A frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology

Abstract

The application discloses a frequency domain anti-interference method based on a multi-interval dynamic threshold reconstruction technology and belongs to the field of communication signal processing.The application realizes the method as follows: after frequency domain information of a received signal is obtained, a plurality of decision thresholds are used to classify signal frequency points, and different weighting values are used to perform attenuation processing on the frequency points in each interval set; after multiple cycles, a residual signal frequency point set is output to obtain a received signal subjected to interference suppression processing; a plurality of decision thresholds are used to perform multi-interval dynamic division, and the signal frequency point set can be more carefully discriminated in a single cycle; the attenuation processing of the interference signal frequency points according to different weighting values can also reduce the misjudgment probability of the decision value edge signal frequency points, and the interference suppression effect is improved under a fixed cycle number.The application can guarantee the interference suppression precision, improve the interference suppression efficiency and improve the robustness and applicability of the interference suppression method under the condition of limited hardware resource consumption.

Description

Technical Field

[0001] This invention relates to a frequency domain anti-interference method for spread spectrum communication systems, and particularly to a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology, belonging to the field of communication signal processing. Background Technology

[0002] Because the difference in correlation between the spread spectrum signal and the interference signal is utilized during the baseband signal generation process, spread spectrum communication systems inherently possess a certain degree of anti-interference capability, and their inherent anti-interference performance depends on the system's spreading gain. However, in some extremely harsh electromagnetic environments, due to limitations in channel capacity and resource consumption, when the energy of external interference signals exceeds the system's interference tolerance, communication quality will severely deteriorate. Therefore, given the limited spreading gain of the communication system, the introduction of anti-interference techniques is crucial.

[0003] Depending on the data processing dimension, anti-interference techniques can be categorized into time-domain, frequency-domain, spatial-domain, and code-assisted methods. Among these, the frequency-domain anti-interference method based on a threshold-iterative continuous mean elimination algorithm has been widely used due to its simple design and ideal anti-interference effect. However, under the constraints of hardware resources and processing speed, the anti-interference effect of this method has significant limitations. Therefore, research on anti-interference algorithms that simultaneously possess ideal suppression effects, fast response speed, and low resource consumption is of great importance.

[0004] Traditional frequency domain anti-interference methods typically employ a single decision threshold to distinguish between spread spectrum signal frequencies and interference signal frequencies. Interference signals are eliminated by setting the estimated frequency of interference signals whose frequency domain energy exceeds the decision threshold to zero. However, transform domain processing, represented by Fourier transform, inevitably leads to high hardware resource consumption in frequency domain anti-interference methods. Therefore, optimizing the algorithm and achieving more ideal anti-interference performance within limited hardware resources is a crucial issue that must be considered in frequency domain anti-interference method research. Summary of the Invention

[0005] To address the aforementioned technical shortcomings of existing frequency domain anti-interference methods based on continuous mean elimination strategies, the main objective of this invention is to provide a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology. After obtaining the frequency domain information of the received signal, this method classifies the signal frequency points using multiple decision thresholds and attenuates the frequency points in each interval set using different weighting values. After multiple iterations, the remaining signal frequency point set is output as the received signal after interference suppression. Because multiple decision thresholds are used for multi-interval dynamic division, the signal frequency point set can be more finely discriminated in a single iteration, improving the accuracy of interference signal suppression. The attenuation processing of interference signal frequency points based on different weighting values ​​can also significantly reduce the probability of misjudging signal frequency points at the edge of the decision value, thus improving the interference suppression effect with a fixed number of iterations. This invention can guarantee interference suppression accuracy under limited system hardware resource consumption, and not only improves interference suppression efficiency but also enhances the robustness and applicability of the interference suppression method.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] This invention discloses a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology, comprising the following steps:

[0008] Step 1: Use a window function with a length equal to the number of FFT points to truncate and window the received signal, and perform a Fourier transform to obtain the frequency domain information of the signal segment to be processed.

