Large-scale MIMO non-orthogonal multiple access bidirectional relay method based on unmanned aerial vehicle

Abstract

The invention discloses a large-scale MIMO non-orthogonal multiple access bidirectional relay method based on an unmanned aerial vehicle, and belongs to the field of wireless communication, and the method comprises the steps: constructing a hovering rotor unmanned aerial vehicle working in a full duplex mode as an amplification forwarding relay, and constructing a model of M ground user pairs needing bidirectional communication; the ground user pair sends a superposed signal to the unmanned aerial vehicle by adopting a non-orthogonal multiple access protocol; the unmanned aerial vehicle performs pilot frequency training by using a minimum mean square error method to obtain imperfect channel state information, and adopts maximum ratio merging or maximum ratio transmission linear processing as a relay precoder; and the unmanned aerial vehicle broadcasts the processed signals back to all ground user pairs. The large-scale MIMO non-orthogonal multiple access bidirectional relay system based on the unmanned aerial vehicle is reliable in performance, high in communication capacity, strong in coverage capability and good in anti-noise performance.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communications, and in particular relates to a large-scale MIMO non-orthogonal multiple access bidirectional relay method based on an unmanned aerial vehicle. Background Art

[0002] With the widespread adoption of the Internet of Things and the rapid development of communication technologies, wireless data transmission has exploded. The demand for efficient transmission and massive connections has become a major challenge for next-generation wireless communication networks. Drone communications, with their advantages such as low cost, high mobility, on-demand deployment, and improved channel quality due to line-of-sight links, are considered a promising technology for expanding network space and achieving high-density, high-speed wireless transmission. They are often used as aerial base stations or relays to assist ground communications. However, while drones can effectively alleviate the pressure on ground communications caused by massive and dense access, spectrum shortages remain a pressing issue.

[0003] Non-orthogonal multiple access has received widespread attention in recent years due to its potential to improve spectrum efficiency and support massive connections. Compared with traditional orthogonal multiple access, non-orthogonal multiple access achieves better spectrum utilization and greater system capacity through multi-user power domain multiplexing. The advantages of non-orthogonal multiple access are also reflected in its transmission and reception algorithms, namely superposition coding and serial interference cancellation. At the same time, full-duplex is another important technology for achieving high spectrum efficiency, which can reduce the additional bandwidth cost of cooperative non-orthogonal multiple access systems. Full-duplex allows signals to be sent and received simultaneously, and theoretically can achieve a transmission rate double that of half-duplex;

[0004] The basic task of the target signal angle estimation algorithm is to recover the direction parameters of the unknown target signal in the airspace through the signals received by the sensor array. Its main technology is to use a small number of sensors to complete the estimation process of the target signal direction in space in a shorter time, obtain more accurate estimation results, and have a certain degree of robustness to Gaussian noise environments. The most popular technology is based on the sparse reconstruction method, which is generally used Norm constructs the cost function, combined with Norm to reconstruct the representation, or combine The sparse reconstruction model is constructed based on the norm and atomic norm. However, these models have some defects, such as not considering the problem of off-grid or not being robust to angle estimation in non-Gaussian noise environments, which ultimately seriously affects the estimation accuracy of the reconstructed signal. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a large-scale MIMO non-orthogonal multiple access bidirectional relay method based on drones, comprising:

[0006] A hovering rotor drone is provided, the drone operates in full-duplex mode and is equipped with 2N antennas, wherein N antennas are used for receiving signals and the other N antennas are used for transmitting signals;

[0007] Provide M pairs of ground users that need two-way communication. The two users in each ground user pair need to exchange information, and each ground user is equipped with a single antenna.

[0008] The ground user pair sends a superimposed signal to the UAV according to a non-orthogonal multiple access protocol;

[0009] After receiving the superimposed signal, the UAV performs pilot training using a minimum mean square error method to obtain non-perfect channel state information;

[0010] According to the imperfect channel state information, the UAV uses maximum ratio combining or maximum ratio transmission linear processing as a relay precoder to process the received signal to obtain a processed signal;

[0011] The UAV transmits the processed signal to all ground user pairs;

[0012] After receiving the processed signal, the ground user pair decodes the target signal using a serial interference cancellation technique.

