Adaptive anti-jamming system and method based on transient error control and jitter suppression

Abstract

The application discloses an adaptive anti-interference method based on transient error control and jitter suppression, comprising the following steps: S1, receiving antenna M-way baseband digital signals in real time, and combining M-channel data of the latest continuous N time domains into a space-time snapshot vector at the current moment; S2, based on the space-time snapshot vector at the current moment and a current weight vector, an anti-interference signal at the current moment is obtained through complex inner product operation, and a current error signal is obtained; wherein the current weight vector is a preset initial weight vector at the first iteration; S3, based on the space-time snapshot vector at the current moment and the current error signal, a next moment original weight value is generated through a transient error control and normalized least mean square iteration method; S4, inertia filtering processing is performed on the next moment original weight value, and a next moment smoothed weight value is output; and S5, based on the next moment smoothed weight value and the space-time snapshot vector, steps S2-S4 are cycled to obtain an anti-interference signal in real time.

Description

Technical Field

[0001] This invention relates to the field of communication and array signal processing technology, and in particular to an adaptive anti-interference system and method based on transient error control and jitter suppression. Background Technology

[0002] In recent years, with the rapid development of low-Earth orbit (LEO) satellite communication and high-speed mobile platform wireless transmission, the robustness of satellite-to-ground communication links has faced severe challenges. Ground receiving devices operate in complex electromagnetic environments, making them highly susceptible to malicious interference (such as high-power broadband suppression interference) and the intermingling of ground multipath reflection clutter. Simultaneously, the extremely high maneuverability of LEO satellites and other platforms, relative to ground receiving antennas, results in highly variable Doppler frequency shift rates and rapidly changing spatial manifold characteristics in satellite-to-ground links. Space-Time Adaptive Processing (STAP) technology, a two-dimensional joint filtering technique that utilizes both spatial and temporal degrees of freedom, has become crucial for ensuring the robustness of ground station and terminal receiving links due to its strong broadband dispersion interference suppression and multipath suppression capabilities.

[0003] In the initial blind acquisition phase, before ground receivers achieve signal time-frequency synchronization and steering vector demodulation, the space-time adaptive processing (PI-STAP) architecture based on the power inversion (PI) criterion is widely used. This architecture locks the first channel tap weight, achieving omnidirectional interference cancellation directly in the time domain without prior information such as the desired signal direction. To solve for the optimal space-time weight vector in real-time and at high frequency in ground baseband processing hardware (such as FPGAs), the Normalized Least Mean Square (NLMS) algorithm has become the preferred choice for engineering deployment due to its simple structure and the fact that it does not require high-dimensional matrix inversion.

[0004] However, traditional adaptive iterative schemes face the following significant challenges when receiving high-dynamic, high-bandwidth satellite signals from the ground:

[0005] First, there is an inherent contradiction between the rapid changes in high-dynamic space and the convergence speed. Traditional algorithms mostly use fixed step sizes. When low-Earth orbit satellites move at high speeds or interference sources change dynamically, the spatial manifold mismatch becomes severe. If a large step size is used, although the algorithm can keep up with high-dynamic tracking, the steady-state weights fluctuate violently, causing the demodulation error rate to soar. If a small step size is used, the convergence speed is slow, making it impossible to suppress sudden strong interference in real time within the limited satellite transit window. Existing variable step size schemes are mostly based on indirect estimation of interference power, and their step size adjustment has an inherent delay, making it difficult to achieve instantaneous response.

[0006] Second, there is a contradiction between steady-state weight jitter and null depth maintenance. Due to the relatively high-speed maneuvering between the receiver and transmitter, the temporarily iterated weight vector is highly susceptible to interference from both rapidly changing spatial characteristics and noise gradients, exhibiting a phenomenon of "weight tearing and steady-state high-frequency jitter." Existing anti-interference schemes often lack a mechanism to effectively process the weights in the hardware feedback loop to suppress jitter, resulting in distortion of the synthesized space beam null and an inability to maintain a stable null depth throughout the high-speed satellite transit.