[0009] Let the signal received by the spread spectrum receiver be r(t). This signal can be considered as consisting of the transmitted signal d(t), Gaussian white noise n(t), and interference signal J(t), that is:

[0010] r(t)=d(t)+n(t)+J(t) (1)

[0011] The frequency domain response of the received signal obtained by performing a Fourier transform is as follows:

[0012]

[0013] Where, r ω (n) represents the signal obtained after windowing to mitigate spectral leakage, where N is the signal length. FFT N is the number of points in the Fast Fourier Transform. FFT It does not have to be equal to N, by N FFT Set to the closest 2 to N m Positive integers can maximize the processing efficiency of hardware resources.

[0014] The modulus of the above signal spectrum is calculated to obtain the frequency domain information of the signal segment, which is represented as a set of frequency point amplitudes as follows:

[0015]

[0016] Step 2: Calculate the decision threshold based on the mean amplitude of the frequency point set. Use the decision threshold to classify the signal frequencies, obtaining estimated sets for spread spectrum signal frequencies, interference signal frequencies, and suspected interference signal frequencies. By using multiple decision thresholds for dynamic multi-interval division, the signal frequency point set can be more finely discriminated within a single loop, improving the accuracy of interference signal suppression. Different weighting values ​​are used to attenuate the signal frequencies in each interval set, reducing the probability of misjudging signal frequencies at the decision threshold edges. This process is repeated multiple times until the preset maximum number of loops is reached or the estimated set of interference signal frequencies becomes empty. The remaining signal frequencies are output to obtain the frequency domain information of the interference-suppressed signal.

[0017] Step 2.1: Based on the statistical characteristics of the received signal, calculate the decision threshold according to the mean amplitude of the frequency point set.

[0018] In the threshold iterative generation continuous mean elimination algorithm, a decision threshold is set based on the expected amplitude of the signal spectrum, dividing the entire set of all spectrum amplitude points into two subsets: spread spectrum signal frequency points and interference signal frequency points.

[0019]

[0020] Among them, S m and N m These represent the total amplitude and the number of frequency points of the spread spectrum signal obtained in the m-th iteration, respectively, and T is the average spectral amplitude S of the spread spectrum signal obtained in this iteration. m N m Iterative adjustment to the target value φ target The step coefficient, T, is expressed as follows:

[0021]

[0022] Wherein, F(φ) target ) is the target distribution function value, which is the confidence probability that the observed frequency point is correctly identified as a signal frequency point, and represents the relative number of observations included in the required set.

[0023] Step 2.2 involves classifying signal frequency points using decision thresholds to obtain estimated sets of spread spectrum signal frequency points, interference signal frequency points, and suspected interference signal frequency points. By using multiple decision thresholds for dynamic multi-interval division, the signal frequency point set can be more finely identified in a single loop, improving the accuracy of interference signal suppression.

[0024] Multiple decision intervals are set to distinguish the signal frequency points, and the decision thresholds are expressed as follows:

[0025]

[0026]

[0027] Frequency points with amplitudes greater than the threshold thre1 are identified as interference signal frequency points; frequency points with amplitudes less than the threshold thre2 are identified as spread spectrum signal frequency points; frequency points with amplitudes between the two thresholds are identified as suspected interference signal frequency points, and corresponding attenuation processing is performed in step 2.3.

[0028] Step 2.3: Attenuate the frequency elements in the above estimation set using different weighting values ​​to reduce the probability of misjudging the frequency points of the signal edge of the decision value.

[0029] After obtaining different sets of estimated signal frequency points, the signal frequency points are processed according to the attenuation formula shown below:

[0030]

[0031] Where φ(k) represents the signal frequency amplitude before attenuation processing, and φ'(k) represents the signal frequency amplitude after attenuation processing; when the signal frequency amplitude is greater than the high threshold thre1, the frequency is determined to be an interference signal frequency and is zeroed; when the signal frequency amplitude is less than the low threshold thre2, the frequency is determined to be a spread spectrum signal frequency and remains unchanged; when the signal frequency amplitude is between the two thresholds, the frequency amplitude is attenuated to near the low threshold while keeping the frequency phase information unchanged.

