

[0056] Step 203: Use a computer to establish a drone The observation equation at each moment ;in, Indicates that the drone The observed value of the state at time instant, and , They represent the The predicted value of the distance measurement, azimuth measurement and elevation measurement of the UAV relative to the ground terminal at each moment; They represent the first The predicted value of the yaw angle, pitch angle and roll angle of the drone at each moment; represents the measurement noise, represents a nonlinear observation function;

[0057] Step 3: Use unscented Kalman filtering to obtain the UAV The optimal state vector at each moment is:

[0058] Step 301: determine whether the RTK mobile station outputs three-axis coordinates and three-axis speeds. If the RTK mobile station does not output three-axis coordinates and three-axis speeds, the drone is The state equation at the moment is estimated by using the unscented Kalman filter, and the prior estimate of the state vector predicted in the unscented Kalman filter is used as the first The optimal state vector at the moment ;

[0059] If the RTK mobile station outputs three-axis coordinates and three-axis speeds, then the inertial measurement unit IMU and the measurement data monitored by the RTK mobile station are input, based on the established UAV The state equation of the UAV at the moment The observation equation at the moment is estimated by using the unscented Kalman filter, and the posterior estimate of the state vector updated in the unscented Kalman filter is used as the UAV’s The optimal state vector at the moment ;

[0060] Step 302: From the drone The optimal state vector at the moment Get the drone The three-axis coordinate optimization value at the moment And drone Quaternion optimization value at time ;

[0061] Step 303: According to the drone Quaternion optimization value at time , get the The optimal distance between the drone and the ground terminal at a certain moment , azimuth optimization value and elevation optimization value ;

[0062] Step 4: Establish the ISAC channel model between the drone and the ground terminal:

[0063] According to the drone The optimal state vector and distance optimization value at each moment , azimuth optimization value and elevation optimization value , establish the ISAC channel model between the UAV and the ground terminal.

[0064] In this embodiment, step 201, the specific process is as follows:

[0065] Step 2011: Obtain the first The status information of the drone at each moment , , ;in, , , Indicates the output of the inertial measurement unit IMU in the carrier coordinate system. The yaw angle, pitch angle, and roll angle of the drone at each moment;

[0066] Step 2012: According to , get the Quaternion at time , , , ;

[0067] Step 2013: According to , get the upper left coefficient matrix ;

[0068] according to , and get the lower right coefficient matrix ;in, represents the fourth-order identity matrix, represents the intermediate coefficient matrix, , They represent the output of the inertial measurement unit IMU. The three-axis angular velocity of the UAV at the moment; among them, Indicates The moment and The time difference between the moments;

[0069] Step 2014: , and obtain the state transition coefficient matrix ;in, represents a 6×4 zero matrix, Represents a 4×6 zero matrix.

[0070] In this embodiment, the measurement prediction value is obtained in the unscented Kalman filter estimation in step 301, which is specifically as follows:

[0071] Step A: Mark the three-axis coordinates of the input observation equation , the quaternion of the input observation equation is recorded as , and get the rotation matrix caused by the jitter effect ;in, , Represents the rotation matrix Middle Line The element value of the column, and and The value range of is a positive integer from 1 to 3;

[0072] Step B: According to , get the The predicted value of the distance measurement of the drone relative to the ground terminal at a certain moment , azimuth angle measurement prediction value and elevation angle measurement prediction ;in, is the distance measurement noise, is the azimuth measurement noise, is the elevation angle measurement noise;

[0073] Step C: According to , get the The predicted value of the UAV's yaw angle measurement at a moment ;

[0074] according to , get the The predicted value of the pitch angle measurement of the UAV at a moment ;

[0075] according to , get the The predicted value of the UAV's roll angle measurement at a moment .

[0076] In this embodiment, step 4, the specific process is as follows:

[0077] Step 401: According to the drone Quaternion optimization value at time , get the rotation matrix caused by the optimized jitter effect ;

[0078] Step 402: ;in, represents the channel response matrix, represents pi, represents a natural constant, represents the imaginary unit, Indicates the carrier wavelength; represents the first steering vector, and , represents the second steering vector, and , represents the third steering vector, and , represents the first intermediate vector, and , Indicates that the drone and the steering vector between the ground terminal at each moment, and , represents the fourth steering vector, and , represents the second intermediate vector, , represents transpose, represents the Kronecker product operation, represents the conjugate transpose.

