A phased array antenna beamforming control method, system, medium and product

Abstract

This application provides a phased array antenna beamforming control method, system, medium, and product. In this method, the communication system monitors the instantaneous received power in real time at the RF receiving front-end. When the instantaneous received power exceeds a preset power threshold, the communication system extracts interference feature signatures and performs a lookup table mapping to obtain initial anti-interference beam weights, thereby generating an initial spatially suppressed beam. Subsequently, after the analog-to-digital converter resumes normal linear sampling, the communication system acquires digitized received signal samples and calculates the target beam weights using a deep reinforcement learning network model. Finally, the communication system calculates the weight difference and divides it into smooth interpolation steps, sequentially distributing intermediate beam weights over multiple preset periods for gradual transition, avoiding RF amplitude and phase jumps caused by a single full weight jump. This method alleviates the technical problem of communication service interruption under sudden high-power interference and improves the stability of signal demodulation.

Description

Technical Field

[0001] This application relates to the field of large-scale antenna technology, and in particular to a phased array antenna beamforming control method, system, medium and product. Background Technology

[0002] Phased array antennas are widely used in communication systems to enhance the physical layer's anti-interference capabilities. Systems typically utilize beamforming technology, adjusting the amplitude and phase weights of the array antenna's receiving channels to create high gain in the desired signal direction and spatial nulls in the interference direction, thereby suppressing ineffective interference and ensuring communication quality.

[0003] In related technologies, phased array receiving systems employ an adaptive beam generation scheme. During implementation, the RF front-end sends the acquired analog signal to an analog-to-digital converter for full sampling; the digital signal processor uses the digital signal output from the analog-to-digital converter to perform algorithmic estimation to obtain the corresponding beam weights; subsequently, the control unit sends the full beam weights to the RF channel at once, replacing the current working weights to achieve beam switching.

[0004] However, when dealing with sudden high-power interference, the existing beamforming system relies on direct analog-to-digital converter (ADC) calculations and a single rewriting of beam weights. When high-power interference enters the system, the saturation distortion of the sampled data caused by the ADC's limited range often merges with the amplitude and phase jumps induced by the sudden reduction of new weights in the RF channel. When using this method for intervention, the cross-effects and transient impacts caused by sampling pauses and hardware abrupt changes disrupt the stable environment required for continuous signal demodulation, ultimately leading to service interruption. Summary of the Invention

[0005] This application provides a phased array antenna beamforming control method, system, medium, and product to alleviate the technical problem of communication service interruption under sudden high-power interference and improve the stability of signal demodulation.

[0006] In a first aspect, this application provides a beamforming control method for a phased array antenna, applied in a communication system. The communication system includes a radio frequency (RF) receiver front-end and an analog-to-digital converter (ADC). The method includes: real-time monitoring of the instantaneous received power of the RF received signal at the RF receiver front-end; when the instantaneous received power exceeds a preset power threshold, extracting an interference feature signature of the RF received signal, wherein the preset power threshold is less than the saturation power of the ADC; performing a lookup mapping in a preset beam immunity map library to obtain an initial anti-interference beam weight corresponding to the interference feature signature; and, based on the initial anti-interference beam weight, driving the RF receiver front-end to generate an initial spatial suppression beam, thereby attenuating the RF received signal power input to the ADC to the linearity of the ADC. Within the sampling interval; when the analog-to-digital converter is in the linear sampling interval, the RF received signal is sampled using the RF converter to obtain a digital received signal sample. The digital received signal sample is input into a pre-trained deep reinforcement learning network model to obtain the target beam weights. The complex weight difference between the initial anti-interference beam weights and the target beam weights is calculated, and the complex weight difference is divided into smooth interpolation steps. In multiple consecutive preset periods, intermediate beam weights transitioning from the initial anti-interference beam weights to the target beam weights are generated sequentially according to the smooth interpolation steps. The intermediate beam weights are sent to the RF receiving front end sequentially according to the preset period until the currently effective beam weights are completely updated to the target beam weights.

[0007] By adopting the above technical solution, the communication system monitors the instantaneous received power in real time at the RF receiving front-end. Under the premise that the preset power threshold is less than the saturation power of the analog-to-digital converter (ADC), a protection mechanism is triggered in time before the ADC enters saturation distortion. Furthermore, the communication system extracts interference feature signatures and obtains initial anti-interference beam weights by looking up a table, generating an initial spatially suppressed beam. This reduces the interference power entering the ADC at the physical layer, controlling it within the linear sampling range. Subsequently, after the ADC resumes normal linear sampling, the communication system acquires digital received signal samples and calculates the target beam weights using a deep reinforcement learning network model. Finally, the communication system calculates the weight difference and divides it into smooth interpolation steps, sequentially issuing intermediate beam weights over multiple preset periods for gradual transition, avoiding RF amplitude and phase jumps caused by a single full-weight jump. This method alleviates the technical problem of communication service interruption under sudden high-power interference and improves the stability of signal demodulation.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, extracting the interference feature signature of the radio frequency received signal specifically includes: acquiring the analog power detection values ​​of multiple receiving channels in the radio frequency receiving front end, each receiving channel having different spatial response characteristics; determining the interference angle of arrival of the interference component in the radio frequency received signal based on the relative power distribution characteristics among the multiple analog power detection values; performing gain compensation calculation on each analog power detection value based on the interference angle of arrival and the pre-calibrated spatial response gain value of each receiving channel in the direction of the interference angle of arrival to obtain the instantaneous power value of the interference component; and generating the interference feature signature of the radio frequency received signal based on the interference angle of arrival and the instantaneous power value.

[0009] By adopting the above technical solution, the communication system acquires simulated power detection values ​​from multiple receiving channels with different spatial response characteristics. This is combined with determining the interference angle of arrival based on relative power distribution characteristics to achieve spatial direction finding. Furthermore, coupling the interference angle of arrival with a pre-calibrated spatial response gain value for gain compensation calculations mitigates the detection bias introduced by the inconsistency in the directivity of the antenna array itself. This method recreates the true instantaneous power at the simulated radio frequency end, ensuring that the communication system maintains high-precision power sensing capabilities even under strong interference.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the construction process of the preset beam immune spectrum library specifically includes: determining the interference parameter space corresponding to the dimension contained in the interference feature signature; performing grid-based discretization on the interference parameter space to obtain multiple interference scene grid points, wherein the interference parameter space includes at least the interference angle of arrival dimension; calculating the anti-interference beam weights corresponding to each interference scene grid point based on the array manifold of the radio frequency receiving front end; establishing an index mapping relationship between the parameter combination represented by each interference scene grid point and the anti-interference beam weights to obtain the preset beam immune spectrum library.

[0011] By adopting the above technical solution, the communication system will discretize the interference parameter space, including the interference angle of arrival dimension, into a grid, and integrate it with the anti-interference beam weights calculated in advance for each discrete grid point based on the RF receiver front-end array manifold, and establish an index mapping relationship between parameter combinations and weights. This method transforms the complex spatial array matrix calculation that originally needed to be performed after a sudden interference intrusion into an offline static address lookup, reducing the delay time for the communication system to generate the initial anti-interference beam.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, after calculating the anti-interference beam weights corresponding to each interference scene grid point, the method further includes: for each interference scene grid point, determining the interference angle of arrival corresponding to the interference scene grid point in the interference angle of arrival dimension; calculating the beam gain value of the anti-interference beam weights corresponding to the interference scene grid point in the interference angle of arrival; when the beam gain value is greater than or equal to a preset beam gain threshold, performing constraint optimization correction on the anti-interference beam weights corresponding to the interference scene grid point until the corrected beam gain value is less than the beam gain threshold, obtaining the corrected anti-interference beam weights; and replacing the corresponding anti-interference beam weights with the corrected anti-interference beam weights.