[0013] Preferably, according to the hovering requirements of the hovering rotor drone, the drone is deployed at a fixed height and placed on a movable vehicle via a tethered cable.

[0014] Preferably, the process of the ground user sending the superimposed signal to the UAV according to the ground user two-way communication model and the non-orthogonal multiple access protocol further includes:

[0015] Establishing an air-to-ground channel model based on the communication requirements between the UAV and ground users;

[0016] The air-to-ground channel model is modeled as line-of-sight probability, which is determined by the communication environment, ground user position, drone position and elevation angle;

[0017] The large-scale attenuation of the air-to-ground channel is modeled as a random variable with line-of-sight and non-line-of-sight probabilities;

[0018] The small-scale fading of the air-to-ground channel is modeled as a Rician fading channel.

[0019] Preferably, the process of the ground user sending the superimposed signal to the UAV according to the non-orthogonal multiple access protocol further includes:

[0020] The ground user pairs simultaneously send orthogonal pilot sequences of length τ; based on the orthogonal pilot sequences, the UAV receives the pilot signals and performs pilot training using a minimum mean square error channel estimation method to obtain a channel estimation matrix and an estimation error matrix.

[0021] Preferably, the process of processing the received signal to obtain the processed signal includes:

[0022] The UAV constructs a precoding matrix based on the obtained imperfect channel state information, which is a diagonal permutation matrix of information exchange between each ground user pair; according to the precoding matrix, the received signal is relayed and amplified and a power constraint coefficient is applied to generate a transmit signal.

[0023] Preferably, the drone adopts a self-interference cancellation technique, according to which the elements of the residual self-interference coefficient matrix are modeled as independent and identically distributed random variables.

[0024] Preferably, the process of the ground user decoding the target signal using the serial interference cancellation technology after receiving the processed signal includes:

[0025] After receiving the processed signal, the ground user uses known self-signal and channel information to perform self-interference elimination; and decodes the target signal using serial interference cancellation technology based on the signal after self-interference elimination.

[0026] On the other hand, the present invention also provides a large-scale MIMO non-orthogonal multiple access bidirectional relay system based on a drone, comprising:

[0027] There are M pairs of ground user devices, where the two users in each ground user pair need to exchange information, and each ground user device is equipped with a single antenna;

[0028] A hovering rotor drone acts as a shared bidirectional amplification and forwarding relay. The drone is equipped with 2N antennas, N of which are used to receive signals and the other N are used to transmit signals, and operates in full-duplex mode.

[0029] The ground user pair sends a superimposed signal to the UAV using a non-orthogonal multiple access protocol;

[0030] The UAV uses the minimum mean square error method to perform pilot training to obtain non-perfect channel state information, and adopts maximum ratio combining or maximum ratio transmission linear processing as a relay precoder;

[0031] The UAV transmits the processed signal to all ground user pairs;

[0032] After receiving the signal, the ground user uses the serial interference cancellation technology to decode the target signal.

[0033] On the other hand, the present invention further provides an electronic device, comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the method is implemented when the processor executes the computing program.

[0034] On the other hand, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the method when executed by a processor.

[0035] Compared with the prior art, the present invention has the following advantages and technical effects:

[0036] This application discloses a large-scale MIMO non-orthogonal multiple access bidirectional relay system based on drones. Compared with traditional ground-based large-scale MIMO relay systems, the strong line-of-sight links provided by drones can significantly improve spectrum efficiency. Under the same configuration parameters, due to power reuse technology and serial cancellation technology, non-orthogonal multiple access schemes are always better than orthogonal multiple access schemes. Combined with distance-based power allocation, more power is allocated to distant ground users, effectively ensuring communication fairness. The array gain brought by large-scale antennas improves the overall system efficiency. The application of power scaling laws reveals significant energy-saving potential. By deploying large-scale antennas, the transmission power of ground users or drones can be greatly reduced while maintaining basically unchanged spectrum efficiency.