[0032] Step 2.4: Repeat the processing in steps 2.1 to 2.3 until the number of loops reaches the preset maximum number of loops or the frequency point estimation set of the interference signal becomes an empty set, thereby achieving the effect of interference suppression.

[0033] Step 3: When performing transform domain processing on the received signal, the truncated window is slid with a step size equal to half the number of FFT points and the received signal is processed in segments in sequence. The output signal of the frequency domain interference suppression system is obtained by overlapping and adding the sliding windowed and interference-suppressed signals to reduce the signal-to-noise ratio loss caused by windowing processing.

[0034] When performing a Fourier transform on a signal, windowing is applied before the transform to avoid unnecessary spectral leakage. The introduction of the window function results in additional signal-to-noise ratio loss. The discrete-time received sequence x(n) can be expressed as:

[0035] x(n)=Ap(n)+w(n) (9)

[0036] Where A is the signal amplitude, p(n) is a pseudocode sequence of length N, and w(n) is a zero-mean, zero-variance sequence. 2 Given an additive white Gaussian noise sequence; define the window function as h(n), then the windowed signal is represented as follows:

[0037] x h (n)=Ap(n)h(n)+w(n)h(n) (10)

[0038] By performing correlation despreading and integration on the above signal sequence, we obtain:

[0039]

[0040] Based on the above formula, the signal-to-noise ratio of the despread system output is:

[0041]

[0042] The signal-to-noise ratio of the received signal input to the despreading system is:

[0043]

[0044] A comparison of equations (12) and (13) shows that windowing causes a non-negligible loss in the signal-to-noise ratio (SNR). Overlap compensation is applied to the windowed signal to reduce the SNR loss and improve the SNR of the windowed signal. A larger overlap ratio results in a more significant compensation effect and a smaller SNR loss, but also increases computational load and hardware resource requirements. Preferably, using a 50% overlap windowing operation to segment the received signal can reduce the computational load and hardware resource requirements of the windowing compensation process while ensuring the compensation effect.

[0045] Beneficial effects:

[0046] 1. This invention discloses a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology. After obtaining the frequency domain information of the received signal, multiple decision thresholds are used to classify the signal frequency points, and different weighting values ​​are used to attenuate the frequency points in each interval set. After multiple iterations, the remaining signal frequency point set is output to obtain the received signal after interference suppression. Compared with traditional frequency domain interference suppression methods, because multiple decision thresholds are used for multi-interval dynamic division, the signal frequency point set can be more finely identified in a single iteration, improving the suppression accuracy of interference signals. The attenuation processing of interference signal frequency points according to different weighting values ​​can also greatly reduce the probability of misjudging signal frequency points at the edge of the decision value, and can improve the interference suppression effect with a fixed number of iterations.

[0047] 2. The present invention discloses a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology. When processing the received signal in the transform domain, the truncation window is slid with a step size equal to half the number of FFT points and the received signal is processed in segments in sequence. The output signal of the system is obtained by overlapping and adding the sliding windowed and interference-suppressed signals to compensate for the signal. This avoids unnecessary spectrum leakage and reduces the signal-to-noise ratio loss caused by windowing.

[0048] 3. The frequency domain anti-interference method disclosed in this invention based on multi-interval dynamic threshold reconstruction technology, on the basis of achieving the above-mentioned beneficial effects 1 and 2, can achieve more effective elimination of interference signal frequency points in fewer loops, and ensure interference suppression accuracy under limited system hardware resource consumption conditions. It not only improves the interference suppression efficiency, but also improves the robustness and applicability of the interference suppression method, so as to solve the problem that it is impossible to effectively suppress the received signal in complex channel environment under limited hardware resources and computing time. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart of a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology disclosed in this invention;

[0051] Figure 2 This is a schematic diagram illustrating the implementation principle of a multi-interval dynamic threshold reconstruction technology disclosed in this invention.

