Deep learning auxiliary interference suppression method and system based on movable antenna array

Abstract

The invention discloses a deep learning auxiliary interference suppression method and system based on a movable antenna array, belongs to the technical field of communication signal processing, considers a legal communication link with a plurality of jammers, and suggests to deploy the movable antenna array on a receiver to resist interference attacks. The method comprises the following steps: firstly, constructing a scene in which a plurality of jammers and a single-source node communicate with a base station, analyzing to obtain an instantaneous signal to interference plus noise ratio of a receiver, and forming a problem model which takes maximization of the signal to interference plus noise ratio of the system as a target and takes receiving beam forming and antenna unit positioning as variables; then, a deep learning-based framework is developed, beam forming is solved as a Rayleigh entropy problem, and antenna unit positioning is solved through multi-layer perceptron training; neural network parameters are optimized by adopting a stochastic gradient descent method, and the purpose of effectively suppressing interference is achieved.

Description

Background Art

[0002] Widespread malicious interference poses a major challenge to wireless information security, and effective anti-interference technology has always been a key condition for wireless systems. Existing technologies for countering interference attacks often rely on mitigating or avoiding interference by jointly utilizing multi-dimensional resources. However, as wireless environments become more complex and attackers become more sophisticated, the implementation of these methods requires more intensive resources or calculations. This has prompted researchers to strive to seek new resource dimensions or degrees of freedom to mitigate interference attacks for information security supply. On the other hand, the network in the 6G era is an integrated space-air-ground network, in which unmanned aerial vehicles (UAVs) are playing an increasingly important role. The flexible deployment of drones can provide coverage for higher and wider areas to expand the network range. In addition, drones have promoted various applications in military and civilian fields. Especially in the field of the Internet of Things, drones can easily establish communications in the absence of traditional network infrastructure and approach low-power IoT devices for efficient data collection. Drones not only save capital and operating expenses, but also save energy for the Internet of Things, because drones can easily reach the vicinity of IoT devices to establish effective communications.

[0003] Recently, movable antennas have emerged as a promising technology with the potential to enhance wireless communications. Unlike conventional fixed antenna systems, movable antennas can dynamically adjust their spatial positions to actively intervene in the radiation pattern in the desired direction. As a result, movable antenna technology has been widely studied in transmission improvement, sensing enhancement, computational assistance, security assurance, etc. In these studies, the flexible configuration of movable antennas is combined with conventional size resources to obtain superior system performance, which shows that there are significant advantages in exploiting antenna movement as a new degree of freedom.

[0004] Despite the extensive attention and emerging research efforts on movable antenna-assisted communications, its potential in interference mitigation remains unrevealed. Inspired by existing results, the spatial adaptability of movable antennas enables reliable communications even in highly dynamic and hostile environments, realizing effective interference mitigation beyond the capabilities of traditional fixed antenna systems. Moreover, by jointly leveraging antenna motion with other resources, the improved spatial diversity allows finding the best path to circumvent malicious attacks, thereby mitigating interference even when multiple adversaries are involved.

[0005] In addition, existing mobile antenna communications mostly use optimization-based analytical methods. However, the antenna position is incorporated into the radiation, so the processing can be quite complex. Therefore, recent studies have also mentioned learning techniques to obtain more effective solutions. In this regard, machine learning, as a powerful tool for solving complex non-convex optimization, can effectively determine the optimal configuration of antenna positions and transmission strategies, providing real-time adaptability that is difficult to achieve with traditional optimization methods. Therefore, by combining movable antenna transmission with learning-based scheme design, it can be expected that a highly adaptive and resilient anti-interference system can effectively adapt to dynamic wireless environments.

[0006] Widespread malicious interference poses a major challenge to the security of wireless information. As the wireless environment becomes more and more complex, the implementation of existing methods requires more intensive resources or calculations, which has prompted researchers to strive to find new resource dimensions or degrees of freedom to mitigate interference attacks. In recent years, movable antennas have emerged as a promising technology with potential enhancements in all aspects of wireless communications. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a deep learning-assisted interference suppression method and system based on a movable antenna array in response to the above-mentioned deficiencies in the prior art, to study the potential of movable antennas in anti-interference communications, and to use deep learning technology to achieve adaptive and efficient strategy design, so as to solve the technical problem that existing anti-interference attack technologies rely on the joint use of multi-dimensional resources to mitigate or avoid interference, thereby improving wireless information security.

[0008] The present invention adopts the following technical solutions:

[0009] A deep learning-assisted interference suppression method based on a movable antenna array includes the following steps:

[0010] S1. Construct a confrontation scenario where there are several jammers and a single legitimate source node communicating with the base station, and deploy a movable antenna array at the base station;

[0011] S2. Based on the given direction of arrival of the known transmitter signal in the confrontation scenario constructed in step S1, obtain the steering vector of the movable antenna array from the transmitter to the base station; model the channel gain between the source node and the jammer and the base station, obtain the expression of the received signal at the base station, and calculate the instantaneous signal to interference and noise ratio;

[0012] S3, according to the given direction of arrival of the transmitter signal, construct a problem model with receiving beamforming and antenna unit position as variables and maximizing the instantaneous signal to interference and noise ratio obtained in step S2 as the objective function;

[0013] S4. Based on the problem model obtained in step S3, deep learning technology is used to implement adaptive and efficient strategy design, extract arrival directions from legitimate source nodes and jammers, and input them into the designed deep learning system; obtain a set of ratios of minimum antenna unit spacing to actual antenna unit spacing, and obtain actual antenna unit positions by remapping;

[0014] S5. Based on the actual antenna unit position obtained in step S4, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-noise ratio under the current antenna unit deployment conditions;

[0015] S6, the actual antenna unit position obtained in step S4 and the optimal beamforming obtained in step S5 are input into the loss function for neural network training;

[0016] S7. Repeat steps S4 to S6 until the loss function converges, and obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal to interference noise ratio of the anti-interference system.