[0037] Furthermore, minimum mean square error channel estimation is used to handle imperfect channel state information, ensuring reliable channel information despite limited pilot resources and multi-user interference. The proposed three power scaling scenarios and their analysis provide clear guidance for system designers. Based on the specific priorities for spectral efficiency, overall performance, antenna cost, and power budget, scaling strategies and parameters can be flexibly selected to achieve the optimal balance between performance and energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0039] Figure 1 This is a schematic diagram of a method flow of a large-scale MIMO non-orthogonal multiple access bidirectional relay system based on a drone according to an embodiment of the present application;

[0040] Figure 2 This is a schematic diagram of the spectrum efficiency and signal-to-noise ratio curves of non-orthogonal multiple access and orthogonal multiple access schemes under different numbers of ground user pairs and numbers of drone antennas in an embodiment of the present application;

[0041] Figure 3 This is a schematic diagram of the spectrum efficiency and the curve of changes with the drone altitude under different drone antenna numbers in the embodiment of the present application. DETAILED DESCRIPTION

[0042] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0043] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0044] Example 1

[0045] This embodiment provides a large-scale MIMO non-orthogonal multiple access bidirectional relay method based on a drone, including:

[0046] There are M pairs of ground user devices, where the two users in each ground user pair need to exchange information, and each ground user device is equipped with a single antenna;

[0047] A hovering rotor drone acts as a shared bidirectional amplification and forwarding relay. The drone is equipped with 2N antennas, N of which are used to receive signals and the other N are used to transmit signals, and operates in full-duplex mode.

[0048] The ground user sends a superimposed signal to the UAV using a non-orthogonal multiple access protocol;

[0049] The UAV adopts the minimum mean square error channel estimation method and uses orthogonal pilot sequences for pilot training to obtain imperfect channel state information of the air-to-ground channel;

[0050] The UAV processes the received signal using maximum ratio combining / maximum ratio transmission linear processing as a relay precoder based on the obtained imperfect channel state information;

[0051] The drone broadcasts the processed signal to all ground users;

[0052] After receiving the signal, the ground user uses the serial interference cancellation technology to decode the target signal.

[0053] The drone is deployed at a fixed height H and is placed on a movable vehicle via a tethered cable to achieve flexible hovering area coverage.

[0054] The air-to-ground channel model considers the line-of-sight probability, which is determined by the communication environment, the ground user's position, the UAV's position and the elevation angle, and is expressed as:

[0055] ;

[0056] in, is a constant that depends on the type of environment, is the elevation angle of the UAV to the i-th ground user, represents the viewing distance probability of the i-th user.

[0057] The large-scale attenuation of the air-to-ground channel is modeled as a random variable that depends on the line-of-sight and non-line-of-sight probabilities. The average large-scale attenuation channel gain is expressed as:

[0058] ;

[0059] in, is the average large-scale attenuation channel gain, and the line-of-sight additional attenuation factor , non-line-of-sight additional attenuation factor , is the channel power gain constant at a reference distance of 1 meter, is the path loss exponent, is the distance from the UAV to the i-th ground user, represents the non-line-of-sight probability of the i-th user, Indicates the signal attenuation factor caused by distance.

[0060] The small-scale fading of the air-to-ground channel is modeled as a Rician fading channel, which is expressed as:

[0061] ;

[0062] in, represents the small-scale fading of the air-to-ground channel, A unified mathematical expression for the channel direction, represents the link from ground user to UAV, represents the link from the UAV to the ground user, represents the identity matrix, is a Rice factor diagonal matrix, is a random scattering component, and its elements obey a cyclically symmetric complex Gaussian distribution with zero mean and unit variance. The deterministic direct component is composed of the steering vectors of the UAV antenna array.

[0063] After the UAV adopts the self-interference elimination technology, the residual self-interference coefficient matrix The elements of are modeled as independent and identically distributed Random variables, where represents a zero-mean complex Gaussian random variable.