[0052] Figure 3 This is a schematic diagram illustrating the time-domain waveform changes of a signal before and after overlapping windowing processing, as disclosed in this invention.

[0053] Figure 4 The figure shows a comparison of the bit error rate performance curves of the present invention and two traditional threshold iterative generation continuous mean elimination algorithms in a radio frequency noise interference channel with an interference ratio of 50dB for frequency domain interference suppression. Detailed Implementation

[0054] To better illustrate the purpose and advantages of the present invention, the invention will be further described below in conjunction with the accompanying drawings and examples.

[0055] Example 1:

[0056] like Figure 1As shown in the figure, this embodiment discloses a frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology, which includes the following steps:

[0057] Step 1: Use a window function with a length equal to the number of FFT points to truncate and window the received signal, and perform a Fourier transform to obtain the frequency domain information of the signal segment to be processed.

[0058] In this embodiment, the sampling rate f s 100.00MHz; Spreading ratio SSR is 256; Oversampling rate OSR is 8; Carrier frequency f c The chip rate is doubled to 25.00MHz; the roll-off factor for the root-raised cosine shaping filter and the matched filter is 0.35.

[0059] Let the signal received by the spread spectrum receiver be r(t). This signal can be considered as consisting of the transmitted signal d(t), Gaussian white noise n(t), and interference signal J(t), that is:

[0060] r(t)=d(t)+n(t)+J(t) (14)

[0061] Considering signal length, processing time, and hardware resource consumption, a 256-point FFT and a 256-point Bohman window function are used to perform time-domain truncation, windowing, and transform-domain processing on the received signal, with a length of 256 points. The frequency domain response of the segmented received signal is shown below after performing a Fourier transform:

[0062]

[0063] Where, r ω (n) is the signal obtained after windowing to mitigate the spectral leakage effect.

[0064] The modulus of the above signal spectrum is calculated to obtain the frequency domain information of the signal segment, which is represented as a set of frequency point amplitudes as follows:

[0065]

[0066] Step 2: Calculate the decision threshold based on the mean amplitude of the frequency point set. Use the decision threshold to classify the signal frequencies, obtaining estimated sets for spread spectrum signal frequencies, interference signal frequencies, and suspected interference signal frequencies. By using multiple decision thresholds for dynamic multi-interval division, the signal frequency point set can be more finely discriminated within a single loop, improving the accuracy of interference signal suppression. Different weighting values ​​are used to attenuate the signal frequencies in each interval set, reducing the probability of misjudging signal frequencies at the decision threshold edges. This process is repeated multiple times until the preset maximum number of loops is reached or the estimated set of interference signal frequencies becomes empty. The remaining signal frequencies are output to obtain the frequency domain information of the interference-suppressed signal.

[0067] like Figure 2 As shown, the main processing flow of a multi-interval dynamic threshold reconstruction technology disclosed in this invention includes the following four sub-steps:

[0068] Step 2.1: Based on the statistical characteristics of the received signal, calculate the decision threshold according to the mean amplitude of the frequency point set.

[0069] In the threshold iterative generation continuous mean elimination algorithm, a decision threshold is set based on the expected amplitude of the signal spectrum, dividing the entire set of all spectrum amplitude points into two subsets: spread spectrum signal frequency points and interference signal frequency points.

[0070]

[0071] Among them, S m and N m These represent the total amplitude and the number of frequency points of the spread spectrum signal obtained in the m-th iteration, respectively, and T is the average spectral amplitude S of the spread spectrum signal obtained in this iteration. m N m Iterative adjustment to the target value φ target The step coefficient, T, is expressed as follows:

[0072]

[0073] Wherein, F(φ) target F(φ) is the target distribution function value, which represents the confidence probability of correctly identifying the observed frequency as a signal frequency, and indicates the relative number of observations included in the required set. target Different values ​​of the threshold coefficient T correspond to different step coefficients. Based on the statistical characteristics of the narrowband interference signal after FFT, the optimal threshold coefficient T = 4 is adopted.