[0017] Preferably, in step S1, in the presence of an interference attack, in a wireless system with a legitimate communication pair, a single-antenna legitimate source node sends a desired signal to a multi-antenna base station; at the same time, there are K jammers deployed with single antennas transmitting interference signals on the same frequency band, which is expressed as An antenna array with movable antenna units is deployed at the base station. The movable antenna array consists of N antenna units, denoted as The position is recorded as The moving range is [0, L], where L represents the linear size of the antenna array. The goal is to maximize the signal-to-interference-noise ratio of the system by seeking a combination of beamforming and antenna unit position strategies.

[0018] Preferably, in step S2, according to the positions of the legitimate source node and the jammer node in the system under consideration, the directions in which the legitimate source node and the jammer node transmit signals to reach the base station are respectively recorded as Get the steering vector a(x,θ) from the transmitter to the movable antenna array at the base station k ); The base station based on the movable antenna array performs beamforming operation on the received signal to optimize the receiving direction of the signal. The beamforming vector is represented by w, and the beam gain is represented by |w H a(x,θ k )| 2 , The signal-to-interference-and-noise ratio η of the signal received at the base station is determined.

[0019] Preferably, the signal to interference and noise ratio η of the signal received at the base station is expressed as:

[0020]

[0021] in, Indicates the background noise power at the receiving end.

[0022] Preferably, in step S3, the problem model with maximizing the signal to interference noise ratio of the anti-interference system as the objective function is:

[0023]

[0024]

[0025]

[0026] ||w||=1

[0027] Among them, d min The minimum antenna spacing to prevent coupling effects between adjacent antenna units, η is the signal-to-interference-to-noise ratio of the received signal at the base station, x n is the position of the nth antenna element, w represents the beamforming at the receiving end.

[0028] Preferably, in step S4, the problem formed in step S3 is decomposed into a receiving beam sub-problem and an antenna unit positioning sub-problem; the output based on the neural network is The actual antenna unit position is obtained as follows:

[0029]

[0030]

[0031] Among them, x1 is the position of the first antenna unit, x n is the position of the nth antenna element, d min is the minimum spacing between adjacent antenna elements, Indicates that the distance between adjacent antenna elements exceeds the specified spacing constraint. x N is the position of the Nth antenna unit, L is the movable area of ​​the antenna unit, N is the number of antenna units, r n is the output of the neural network.

[0032] Preferably, the receive beam subproblem:

[0033]

[0034] st||w||=1

[0035] Antenna unit positioning sub-problem:

[0036]

[0037]

[0038]

[0039] Wherein, η is the signal-to-interference-noise ratio of the signal received at the base station.

[0040] Preferably, in step S5, for the receive beamforming subproblem, matrix A and matrix B are Hermitian matrices, eigenvalue decomposition is performed, and the normalized eigenvector corresponding to the maximum eigenvalue is used as the desired receive beamformer. The receive beamforming optimizer is realized by solving the generalized Rayleigh entropy problem. Matrix A and matrix B are respectively:

[0041] A=a(x,θ0)a H (x,θ0)

[0042]

[0043] Among them, I N is the N-dimensional identity matrix, Indicates the background noise power at the receiving end.

[0044] Preferably, in step S6, the loss function is defined as:

[0045]

[0046] Among them, Θ is the collective parameter of the neural network, In order to input the obtained receive beamforming and antenna unit position into the loss function for neural network training and obtain a set of anti-interference strategies, θ k is the direction from which the kth jammer node transmits the signal to the base station, η is the signal-to-interference-noise ratio of the signal received at the base station.

[0047] In a second aspect, an embodiment of the present invention provides a deep learning-assisted interference suppression system based on a movable antenna array, comprising:

[0048] The scenario module constructs a confrontation scenario in which there are several jammers and a single legitimate source node communicating with the base station, and deploys a movable antenna array at the base station;

[0049] The calculation module obtains the steering vector of the movable antenna array from the transmitter to the base station based on the given direction of arrival of the known transmitter signal in the constructed confrontation scenario; models the channel gain between the source node and the jammer and the base station, obtains the expression of the received signal at the base station, and calculates the instantaneous signal-to-interference-noise ratio;

[0050] The function module constructs a problem model with receiving beamforming and antenna unit position as variables and maximizing instantaneous signal-to-interference-noise ratio as the objective function according to the given direction of arrival of the transmitter signal;

[0051] The mapping module uses deep learning technology to achieve adaptive and efficient strategy design based on the problem model, extracts the arrival direction from the legitimate source nodes and jammers, and inputs it into the designed deep learning system; obtains the set of ratios of the minimum antenna unit spacing to the actual antenna unit spacing, and obtains the actual antenna unit position through remapping; based on the actual antenna unit position, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-to-noise ratio under the current antenna unit deployment conditions;

[0052] The suppression module inputs the actual antenna unit position and the optimal beamforming into the loss function for neural network training; repeat until the loss function converges, and obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal-to-interference-noise ratio of this anti-interference system.