[0007] Third, hardware timing and resource bottlenecks under ultra-high bandwidth. Existing technologies often employ variable step-size mechanisms or traditional frequency domain matrix inversion (such as Cholesky decomposition) schemes to pursue convergence. However, ground stations face high throughput and broadband data streams. The floating-point square root and high-dimensional division operations involved in frequency domain matrix inversion can lead to severe pipeline truncation and timing deterioration within the FPGA. Furthermore, direct hardware mapping of existing nonlinear variable step-size functions (such as exponential functions) can result in unacceptable computational latency.

[0008] In summary, existing FPGA-based anti-interference solutions cannot simultaneously address the challenges of rapid tracking under high dynamic interference, suppression of steady-state weight jitter, and low-latency hardware closed-loop implementation at high bandwidth and ultra-high throughput for ground receivers. This represents a common technical challenge in the field. Therefore, designing a spatial multi-dimensional joint anti-interference hardware method and device that eliminates the need for matrix inversion, operates in a pipelined manner across the entire time domain, exhibits ultra-fast convergence, and effectively suppresses weight jitter has become a pressing technical problem to be solved in the field of low-Earth orbit satellite ground receiving engineering. Summary of the Invention

[0009] The present invention aims to at least partially solve one of the technical problems existing in the related art.

[0010] The purpose of this invention is to provide an adaptive anti-interference system and method based on transient error control and jitter suppression, which solves the problem of the contradiction between convergence speed and steady-state accuracy caused by fixed step size or indirect variable step size in traditional adaptive algorithms in high dynamic communication scenarios, as well as the problem of zero-trapped distortion caused by high-frequency jitter of iterative weights in dynamic environments.

[0011] To achieve the above objectives, the present invention provides an adaptive anti-interference system based on transient error control and jitter suppression, comprising:

[0012] Multi-channel synchronous acquisition interface, used to receive M-channel baseband digital signals in real time;

[0013] The space-time snapshot construction module is used to combine the data from the current time and the M channels from the previous N-1 times to generate the current space-time snapshot vector;

[0014] The transient error control module is used to calculate the step size factor based on the current error signal;

[0015] The weight iteration update module is used to perform normalized minimum mean square iteration based on the step size factor output by the transient error control module to generate the original weights.

[0016] The jitter suppression module is used to perform inertial filtering post-processing on the original weights and output smooth weights;

[0017] The beamforming output module is used to perform complex inner product operations on the space-time snapshot vector using smoothing weights, and output anti-interference signals and error signals.

[0018] The convergence monitoring module is used to continuously monitor the power change of the error signal and the convergence status of the weight vector. Based on the monitored steady-state error fluctuations, interference environment and preset thresholds, it dynamically adjusts the system parameters.

[0019] As a preferred method, the current spatiotemporal snapshot vector is generated as follows: based on the M channel data of the current time and the previous N-1 times, the spatiotemporal snapshot vector of the current time is generated by combining them in the order of "array element priority, time subsequent". M and N are both positive integers, and M≥2 and N≥2.

[0020] Preferably, the transient error control module includes a normalized error calculation unit, a mapping function generation unit, and a step size limiting unit; wherein: the normalized error calculation unit is used to calculate the ratio of the error amplitude to the noise reference; and the mapping function generation unit is used to map the normalized error amplitude to a step size factor.

[0021] As a preferred option, the specific calculation process for the step size factor is as follows:

[0022] Step 1: Calculate the normalized power norm based on the space-time snapshot vector:

[0023]

[0024] in, Let n be the normalized power norm at time n. Let n be the MN×1 dimensional spacetime snapshot vector at time n. Let i be the i-th element of the spacetime snapshot vector at time n;

[0025] Step 2: Calculate the normalized error amplitude based on the error signal and the preset noise benchmark. :

[0026]

[0027] in, Let be the error signal at time n. The preset noise standard;

[0028] Step 3: Based on the obtained error magnitude, calculate the current step size factor using the mapping function of the transient error control unit:

[0029]

[0030] in, Let A be the step size factor at time n, A be the scaling factor, and m be the exponential factor.

[0031] Preferably, the specific iterative formula for the weights in the weight iteration update module is as follows:

[0032] i=2,3,…,MN

[0033] in, This represents the i-th element of the original iterative weight vector at time n+1. , The initial weight is 1. The conjugate of the error signal at time n is... This is the regularization factor.

[0034] Preferably, if the current normalized power norm is less than the preset lower limit threshold of the power norm, the weights for this time remain unchanged from the weights for the previous time.