[0064] The power distribution coefficient satisfies: and , and meet the requirements based on the distance between the drone and the users on the ground . represents the power allocation coefficient of the mth terrestrial user, represents the power allocation coefficient of the nearest user, represents the Euclidean distance from the drone to the nearest user. The calculation formula for the sum of spectrum efficiency is:

[0065] ;

[0066] in, represents the spectrum efficiency and, is the coherence interval length, τ is the number of pilot symbols, Decode the paired user for the mth ground user The received signal-to-interference ratio of the signal. The calculation formula for the total energy efficiency is:

[0067] ;

[0068] in, represents the total energy efficiency, , is the total transmission power of all terrestrial users, is the total transmission power of the UAV, is the system fixed circuit power consumption, is the propulsion power consumption of the hovering drone, is the blade profile power, is the induced power.

[0069] The system applies power scaling laws to achieve energy efficiency optimization when deploying large-scale antennas (N is large enough), specifically in the following cases:

[0070] Case (1): Fixed terrestrial user transmission power , UAV transmission power and pilot energy , the pilot symbol power is based on Zoom, where Represents the pilot power scaling factor. This factor is applicable to scenarios where channel estimation resources are limited, such as high-mobility environments or battery-constrained drones, where pilot training overhead needs to be reduced. When rapid deployment is required in emergency communications, pilot power can be reduced to maintain spectral efficiency and stability, saving training energy.

[0071] Case (2): Fixed pilot symbol power , ground user energy and drone energy , Ground user transmission power , UAV transmission power ( ),in represents the UAV power scaling factor, Indicates the user power scaling factor. Applicable to energy-sensitive scenarios, such as when the drone battery is limited and the power consumption of transmission between the user and the drone needs to be reduced;

[0072] Case (3): Fixed ground user energy , drone energy , and pilot energy , ground user transmission power , UAV transmission power , pilot symbol power It is applicable to global energy efficiency optimization scenarios, such as solar-powered drones, which need to jointly optimize communication and hovering energy consumption to achieve a balance between energy efficiency and spectrum efficiency.

[0073] Example 2

[0074] This embodiment provides a large-scale MIMO non-orthogonal multiple access bidirectional relay method based on a drone, including:

[0075] A UAV-assisted full-duplex massive MIMO non-orthogonal multiple access (NOMA) bidirectional relay system model is established. A hovering rotorcraft UAV, equipped with 2N antennas and operating in a half-transmit, half-receive, full-duplex mode, acts as a bidirectional amplification and forwarding relay, assisting M in bidirectional communication with ground users. Ground users use a NMA protocol to transmit superimposed signals to the UAV.

[0076] The Rician fading model is used to analyze line-of-sight air-to-ground links with imperfect channel state information. The drone uses a minimum mean square error (MMSE) channel estimation method and orthogonal pilot sequences for pilot training to obtain the imperfect channel state information of the air-to-ground channel. Based on this channel state information, the drone uses maximum ratio combining (MRC) or maximum ratio transmit linear processing (MRTL) as a relay precoder to process the received signal.

[0077] The drone broadcasts the processed signal to all ground users, who then use serial interference cancellation technology to decode the target signal.

[0078] Closed-form and asymptotic expressions for the system's spectral efficiency and total energy efficiency under imperfect channel state information are derived. A power scaling law is proposed, and by selecting different combinations of power scaling parameters, the relationship between ground user transmit power, UAV transmit power, pilot symbol power, and system spectral efficiency is explored.

[0079] Optional, based on drone location and ground user locations , calculate the distance and elevation . Using line-of-sight probability and the average large-scale fading channel gain Calculate the large-scale channel attenuation.

[0080] Optionally, use the Rice fading model Calculate the channel small-scale fading, where the steering vector The UAV antenna array structure and the ground user arrival angle The residual self-interference is modeled as elements obeying Matrix .