[0013] By adopting the above technical solution, the communication system combines the calculated beam gain value of the beam weights corresponding to the grid points in the interference scene at the interference arrival angle with the constraint optimization correction triggered when the beam gain value is greater than or equal to a preset threshold. When the discretely calculated weights have poor suppression, the communication system reduces the response gain in that direction until the closed-loop constraint requirements are met. This method ensures that each mapping weight in the pre-stored beam immunity map library has a spatial null of a defined depth.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, a lookup mapping is performed in a preset beam immune map library to obtain the initial anti-interference beam weights corresponding to the interference feature signature. Specifically, this includes: calculating the feature distance between the interference feature signature and each interference scene grid point in the preset beam immune map library, and determining the minimum feature distance; when the minimum feature distance is less than a preset matching threshold, using the anti-interference beam weight of the interference scene grid point corresponding to the minimum feature distance as the initial anti-interference beam weight; when the minimum feature distance is greater than or equal to the matching threshold, selecting multiple adjacent interference scene grid points with the minimum feature distance; and weighting and fusing the anti-interference beam weights of adjacent interference scene grid points with the reciprocal of the feature distance between each adjacent interference scene grid point and the interference feature signature as the weight to obtain the initial anti-interference beam weight.

[0015] By adopting the above technical solution, the communication system calculates the minimum feature distance between the interference feature signature and the grid points in the spectrum library. When the minimum feature distance is greater than or equal to the matching threshold, adjacent grid points are selected and weighted by the reciprocal of the distance. When the interference feature falls into the preset grid gap, the communication system reconstructs the response weight ratio of the surrounding spatial grid, alleviating the inherent truncation error of manual quantization segmentation of the parameter space and improving the beam suppression reconstruction capability.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, intermediate beam weights that transition from the initial anti-interference beam weights to the target beam weights are generated sequentially according to a smooth interpolation step size. Specifically, this includes: generating candidate intermediate beam weights according to the current smooth interpolation step size in each preset period; determining the angle of arrival (AHA) of the interference component in the RF received signal based on the interference feature signature, and calculating the beam response value of the candidate intermediate beam weights in the direction corresponding to the AHA; when the beam response value is less than a preset response threshold, determining the candidate intermediate beam weights as the intermediate beam weights for the current preset period; when the beam response value is greater than or equal to the response threshold, adjusting the weights of the candidate intermediate beam weights with the constraint that the beam response value at the AHA is less than the response threshold, to obtain the corrected intermediate beam weights; and determining the corrected intermediate beam weights as the intermediate beam weights for the current preset period.

[0017] By adopting the above technical solution, the communication system will generate candidate intermediate beam weights according to a smooth interpolation step size in each preset period, and integrate this with the calculation of the beam response value corresponding to the interference arrival angle of the candidate weights and the directional constraint adjustment when the value is greater than or equal to the response threshold. In addressing the potential for physical space null distortion degradation caused by mathematical linear interpolation of complex domain weights, this method reduces the risk of transient high-power interference penetrating physical defenses during the interpolation conversion evolution process.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the complex weight difference is divided into smooth interpolation step sizes, specifically including: determining the first spatial suppression angle interval corresponding to the initial anti-interference beam weight and the second spatial suppression angle interval corresponding to the target beam weight based on the interference feature signature; calculating the angular offset between the first spatial suppression angle interval and the second spatial suppression angle interval; determining the number of divisions of the complex weight difference based on the angular offset and the minimum value between the angular width of the first spatial suppression angle interval and the angular width of the second spatial suppression angle interval; and dividing the complex weight difference equally according to the number of divisions to obtain the smooth interpolation step size.

[0019] By employing the above technical solution, the communication system calculates the angular offset of the spatial suppression angle interval corresponding to the initial and target beam weights, and combines this offset with the determination of the number of complex weight difference divisions based on the minimum angular width of the two intervals. The communication system establishes a mathematically constrained closed loop by integrating the segmentation of the complex weight domain with the geometric characteristics of the suppression region in the spatial radiation domain. This method alleviates the problem of spatial boundary protection faults caused by excessively large iteration spans of the weight step size.

[0020] In a second aspect, this application provides a communication system including a radio frequency receiving front end and an analog-to-digital converter; the communication system includes one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the communication system to perform the methods described in the first aspect and any possible implementation thereof.

[0021] Thirdly, this application provides a computer-readable storage medium including computer instructions that, when executed on a communication system, cause the communication system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, this application provides a computer program product, including a computer program / instruction that, when run on a communication system, causes the communication system to perform the method described in the first aspect and any possible implementation thereof.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By employing a technique of rapidly generating the initial beam by looking up a table before the analog-to-digital converter saturates to suppress interference and restore linear sampling, and combining this with reinforcement learning to calculate the target weights and smoothly evolve and transmit them in multiple cycles according to the step size, the technical problems of receiver overload shutdown caused by high-power burst interference and RF transient impact caused by single full-scale phase modulation are effectively alleviated. Thus, the technical effect of ensuring beam effectiveness and maintaining continuous and stable communication demodulation is achieved while avoiding the failure of the underlying hardware.

[0024] 2. By adopting a technique that uses a threshold-based feature matching degree determination, extracts adjacent discrete grid points for non-matching features, and combines the reciprocal of the feature distance as a weight for spatial weighted fusion reconstruction of beam weights, the inherent technical problems of parameter spatial quantization truncation error and insufficient non-node direction suppression accuracy of a single static lookup table mapping are effectively alleviated. This achieves the technical effect of giving the discrete spectrum library a smooth output of high-precision continuous suppressed beams in any irregular direction.

[0025] 3. By adopting the technical means of continuously verifying the physical response of the spatial radiation domain during the weight interpolation transition period and performing single-point directional correction constraint of the main interference direction based on the preset suppression extreme value for out-of-bounds candidate weights, the technical problems of physical space zero-depression distortion and transient high-power interference penetration are effectively alleviated. This achieves the technical effect of forcibly locking the lower limit of spatial suppression of the transition dynamic beam and ensuring that the protection is not degraded throughout the beam switching process. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of an application scenario of the phased array antenna beamforming control method in the embodiments of this application; Figure 2 This is a flowchart illustrating a phased array antenna beamforming control method in an embodiment of this application. Figure 3 This is another schematic flowchart of the phased array antenna beamforming control method in the embodiments of this application; Figure 4 This is a schematic diagram of the physical device structure of a communication system in an embodiment of this application. Detailed Implementation

[0027] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0028] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0029] Below is an application scenario of the method provided in this implementation. Please refer to [link / reference]. Figure 1 This is a schematic diagram of an application scenario of the phased array antenna beamforming control method in the embodiments of this application.

[0030] like Figure 1 As shown, when external radio frequency signals (including desired signals and sudden strong interference) arrive at the phased array antenna, the communication system as a whole performs the following beamforming control process: First, preliminary anti-interference suppression is performed. The communication system monitors the instantaneous received power of the signal received by the RF receiving front-end in real time. When this power exceeds a preset power threshold (and this threshold is less than the saturation power of the analog-to-digital converter), the communication system extracts the interference feature signature of the signal and performs a lookup mapping in the beam immune spectrum library to quickly obtain the initial anti-interference beam weights. As shown in step A, the communication system drives the RF receiving front-end to generate an initial spatial suppression beam based on these initial weights, causing the input signal power to attenuate and safely fall back into the linear sampling range of the analog-to-digital converter.

[0031] Next, precise optimization is performed. After ensuring that the signal is within the linear sampling range, the communication system uses an analog-to-digital converter to securely sample the signal, obtaining a digital received signal sample. The communication system then inputs this sample into a pre-trained deep reinforcement learning network model for inference, calculating a more accurate target beam weight.

[0032] Finally, a smooth beam transition is performed. To avoid link oscillation caused by beam abrupt changes, the communication system calculates the complex weight difference between the initial anti-interference beam weight and the target beam weight, and divides it into smooth interpolation steps. As shown in step B, the communication system sequentially generates smooth transition intermediate beam weights within multiple preset periods and sends them to the RF receiving front end periodically until the currently effective beam weight is smoothly and completely updated to the target beam weight.

[0033] The following describes the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2 This is a flowchart illustrating a phased array antenna beamforming control method in an embodiment of this application.