[0053] In a third aspect, a computer device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned deep learning assisted interference suppression method based on a movable antenna array when executing the computer program.

[0054] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, including a computer program, which, when executed by a processor, implements the steps of the above-mentioned deep learning assisted interference suppression method based on a movable antenna array.

[0055] In a fifth aspect, a chip comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the above-mentioned deep learning assisted interference suppression method based on a movable antenna array are implemented.

[0056] In a sixth aspect, an embodiment of the present invention provides an electronic device, comprising a computer program, which, when executed by the electronic device, implements the steps of the above-mentioned deep learning assisted interference suppression method based on a movable antenna array.

[0057] Compared with the prior art, the present invention has at least the following beneficial effects:

[0058] A deep learning-assisted interference suppression method based on a movable antenna array considers legitimate communication links with multiple jammers and proposes to deploy a movable antenna array on the receiver to counter interference attacks. The problem is formulated as maximizing the system signal-to-interference-to-noise ratio, and the receive beamforming and antenna element position are taken as variables to be optimized. From the optimization point of view, this problem is a non-convex problem and is difficult to solve effectively. Specifically, in addition to the non-convex objective function, the constraints of the spacing between adjacent antennas and the constraint that the modulus of the beam vector is 1 also induce some difficult-to-handle non-convex regions. In addition, the positioning of the antenna element involves complex exponential functions, which are difficult to handle. In addition, the beamforming and the antenna element position are mutually coupled and affect each other. Based on this, the present invention develops a deep learning-based framework, in which the beamforming is solved as a Rayleigh entropy problem, and the antenna element positioning is solved by multi-layer perceptron training. The stochastic gradient descent method is used to optimize the neural network parameters to achieve the purpose of effectively suppressing interference. Numerical results show that this method achieves near-optimal anti-interference performance while significantly improving the efficiency of strategy determination.

[0059] Furthermore, we first construct an adversarial model in which there are several jammers and a single legitimate source node communicating with the base station. At the same time, a movable antenna array is deployed at the base station to achieve the function of mitigating interference, which will serve as the communication scenario for our subsequent analysis and modeling.

[0060] Furthermore, according to the communication scenario that has been constructed, based on the given direction of arrival of the known transmitter signal, the steering vector from the transmitter to the movable antenna array at the base station is obtained. At the same time, the channel gain between the source node and the jammer and the base station is modeled. Based on this, we can derive the expression of the received signal at the base station and further calculate the instantaneous signal-to-interference-to-noise ratio at the base station.

[0061] Furthermore, according to the derived instantaneous signal to interference and noise ratio at the base station and the given direction of arrival of the transmitter signal, a problem model is constructed with receive beamforming and antenna unit position as variables to maximize the signal to interference and noise ratio of this anti-interference system as the objective function.

[0062] Furthermore, by analyzing the problem model established above, it can be seen that the problem is non-convex. Solving the problem with traditional methods usually requires some iterative approximation processes, which may be quite time-consuming. Therefore, we use deep learning technology to achieve adaptive and efficient strategy design, extract the arrival direction from the legitimate source nodes and jammers, and input it into the designed deep learning system. The set of ratios of the minimum antenna unit spacing to the actual antenna unit spacing is obtained, and the actual antenna unit position is obtained by remapping.

[0063] Furthermore, since it is also necessary to seek the optimal beamforming, by analyzing the problem with beamforming as the optimization variable, it can be seen that this problem conforms to the architecture of the Rayleigh entropy problem. Therefore, based on the antenna unit position obtained in the previous step, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-noise ratio under the current antenna unit deployment conditions.

[0064] Furthermore, based on the obtained antenna unit position and receive beamforming, they are input together into the loss function for neural network training until the loss function converges, and the optimal strategy combination of beamforming and antenna element position is obtained to maximize the signal-to-interference-noise ratio of this anti-interference system.

[0065] It can be understood that the beneficial effects of the second to sixth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0066] In summary, the present invention applies the movable antenna array technology to communication scenarios with interference and adopts machine learning methods to solve complex problems. In this regard, regardless of the complex mathematical form, the neural network can approximate the intrinsic relationship between the input and output of the system. In addition, the learning dynamically simulates the interactive strategy update to achieve the best anti-interference performance.

[0067] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0069] Figure 1 A model diagram of a wireless transmission system constructed for the present invention;

[0070] Figure 2 is a flow chart of the method of the present invention;

[0071] Figure 3 A schematic diagram showing the relationship between the signal-to-interference-and-noise ratio of the receiver of the system of the present invention and the number of movable antenna units;

[0072] Figure 4 A schematic diagram showing the relationship between the signal-to-interference-noise ratio of the receiver and the number of jammers in the system of the present invention;

[0073] Figure 5 A schematic diagram of the relationship between the signal-to-interference-noise ratio of the receiver of the system of the present invention and the movement range of the antenna unit under different schemes;

[0074] Figure 6 A schematic diagram of a computer device provided by an embodiment of the present invention;

[0075] Figure 7 The block diagram of a chip provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0076] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0077] In the description of the present invention, it should be understood that the terms “include” and “comprises” indicate the presence of described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0078] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0079] It should be further understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects are in an "or" relationship.