[0079] In this embodiment, when the unscented Kalman filter is used for estimation in step 301, the initialization settings are as follows:

[0080] according to , get the scaling parameter in the unscented Kalman filter ;in, represents the first parameter, and its value is 0.01; Represents the second parameter, and its value is 0. represents the dimension of the state vector, and The value is 10;

[0081] Set the initial state of the state equation and the initial covariance matrix ;in, , is the three-axis coordinate of the drone at the initial moment, is the three-axis speed of the drone at the initial moment, and the initial value of the quaternion is 0. is a 10×10 diagonal matrix, and .

[0082] In this embodiment, during specific implementation, the specific process of step 303 may refer to step B, and the specific process of step 401 may refer to step A.

[0083] In this embodiment, it should be noted that the nonlinear observation function , which refers to the acquisition process from step A to step C.

[0084] In this embodiment, during the specific implementation, , The RTK mobile station is the D-RTK 2 high-precision GNSS mobile station, and the inertial measurement unit IMU is the industrial-grade high-performance XR290 IMU.

[0085] In this embodiment, step 102 is specifically as follows:

[0086] A Cartesian coordinate system is established at the ground terminal; wherein the upper left corner of the uniform planar array antenna of the ground terminal is the origin O, the X-axis passing through the origin O points to the due east, the Y-axis passing through the origin O points to the due north, and the Z-axis passing through the origin O is perpendicular to the X-axis and the Y-axis and points to the sky;

[0087] A carrier coordinate system is established at the drone; wherein the center of the uniform planar array antenna of the carrier coordinate system is the origin O, the X-axis passing through the origin O coincides with the longitudinal axis of the drone and points to the forward direction of the drone, the Y-axis passing through the origin O coincides with the transverse axis of the drone and points to the right side of the forward direction of the drone, and the Z-axis passing through the origin O is perpendicular to the X-axis and the Y-axis and points to the bottom of the drone;

[0088] In this embodiment, during specific implementation, The value is 1 second and can be adjusted as needed without specific limitation.

[0089] In this embodiment, during the specific implementation, , and are all zero-mean Gaussian white noise.

[0090] In this embodiment, when the system noise is implemented, It obeys a Gaussian distribution with a mean of 0 and a variance of 1; the measurement noise It follows a Gaussian distribution with a mean of 0 and a variance of 1.

[0091] like Figure 2As shown, in this embodiment, the method of the present invention is used to estimate the error between the optimal state vector of the drone and the actual trajectory of the drone on the three coordinate axes. It can be seen that the errors in the three dimensions are all less than 1 meter, which meets the requirements.

[0092] like Figure 3 As shown, in this embodiment, the estimated value of the attitude data obtained by the optimal state vector of the drone and the estimated errors of the three attitude angles generated by the actual output of the inertial measurement unit IMU are all within 2 degrees, which meets the requirements.

[0093] In this embodiment, the information of the optimal state vector is used to perform channel reconstruction, which realizes the integration of communication and perception functions, thereby improving the overall performance of UAV communication, greatly simplifying the communication process, and reducing the complexity of communication.

[0094] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A UAV ISAC channel estimation method based on multimodal fusion, characterized in that: The method comprises the following steps: Step 1: Build the drone ISAC base station system: Step 101: The drone ISAC base station system includes a drone ISAC base station and a ground terminal. The drone ISAC base station includes a drone and an ISAC base station integrated on the drone. The bottom of the drone is installed The uniform planar array antenna, the ground terminal is installed Uniform planar array antenna; Indicates the number of rows of uniform planar array antennas at the bottom of the drone, Represents the number of columns of uniform planar array antenna at the bottom of the drone, represents the number of rows of the uniform planar array antenna of the ground terminal, Indicates the number of columns of the uniform planar array antenna of the ground terminal; wherein the drone is equipped with an inertial measurement unit IMU and an RTK mobile station; Step 102: Establish a Cartesian coordinate system at the ground terminal; establish a carrier coordinate system at the UAV; Step 2: Establish the state equation and observation equation of the drone: Step 201: Obtain the quaternion of the state equation through the inertial measurement unit IMU , , , and the state transition coefficient matrix ; Step 202: Use a computer to establish a drone The state equation at the moment ;in, Indicates that the drone The state vector at the moment, and , In the Cartesian coordinate system, The three-axis coordinates of the drone at the moment, In the Cartesian coordinate system, The three-axis speed of the drone at this moment, Indicates that the drone The state vector at the moment, represents the system noise; is a positive integer, and ; represents transpose; Step 203: Use a computer to establish a drone The observation equation at each moment ;in, Indicates that the drone The observed value of the state at time instant, and , They represent the The predicted value of the distance measurement, azimuth measurement and elevation measurement of the UAV relative to the ground terminal at each moment; They represent the first The predicted value of the yaw angle, pitch angle and roll angle of the drone at each moment; represents the measurement noise, represents a nonlinear observation function; Step 3: Use unscented Kalman filtering to obtain the UAV The optimal state vector at each moment is: Step 301: determine whether the RTK mobile station outputs three-axis coordinates and three-axis speeds. If the RTK mobile station does not output three-axis coordinates and three-axis speeds, the drone is The state equation at the moment is estimated by using the unscented Kalman filter, and the prior estimate of the state vector predicted in the unscented Kalman filter is used as the first The optimal state vector at the moment ; If the RTK mobile station outputs three-axis coordinates and three-axis speeds, then the inertial measurement unit IMU and the measurement data monitored by the RTK mobile station are input, based on the established UAV The state equation of the UAV at the moment The observation equation at the moment is estimated by using the unscented Kalman filter, and the posterior estimate of the state vector updated in the unscented Kalman filter is used as the UAV’s The optimal state vector at the moment ; Step 302: From the drone The optimal state vector at the moment Get the drone The three-axis coordinate optimization value at the moment And drone Quaternion optimization value at time ; Step 303: According to the drone Quaternion optimization value at time , get the The optimal distance between the drone and the ground terminal at a certain moment , azimuth optimization value and elevation optimization value ; Step 4: Establish the ISAC channel model between the drone and the ground terminal: According to the drone The optimal state vector and distance optimization value at each moment , azimuth optimization value and elevation optimization value , establish the ISAC channel model between the UAV and the ground terminal.