[0081] Optionally, the ground user sends a message of length Orthogonal pilot sequence The drone receives the pilot signal and , represents the additive white Gaussian noise at the transmitting antenna array, represents the additive white Gaussian noise at the drone, Represents the pilot power scaling coefficient matrix. The drone uses the minimum mean square error channel estimation method to estimate the random scattering component by subtracting the known deterministic component and the Rice factor term. and , and obtain the channel estimation matrix and The actual channel is decomposed into and ,in and Denotes the estimation error matrix, which are independent of each other. The estimation error variance is , ,in, represents the Rice factor, represents the transmit power of a single pilot symbol, represents the pilot power coefficient of the user, represents the large-scale channel gain of the user.

[0082] Optionally, the ground user sends a superimposed signal to the UAV at the same time according to the non-orthogonal multiple access protocol at time u , where M represents the number of ground user pairs, represents the user's data symbol power allocation coefficient, represents the total transmission power of all ground users, Indicates the data symbol sent by the user, and the power allocation coefficient satisfies and The drone receives the signal ,in, represents the diagonal matrix of power allocation coefficients, represents the residual self-interference channel matrix, represents the residual self-interference signal vector, Indicates that the precoding matrix is ​​constructed based on the estimated channel ,in, express, represents the additive white Gaussian noise at the drone receiver, The diagonal permutation matrix of information exchange between each ground user pair. The drone performs relay amplification processing on the received signal and imposes a power constraint coefficient , generates a sending signal , To handle delay, is the power constraint coefficient of the UAV relay, is the beamforming matrix, is the signal vector received by the drone. The drone broadcasts the signal through the transmitting antenna array To all ground users. The mth ground user receives the signal ,in is the channel row vector from the drone to the user, The additive white Gaussian noise at the user receiving end is used. After self-interference cancellation using the known self-signal and channel information, the target signal is decoded using the serial interference cancellation technology.

[0083] Example 3

[0084] This embodiment provides a large-scale MIMO non-orthogonal multiple access bidirectional relay method based on a drone, including:

[0085] Due to unexpected damage to ground base stations, information transmission between pairs of ground users during an emergency requires drones. This example assumes that the drones are deployed at a fixed altitude H and are placed on a movable aircraft via tethers to achieve a flexible and continuous hovering range. It is also assumed that all ground users are located in microcells and remain within the drone's coverage area. We also assume that the ground user devices are equipped with a single antenna, while the drones are equipped with 2N antennas, half of which are used for transmission and the other half for reception.

[0086] Set the coordinates of the drone's hovering position to , where the ground plane is Indicates that z is the height. Meanwhile, the i-th ground user is located at and the height of the ground user equipment is negligible compared to the height of the UAV, assuming that the position of each ground user equipment is fixed during the communication period. In addition, the distance between the UAV and the i-th ground user is given by Indicates that the elevation angle of the UAV toward the i-th ground user in the horizontal plane is .therefore, and Can be respectively and Calculate, where represents the horizontal Euclidean distance between the UAV projection point and the ground user i.

[0087] like Figure 1 As shown, a large-scale MIMO non-orthogonal multiple access bidirectional relay system and method based on drones has the following process structure:

[0088] A hovering rotor drone at a height of H, which can be flexibly hovered by a tethered vehicle, locates M groups of ground user pairs. The two users in each ground user pair need to exchange information. Each ground user device is equipped with a single antenna. All ground users transmit orthogonal pilot sequences synchronously. After receiving the pilot signal, the UAV uses the minimum mean square error channel estimation method to estimate the random scattering component by subtracting the known deterministic component and the Rice factor term. and , and obtain the channel estimation matrix and .

[0089] According to the non-orthogonal multiple access protocol, the ground user sends superimposed signals to the UAV at the same time at time u, and the UAV receives the signal , and construct the precoding matrix based on the estimated channel ,in The diagonal permutation matrix of information exchange between each ground user pair. The drone performs relay amplification processing on the received signal and imposes a power constraint coefficient , generates a sending signal , To handle latency, the drone broadcasts signals via a transmitting antenna array. To all ground users. The mth ground user receives the signal , using the known self-signal and channel information, after self-interference elimination, the serial interference cancellation technology is used in sequence to decode the target signal.