[0080] It should be understood that, although the terms first, second, third, etc. may be used to describe preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0081] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0082] Various structural schematic diagrams according to the disclosed embodiments of the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified for the purpose of clear expression, and some details may be omitted. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships between them are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs. The present invention provides a deep learning-assisted interference suppression method based on a movable antenna array, revealing the potential of movable antennas in enhancing anti-interference communications. A legitimate communication link with multiple jammers is considered, and it is recommended to deploy a movable antenna array on the receiver to counter interference attacks. First, a scenario in which multiple jammers and a single source node communicate with a base station is constructed, and the instantaneous signal to noise ratio of the receiver is analyzed and obtained, and a problem model is formed with the goal of maximizing the signal to noise ratio of this system, and the receive beamforming and antenna unit positioning are variables; then a deep learning-based framework is developed, in which beamforming is solved as a Rayleigh entropy problem, and antenna unit positioning is solved by multi-layer perceptron training. The stochastic gradient descent method is used to optimize the neural network parameters to effectively suppress interference.

[0083] See also Figure 2 The present invention provides a deep learning-assisted interference suppression method based on a movable antenna array, comprising the following steps:

[0084] S1. Construct a confrontation scenario where there are several jammers and a single legitimate source node communicating to the base station, and deploy a movable antenna array at the base station to mitigate interference;

[0085] See also Figure 1 , consider a wireless system with legitimate communication pairs in the presence of jamming attacks. A single-antenna legitimate source node sends the desired signal to a multi-antenna base station. At the same time, there are K jammers deployed with single antennas, represented as Interference signals are transmitted on the same frequency band to disrupt legitimate transmissions. In order to defend against interference attacks, an antenna array with movable antenna units is deployed at the base station to facilitate effective interference mitigation. The movable antenna array consists of N antenna units, denoted as Its location is recorded as The moving range is [0, L], where L represents the linear size of the antenna array. The combination of beamforming and antenna unit position strategies is sought to maximize the signal-to-interference-noise ratio of the system to counter interference attacks.

[0086] S2. Based on the given direction of arrival of the known transmitter signal, obtain the steering vector of the movable antenna array from the transmitter to the base station; model the channel gain between the source node and the jammer and the base station, obtain the expression of the received signal at the base station, and calculate the instantaneous signal-to-interference-to-noise ratio;

[0087] According to the locations of the legitimate source nodes and jammer nodes in the system under consideration, the directions in which the legitimate source nodes and jammer nodes transmit signals to the base station are recorded as In particular, the problem of interference mitigation based on movable antennas is dealt with, focusing on the signal direction without explicitly considering the distance between the signal source and the receiver.

[0088] Based on this, the steering vector from the transmitter to the movable antenna array at the base station is obtained as:

[0089]

[0090] Among them, λ represents the wavelength of the signal carrier, and the set element 0 represents the legal source node.

[0091] The base station based on the movable antenna array performs beamforming on the received signal to optimize the receiving direction of the signal. The beamforming vector is represented by w. Then, the beam gain can be expressed as |w H a(x,θ k )| 2 ,

[0092] Based on this, the signal-to-interference-to-noise ratio of the signal received at the base station is expressed as:

[0093]

[0094] in, Indicates the background noise power at the receiving end.

[0095] S3. According to the given direction of arrival of the transmitter signal, a problem model is constructed with receiving beamforming and antenna unit position as variables and maximizing the signal to interference noise ratio of this anti-interference system as the objective function;

[0096] For the system under consideration, it is desirable to defend against malicious jamming attacks in order to protect the desired legitimate transmission links. Therefore, a joint consideration is taken of receive beamforming and antenna element positions so that legitimate reception can be well determined while adequately suppressing interference. In this regard, the flexible movement of antenna elements enables the present invention to achieve this goal by exploiting complex radiation patterns beyond traditional spatial diversity.

[0097] Therefore, according to the given direction of arrival of the transmitter signal, a problem model is constructed with receive beamforming and antenna unit position as variables to maximize the signal-to-interference-noise ratio of this anti-interference system as the objective function. The problem description is as follows:

[0098]

[0099]

[0100]

[0101] ||w||=1

[0102] Among them, d min is the minimum antenna spacing to prevent coupling effects between adjacent antenna units, satisfying (N-1)d min ≤L. Since w represents the beamforming at the receiving end, which is usually a complex vector, in order to ensure that the beamforming does not amplify or attenuate the power of the signal, but only changes the phase direction of the signal, and it is hoped that the signal after beamforming maintains the power of the original signal, the modulus of the beam vector is set to 1, indicating that only the direction and phase of the signal are adjusted without changing the amplitude or energy of the signal.

[0103] S4. Use deep learning technology to achieve adaptive and efficient strategy design, extract arrival directions from legitimate source nodes and jammers, and input them into the designed deep learning system; obtain the set of ratios of minimum antenna unit spacing and actual antenna unit spacing, and obtain the actual antenna unit position by remapping;

[0104] The problem formed in step S3 is a non-convex problem and is difficult to solve effectively. Specifically, in addition to the non-convex objective function, the constraints of the spacing between adjacent antennas and the beam vector modulus of 1 also induce some non-convex areas that are difficult to handle. In addition, the positioning of the antenna unit involves a complex exponential function that is difficult to handle. In addition, beamforming and antenna unit position are mutually coupled and affect each other. In order to optimize performance, it is necessary to analyze the combined effect of the two.