[0035] Preferably, the jitter suppression module employs a first-order recursive filtering structure, performing weight smoothing calculations within the same clock domain to obtain the jitter-suppressed weight vector. The specific calculation formula is as follows:

[0036]

[0037] in, Let i be the i-th element of the weight vector after jitter suppression at time n. It is a smoothing factor;

[0038] Based on the current original iterative weights and the weights after jitter suppression at the previous time step, calculate the weight fluctuation amplitude at the current time step:

[0039]

[0040] in, The weight fluctuation amplitude at time n represents the maximum deviation between the current original weight and the weight after jitter suppression at the previous time.

[0041] Preferably, the method for obtaining the anti-interference signal is as follows: The anti-interference output signal at the current moment is obtained by performing a complex inner product operation between the current jitter-suppressed weight vector and the spatiotemporal snapshot vector. The specific calculation formula is as follows:

[0042]

[0043] in, Here, H represents the anti-interference output signal at time n, and H denotes the conjugate transpose.

[0044] The error signal is the negative value of the anti-interference signal.

[0045] Preferably, the convergence monitoring module uses a sliding window to statistically analyze the mean of the normalized error amplitude. and variance ,when When the value remains consistently high, decrease the scaling factor A; when As the value continues to increase, the exponential factor m is decreased; simultaneously, the fluctuation range of the weights is monitored in real time and compared with the preset weight fluctuation range threshold. When the fluctuation range exceeds the preset weight threshold, the smoothing factor increases and the scaling factor decreases.

[0046] Another aspect of the present invention provides an adaptive anti-interference method based on transient error control and jitter suppression, comprising the following steps:

[0047] S1. Receive M baseband digital signals from antennas in real time and combine the data from the M channels of the most recent N consecutive time domains into a space-time snapshot vector for the current moment.

[0048] S2. Based on the current time-space snapshot vector and the current weight vector, the anti-interference signal at the current time is obtained through complex inner product operation, and the current error signal is acquired; wherein the current weight vector is the preset initial weight in the first iteration;

[0049] S3. Based on the current time-space snapshot vector and the current error signal, the original weights for the next time step are generated through transient error control and normalized minimum mean square iteration method.

[0050] S4. Perform inertial filtering on the original weights at the next time step and output the smoothed weights at the next time step.

[0051] S5. Based on the smoothing weight and spatiotemporal snapshot vector of the next moment, repeat steps S2-S4 to obtain anti-interference signals in real time.

[0052] Beneficial Effects: This invention introduces a collaborative mechanism of transient error control and jitter suppression within the PI-STAP framework. On one hand, the step size factor is directly driven by the instantaneous relative error at the current moment, avoiding the inherent delay caused by indirect estimation based on interference power, and achieving immediate response and rapid reconvergence when interference direction changes abruptly. On the other hand, post-processing of the iterative weights through inertial filtering effectively suppresses high-frequency random fluctuations of the weights in dynamic environments, improving the depth consistency and positional stability of the nulls. These two mechanisms act on the pre-iteration and post-processing stages of weight update, respectively, forming a collaborative working mode of "fast tracking + robust output," fundamentally resolving the inherent contradiction between convergence speed and steady-state accuracy in highly dynamic scenarios. Attached Figure Description

[0053] Figure 1 The flowchart of an adaptive anti-interference method based on transient error control and jitter suppression is provided in Example 2.

[0054] Figure 2 The image shows the interference null plot of the adaptive anti-interference method based on transient error control and jitter suppression of the present invention in the simulation experiment of Example 2. Figure 2 (a) is a 3D diagram of the space-time response after anti-interference in the simulation experiment of the method of the present invention; Figure 2 (b) is a two-dimensional projection diagram of the spacetime in the simulation experiment of the method of the present invention.

[0055] Figure 3 The diagram shows the interference null traps of the existing fixed-step NLMS scheme in the simulation experiment of Example 2; Figure 3 (a) is a 3D diagram of the space-time response after anti-interference in the simulation experiment of the existing fixed-step NLMS scheme; Figure 3 (b) is a two-dimensional projection diagram of the space-time in the simulation experiment of the existing fixed step size NLMS scheme. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, embodiments of this invention, and should not be construed as limiting the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. In the description of this invention, it should be understood that the terminology used is for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0057] The following is combined with Figures 1-3 This invention describes an adaptive anti-interference system and method based on transient error control and jitter suppression.