[0034] 201. Real-time monitoring of the instantaneous received power of the radio frequency receiving signal at the radio frequency receiving front end.

[0035] The radio frequency (RF) receiver front end refers to the physical circuit module located between the antenna element and the analog-to-digital converter, which is responsible for amplifying and filtering microwave electrical signals; the RF received signal refers to the analog electromagnetic wave received by the antenna array from the spatial electromagnetic field and fed into the communication system; the instantaneous received power refers to the electromagnetic energy carried by the RF received signal within a preset time window.

[0036] Specifically, the communication system incorporates a power detection circuit in the analog signal path of the RF receiving front-end to sample the continuously flowing RF received signal during the high-frequency analog phase. The communication system calculates the square of the current envelope voltage of the RF received signal to obtain the instantaneous received power, characterizing the energy level. The communication system continuously collects and outputs the instantaneous received power data at preset time intervals, thereby acquiring the energy fluctuation state in the RF link before the analog-to-digital converter performs digitization, and thus determining whether the RF receiving front-end experiences sampling distortion due to power overload.

[0037] The communication system sets a preset power threshold value lower than the analog-to-digital converter's (ADC) saturation power. Within this power margin range—that is, when the instantaneous received power exceeds the preset power threshold but has not yet reached or has exceeded the ADC's saturation power—the analog signal path of the RF receiver front-end and its analog power detection circuit continue to operate normally, providing analog-domain physical quantity measurement inputs for subsequent interference signature extraction. This design ensures that the communication system still has the physical basis for interference sensing and response relying on analog RF hardware during the time window when the ADC may lose its effective sampling capability.

[0038] 202. When the instantaneous received power exceeds the preset power threshold, the interference feature signature of the radio frequency received signal is extracted. The preset power threshold is less than the saturation power of the analog-to-digital converter.

[0039] The preset power threshold refers to the comparison reference constant that is fixed in the communication system and used as the trigger condition for the beam lookup intervention mechanism; the interference feature signature represents the array data of the reduced-dimensional representation of the interference signal in the radio frequency received signal in various physical attributes; the analog-to-digital converter refers to the internal semiconductor device that converts the time-continuous voltage signal output by the radio frequency receiving front end into a discrete digital logic signal sequence; the saturation power refers to the critical signal power when the input voltage signal of the analog-to-digital converter exceeds the range of its internal linear quantizer.

[0040] Specifically, the communication system compares the measured instantaneous received power value with a preset power threshold. When the comparison result indicates that the instantaneous received power value is greater than the preset power threshold, the communication system initiates the spatial feature extraction and analysis procedure. Since the preset power threshold is set to be less than the saturation power, the communication system performs feature extraction at the time point before the analog-to-digital converter falls into a nonlinear saturation distortion state due to overload. The communication system performs multi-channel dimension physical parameter calculation and direction finding on the RF received signal, and extracts the parameter parameters of the RF received signal in the dimensions of spatial distribution and energy intensity. The feature parameter arrays of different dimensions are combined to construct an interference feature signature that can uniquely identify the physical characteristics of the strong interference.

[0041] Since the preset power threshold is set to be less than the saturation power of the analog-to-digital converter (ADC), the interference signature extraction process triggered when the instantaneous received power exceeds the preset power threshold relies on the power detection circuit in the analog signal path of the RF receiver front-end, rather than the digital sampling output of the ADC, for both signal sensing and calculation. Multiple receiving channels in the RF receiver front-end are each equipped with an independent analog power detection circuit. Each receiving channel exhibits different spatial response characteristics due to differences in the spatial position of the antenna elements or the signal distribution characteristics of the analog beamforming network. The communication system can estimate the interference direction and extract features based on the relative distribution relationship between the analog power detection values ​​of each channel. Therefore, regardless of whether the ADC is in normal sampling mode or has approached or entered the saturation range at that moment, the interference signature extraction process can be completed independently of the ADC and is unaffected by its operating state.

[0042] In some embodiments, interference signatures of radio frequency received signals can be extracted by the following method: The communication system acquires the analog power detection values ​​output by multiple receiving channels in the radio frequency receiving front-end, wherein each receiving channel has different and known spatial response physical parameter characteristics. At this time, the radio frequency receiving front-end is configured with direction-specific antenna array elements (such as heterogeneous antennas with radiation patterns tilted at a specific angle) or a multi-beam analog matrix network (such as a Butler matrix) is cascaded in its internal analog link, so that the directional information originally carried in the spatial phase difference is mapped into obvious amplitude differences between different analog receiving channels after the above hardware construction.

[0043] Specifically, taking a cascaded multi-beam analog matrix network (such as a Butler matrix) as an example, it is essentially a passive microwave network composed of microwave directional couplers and fixed phase shifters connected in an alternating manner. The number of its input ports corresponds to the number of elements in the antenna array, and the number of its output ports corresponds to the subsequent RF receiving channels. When a sudden high-power interference signal with a specific angle of arrival exists in the external space, the interference signal will exhibit a physical spatial phase difference that is strictly trigonometrically related to the angle of arrival when it reaches each element of the uniformly arranged phased array antenna. When this set of RF received signals carrying the spatial phase difference is simultaneously fed into the multi-beam analog matrix network, the passive network actually performs an equivalent spatial discrete Fourier transform at the simulated physical transmission layer.

[0044] This physical transformation process causes the interference signal energy, which was originally uniformly distributed but phase-progressive, to undergo coherent constructive and destructive interactions within the network. Ultimately, its output energy is highly focused and converges at one or several adjacent specific output ports of the matrix network. Therefore, the envelope detectors or logarithmic amplifiers set in different receiving channels will capture analog power envelope voltages with significant step differences. In other words, this hardware architecture pre-maps the phase difference direction finding, which in conventional communication systems relies solely on digital baseband or analog-to-digital converters, to the amplitude distribution determination in the analog domain. Based on this, the communication system only needs to extract the analog power detection values ​​fed back from each receiving channel at the same time. By using a multi-port amplitude comparison direction finding algorithm to perform ratio fitting interpolation on the analog power of the adjacent ports with the highest convergence energy, even in the extreme case where the analog-to-digital converter loses its digital sampling and calculation capabilities due to complete saturation, it can solve and lock the physical incident direction (interference angle of arrival) of sudden strong interference signals solely based on the physical response distribution law of the analog hardware. This provides a solid physical parameter basis for subsequent table lookup and shaping.

[0045] Based on the aforementioned amplitude mapping relationship, and according to the distribution pattern and characteristics of the relative power ratio differences among multiple analog power detection values, the communication system employs a multi-port amplitude comparison direction finding algorithm to determine the angle of arrival (AHA) of the interference component in the RF received signal. Based on the obtained AHA and the calibrated spatial response gain values ​​of each receiving channel in the pre-stored AHA direction, the communication system performs gain compensation calibration calculations by dividing each analog power detection value by the corresponding channel's calibrated spatial response gain value, thereby obtaining the instantaneous power value of the interference component after eliminating hardware-related interference. The communication system extracts the AHA data and the instantaneous power value, combines them into a vector, and concatenates them to establish and output the interference feature signature of the RF received signal. The gain calibration calculation eliminates signal direction perception deviations caused by inconsistencies in antenna fabrication processes.

[0046] 203. Perform a lookup mapping in the preset beam immune spectrum library to obtain the initial anti-interference beam weights corresponding to the interference feature signature.

[0047] The beam immunity map library refers to the data form architecture that maps and establishes the relationships between the discrete interference scene grid data and the corresponding optimal beam suppression array parameters; the initial anti-interference beam weights represent a set of complex parameters that are issued and executed by the communication system in the early stages of prevention and intervention to present a spatial suppression and reduction effect on the direction of interference.

[0048] Specifically, the communication system extracts numerical variables from the buffer medium storing interference signatures, scales and transcodes these variables to construct address retrieval condition information that conforms to the underlying construction format and attribute rules of the target beam immune map library. The communication system then calls the built-in beam immune map library. It traverses the discrete data structure to find object class parameter node structures that match the address retrieval condition information. The communication system reads the complex matrix weight parameter combinations stored under the confirmed object class parameter node structures. It copies and extracts these as the system output for this stage, serving as the initial anti-interference beam weights. This allows the use of pre-calculated data from a pre-set database to replace the current hard calculation of the spatial correlation matrix equations.