[0105] Due to the non-convex nature of the problem formed in step S3, solving the problem using traditional methods usually requires some iterative approximation processes, which can be quite time-consuming. To address this problem, a deep learning approach is used for efficient algorithm design. In this regard, neural networks are able to approximate the intrinsic relationship between system inputs and outputs regardless of the complex mathematical form. In addition, learning dynamically simulates interactive strategy updates to achieve optimal anti-interference performance. Therefore, a deep learning-assisted approach is proposed as follows.

[0106] Looking back at the interference mitigation problem formulation formed in step S3, a combination of beamforming and antenna unit positioning strategies is sought to achieve the highest signal-to-noise ratio, given the direction of arrival of the transmitter signal. Therefore, the overall architecture for solving the problem takes the direction of arrival of the signal as input and generates an anti-interference strategy. In addition, since the strategy contains two parts, receive beamforming and antenna unit positioning, the properties of the corresponding sub-problems are analyzed and the optimizer is designed accordingly. Based on the design of a single optimizer, the problem formed in step S3 is decomposed into two sub-problems:

[0107] Solve the receive beam subproblem:

[0108]

[0109] st||w||=1

[0110] Solve the antenna unit positioning subproblem:

[0111]

[0112]

[0113]

[0114] For the subproblem of solving the antenna unit positioning, the problem is relatively complex due to the presence of complex exponential functions and non-convex objectives and feasible domains. In order to solve the difficulties from an optimization perspective, a deep neural network is used to learn the complex relationship between the known arrival direction and the desired antenna unit position. In this regard, the general representation ability of the neural network can approximate the objective function to approach the highest SINR. The main challenge here is the non-convex feasible region in this subproblem because the minimum antenna spacing needs to be met.

[0115] For this problem, a series of intermediate variables are introduced To express the ratio of the minimum spacing between adjacent antenna elements to the actual adjacent spacing. Naturally, we get This is consistent with the output range of common neural networks. Therefore, the tentative position of the antenna unit is:

[0116]

[0117] Therein, the first antenna unit is arranged at an initial position without losing versatility.

[0118] It can be seen that the above obtained position obviously satisfies the constraint that the spacing between adjacent antenna units given in the sub-problem needs to satisfy. At the same time, if the maximum allowable movable range L of the antenna unit given in the sub-problem is satisfied, the above result can be used as the actual antenna unit position. Otherwise, in order to ensure the maximum allowable movable range constraint of the antenna unit, the following processing is required:

[0119] For this case, the antenna spacing between adjacent antenna elements is rewritten as:

[0120]

[0121] in, It represents the distance between adjacent antenna elements beyond the specified spacing constraint. Obviously, when δ = d min , it means that the tentative position of the antenna unit shown above is the final position that satisfies all constraints.

[0122] However, if the position of the last antenna unit exceeds the movable range, the antenna units are rearranged by setting the position of the last antenna unit to L. At this time, it can be deduced that:

[0123]

[0124] Through the above analysis, the output based on the neural network is The position of the antenna unit can be uniformly expressed as:

[0125]

[0126] in

[0127]

[0128] S5. Based on the antenna unit position obtained in step S4, the Rayleigh entropy algorithm is used to find the best beamforming that maximizes the system signal to interference noise ratio under the current antenna unit deployment conditions;

[0129] In order to facilitate the study of the receive beamforming problem, the following transformation is introduced:

[0130] A=a(x,θ0)a H (x,θ0)

[0131]

[0132] Among them, I Nis the N-dimensional identity matrix. For the receive beamforming subproblem, the matrices A and B are obviously Hermitian matrices, and the objective function has the form of Rayleigh entropy. Therefore, in order to find the beamforming strategy with the best signal-to-noise ratio, eigenvalue decomposition is performed, and the normalized eigenvector corresponding to the maximum eigenvalue is used as the desired receive beamformer. Therefore, the receive beamforming optimizer can be implemented by solving the generalized Rayleigh entropy problem.

[0133] S6, based on the antenna unit positions and receive beamforming obtained in steps S4 and S5, input them together into the loss function to train the neural network;

[0134] In step S6, the loss function of the network is defined as:

[0135]

[0136] Among them, Θ represents the set parameters of the neural network, Φ(·) means that the obtained receive beamforming and antenna unit position are input into the loss function to train the neural network and obtain a set of anti-interference strategies.

[0137] S7. Repeat steps S4-S6 until the loss function converges, and obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal-to-interference-noise ratio of this anti-interference system.

[0138] Since the original problem is decomposed into two sub-problems and the optimizer is designed accordingly, steps S4-S6 need to be repeated until the loss function converges and the anti-interference strategy is finally obtained.

[0139] Therefore, the entire designed network framework should be:

[0140] First, the arrival direction is extracted from the legitimate source node and the jammer: And input it into the learning system. Through a fully connected feature embedding module, the obtained high-dimensional features are used as the input of the multi-layer perceptron network, and the set of ratios of the minimum antenna spacing to the actual antenna spacing is obtained as According to the physical constraints of antenna spacing, ReLU function is used as the activation function of the input layer and hidden layer, and Sigmoid is used as the activation function of the output layer.

[0141] Note that the range of the neural network output is consistent with the antenna spacing ratio requirement, e.g.

[0142] Then, the actual antenna unit position is obtained by remapping; the obtained antenna unit position is substituted into the calculation matrix A and B, and the matrix A and B are further used to solve the generalized Rayleigh quotient problem to derive the receiving beamforming;

[0143] Finally, the obtained receive beamforming together with the antenna unit position is input into the loss function for neural network training, and finally the anti-interference strategy is obtained.