[0074] Step 2.2 involves classifying signal frequency points using decision thresholds to obtain estimated sets of spread spectrum signal frequency points, interference signal frequency points, and suspected interference signal frequency points. By using multiple decision thresholds for dynamic multi-interval division, the signal frequency point set can be more finely identified in a single loop, improving the accuracy of interference signal suppression.

[0075] Multiple decision intervals are set to distinguish the signal frequency points, and the decision thresholds are expressed as follows:

[0076]

[0077]

[0078] Frequency points with amplitudes greater than the threshold thre1 are identified as interference signal frequency points; frequency points with amplitudes less than the threshold thre2 are identified as spread spectrum signal frequency points; frequency points with amplitudes between the two thresholds are identified as suspected interference signal frequency points, and corresponding attenuation processing is performed in step 2.3.

[0079] Step 2.3: Attenuate the frequency elements in the above estimation set using different weighting values ​​to reduce the probability of misjudging the frequency points of the signal edge of the decision value.

[0080] Based on different weighting methods, three different frequency domain interference removal algorithms are used for frequency elements within each decision interval: For spread spectrum signal frequencies, their amplitude and phase information are kept unchanged; for interference signal frequencies, the threshold zeroing method is used to estimate the location of the interference signal frequency and set the corresponding frequency weighting coefficient to zero to remove the interference signal; for suspected interference signal frequencies, a partial pruning method is used to correct the amplitude of the suspected interference signal frequency to a fixed constant while keeping the signal phase information unchanged, so as to reduce its impact on subsequent data processing.

[0081] After obtaining different sets of estimated signal frequency points, the signal frequency points are processed according to the attenuation formula shown below:

[0082]

[0083] Where φ(k) represents the signal frequency amplitude before attenuation processing, and φ'(k) represents the signal frequency amplitude after attenuation processing; when the signal frequency amplitude is greater than the high threshold thre1, the frequency is determined to be an interference signal frequency and is zeroed; when the signal frequency amplitude is less than the low threshold thre2, the frequency is determined to be a spread spectrum signal frequency and remains unchanged; when the signal frequency amplitude is between the two thresholds, the frequency amplitude is attenuated to near the low threshold while keeping the frequency phase information unchanged.

[0084] Step 2.4: Repeat the processing in steps 2.1 to 2.3 until the number of loops reaches the preset maximum number of loops max(CME_Loop) = 3 or the frequency point estimation set of the interference signal becomes an empty set, thereby achieving the effect of interference suppression.

[0085] Step 3: When performing transform domain processing on the received signal, the truncated window is slid with a step size equal to half the number of FFT points and the received signal is processed in segments in sequence. The output signal of the frequency domain interference suppression system is obtained by overlapping and adding the sliding windowed and interference-suppressed signals to reduce the signal-to-noise ratio loss caused by windowing processing.

[0086] When performing a Fourier transform on a signal, windowing is applied before the transform to avoid unnecessary spectral leakage. The introduction of the window function results in additional signal-to-noise ratio loss. The discrete-time received sequence x(n) can be expressed as:

[0087] x(n)=Ap(n)+w(n) (22)

[0088] Where A is the signal amplitude, p(n) is a pseudocode sequence of length N, and w(n) is a zero-mean, zero-variance sequence. 2 Given an additive white Gaussian noise sequence; define the window function as h(n), then the windowed signal is represented as follows:

[0089] x h (n)=Ap(n)h(n)+w(n)h(n) (23)

[0090] By performing correlation despreading and integration on the above signal sequence, we obtain:

[0091]

[0092] Based on the above formula, the signal-to-noise ratio of the despread system output is:

[0093]

[0094] The signal-to-noise ratio of the input signal to the despreading system is:

[0095]