[0090] Figure 2 The spectral efficiency and channel ratio curves of non-orthogonal and orthogonal multiple access schemes with different numbers of ground user pairs and different numbers of UAV antennas are compared. In addition, the spectral efficiency and comparison of the proposed non-orthogonal multiple access scheme with perfect channel state information and imperfect channel state information are provided. Figure 2It can be observed that as the number of ground user pairs M increases, the spectral efficiency and spectral efficiency of both non-orthogonal and orthogonal multiple access (NOMA) systems increase. This is because the system's spectral efficiency is contributed by all ground users, and the performance improvement from more ground users dominates the performance degradation caused by increased interference between user pairs. Furthermore, increasing the number of UAV antennas N leads to even greater spectral efficiency gains. This indicates that more antennas can significantly improve spectrum utilization. Furthermore, the proposed system under NMA outperforms OMA, primarily due to the utilization of power multiplexing and serial cancellation techniques. Furthermore, a comparison of the curves reveals that while imperfect channel state information does not significantly impact spectral efficiency performance, multiple ground users in real systems lead to greater inter-user interference, necessitating channel estimation. Finally, the spectral efficiency and spectral efficiency improve with increasing signal-to-noise ratio (SNR) and converge to a constant at high SNRs. In other words, the spectral efficiency and spectral efficiency of both NMA and OMA saturate at high SNRs. The higher saturation value for NMA indicates that NMA can achieve a closer approach to the capacity limit of multi-ground user systems.

[0091] Figure 2 Three curves are involved: the first is the theoretical upper bound, or asymptote; the second is the theoretical analysis curve; and the third is the curve obtained from actual simulation experiments. Therefore, when the theoretical analysis is correct, the theoretical analysis curve and the actual simulation curve will coincide, and they will gradually approach the asymptote as the signal-to-noise ratio increases. The simulation results in the figure confirm the correctness of the analysis results.

[0092] In UAV communication systems, the effects of signal blocking and scattering vary with the altitude of the UAV. Figure 3 This relationship was further explored. Compared with ground-based massive MIMO relay communications, deploying UAVs at a suitable altitude can achieve better spectrum efficiency performance, and the performance gain can be enhanced when the number of antennas is greater. This shows that the multi-antenna array gain has a great benefit to the system. In addition, as the UAV altitude H increases, the spectrum efficiency and first increase until it reaches the optimal hovering altitude and then begin to decrease. This is because, as H increases, becomes larger, leading to the Rice factor The performance improvement due to line-of-sight propagation is more significant than the performance degradation due to increased distance. However, as H continues to increase above the optimal height, performance is limited by the increased distance. Finally, the optimal hovering height of the UAV increases with the number of antennas, so for UAVs with different numbers of antennas, the maximum performance gain can be achieved by flexibly adjusting the corresponding UAV height.

[0093] On the other hand, this embodiment also provides a UAV-based massive MIMO non-orthogonal multiple access bidirectional relay system, including:

[0094] There are M pairs of ground user devices, where the two users in each ground user pair need to exchange information, and each ground user device is equipped with a single antenna;

[0095] A hovering rotor drone acts as a shared bidirectional amplification and forwarding relay. The drone is equipped with 2N antennas, N of which are used to receive signals and the other N are used to transmit signals, and operates in full-duplex mode.

[0096] The ground user pair sends a superimposed signal to the UAV using a non-orthogonal multiple access protocol;

[0097] The UAV uses the minimum mean square error method to perform pilot training to obtain non-perfect channel state information, and adopts maximum ratio combining or maximum ratio transmission linear processing as a relay precoder;

[0098] The drone broadcasts the processed signal back to all ground user pairs;

[0099] After receiving the signal, the ground user uses the serial interference cancellation technology to decode the target signal.