[0049] In some embodiments, the initial anti-interference beam weights can be obtained by looking up a table in a preset beam immune map library in the following way: The communication system converts the parameter values ​​of each dimension contained in the interference feature signature into the corresponding storage address offset according to the index encoding rules of the beam immune map library, adds them to the base address of the beam immune map library to obtain the storage address of the target data, and reads the corresponding anti-interference beam weight data from the storage address as the initial anti-interference beam weights.

[0050] 204. Based on the initial anti-interference beam weight, drive the RF receiving front end to generate an initial spatial suppression beam, so that the RF received signal power input to the analog-to-digital converter is attenuated to within the linear sampling range of the analog-to-digital converter.

[0051] The initial spatial suppression beam indicates that after the communication system loads the initial anti-interference beam weights onto the RF receiving front end, the signals of each array element channel are weighted and synthesized to form a spatially directional receiving pattern with deep power attenuation in the direction of interference incident. The linear sampling interval refers to the signal power segment where the input signal power of the analog-to-digital converter is within the linear response range of its quantization circuit and the output digital code is linearly proportional to the input voltage.

[0052] Specifically, the communication system sends the initial anti-interference beam weight vector element-by-element to the control modules of each array element channel in the RF receiving front-end. Each array element channel adjusts the attenuation state of the variable attenuator according to the amplitude control value and adjusts the phase offset of the phase shifter according to the phase control value. The communication system thus drives each channel of the RF receiving front-end to apply corresponding amplitude and phase weights to the RF received signal. The weighted signals from each channel are spatially superimposed by the array combining network. In the direction of the incoming interference signal, the signals from each channel form spatial nulls due to phase cancellation after weighting, and the power of the interference component is significantly attenuated during spatial combining. Because the interference component is sufficiently suppressed, the total power of the RF received signal input to the analog-to-digital converter (ADC) falls back to within the linear sampling range of the ADC, ensuring that subsequent sampling by the ADC does not experience digital truncation distortion.

[0053] In some embodiments, the generation of an initial spatially suppressed beam based on the initial anti-interference beam weights can be achieved by driving the RF receiving front-end to generate an initial spatially suppressed beam: The communication system decomposes the complex weights corresponding to each array element channel in the initial anti-interference beam weight vector into amplitude and phase components. The amplitude components are converted into attenuation control words for the variable attenuator of that channel according to the attenuator level mapping relationship, and the phase components are converted into phase bias control words for the phase shifter of that channel according to the phase shifter coding relationship. The communication system writes the control words of each channel into the control register corresponding to the RF receiving front-end channel by channel through the control bus. Each channel of the RF receiving front-end completes the amplitude and phase state adjustment according to the control words. The output signals of each channel form a destructive superposition at the synthesis point, generating a spatial null in the interference direction, so that the power of the RF received signal input to the analog-to-digital converter is attenuated to within the linear sampling range.

[0054] In some embodiments, after the RF receiving front-end generates an initial spatial suppression beam and before sampling the RF received signal using an analog-to-digital converter (ADC), the communication system re-acquires the instantaneous received power after applying the initial spatial suppression beam through the power detection circuit of the RF receiving front-end. The communication system compares this instantaneous received power with the upper limit of the ADC's linear sampling interval. When the instantaneous received power is less than the upper limit of the ADC's linear sampling interval, the communication system confirms that the ADC has returned to linear sampling and proceeds to subsequent sampling steps. When the instantaneous received power is still greater than or equal to the upper limit of the ADC's linear sampling interval, the communication system re-looks up the beam immune spectrum in the beam immune spectrum library or further applies gain reduction correction to the current initial anti-interference beam weights in the direction of interference arrival angle until the input power of the ADC is confirmed to fall within the linear sampling interval before proceeding to subsequent sampling steps.

[0055] 205. When the analog-to-digital converter is in the linear sampling range, the RF received signal is sampled by the analog-to-digital converter to obtain digital received signal samples. The digital received signal samples are then input into the pre-trained deep reinforcement learning network model to obtain the target beam weights.

[0056] Digital received signal sample refers to the discrete digital sequence output by the analog-to-digital converter after linearly quantizing the radio frequency received signal according to the sampling clock cycle; deep reinforcement learning network model refers to a parameterized neural network in the communication system that completes parameter training offline, takes digital received signal samples as input and outputs beam control parameters, such as a policy inference network trained with an Actor-Critic architecture; target beam weight refers to the complex weight vector used to configure the accurate anti-interference beamforming state of the radio frequency receiving front end.

[0057] Specifically, after confirming that the input signal power of the analog-to-digital converter (ADC) is within the linear sampling range, the communication system synchronously samples the RF received signal output from the RF receiving front-end using the ADC. The ADC converts the analog voltage value at each sampling moment into binary digital code according to linear quantization encoding rules. The communication system then arranges the output digital codes from multiple consecutive sampling points in time sequence to obtain digitized received signal samples. The communication system then frames these digitized received signal samples according to a preset frame length and inputs them into a pre-trained deep reinforcement learning network model for forward inference calculation. Based on the channel characteristics and interference features reflected in the signal samples, the deep reinforcement learning network model outputs the target beam weights that optimize the communication system's receiving performance in the current scenario, according to the trained policy function.

[0058] In the inference and training framework of deep reinforcement learning network models (such as the Actor-Critic architecture), the physical elements of the model are defined as follows: the model's state space is defined as a one-dimensional real eigenvector of the sample covariance matrix containing the spatial distribution of interference in the current digital signal solution, and the beam weight state implemented by the current communication system; the model's action space is defined as the amplitude and phase adjustment offset of the combination of complex weight parameters output by each element channel of the phased array; the model's reward function is defined as the weighted sum of the increment of the signal-to-interference-plus-noise ratio of the final output combined signal and the penalty term of the preset null depth threshold in the direction of the interference arrival angle. Through the complete definition of the Markov decision process architecture described above, the model can output the target beam weights that maximize the communication reception quality based on the rapidly changing sampling characteristics.

[0059] In some embodiments, the deep reinforcement learning network model adopts an Actor-Critic architecture. Both the Actor policy network and the Critic value network contain an input layer, three fully connected hidden layers, and an output layer. The input layer receives a concatenation of a one-dimensional real-valued feature vector and the current beam weight state vector; the three hidden layers have 512, 256, and 128 neurons respectively, with ReLU activation functions used between layers. The Actor network's output layer has 2*M neurons (M being the number of array elements), corresponding to the amplitude and phase adjustment offset of each channel weight, and the output range is constrained by the Tanh activation function; the Critic network's output layer is a single neuron, outputting a scalar value assessment of the state-action pair. During the training phase, the communication system constructs a simulation environment based on an array manifold model, randomly generating interference angle of arrival, interference power, and desired signal direction, and synthesizing digital received signal samples as training input. The Actor network updates its parameters by maximizing the cumulative discount reward using the policy gradient method, while the Critic network updates its parameters by minimizing the temporal difference error. After training, the communication system stores the Actor network parameters. During online inference, the Actor network is only invoked to perform forward propagation and output the target beam weights.

[0060] In some embodiments, the input of digitized received signal samples into a deep reinforcement learning network model and the obtaining of target beam weights can be achieved by the following method: the communication system sends the digitized received signal samples into the covariance matrix estimation module, according to... =(1 / L)*X*X H The sample covariance matrix of the received signal for each array element channel is calculated, where X is the data matrix composed of sampled data from each channel, L is the number of sampling snapshots, and H represents the conjugate transpose. The communication system expands the real and imaginary parts of each element of the covariance matrix row by row and concatenates them into a one-dimensional real eigenvector. The communication system inputs the one-dimensional real eigenvector into the input layer of the deep reinforcement learning network model. After forward propagation calculation through a multi-layer neural network, the model output layer outputs the complex weight array of each array element channel. The communication system uses this complex weight array as the target beam weight.