[0096] A comparison of equations (25) and (26) shows that windowing causes a non-negligible loss in the signal-to-noise ratio (SNR). By performing overlap compensation on the windowed signal, the SNR loss caused by windowing can be reduced, thereby improving the SNR of the windowed signal. Figure 3 As shown, after overlapping compensation, the energy loss of the windowed signal at the edges of the window function is compensated, the distortion is reduced, and the impact on the signal threshold decision is reduced accordingly. Theoretical calculations and practical verification show that a larger overlap ratio results in a more significant compensation effect and a smaller signal-to-noise ratio loss, but also increases the computational load and hardware resource requirements. Preferably, a 50% overlap windowing operation is used to segment the received signal. Specifically, the first 128 points of each 256-point signal segment are overlapped and added to the last 128 points of the previous segment after interference suppression processing. This reduces the computational load and hardware resource requirements of the windowing compensation process while maintaining the aforementioned compensation effect.

[0097] like Figure 4As shown, the bit error rate (BER) curves depict the system BER performance of three frequency domain interference suppression methods after the channel is exposed to RF noise interference with a signal-to-interference ratio (SNR) of 50 dB: the traditional single-threshold zero-reset threshold iterative generation continuous mean elimination algorithm, the multi-interval dynamic threshold reconstruction threshold iterative generation algorithm, and the single optimal threshold continuous mean elimination algorithm. The frequency domain anti-interference method based on multi-interval dynamic threshold reconstruction technology disclosed in this invention exhibits excellent performance under different SNRs. Compared to the traditional single-threshold zero-reset threshold iterative generation continuous mean elimination algorithm, this invention achieves a lower BER and performance closer to that of an ideal BPSK modulated spread spectrum communication system.

[0098] The above detailed description further illustrates the purpose, technical solution, and beneficial effects of the invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A frequency domain anti-jamming method based on multi-interval dynamic threshold reconstruction technology, characterized in that: Includes the following steps, Step 1: Use a window function with a length equal to the number of FFT points to truncate and window the received signal, and perform a Fourier transform to obtain the frequency domain information of the signal segment to be processed; Step 2: Calculate the decision threshold based on the mean amplitude of the frequency point set, and classify the signal frequency points using the decision threshold to obtain the estimated sets of spread spectrum signal frequency points, interference signal frequency points, and suspected interference signal frequency points respectively; by using multiple decision thresholds to dynamically divide multiple intervals, the signal frequency point set can be more finely identified in a single round of loop, improving the suppression accuracy of interference signals. Different weighting values ​​are used to attenuate the signal frequency points in each interval set to reduce the probability of misjudging signal frequency points at the edge of the decision value. This process is repeated multiple times until the number of loops reaches the preset maximum number of loops or the estimated set of interference signal frequency points becomes empty. The signal frequency points of the remaining set are then output to obtain the frequency domain information of the signal after interference suppression. Step 3: When processing the received signal in the transform domain, the truncated window is slid with a step size equal to half the number of FFT points and the received signal is processed in segments in sequence. The output signal of the frequency domain interference suppression system is obtained by overlapping and adding the sliding windowed and interference-suppressed signals to reduce the signal-to-noise ratio loss caused by windowing.

2. The frequency domain anti-jamming method based on multi-interval dynamic threshold reconstruction technique according to claim 1, characterized in that: Step 1 is implemented as follows: Let the signal received by the spread spectrum receiver be r(t). This signal is equivalent to a signal consisting of the transmitted signal d(t), Gaussian white noise n(t), and interference signal J(t), that is: r(t)=d(t)+n(t)+J(t) (1) The frequency domain response is obtained by performing a Fourier transform on the received signal. Where, r ω (n) represents the signal obtained after windowing to mitigate spectral leakage, where N is the signal length. FFT N is the number of points in the Fast Fourier Transform. FFT It does not have to be equal to N, by N FFT Set to the closest 2 to N m Positive integers; The signal spectrum is modulated to obtain signal segment frequency domain information and is expressed as a frequency point amplitude set as follows: .