[0144] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as systems, methods or program products. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which may be collectively referred to herein as "circuits", "modules" or "platforms".

[0145] In another embodiment of the present invention, a deep learning assisted interference suppression system based on a movable antenna array is provided. The system can be used to implement the above-mentioned deep learning assisted interference suppression method based on a movable antenna array. Specifically, the deep learning assisted interference suppression system based on a movable antenna array includes a scene module, a calculation module, a function module, a mapping module and a suppression module.

[0146] Among them, the scenario module constructs a confrontation scenario in which there are several jammers and a single legitimate source node communicating with the base station, and deploys a movable antenna array at the base station;

[0147] The calculation module obtains the steering vector of the movable antenna array from the transmitter to the base station based on the given direction of arrival of the known transmitter signal in the constructed confrontation scenario; models the channel gain between the source node and the jammer and the base station, obtains the expression of the received signal at the base station, and calculates the instantaneous signal-to-interference-noise ratio;

[0148] The function module constructs a problem model with receiving beamforming and antenna unit position as variables and maximizing instantaneous signal-to-interference-noise ratio as the objective function according to the given direction of arrival of the transmitter signal;

[0149] The mapping module uses deep learning technology to achieve adaptive and efficient strategy design based on the problem model, extracts the arrival direction from the legitimate source nodes and jammers, and inputs it into the designed deep learning system; obtains the set of ratios of the minimum antenna unit spacing to the actual antenna unit spacing, and obtains the actual antenna unit position through remapping; based on the actual antenna unit position, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-to-noise ratio under the current antenna unit deployment conditions;

[0150] The suppression module inputs the actual antenna unit position and the optimal beamforming into the loss function for neural network training; repeat until the loss function converges, and obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal-to-interference-noise ratio of this anti-interference system.

[0151] In another embodiment of the present invention, a terminal device is provided, which includes a processor and a memory, the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, graphics processors (GPU), tensor processors (TPU), digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, which are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding functions; the processor described in the embodiment of the present invention can be used for the operation of the deep learning-assisted interference suppression method based on the movable antenna array, including:

[0152] Construct an adversarial scenario in which there are several jammers and a single legitimate source node communicating with the base station, and deploy a movable antenna array at the base station; based on the given direction of arrival of the known transmitter signal in the constructed adversarial scenario, obtain the steering vector from the transmitter to the movable antenna array at the base station; model the channel gain between the source node and the jammer and the base station, obtain the expression of the received signal at the base station, and calculate the instantaneous signal to interference and noise ratio; according to the given direction of arrival of the transmitter signal, construct a problem model with the receive beamforming and antenna unit position as variables and the objective function of maximizing the instantaneous signal to interference and noise ratio; based on the problem model, use deep learning technology to achieve adaptive and high An effective strategy design is implemented to extract arrival directions from legitimate source nodes and jammers and input them into the designed deep learning system; a set of ratios of the minimum antenna unit spacing to the actual antenna unit spacing is obtained, and the actual antenna unit position is obtained by remapping; based on the actual antenna unit position obtained, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-noise ratio under the current antenna unit deployment conditions; the actual antenna unit position and the optimal beamforming are input into the loss function for neural network training; this is repeated until the loss function converges to obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal-to-interference-noise ratio of this anti-jamming system.

[0153] See also Figure 6, the terminal device is a computer device, and the computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the computer program 63 is executed by the processor 61, the deep learning-assisted interference suppression method based on the movable antenna array in the embodiment is implemented. To avoid repetition, it is not described one by one here. Alternatively, when the computer program 63 is executed by the processor 61, the functions of each model / unit in the deep learning-assisted interference suppression system based on the movable antenna array in the embodiment are implemented. To avoid repetition, it is not described one by one here.

[0154] The computer device 60 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will appreciate that Figure 6 This is only an example of the computer device 60 and does not constitute a limitation of the computer device 60. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0155] The processor 61 may be a central processing unit (CPU), or other general-purpose processors, graphics processing units (GPU), tensor processing units (TPU), digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0156] The memory 62 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the computer device 60.

[0157] Furthermore, the memory 62 may include both an internal storage unit of the computer device 60 and an external storage device. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is to be output.

[0158] See also Figure 7 The terminal device 600 is an electronic device, and the electronic device is in the form of a general computing device. The components of the electronic device may include but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0159] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 performs the steps according to various exemplary embodiments of the present invention described in the above method section of this specification. For example, the processing unit 610 can perform the following steps: Figure 2 Follow the steps shown in .

[0160] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .

[0161] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0162] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0163] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.

[0164] In another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium, which is a memory device in a terminal device for storing programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs. It should be noted that the computer-readable storage medium here can be a high-speed RAM memory, or a non-volatile memory (Non-VolatileMemory), such as at least one disk memory.