[0061] 206. Calculate the complex weight difference between the initial anti-interference beam weight and the target beam weight, and divide the complex weight difference into smooth interpolation steps.

[0062] The complex weight difference refers to the difference vector obtained by performing complex subtraction on the complex elements of each corresponding channel in the initial anti-jamming beam weight vector and the target beam weight vector in the communication system. For example, for the i-th channel, the target weight element w i (T) Subtract the initial weight element w i (0) The difference Δw is obtained i The differences between all channels form a complex weight difference vector; the smooth interpolation step size refers to the complex increment vector corresponding to each part after the communication system divides the complex weight difference vector into equal parts, which is used to apply a fixed increment to the current beam weight in each preset period.

[0063] Specifically, the communication system performs complex subtraction on the complex elements of each channel in the target beam weight vector and the corresponding complex elements in the initial anti-interference beam weight vector. This involves subtracting the real part of the initial weight from the real part of the target weight, and subtracting the imaginary part of the initial weight from the imaginary part of the target weight, resulting in a complex difference for each channel. The communication system then arranges these complex differences in channel index order to form a complex weight difference vector. The communication system then determines the number of divisions N according to a preset criterion, dividing the complex weight difference vector into N equal parts. Each part serves as a smooth interpolation step, representing a fixed incremental weight change from the current value to the target value within each preset period. The value of N directly determines the number of transition steps: a larger N results in smaller amplitude and phase changes per step, a smoother transition, and a larger number of preset periods required; a smaller N results in larger changes per step, a faster transition speed, and a higher risk of amplitude and phase jumps.

[0064] In some embodiments, the complex weight difference is divided into smooth interpolation steps using the following method: Based on the interference feature signature, the communication system determines the angular coverage range of the spatial beam pattern formed by the initial anti-interference beam weights in the direction of effective interference signal suppression, obtaining a first spatial suppression angle interval, denoted as θ1 and W1. Simultaneously, it determines the angular coverage range of the spatial beam pattern formed by the target beam weights in the direction of effective interference signal suppression, obtaining a second spatial suppression angle interval, denoted as θ2 and W2. The communication system calculates the absolute value of the difference between the two center angles, obtaining the angle offset Δθ = |θ1 - θ2|, reflecting the degree of deviation between the initial beam and the target beam in the direction of spatial suppression. The communication system takes the smaller value W1 of the first spatial suppression angle interval and the smaller value W2 of the second spatial suppression angle interval. min =min(W1, W2), serving as the minimum reference width for the spatial coverage capability of the two sets of beams. The communication system divides the angular offset Δθ by the minimum angular width W. min Round up to the nearest integer to get the number of divisions N = ⌈Δθ / W min ⌉. The physical meaning of this calculation method and the hardware protection method for preventing loss of lock are as follows: The larger the angular offset, the greater the deviation in the suppression directions of the two sets of beams, and the more transition steps are required; the smaller the minimum angular width, the weaker the beam space fault tolerance, and the smaller the step size is required. The angular offset is then compared with the minimum effective null protection span (W). min Dividing by the step number is intended to enforce a constraint from the physical radiation boundary: the absolute span of the spatial pattern null drift or distortion angle caused by arbitrary single-step linear interpolation in the complex weight domain is locked at W. min Within this range. This constraint method not only avoids the sidelobe protrusion caused by conventional large leap interpolation, which leads to transient high-power interference penetration, but also ensures that in the transition state radiation pattern of adjacent interpolation periods, there must be a spatial null physical overlap region sufficient to cover the real interference source. Thus, even in the worst case, the anti-saturation red line of the analog-to-digital converter is still maintained, ensuring that continuous communication is not interrupted by physical impacts.

[0065] Taking an initial beam suppression interval centered at 30° and with a width of 5°, and a target beam suppression interval centered at 40° and with a width of 4° as an example, then Δθ = 10°, W min =4°, N=⌈10 / 4⌉=3, the communication system will use the complex weight difference vector Δw (where the channel difference Δw) i =w i (T) -w i (0) The difference between each channel is divided by 3 to obtain the smooth interpolation step size Δw / N, which is then added to the current weight in subsequent preset periods to gradually approach the target beam weight from the initial anti-interference beam weight.

[0066] 207. Within a series of preset periods, intermediate beam weights are generated sequentially from the initial anti-interference beam weights to the target beam weights according to the smooth interpolation step size.

[0067] The preset period refers to the smallest scheduling unit for updating beam weights in a communication system at fixed time intervals; the intermediate beam weights refer to the complex weight vectors of each channel actually generated and to be transmitted to the radio frequency receiving front end within a certain preset period during the smooth transition from the initial anti-interference beam weights to the target beam weights in the communication system, such as the weight vector w in the kth preset period. (k) =w (0) +k*Δw, where w (0) Δw is the initial anti-interference beam weight, Δw is the smooth interpolation step size, and k is the current cycle number.

[0068] Specifically, starting from the first preset period, the communication system performs complex addition on the complex elements of each channel of the initial anti-interference beam weight vector with the complex increment of the corresponding channel of the smooth interpolation step size to obtain the intermediate beam weight for the first preset period; in the kth preset period, w is calculated. (k) =w (0) +k*Δw, where Δw=(w (T) -w (0) The interpolation step size is ) / N, which yields the intermediate beam weights for the current preset period. The communication system generates all intermediate beam weights within N consecutive preset periods. The intermediate beam weight w generated in the Nth preset period... （N） =w (0) +N*Δw=w （T） The intermediate beam weights are numerically equal to the target beam weights, completing a full linear transition. Within each preset cycle, the communication system must also verify whether the generated intermediate beam weights meet spatial suppression requirements to ensure uninterrupted interference suppression throughout the transition.

[0069] In some embodiments, the specific process of generating intermediate beam weights periodically is as follows: In each preset period, the communication system multiplies the current period number k by the smooth interpolation step size vector Δw and then combines it with the initial anti-interference beam weight vector w. (0) Perform complex addition element by element to obtain the candidate intermediate beam weights w (k) 1=w (0) +k*Δw. The communication system reads the interference arrival angle θ3 of the interference component from the interference signature, constructs the array steering vector a(θ3) corresponding to this direction, and its i-th element is e. （j *2π*d*（i-1）·sin（θ3） / λ）, where d is the element spacing and λ is the operating wavelength; the communication system calculates the squared modulus of the inner product of the conjugate transpose of the candidate intermediate beam weight vector and the array steering vector to obtain the beam response value R = |(w (k) 1) * a(θ3)|². R represents the reception gain magnitude of the candidate intermediate beam weight for the interference signal in the direction of the interference arrival angle. The communication system compares R with a preset response threshold R0: If R < R0, the communication system determines the candidate intermediate beam weight w (k) 1 as the intermediate beam weight for the current preset period without additional adjustment; if R ≥ R0, it indicates that the spatial suppression depth of the candidate intermediate beam weight in the direction of the interference arrival angle is insufficient. The communication system performs constrained optimization adjustment on the candidate intermediate beam weight with the constraint that the beam response value in the θ3 direction is less than R0, and obtains the corrected intermediate beam weight by solving under the premise of satisfying this constraint condition. The communication system determines the corrected intermediate beam weight as the intermediate beam weight for the current preset period. The above process is cyclically executed within N preset periods to ensure that the intermediate beam weights generated in each preset period meet the preset spatial suppression requirements in the direction of the interference arrival angle and prevent interference penetration during the weight transition period.

[0070] 208. Sequentially send the intermediate beam weights to the RF receiving front end according to the preset period until the currently effective beam weight is completely updated to the target beam weight.

[0071] The currently effective beam weight refers to the complex weight vector that the communication system has sent to the RF receiving front end at the current moment and is currently controlling the amplitude and phase working states of each element channel.