[0100] On the other hand, this embodiment further provides an electronic device, comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the method is implemented when the processor executes the computing program.

[0101] On the other hand, this embodiment further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the method when executed by a processor.

[0102] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A large-scale MIMO non-orthogonal multiple access bidirectional relay method based on drones, characterized in that: include: Provided is a drone, the drone operating in full-duplex mode and equipped with 2N antennas, wherein N antennas are used for receiving signals and the other N antennas are used for transmitting signals; Provide M pairs of ground users that need two-way communication. The two users in each ground user pair need to exchange information, and each ground user is equipped with a single antenna. The ground user pair sends a superimposed signal to the UAV according to a non-orthogonal multiple access protocol; After receiving the superimposed signal, the UAV performs pilot training using a minimum mean square error method to obtain non-perfect channel state information; According to the imperfect channel state information, the UAV uses maximum ratio combining or maximum ratio transmission linear processing as a relay precoder to process the received signal to obtain a processed signal; The UAV transmits the processed signal to all ground user pairs; After receiving the processed signal, the ground user pair decodes the target signal using a serial interference cancellation technique.

2. The method according to claim 1, characterized in that The drone is a hovering rotor drone, which hovers above the movable vehicle when the hovering rotor drone is working. The hovering rotor drone and the movable vehicle are connected by a tethered cable.

3. The method according to claim 1, characterized in that The process of the ground user sending the superimposed signal to the UAV according to the non-orthogonal multiple access protocol further includes: Establishing an air-to-ground channel model based on the communication requirements between the UAV and ground users; The air-to-ground channel model is modeled as line-of-sight probability, which is determined by the communication environment, ground user position, drone position and elevation angle; The large-scale attenuation of the air-to-ground channel is modeled as a random variable with line-of-sight and non-line-of-sight probabilities; The small-scale fading of the air-to-ground channel is modeled as a Rician fading channel.

4. The method according to claim 1, wherein The process of the ground user sending the superimposed signal to the UAV according to the non-orthogonal multiple access protocol further includes: The ground user pairs simultaneously send orthogonal pilot sequences of length τ; based on the orthogonal pilot sequences, the UAV receives the pilot signals and performs pilot training using a minimum mean square error channel estimation method to obtain a channel estimation matrix and an estimation error matrix.

5. The method according to claim 1, wherein The process of processing the received signal to obtain the processed signal includes: The UAV constructs a precoding matrix based on the obtained imperfect channel state information, which is a diagonal permutation matrix of information exchange between each ground user pair; according to the precoding matrix, the received signal is relayed and amplified and a power constraint coefficient is applied to generate a transmit signal.

6. The method according to claim 1, characterized in that The drone employs a self-interference cancellation technique, according to which elements of a residual self-interference coefficient matrix are modeled as independent and identically distributed random variables.

7. The method according to claim 1, characterized in that After receiving the processed signal, the ground user pair decodes the target signal using the serial interference cancellation technology, including: After receiving the processed signal, the ground user uses known self-signal and channel information to perform self-interference elimination; and decodes the target signal using serial interference cancellation technology based on the signal after self-interference elimination.

8. A large-scale MIMO non-orthogonal multiple access bidirectional relay system based on drones, characterized in that: include: There are M pairs of ground user devices, where the two users in each ground user pair need to exchange information, and each ground user device is equipped with a single antenna; A hovering rotor drone acts as a shared bidirectional amplification and forwarding relay. The drone is equipped with 2N antennas, N of which are used to receive signals and the other N are used to transmit signals, and operates in full-duplex mode. The ground user pair sends a superimposed signal to the UAV using a non-orthogonal multiple access protocol; The UAV uses the minimum mean square error method to perform pilot training to obtain non-perfect channel state information, and adopts maximum ratio combining or maximum ratio transmission linear processing as a relay precoder; The UAV transmits the processed signal to all ground user pairs; After receiving the signal, the ground user uses the serial interference cancellation technology to decode the target signal.

9. An electronic device comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein: When the processor executes the computing program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images