[0058] Example 1: This example provides an adaptive anti-interference system based on transient error control and jitter suppression, including:

[0059] Multi-channel synchronous acquisition interface, used to receive M-channel baseband digital signals in real time;

[0060] The space-time snapshot construction module is used to combine the data from the current time and the M channels from the previous N-1 times to generate the current space-time snapshot vector;

[0061] The space-time snapshot construction module acquires the sampled values ​​of M channels at the current moment in each clock cycle, and simultaneously reads the historical values ​​of the previous N-1 moments from the shift register chain. Following the order of "array element priority, time subsequent," it combines the M×N data into a space-time snapshot vector for the current moment. This vector contains both the spatial distribution information and temporal evolution information of the signal, where M and N are both positive integers, and M≥2 and N≥2.

[0062] The transient error control module is used to calculate the step size factor based on the current error signal and drive the weight update. The transient error control module includes a normalized error calculation unit and a mapping function generation unit. Specifically, the normalized error calculation unit calculates the ratio of the error amplitude to the noise reference, specifically by calculating the normalized power norm, whose formula is:

[0063]

[0064] in, Let n be the normalized power norm at time n. Let be the MN×1 dimensional spacetime snapshot vector at time n. The i-th element of the space-time snapshot vector at time n is used to normalize the step size so that the weight update is not affected by the amplitude change of the input signal.

[0065] The mapping function generation unit is used to map the normalized error amplitude to a step size factor, specifically: calculating the normalized error amplitude based on the amplitude of the error signal and a preset noise reference. ,in, Let be the error signal at time n. The noise baseline is preset. The current step size factor is calculated using the mapping function of the transient error control unit. ,in, Let A be the step size factor at time n, A be the scaling factor, and m be the exponential factor. This function is a monotonically increasing function. When the error is large, the step size automatically increases to accelerate convergence, and when the error is small, the step size automatically decreases to improve steady-state accuracy. This mapping relationship allows the step size factor to be directly driven by the instantaneous relative error at the current time, without relying on indirect steps such as interference power estimation.

[0066] The weight iteration update module is used to perform normalized minimum mean square iteration based on the step size factor output by the transient error control module to generate the original weights. Specifically, according to the power inversion criterion, the first weight is fixed to be 1 (constraint condition). Only the remaining MN-1 weights are updated, and the iteration formula is as follows: ,i=2,3,…,MN. Where, This represents the i-th element of the original iterative weight vector at time n+1. , The i-th element of the snapshot vector when it is empty. The conjugate of the error signal at time n is... As a regularization factor, Let n be the step size factor at time n. If the current normalized power norm is less than the preset lower threshold (determined as Y), then skip this weight update to maintain numerical stability; if it is normal (determined as N), then complete the iterative update normally.

[0067] The jitter suppression module is used to perform inertial filtering post-processing on the original weights and output smooth weights; specifically, it adopts a first-order recursive weighting structure, the construction form of which is as follows: , Let i be the i-th element of the weight vector after jitter suppression at time n. The smoothing factor is the weighting coefficient of the historical weights. This structure is essentially a first-order low-pass filter, suppressing high-frequency random fluctuations in the weights. It's equivalent to adding an inertial element in series in the weight update path: a larger smoothing factor results in smoother weight output and stronger anti-jitter capability, but a correspondingly slower response to disturbance changes; a smaller smoothing factor results in a larger proportion of the current weights, faster system response, but a weaker smoothing effect. This processing is independent of the weight iteration process, does not change the steady-state position of weight convergence, and only suppresses random high-frequency jitter components. Furthermore, the jitter suppression module monitors the fluctuation state of the weight vector in real time. The current weight fluctuation amplitude is defined as: , The weight fluctuation amplitude at time n represents the maximum deviation between the current original weight and the weight after jitter suppression at the previous time, i.e., the maximum absolute value of the difference between the current original iterative weight and the filtered weight at the previous time. This value reflects the severity of the instantaneous fluctuation of the weight. If detected... If the fluctuation exceeds a preset threshold (determined as Y), a temporary increase in the smoothing factor is triggered to enhance the jitter suppression effect; if the fluctuation is normal (determined as N), the current value of the smoothing factor is maintained. The weight vector after jitter suppression processing. As the final weights, they are sent to the beamforming output module for beamforming calculations at the next moment.