[0072] Specifically, when each preset period arrives, the communication system reads the intermediate beam weight vector corresponding to this preset period from the storage buffer. The communication system sends this intermediate beam weight vector to the control modules of each element channel of the RF receiving front end through the control bus. Each channel of the RF receiving front end updates the amplitude and phase control states of the phase shifter and attenuator according to the received intermediate beam weight vector, and the currently effective beam weight is immediately updated to the intermediate beam weight of this preset period. When the next preset period arrives, the communication system repeats the above process and sends the intermediate beam weight of the next preset period to the RF receiving front end. The communication system sequentially completes the sending of all intermediate beam weights within N preset periods until the currently effective beam weight is numerically exactly the same as the target beam weight, and the communication system ends the cyclic execution of this step.

[0073] In some embodiments, the intermediate beam weights can be sent to the RF receiving front end sequentially according to a preset cycle in the following manner: At the beginning of each preset cycle, the processor reads the intermediate beam weight vector of the current preset cycle from the memory, decomposes the complex weight of each channel into amplitude components and phase components, and converts the amplitude components into variable attenuator control words according to the attenuator level mapping relationship (such as the limited control bit resolution corresponding to the digital step attenuator integrated in the RF chip) using an internally set nearest-neighbor discretization rounding algorithm. The phase components are converted into phase shifter control words according to the phase shifter encoding relationship (such as the discrete phase state space corresponding to a 5-bit or 6-bit digital phase shifter) using the minimum Euclidean distance optimization mechanism. The communication system sequentially writes the control words of each channel into the control register of the corresponding channel of the RF receiving front end through the serial peripheral interface bus. After receiving the control words, each channel of the RF receiving front end immediately updates its amplitude and phase working status. After confirming that all channels have been written, the communication system waits for the next preset cycle to be triggered, and repeats the cycle until the currently effective beam weights are completely updated to the target beam weights.

[0074] The phased array antenna beamforming control method described in this application involves the communication system first monitoring the instantaneous received power of the RF receiving front-end in real time. When the power exceeds a preset power threshold less than the saturation power of the analog-to-digital converter (ADC), interference signature features are extracted. Then, the communication system obtains initial anti-interference beam weights by looking up a table in a beam immunity map library, driving the RF receiving front-end to generate an initial spatially suppressed beam, attenuating the signal power input to the ADC to the linear sampling range. Next, the communication system uses digitized received signal samples and a deep reinforcement learning network model to obtain the target beam weights, calculates the complex weight difference between the initial and target weights, and divides it into smooth interpolation steps, providing a basis for smooth weight transition. Finally, the communication system generates intermediate beam weights within multiple preset periods according to the smooth interpolation steps and sends them to the RF receiving front-end periodically, avoiding RF amplitude and phase jumps caused by sudden full weight jumps. This method alleviates the communication service interruption problem caused by both ADC sampling distortion and weight jumps under sudden high-power interference, maintaining the stability of the signal demodulation process.

[0075] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 3 This is another flowchart illustrating the phased array antenna beamforming control method in this application embodiment.

[0076] 301. Monitor the instantaneous received power of the RF signal received by the RF receiving front-end in real time. (Refer to the execution process of step 201) 302. When the instantaneous received power exceeds a preset power threshold, extract the interference feature signature of the RF received signal. The preset power threshold is less than the saturation power of the analog-to-digital converter. (Refer to the execution process of step 202) 303. Determine the interference parameter space corresponding to the dimensions contained in the interference feature signature, and perform grid-based discretization on the interference parameter space to obtain multiple interference scene grid points. The interference parameter space shall include at least the interference angle of arrival dimension.

[0077] The interference parameter space represents a multidimensional mathematical abstract set that covers the range of all expected interference feature values; the gridded discretization represents the physical partitioning mathematical process of dividing the continuous parameter space into a finite number of discrete numerical nodes according to a preset quantization step size; the interference scene grid point refers to the discrete quantized parameter combination with a specific multidimensional coordinate position in the partitioned parameter space; the interference angle of arrival dimension represents the incident spatial direction attribute of the interference signal relative to the receiving phased array antenna array.

[0078] Specifically, to transform real-time complex matrix solving under sudden conditions into ultra-fast static querying, the communication system predetermines an interference parameter space that includes and is equivalent to all feature signature dimensions. This space is essentially a multi-dimensional continuous coordinate range. Subsequently, the communication system performs a gridded discretization operation on each dimension of the interference parameter space according to a fixed scan step size, transforming the infinite state continuum into a finite number of mutually independent and uniformly distributed interference scene grid points, thereby establishing an offline skeleton node set that fully covers the estimated constrained spatial range.

[0079] In some embodiments, the interference parameter space can be determined and its gridded discretization can be achieved in the following ways: the communication system determines the parameter range of the interference arrival angle dimension to be from -90 degrees to +90 degrees based on the physical aperture boundary of the phased array, and sets a fixed angle step size of one degree; the communication system determines the interference power dimension range based on the hardware saturation capability limit, and sets an arithmetic step size to perform power numerical slicing; the communication system performs a Cartesian product cross combination of the angle step size and the power step size to divide the entire continuous parameter space into a uniform orthogonal grid in a plane, thereby obtaining uniformly distributed discrete interference scene grid points.

[0080] 304. Based on the array manifold of the RF receiver front end, calculate the anti-interference beam weights corresponding to the grid points of each interference scenario.

[0081] The array manifold represents the mathematical and physical matrix model of the inherent characteristics of the spatial phase difference and amplitude response of each antenna element in the receiving front end in all directions; the anti-interference beam weight represents the set of complex control parameters that control the attenuation and phase shift states of each RF channel to form spatial suppression attenuation in a specific direction of incoming wave; the beam gain value refers to the physical amplification or attenuation factor of the RF signal energy when the antenna array is pointed at a certain angle after the weight is applied; the beam gain threshold represents a preset constant coefficient that represents the requirements of the null suppression standard.

[0082] Specifically, the communication system first extracts an array manifold matrix model that includes physical differences in channel hardware spacing, initial phase offset, and mutual coupling effects. Then, the system sequentially traverses all previously discretized interference scenario grid points, substituting the coordinate pointing parameters configured for each grid point into the array manifold matrix to form the spatial constraint base matrix for that scenario. The communication system then calculates the optimal solution for constraint noise reduction by inverting and decomposing the theoretical spatial covariance model constructed based on the array manifold. It then analyzes the complex weighted solution array of channel elements capable of creating physical spatial nulls in the direction of that grid point and uses this array as the anti-interference beam weight for that grid point.

[0083] In some embodiments, after calculating the anti-interference beam weights corresponding to each interference scenario grid point, the communication system performs pre-verification of its own suppression effectiveness and physical nulling enhancement: First, the communication system sequentially extracts each interference scenario grid point and parses its recorded directional parameter attributes, i.e., determines the corresponding interference angle of arrival value. Second, the communication system performs complex conjugate transpose processing on the original anti-interference beam weight of the grid point, performs an inner product dot product operation with the corresponding ideal space steering vector of the interference angle of arrival, and immediately takes the modulus and squares it to calculate the actual physical power amplification factor that the weight can generate at this interference angle of arrival, i.e., the beam gain value.

[0084] When the beam gain value is greater than or equal to a preset beam gain threshold, the communication system determines that the spatial attenuation null depth generated by the static pre-calculation is too shallow and insufficient to suppress the burst interference signal of this strength at the physical layer. At this point, the communication system uses the final generated beam gain value of the aforementioned interference angle of arrival being less than the preset beam gain threshold as the boundary constraint calculation condition for the mathematical inequality, and applies a convex optimization gradient descent algorithm to forcibly apply multi-bit bias optimization and update constraint correction actions to the channel weights of the original complex matrix. The communication system iterates multiple times in the correction optimization loop until the beam gain value corresponding to the new matrix is ​​completely compressed and is less than the beam gain threshold, ultimately obtaining and generating the corrected anti-interference beam weights that meet the requirements. Finally, the communication system physically overwrites the non-compliant anti-interference beam weights based on the absolute address at the memory layer, and safely replaces the corresponding original anti-interference beam weights with the corrected anti-interference beam weights that have better suppression capabilities.