2. The method for estimating the ISAC channel of a UAV based on multimodal fusion according to claim 1, characterized in that: Step 201, the specific process is as follows: Step 2011: Obtain the first The status information of the drone at each moment , , ;in, , , Indicates the output of the inertial measurement unit IMU in the carrier coordinate system. The yaw angle, pitch angle, and roll angle of the drone at each moment; Step 2012: According to , get the Quaternion at time , , , ; Step 2013: According to , get the upper left coefficient matrix ; according to , and get the lower right coefficient matrix ;in, represents the fourth-order identity matrix, represents the intermediate coefficient matrix, , They represent the output of the inertial measurement unit IMU. The three-axis angular velocity of the UAV at the moment; among them, Indicates The moment and The time difference between the moments; Step 2014: , and obtain the state transition coefficient matrix ;in, represents a 6×4 zero matrix, Represents a 4×6 zero matrix.

3. The method for estimating the ISAC channel of a UAV based on multimodal fusion according to claim 1, characterized in that: Step 301 obtains the measurement prediction value in the unscented Kalman filter estimation, as follows: Step A: Mark the three-axis coordinates of the input observation equation , the quaternion of the input observation equation is recorded as , and get the rotation matrix caused by the jitter effect ;in, , Represents the rotation matrix Middle Line The element value of the column, and and The value range of is a positive integer from 1 to 3; Step B: According to , get the The predicted value of the distance measurement of the drone relative to the ground terminal at a certain moment , azimuth angle measurement prediction value and elevation angle measurement prediction ;in, is the distance measurement noise, is the azimuth measurement noise, is the elevation angle measurement noise; Step C: According to , get the The predicted value of the UAV's yaw angle measurement at a moment ; according to , get the The predicted value of the pitch angle measurement of the UAV at a moment ; according to , get the The predicted value of the UAV's roll angle measurement at a moment .

4. The method for estimating the ISAC channel of a UAV based on multimodal fusion according to claim 1, characterized in that: Step 4: The specific process is as follows: Step 401: According to the drone Quaternion optimization value at time , get the rotation matrix caused by the optimized jitter effect ; Step 402: Establish a channel model between the drone and the ground terminal ;in, represents the channel response matrix, represents pi, represents a natural constant, represents the imaginary unit, Indicates the carrier wavelength; represents the first steering vector, and , represents the second steering vector, and , represents the third steering vector, and , represents the first intermediate vector, and , Indicates that the drone and the steering vector between the ground terminal at each moment, and , represents the fourth steering vector, and , represents the second intermediate vector, , represents transpose, represents the Kronecker product operation, represents the conjugate transpose.

5. The method for estimating the ISAC channel of a UAV based on multimodal fusion according to claim 1, characterized in that: When the unscented Kalman filter is used for estimation in step 301, the initialization settings are as follows: according to , get the scaling parameter in the unscented Kalman filter ;in, represents the first parameter, and its value is 0.01; Represents the second parameter, and its value is 0. represents the dimension of the state vector, and The value is 10; Set the initial state of the state equation and the initial covariance matrix ;in, , is the three-axis coordinate of the drone at the initial moment, is the three-axis speed of the drone at the initial moment, and the initial value of the quaternion is 0. is a 10×10 diagonal matrix, and .

Images