3. The frequency domain anti-jamming method based on multi-interval dynamic threshold reconstruction technique according to claim 2, characterized in that: The second step is implemented as follows: Step 2.1: Based on the statistical characteristics of the received signal, calculate the decision threshold according to the mean amplitude of the frequency point set; In the threshold iterative generation continuous mean elimination algorithm, a decision threshold is set based on the expected amplitude of the signal spectrum, dividing the entire set of all spectrum amplitude points into two subsets: spread spectrum signal frequency points and interference signal frequency points. Among them, S m and N m These represent the total amplitude and the number of frequency points of the spread spectrum signal obtained in the m-th iteration, respectively, and T is the average spectral amplitude S of the spread spectrum signal obtained in this iteration. m / N m Iterative adjustment to the target value φ target The step coefficient, T, is expressed as follows: Wherein, F(φ) target ) is the target distribution function value, which is the confidence probability that the observed frequency point is correctly identified as a signal frequency point, and represents the relative number of observations included in the required set; Step 2.2: Classify the signal frequency points using decision thresholds to obtain estimated sets of spread spectrum signal frequency points, interference signal frequency points, and suspected interference signal frequency points. By using multiple decision thresholds to dynamically divide multiple intervals, the signal frequency point set can be more finely identified in a single round of loop, improving the suppression accuracy of interference signals. Multiple decision intervals are set to distinguish the signal frequency points, and the decision thresholds are expressed as follows: Frequency points with amplitudes greater than the threshold thre1 are identified as interference signal frequency points; frequency points with amplitudes less than the threshold thre2 are identified as spread spectrum signal frequency points; frequency points with amplitudes between the two thresholds are identified as suspected interference signal frequency points, and corresponding attenuation processing is performed in step 2.

3. Step 2.3: Attenuate the frequency elements in the above estimation set using different weighting values ​​to reduce the probability of misjudging the frequency points of the signal edge of the decision value. After obtaining different sets of estimated signal frequency points, the signal frequency points are processed according to the attenuation formula shown below: Where φ(k) represents the signal frequency amplitude before attenuation processing, and φ'(k) represents the signal frequency amplitude after attenuation processing; when the signal frequency amplitude is greater than the high threshold thre1, the frequency is determined to be an interference signal frequency and is zeroed; when the signal frequency amplitude is less than the low threshold thre2, the frequency is determined to be a spread spectrum signal frequency and remains unchanged; when the signal frequency amplitude is between the two thresholds, the frequency amplitude is attenuated to near the low threshold while keeping the frequency phase information unchanged. Step 2.4: Repeat the processing in steps 2.1 to 2.3 until the number of loops reaches the preset maximum number of loops or the frequency point estimation set of the interference signal becomes an empty set.

4. The frequency domain anti-jamming method based on multi-interval dynamic threshold reconstruction technique according to claim 3, characterized in that: Step 3 is implemented as follows: When performing a Fourier transform on a signal, windowing is applied before the transform to avoid unnecessary spectral leakage; the introduction of the window function causes additional signal-to-noise ratio loss; the discrete-time received sequence x(n) is expressed as: x(n)=Ap(n)+w(n) (9) Where A is the signal amplitude, p(n) is a pseudocode sequence of length N, and w(n) is a zero-mean, zero-variance sequence. 2 Given an additive white Gaussian noise sequence; define the window function as h(n), then the windowed signal is represented as follows: x h (n) = Ap(n)h(n) + w(n)h(n) (10) By performing correlation despreading and integration on the above signal sequence, we obtain: Based on the above formula, the signal-to-noise ratio of the despread system output is: The signal-to-noise ratio of the received signal input to the despreading system is: By performing overlap compensation on the windowed signal, the signal-to-noise ratio loss caused by windowing is reduced, thereby improving the signal-to-noise ratio of the windowed signal.

5. The frequency domain anti-jamming method based on multi-interval dynamic threshold reconstruction technique according to claim 4, characterized in that: The received signal is segmented using an overlap windowing operation with an overlap ratio of 50%.

Images