[0085] 305. Establish an index mapping relationship between the parameter combinations represented by grid points in each interference scenario and the anti-interference beam weights to obtain a preset beam immunity spectrum library.

[0086] Parameter combination refers to the set of physical quantity values ​​describing the specific interference state of grid points in the interference scene in a multi-dimensional interference parameter space; beam immune map library refers to a structured data retrieval engine containing all discrete grid point parameters and their corresponding offline optimization weights.

[0087] Specifically, the communication system serializes and encodes the dimensional values ​​(such as interference angle of arrival, interference power, etc.) contained in each grid point of the interference scenario, using them as unique retrieval keys. The communication system uses the modified anti-interference beam weight matrix as the corresponding data payload. The communication system constructs a one-to-one correspondence between keys and payloads, generating a lookup structure with fast addressing capabilities, thereby completing the beam immune map library and storing it in the high-speed storage medium of the communication system.

[0088] In some embodiments, the index mapping relationship can be established in the following way: the communication system constructs a lookup table of multidimensional array structure, directly maps the integer index after discretization of parameters such as interference angle of arrival and interference power to the subscript coordinates of the multidimensional array, and stores the corresponding anti-interference beam weights in the corresponding array storage unit to achieve memory access with a time complexity of O(1).

[0089] 306. Calculate the feature distance between the interference feature signature and each interference scene grid point in the preset beam immune spectrum library, and determine the minimum feature distance.

[0090] Feature distance refers to the numerical difference between the measured radio frequency signal interference feature signature vector and the interference scene grid point vector pre-stored in the beam immune map library in the multidimensional parameter space; minimum feature distance represents the distance measurement result with the smallest difference value, characterizing the most similar offline scene.

[0091] Specifically, the communication system reads the currently extracted interference feature signature and traverses or searches for interference scene grid points in a pre-set beam immune map library. For each interference scene grid point being compared, the communication system uses a mathematical distance formula to calculate the aggregated difference between the parameter combination it represents and the current interference feature signature across various dimensions. The communication system uses a comparator to select the result with the smallest difference value, locking it as the minimum feature distance, thereby locating the reference point with the closest geometric position in the parameter space.

[0092] In some embodiments, the feature distance can be calculated as follows: the communication system uses the Euclidean distance algorithm to extract the angle value x and power value p from the interference feature signature, as well as the angle value x corresponding to the grid point. iWith power value p i ,calculate The minimum value is selected using the feature distance and either bubble sort or quick sort algorithm.

[0093] 307. When the minimum feature distance is less than the preset matching threshold, the anti-interference beam weight of the interference scene grid point corresponding to the minimum feature distance is used as the initial anti-interference beam weight.

[0094] The matching threshold refers to the pre-set tolerance value limit used to determine whether the current interference feature is sufficiently similar to the grid points in the library.

[0095] Specifically, the communication system compares the minimum feature distance with a preset matching threshold stored internally. If the comparison result shows that the minimum feature distance falls within the allowable error range (i.e., less than the threshold), the communication system determines that the grid point is sufficient to represent the current interference environment. The communication system reads the anti-interference beam weights associated with the interference scene grid point corresponding to the minimum feature distance and marks them as the initial anti-interference beam weights.

[0096] 308. When the minimum feature distance is greater than or equal to the matching threshold, select multiple adjacent interference scene grid points with the minimum feature distance.

[0097] Adjacent interference scene grid points represent a finite number of discrete quantization parameter combination points that are closest to the interference scene grid structure point corresponding to the minimum feature distance and are distributed around the current interference feature signature.

[0098] Specifically, when the minimum feature distance is greater than or equal to a pre-set matching threshold, the communication system confirms that the physical attributes of the current interference signature fall within the grid segmentation gaps of the preset parameter space. Using the interference scene grid point that generates the minimum feature distance as the central absolute coordinate geometric anchor point, the communication system expands outwards across the Euclidean geometric coordinate system of the multidimensional interference parameter space established based on the interference angle of arrival and interference power. It selects multiple other adjacent interference scene grid points that are closely adjacent to the central anchor point in the multidimensional coordinate system, thereby establishing a multidimensional enclosing grid vertex set structure for subsequent spatial multivariate numerical interpolation.

[0099] In some embodiments, efficient retrieval and localization of multiple adjacent interference scene grid points with the minimum feature distance can be achieved in the following way: The communication system constructs and solidifies the data structure of the beam immune map library into a KD-tree branch structure, and then inputs the real-time parsed interference feature signature as the target query comparison node into the KD-tree search algorithm function library. The communication system performs multi-dimensional nearest neighbor node distance traversal and backtracking comparison calculation in the sub-branch node clusters of the KD-tree, sets and truncates the total number of returned nodes to a preset constant K nearest neighbor parameter index at the code level, and then outputs the K adjacent interference scene grid point objects closest to the current interference feature to the central processor.

[0100] 309. Using the reciprocal of the feature distance between each adjacent interference scene grid point and the interference feature signature as the weight, the anti-interference beam weights of adjacent interference scene grid points are weighted and fused to obtain the initial anti-interference beam weights.

[0101] Anti-interference beam weights refer to the complex adjustment parameters of each independent node already stored in the beam immune map library.

[0102] Specifically, the communication system retrieves and extracts the previously calculated feature distance values ​​between each adjacent interference scene grid point and the current real-time interference feature signature. The system inputs each acquired feature distance into a floating-point unit to perform a mathematical inverse division operation, obtaining the reciprocal of the corresponding feature distance, and directly configures it in the cache as the influence weight factor required for the interpolation process. Subsequently, the system extracts the pre-stored anti-interference beam weight multidimensional complex matrix from the database within each adjacent interference scene grid point. The system then performs scalar multiplication of the complex domain by the vector complex scalar multiplication on these originally discrete and isolated matrices according to the calculated and assigned numerical weights. Next, the system performs a weighted multi-stream fusion accumulation operation, adding the inter-matrix channels bit-by-bit. Within the originally blank grid parameter gaps, the system dynamically fits and smooths a beam null amplitude and phase integrated response output curve that matches the high-precision requirements, thus obtaining high-precision initial anti-interference beam weights that fully conform to the current real-world transient strong electromagnetic interference environment and are unaffected by discrete truncation errors.

[0103] 310. Based on the initial anti-interference beam weights, drive the RF receiving front-end to generate an initial spatial suppression beam, so that the RF received signal power input to the analog-to-digital converter is attenuated to within the linear sampling range of the analog-to-digital converter. (Refer to the execution process of step 204) 311. When the analog-to-digital converter (ADC) is within the linear sampling range, the RF received signal is sampled using the ADC to obtain digital received signal samples. These digital received signal samples are then input into a pre-trained deep reinforcement learning network model to obtain the target beam weights. (Refer to the execution process of step 205) 312. Calculate the complex weight difference between the initial anti-jamming beam weight and the target beam weight, and divide the complex weight difference into smooth interpolation steps. (Refer to the execution process of step 206) 313. Within multiple consecutive preset periods, intermediate beam weights are generated sequentially according to a smooth interpolation step size, transitioning from the initial anti-interference beam weights to the target beam weights. (Refer to the execution process of step 207) 314. According to the preset cycle, the intermediate beam weights are sent to the RF receiving front end sequentially until the currently effective beam weights are completely updated to the target beam weights. (Refer to the execution process of step 208) The phased array antenna beamforming control method described in this application first initiates an anti-interference process before the analog-to-digital converter (ADC) reaches saturation. Then, the communication system spatially discretizes the interference parameters, calculates and corrects the anti-interference beam weights at each grid point based on the array manifold, establishes a beam immunity map library, and converts real-time matrix solving into static querying. Next, the communication system calculates the feature distance between the interference signature and the grid points, and obtains the initial anti-interference beam weights by using the matching results or weighted fusion, reducing parameter spatial quantization truncation errors. Subsequently, the communication system generates an initial spatially suppressed beam to bring the signal power back to the linear sampling range, and obtains the target beam weights through a deep reinforcement learning network model. Finally, the communication system divides the weight difference into smooth interpolation steps, generates and distributes intermediate beam weights periodically, and completes a smooth weight transition. This method alleviates the communication service interruption problem caused by ADC sampling distortion and sudden weight jumps under sudden high-power interference, maintaining the stability of signal demodulation.