[0068] The beamforming output module is used to perform a complex inner product operation on the space-time snapshot vector using smoothing weights, and output anti-interference signals and error signals; specifically, let the space-time snapshot vector at the current moment be... The current weight vector after jitter suppression is Then the output signal ,in, Let be the anti-interference output signal at time n. The superscript H indicates the conjugate transpose, which is the complex inner product operation performed on the received signals of each tap of each channel. According to the power inversion criterion, the system aims to minimize the total output power, expecting a zero response; therefore, the error signal is taken as the negative value of the output signal. If the current iteration is the first iteration (determined as Y), the weight vector is initialized as a constraint vector with only the first element set to 1 and the rest to 0; if it is not the first iteration (determined as N), the weight vector updated in the previous time step is used. The error signal is simultaneously sent to the convergence monitoring module and the transient error control module.

[0069] The convergence monitoring module continuously monitors the power changes of the error signal and the convergence status of the weight vector. If the steady-state error fluctuation exceeds a preset threshold or a sudden change occurs in the interference environment (determined as Y), dynamic parameter adjustment is triggered: the system uses a sliding window to statistically calculate the mean of the normalized error amplitude. and variance When insufficient convergence speed is detected ( If the error remains high, reducing the scaling factor A shifts the mapping curve to the left, resulting in a larger output step size and faster convergence; when increased error fluctuations are detected ( The exponential factor m is continuously increased, while the response capability to large errors is reduced; once the error stabilizes, it is gradually restored to the preset value. Simultaneously, the smoothing factor in the jitter suppression module is temporarily adjusted to balance response speed and jitter suppression effect: when a weight fluctuation is detected... If the value exceeds the preset threshold continuously, the smoothing factor will be adjusted. Temporarily increase from the current value To enhance jitter suppression, the scaling factor A is temporarily reduced to accelerate the response. After the weight fluctuations fall below the threshold and remain below it for a preset duration, they are gradually restored to their original values. If the convergence is normal (judged as N), the current parameters are maintained.

[0070] Example 2: As Figure 1 As shown, this embodiment provides an adaptive anti-interference method based on transient error control and jitter suppression, including the following steps:

[0071] S1. Receive M baseband digital signals from antennas in real time and combine the data from the M channels of the most recent N consecutive time domains into a space-time snapshot vector for the current moment.

[0072] S2. Based on the current time-space snapshot vector and the current weight vector, the anti-interference signal at the current time is obtained through complex inner product operation, and the current error signal is acquired; wherein the current weight vector is the preset initial weight in the first iteration;

[0073] S3. Based on the current time-space snapshot vector and the current error signal, the original weights for the next time step are generated through transient error control and normalized minimum mean square iteration method.

[0074] S4. Perform inertial filtering on the original weights at the next time step and output the smoothed weights at the next time step.

[0075] S5. Based on the smoothing weight and spatiotemporal snapshot vector of the next moment, repeat steps S2-S4 to obtain anti-interference signals in real time.

[0076] To verify the effectiveness of the adaptive anti-interference method based on transient error control and jitter suppression proposed in this invention, this embodiment constructs a low-orbit satellite communication ground receiving scenario in the MATLAB simulation environment and compares the present invention with the existing fixed step size NLMS scheme.

[0077] 1. Experimental conditions

[0078] Antenna configuration: Five-element uniform circular array, with an element spacing of half a wavelength;

[0079] Number of time-domain taps: N = 4;

[0080] Snap count: 1000;

[0081] Signal and interference settings: Desired signal direction is 0°, signal-to-noise ratio (SNR) = -20dB (weak signal); two narrowband interferences, with directions of 80° and 160° respectively, normalized frequencies of 0.2 and -0.15 respectively, and interference-to-noise ratio (JNR) = 60dB; one wideband interference, with a direction of 240°, normalized frequency coverage of 0.25~0.45, and interference-to-noise ratio (JNR) = 60dB.