[0165] One or more instructions stored in a computer-readable storage medium may be loaded and executed by a processor to implement the corresponding steps of the deep learning-assisted interference suppression method based on a movable antenna array in the above embodiment; one or more instructions in the computer-readable storage medium are loaded and executed by the processor as follows:

[0166] Construct an adversarial scenario in which there are several jammers and a single legitimate source node communicating with the base station, and deploy a movable antenna array at the base station; based on the given direction of arrival of the known transmitter signal in the constructed adversarial scenario, obtain the steering vector from the transmitter to the movable antenna array at the base station; model the channel gain between the source node and the jammer and the base station, obtain the expression of the received signal at the base station, and calculate the instantaneous signal to interference and noise ratio; according to the given direction of arrival of the transmitter signal, construct a problem model with the receive beamforming and antenna unit position as variables and the objective function of maximizing the instantaneous signal to interference and noise ratio; based on the problem model, use deep learning technology to achieve adaptive and high An effective strategy design is implemented to extract arrival directions from legitimate source nodes and jammers and input them into the designed deep learning system; a set of ratios of the minimum antenna unit spacing to the actual antenna unit spacing is obtained, and the actual antenna unit position is obtained by remapping; based on the actual antenna unit position obtained, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-noise ratio under the current antenna unit deployment conditions; the actual antenna unit position and the optimal beamforming are input into the loss function for neural network training; this is repeated until the loss function converges to obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal-to-interference-noise ratio of this anti-jamming system.

[0167] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention described and shown in the drawings here can usually be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0168] See also Figure 3 , a schematic diagram of the relationship between the signal-to-interference-to-noise ratio of the system receiver and the number of movable antenna units is given. Figure 3It can be seen that when the number of movable antenna units increases, the anti-interference performance of the system gradually improves when the number of jammers remains unchanged, which is due to the enhancement of beamforming gain. However, although the performance of the scheme proposed in the present invention is inferior to that of the AO scheme (i.e., both the receive beamforming and the antenna unit position are optimized), the AO scheme code runs for a long time, which poses a major challenge to practical engineering applications involving delay-sensitive tasks. In addition, the anti-interference performance of the scheme proposed in the present invention and the AO scheme are better than those of the FPA scheme (the antenna unit position is fixed, and only the receive beamforming is optimized) and the RAB (the antenna unit and the receive beamforming are fixed) scheme, which illustrates the potential of movable antennas in enhancing anti-interference communications.

[0169] See also Figure 4 , a schematic diagram of the relationship between the signal-to-interference-noise ratio of the system receiver and the number of jammers is given. When the number of jammers increases, the number of interference signals received in the system also increases. Each jammer introduces additional interference signals to the system, which increases the complexity of system processing and interference suppression. Since the number of antennas is fixed and the degrees of freedom available to the system are limited, each additional jammer will disperse the system's resources for anti-interference. Similarly, although the scheme proposed in the present invention fails to achieve optimal performance, its reasoning time is shorter than that of the AO scheme. In addition, the code running time of the FPA scheme and the RAB scheme is very short, but their anti-interference performance is not as good as that of the AO scheme and the scheme proposed in the present invention. This is mainly due to the low complexity of the algorithms used in these two schemes, resulting in significant performance losses.

[0170] See also Figure 5, a schematic diagram of the relationship between the signal-to-interference-to-noise ratio of the receiver of the system of the present invention and the change of the moving range of the antenna unit under different schemes is given. The core advantage of the movable antenna is that it can change the propagation path of the signal by adjusting the physical position of the antenna. When the movable range increases, the antenna can select the best position in a larger range, so as to better avoid the interference source. This flexibility allows the system to optimize the position layout of the antenna array to maximize signal reception and minimize the impact of interference signals, thereby improving the anti-interference performance. The adjustment of the antenna position will also affect the effect of beamforming. As the movable range increases, the geometric layout of the antenna can better cooperate with the beamforming technology, accurately point to the target signal, and form nulls to suppress the interference source. In this way, the system can effectively enhance the signal quality and reduce interference, thereby improving the ability to separate signals from interference. Therefore, as the movable range increases, the anti-interference performance of the system will be due to the FPA scheme (the antenna unit position is fixed, only the receive beamforming is optimized) and the RAB (the antenna unit and the receive beamforming are fixed) scheme. The antenna unit position is immovable, so when the movable range of the antenna unit is changed, it will not improve the anti-interference performance of the system.

[0171] In summary, the present invention provides a deep learning-assisted interference suppression method and system based on a movable antenna array, which can counter interference attacks by deploying a movable antenna array on the receiver in a legitimate communication link with multiple jammers. The present invention formulates the problem as a signal-to-noise ratio maximization problem, and develops a deep learning framework to solve beamforming and antenna unit positioning, and uses the stochastic gradient descent method to optimize the parameters of the neural network to achieve the purpose of effectively suppressing interference.

[0172] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0173] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0174] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0175] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0176] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0177] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0178] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0179] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices, and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0180] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0181] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0182] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A deep learning-assisted interference suppression method based on a movable antenna array, characterized in that: The following steps are involved: S1. Construct a confrontation scenario where there are several jammers and a single legitimate source node communicating with the base station, and deploy a movable antenna array at the base station; S2, based on the given direction of arrival of the known transmitter signal in the confrontation scenario constructed in step S1, obtain a steering vector from the transmitter to the movable antenna array at the base station; Model the channel gain between the source node and the jammer and the base station, obtain the expression of the received signal at the base station, and calculate the instantaneous signal-to-interference-to-noise ratio; S3, according to the given direction of arrival of the transmitter signal, construct a problem model with receiving beamforming and antenna unit position as variables and maximizing the instantaneous signal to interference and noise ratio obtained in step S2 as the objective function; S4, based on the problem model obtained in step S3, using deep learning technology to achieve adaptive and efficient strategy design, extracting arrival directions from legitimate source nodes and jammers, and inputting them into the designed deep learning system; A set of ratios of the minimum antenna unit spacing and the actual antenna unit spacing is obtained, and the actual antenna unit position is obtained by remapping; S5. Based on the actual antenna unit position obtained in step S4, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-noise ratio under the current antenna unit deployment conditions; S6, the actual antenna unit position obtained in step S4 and the optimal beamforming obtained in step S5 are input into the loss function for neural network training; S7. Repeat steps S4 to S6 until the loss function converges, and obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal to interference noise ratio of the anti-interference system.

2. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 1, characterized in that: In step S1, in the presence of interference attack, in a wireless system with a legitimate communication pair, a single-antenna legitimate source node sends a desired signal to a multi-antenna base station; at the same time, there are K jammers deployed with single antennas transmitting interference signals on the same frequency band, expressed as An antenna array with movable antenna units is deployed at the base station. The movable antenna array consists of N antenna units, denoted as The position is recorded as The moving range is [0, L], where L represents the linear size of the antenna array. The goal is to maximize the signal-to-interference-noise ratio of the system by seeking a combination of beamforming and antenna unit position strategies.

3. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 1, characterized in that: In step S2, according to the positions of the legitimate source node and the jammer node in the system under consideration, the directions in which the legitimate source node and the jammer node transmit signals to the base station are denoted as θ0 and θ1 respectively. Get the steering vector a(x,θ) from the transmitter to the movable antenna array at the base station k ); The base station based on the movable antenna array performs beamforming operation on the received signal to optimize the receiving direction of the signal. The beamforming vector is represented by w, and the beam gain is represented by The signal-to-interference-and-noise ratio η of the signal received at the base station is determined.

4. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 3 is characterized in that: The signal-to-interference-to-noise ratio η of the signal received at the base station is expressed as: in, Indicates the background noise power at the receiving end.

5. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 1, characterized in that: In step S3, the problem model with maximizing the signal to interference noise ratio of the anti-interference system as the objective function is: ||w||＝1 Among them, d min The minimum antenna spacing to prevent coupling effects between adjacent antenna units, η is the signal-to-interference-to-noise ratio of the received signal at the base station, x n is the position of the nth antenna element, w represents the beamforming at the receiving end.

6. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 1, characterized in that: In step S4, the problem formed in step S3 is decomposed into a receiving beam subproblem and an antenna unit positioning subproblem; the output based on the neural network is The actual antenna unit position is obtained as follows: other Among them, x1 is the position of the first antenna unit, x n is the position of the nth antenna element, d min is the minimum spacing between adjacent antenna units, Indicates that the distance between adjacent antenna elements exceeds the specified spacing constraint. x N is the position of the Nth antenna unit, L is the movable area of ​​the antenna unit, N is the number of antenna units, r n is the output of the neural network.

7. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 6, characterized in that: Receive beam subproblem: st||w||=1 Antenna unit positioning sub-problem: Wherein, η is the signal-to-interference-noise ratio of the signal received at the base station.

8. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 1, characterized in that: In step S5, for the receive beamforming subproblem, matrix A and matrix B are Hermitian matrices, and eigenvalue decomposition is performed, and the normalized eigenvector corresponding to the maximum eigenvalue is used as the desired receive beamformer. The receive beamforming optimizer is realized by solving the generalized Rayleigh entropy problem. Matrix A and matrix B are respectively: A=a(x,θ0)a H (x,θ0) Among them, I N is the N-dimensional identity matrix, Indicates the background noise power at the receiving end.

9. The deep learning-assisted interference suppression method based on a movable antenna array according to claim 1, characterized in that: In step S6, the loss function is defined as: Among them, Θ is the collective parameter of the neural network, In order to input the received beamforming and antenna unit position into the loss function for neural network training and obtain a set of anti-interference strategies, θ k is the direction from which the kth jammer node transmits the signal to the base station, η is the signal-to-interference-noise ratio of the signal received at the base station.

10. A deep learning-assisted interference suppression system based on a movable antenna array, characterized in that: include: The scenario module constructs a confrontation scenario in which there are several jammers and a single legitimate source node communicating with the base station, and deploys a movable antenna array at the base station; A calculation module, based on a given direction of arrival of a known transmitter signal in the constructed confrontation scenario, obtains a steering vector from the transmitter to the movable antenna array at the base station; Model the channel gain between the source node and the jammer and the base station, obtain the expression of the received signal at the base station, and calculate the instantaneous signal-to-interference-to-noise ratio; The function module constructs a problem model with receiving beamforming and antenna unit position as variables and maximizing instantaneous signal-to-interference-noise ratio as the objective function according to the given direction of arrival of the transmitter signal; The mapping module, based on the problem model, uses deep learning technology to achieve adaptive and efficient strategy design, extracts arrival directions from legitimate source nodes and jammers, and inputs them into the designed deep learning system; A set of ratios of the minimum antenna unit spacing and the actual antenna unit spacing is obtained, and the actual antenna unit position is obtained by remapping; Based on the actual antenna unit position, the Rayleigh entropy algorithm is used to find the optimal beamforming that maximizes the system signal-to-interference-noise ratio under the current antenna unit deployment conditions; The suppression module inputs the actual antenna unit position and the optimal beamforming into the loss function for neural network training; repeat until the loss function converges, and obtain the optimal strategy combination of beamforming and antenna element position to maximize the signal-to-interference-noise ratio of this anti-interference system.

Images