[0104] The methods provided in the above embodiments can be executed by a communication system. The communication system in the embodiments of this invention is described below from a hardware processing perspective; please refer to [link / reference]. Figure 4 This is a schematic diagram of the physical device structure of a communication system in an embodiment of this application.

[0105] It should be noted that, Figure 4 The structure of the communication system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0106] like Figure 4As shown, the communication system includes a processing unit (CPU) 401, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 402 or a program loaded from storage portion 408 into random access memory (RAM) 403, such as performing the methods described in the above embodiments. The RAM 403 also stores various programs and data required for system operation. The CPU 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0107] The following components are connected to I / O interface 405: input section 406 including audio input devices, push-button switches, etc.; output section 407 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 408 including a hard disk, etc.; and communication section 409 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 409 performs communication processing via a network such as the Internet. Drive 410 is also connected to I / O interface 405 as needed. Removable media 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 410 as needed so that computer programs read from them can be installed into storage section 408 as needed.

[0108] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by central processing unit (CPU) 401, it performs the various functions defined in the present invention.

[0109] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0110] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0111] Specifically, the communication system in this embodiment includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the phased array antenna beamforming control method provided in the above embodiment.

[0112] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the communication system described in the above embodiments; or it may exist independently and not assembled into the communication system. The storage medium carries one or more computer programs that, when executed by a processor of the communication system, cause the communication system to implement the phased array antenna beamforming control method provided in the above embodiments.

[0113] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0114] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0115] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0116] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A beamforming control method for a phased array antenna, characterized in that, Applied to a communication system, the communication system including a radio frequency receiving front-end and an analog-to-digital converter, the method includes: Real-time monitoring of the instantaneous received power of the radio frequency received signal at the radio frequency receiving front end; When the instantaneous received power exceeds a preset power threshold, the interference feature signature of the radio frequency received signal is extracted, and the preset power threshold is less than the saturation power of the analog-to-digital converter; A lookup table is performed in a preset beam immune spectrum library to obtain the initial anti-interference beam weights corresponding to the interference feature signature; Based on the initial anti-interference beam weight, the RF receiving front end is driven to generate an initial spatial suppression beam, so that the power of the RF received signal input to the analog-to-digital converter is attenuated to within the linear sampling range of the analog-to-digital converter; When the analog-to-digital converter is within the linear sampling interval, the radio frequency received signal is sampled using the analog-to-digital converter to obtain a digital received signal sample. The digital received signal sample is then input into a pre-trained deep reinforcement learning network model to obtain the target beam weights. Calculate the complex weight difference between the initial anti-interference beam weight and the target beam weight, and divide the complex weight difference into smooth interpolation step sizes; Within a series of preset periods, intermediate beam weights are generated sequentially from the initial anti-interference beam weights to the target beam weights according to the smooth interpolation step size. According to the preset period, the intermediate beam weights are sent to the radio frequency receiving front end in sequence until the currently effective beam weights are completely updated to the target beam weights.

2. The method according to claim 1, characterized in that, Extracting the interference feature signature of the received radio frequency signal specifically includes: The analog power detection values ​​of multiple receiving channels in the radio frequency receiving front end are obtained respectively, and each receiving channel has different spatial response characteristics; Based on the relative power distribution characteristics among the multiple analog power detection values, the angle of arrival of the interference component in the radio frequency received signal is determined; Based on the interference angle of arrival and the pre-calibrated spatial response gain value of each of the receiving channels in the direction of the interference angle of arrival, gain compensation calculation is performed on each of the analog power detection values ​​to obtain the instantaneous power value of the interference component. An interference signature is generated for the radio frequency received signal based on the interference angle of arrival and the instantaneous power value.

3. The method according to claim 1, characterized in that, The construction process of the preset beam immune spectrum library specifically includes: Determine the interference parameter space corresponding to the dimensions contained in the interference feature signature, and discretize the interference parameter space into a grid to obtain multiple interference scene grid points. The interference parameter space includes at least the interference angle of arrival dimension. Based on the array manifold of the radio frequency receiving front end, calculate the anti-interference beam weights corresponding to each grid point of the interference scenario; An index mapping relationship is established between the parameter combinations represented by each grid point in the interference scenario and the anti-interference beam weights to obtain the preset beam immune spectrum library.

4. The method according to claim 3, characterized in that, After the step of calculating the anti-interference beam weights corresponding to each of the interference scene grid points, the method further includes: For each of the aforementioned interference scene grid points, determine the interference angle of arrival corresponding to the interference scene grid point in the interference angle of arrival dimension; Calculate the beam gain value of the anti-interference beam weight corresponding to the grid point of the interference scene at the interference angle of arrival; When the beam gain value is greater than or equal to the preset beam gain threshold, the anti-interference beam weight corresponding to the interference scene grid point is constrained and optimized until the corrected beam gain value is less than the beam gain threshold, and the corrected anti-interference beam weight is obtained. The corresponding anti-interference beam weights are replaced with the corrected anti-interference beam weights.

5. The method according to claim 3, characterized in that, The step of performing a lookup mapping in a preset beam immune spectrum library to obtain the initial anti-interference beam weights corresponding to the interference feature signature specifically includes: Calculate the feature distance between the interference feature signature and each interference scene grid point in the preset beam immune spectrum library, and determine the minimum feature distance; When the minimum feature distance is less than the preset matching threshold, the anti-interference beam weight of the interference scene grid point corresponding to the minimum feature distance is used as the initial anti-interference beam weight. When the minimum feature distance is greater than or equal to the matching threshold, select multiple adjacent interference scene grid points with the minimum feature distance. The anti-interference beam weights of the adjacent interference scene grid points are weighted and fused using the reciprocal of the feature distance between each adjacent interference scene grid point and the interference feature signature as the weight, to obtain the initial anti-interference beam weights.

6. The method according to claim 1, characterized in that, Intermediate beam weights are generated sequentially according to the smooth interpolation step size, transitioning from the initial anti-jamming beam weights to the target beam weights, specifically including: In each of the preset periods, candidate intermediate beam weights are generated according to the current smooth interpolation step size; Based on the interference feature signature, the interference angle of arrival of the interference component in the radio frequency received signal is determined, and the beam response value of the candidate intermediate beam weight in the direction corresponding to the interference angle of arrival is calculated. When the beam response value is less than the preset response threshold, the candidate intermediate beam weight is determined as the intermediate beam weight of the current preset period. When the beam response value is greater than or equal to the response threshold, the weights of the candidate intermediate beam weights are adjusted with the constraint that the beam response value at the interference arrival angle is less than the response threshold, so as to obtain the corrected intermediate beam weights. The corrected intermediate beam weights are determined as the intermediate beam weights for the current preset period.

7. The method according to claim 1, characterized in that, Dividing the complex weight difference into smooth interpolation step sizes specifically includes: Based on the interference feature signature, the first spatial suppression angle interval corresponding to the initial anti-interference beam weight and the second spatial suppression angle interval corresponding to the target beam weight are determined respectively. Calculate the angular offset between the first spatial suppression angle interval and the second spatial suppression angle interval; Based on the angular offset and the minimum value between the angular width of the first spatial suppression angular interval and the angular width of the second spatial suppression angular interval, the number of divisions of the complex weight difference is determined; The complex weight difference is divided equally according to the number of divisions to obtain the smooth interpolation step size.

8. A communication system, characterized in that, Includes RF receiver front-end and analog-to-digital converter; The communication system includes one or more processors and a memory; The memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the communication system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising computer instructions, characterized in that, When the computer instructions are executed on the communication system, the communication system performs the method as described in any one of claims 1-7.

10. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are run on the communication system, the communication system performs the method as described in any one of claims 1-7.

Images