[0082] Comparison schemes: Scheme A (this invention): an adaptive anti-interference method based on transient error control and jitter suppression; Scheme B: the existing fixed step size NLMS scheme.

[0083] 2. Experimental Data

[0084] Output performance comparison:

[0085] Option A: [Narrowband Interference Suppression]

[0086] Depth in 160° direction: -82.86 dB

[0087] Depth in 80° direction: -87.83 dB

[0088]

Broadband Interference Suppression (240°, bandwidth 0.25~0.45)

[0089] Worst in-band rejection: -30.63 dB (representing the bottleneck in interference immunity)

[0090] In-band average suppression: -35.08 dB (representing overall cancellation level)

[0091] Deepest zero dip within the band: -61.22 dB

[0092] Option B: [Narrowband Interference Suppression]

[0093] Depth in 160° direction: -60.40 dB

[0094] Depth in 80° direction: -57.06 dB

[0095]

Broadband Interference Suppression (240°, bandwidth 0.25~0.45)

[0096] Worst in-band rejection: -28.58 dB (representing the bottleneck in interference immunity)

[0097] In-band average suppression: -31.55 dB (representing overall cancellation level)

[0098] Deepest zero dip within the band: -43.14 dB

[0099] 3. Experimental Results

[0100] Under the same four-tap space-time processing architecture and the exact same interference environment, the comparison results between the present invention (Solution A) and fixed-step NLMS (Solution B) show that:

[0101] (1) Regarding narrowband interference suppression, such as Figure 2 As shown in (a), scheme A forms sharp and significantly deep nulls in two narrowband interference directions (80° and 160°); in contrast, as Figure 3 As shown in (a), scheme B has a significantly shallower null depth in the same direction. Specifically, scheme A has a null depth of -82.86 dB in the 160° direction and -87.83 dB in the 80° direction, while scheme B has -60.40 dB and -57.06 dB, respectively. This invention improves the narrowband null depth by approximately 22-30 dB, indicating that the transient error control mechanism enables more sufficient weight convergence, and the steady-state null depth is significantly better than the fixed step size scheme.

[0102] (2) In terms of broadband interference suppression, such as Figure 2 (b) and Figure 3As shown in the two-dimensional projection diagram (b), both schemes can form null regions of a certain width in the 240° broadband interference direction. However, considering the specific data, Scheme A has an in-band average suppression of -35.08 dB, a deepest in-band null of -61.22 dB, and a worst in-band suppression of -30.63 dB; Scheme B has an in-band average suppression of -31.55 dB, a deepest in-band null of -43.14 dB, and a worst in-band suppression of -28.58 dB. This invention improves the in-band average suppression by approximately 3.5 dB, the deepest in-band null by approximately 18 dB, and the worst in-band suppression by approximately 2 dB. From... Figure 2 As can be observed in (b), scheme A has null points with prominent depths in the broadband interference band. These points correspond to the better suppression effect formed by the algorithm in the band, indicating that the null depth and uniformity of the present invention are better than the fixed step size scheme in the broadband interference band.

[0103] (3) Overall performance: Scheme A significantly improves the output SINR compared to Scheme B, further verifying the effectiveness of the synergistic mechanism of transient error control and jitter suppression in improving convergence speed, steady-state null depth and broadband interference uniformity suppression.

[0104] In summary, compared with the fixed step size NLMS scheme, the present invention has significant advantages in narrowband null depth, broadband interference uniformity suppression, and steady-state output SINR, verifying the effectiveness of the synergistic mechanism of transient error control and jitter suppression.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention 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; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An adaptive anti-interference system based on transient error control and jitter suppression, characterized in that, include: Multi-channel synchronous acquisition interface, used to receive M-channel baseband digital signals in real time; The space-time snapshot construction module is used to combine the data from the current time and the M channels from the previous N-1 times to generate the current space-time snapshot vector; The transient error control module is used to calculate the step size factor based on the current error signal; The weight iteration update module is used to perform normalized minimum mean square iteration based on the step size factor output by the transient error control module to generate the original weights. The jitter suppression module is used to perform inertial filtering post-processing on the original weights and output smooth weights; The beamforming output module is used to perform complex inner product operations on the space-time snapshot vector using smoothing weights, and output anti-interference signals and error signals. The convergence monitoring module is used to continuously monitor the power change of the error signal and the convergence status of the weight vector. Based on the monitored steady-state error fluctuations, interference environment and preset thresholds, it dynamically adjusts the system parameters.

2. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 1, characterized in that, The specific method for generating the current spacetime snapshot vector is as follows: based on the M channel data of the current time and the previous N-1 times, the spacetime snapshot vector of the current time is generated by combining them in the order of "array element priority, time subsequent". M and N are both positive integers, and M≥2 and N≥2.

3. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 2, characterized in that, The transient error control module includes a normalized error calculation unit, a mapping function generation unit, and a step size limiting unit; wherein: the normalized error calculation unit is used to calculate the ratio of the error amplitude to the noise reference; the mapping function generation unit is used to map the normalized error amplitude to a step size factor.

4. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 3, characterized in that, The specific calculation process of the step size factor is as follows: Step 1: Calculate the normalized power norm based on the space-time snapshot vector: ； in, Let n be the normalized power norm at time n. Let be the MN×1 dimensional spacetime snapshot vector at time n. Let i be the i-th element of the spacetime snapshot vector at time n; Step 2: Calculate the normalized error amplitude based on the error signal and the preset noise benchmark. : ； in, Let be the error signal at time n. The preset noise standard; Step 3: Based on the obtained error magnitude, calculate the current step size factor using the mapping function of the transient error control unit: ； in, Let A be the step size factor at time n, A be the scaling factor, and m be the exponential factor.

5. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 4, characterized in that, The specific iterative formula for the weights in the weight iteration update module is as follows: ,i=2,3,…,MN; in, This represents the i-th element of the original iterative weight vector at time n+1. , The initial weight is 1. Let be the conjugate of the error signal at time n. This is the regularization factor.

6. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 5, characterized in that, If the current normalized power norm is less than the preset lower limit threshold of the power norm, the weights for this time will remain unchanged from the weights for the previous time.

7. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 6, characterized in that, The jitter suppression module employs a first-order recursive filtering structure, performing weight smoothing calculations within the same clock domain to obtain the jitter-suppressed weight vector. The specific calculation formula is as follows: ； in, Let i be the i-th element of the weight vector after jitter suppression at time n. It is a smoothing factor; Based on the current original iterative weights and the weights after jitter suppression at the previous time step, calculate the weight fluctuation amplitude at the current time step: ； in, The weight fluctuation amplitude at time n represents the maximum deviation between the current original weight and the weight after jitter suppression at the previous time.

8. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 7, characterized in that, The method for obtaining the anti-interference signal is as follows: The anti-interference output signal at the current moment is obtained by performing a complex inner product operation between the current jitter-suppressed weight vector and the space-time snapshot vector. The specific calculation formula is as follows: ； in, Here, H represents the anti-interference output signal at time n, and H denotes the conjugate transpose. The error signal is the negative value of the anti-interference signal.

9. The adaptive anti-interference system based on transient error control and jitter suppression according to claim 8, characterized in that, The convergence monitoring module uses a sliding window to statistically analyze the mean of the normalized error amplitude. and variance ,when When the value remains consistently high, decrease the scaling factor A; when As the value continues to increase, the exponential factor m is decreased; simultaneously, the fluctuation range of the weights is monitored in real time and compared with a preset weight fluctuation range threshold. When the fluctuation range exceeds the preset weight threshold, the smoothing factor increases and the scaling factor decreases.

10. An adaptive anti-interference method based on transient error control and jitter suppression using the system described in any one of claims 1-9, comprising the following steps: S1. Receive M baseband digital signals from antennas in real time and combine the data from the M channels of the most recent N consecutive time domains into a space-time snapshot vector for the current moment. S2. Based on the current time-space snapshot vector and the current weight vector, the anti-interference signal at the current time is obtained through complex inner product operation, and the current error signal is acquired; wherein the current weight vector is the preset initial weight in the first iteration; S3. Based on the current time-space snapshot vector and the current error signal, generate the original weights for the next time step through transient error control and normalized minimum mean square iteration method; S4. Perform inertial filtering on the original weights at the next time step and output the smoothed weights at the next time step. S5. Based on the smoothing weight and spatiotemporal snapshot vector of the next moment, repeat steps S2-S4 to obtain anti-interference signals in real time.

Images