An adaptive communication anti-jamming method and system based on blind source separation

Abstract

This application relates to the field of wireless communication anti-interference technology, and discloses an adaptive communication anti-interference method and system based on blind source separation. The method includes: acquiring received signals through multiple antennas; pre-identifying the interference type of the received signals to determine if the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference; adaptively selecting a suitable blind separation algorithm based on the pre-identification result; executing the selected algorithm to obtain multiple separated output signals; performing sliding correlation between locally stored pilot sequences and each separated output signal, and identifying the communication signal branch based on the correlation peak value; the locally stored pilot sequences are existing pilot sequences or preambles in the communication signal frame; performing carrier synchronization and channel equalization on the identified communication signals, and outputting the demodulated bit stream. This method can significantly reduce the bit error rate and achieve efficient and robust suppression of various unknown interferences.

Description

Technical Field

[0001] This application relates to the field of wireless communication anti-interference technology, and more specifically, to an adaptive communication anti-interference method and system based on blind source separation. Background Technology

[0002] With the increasing prominence of the strategic position of electromagnetic space, communication systems face severe anti-interference challenges in complex electromagnetic environments. Traditional anti-interference technologies such as direct sequence spread spectrum (DSSS) and frequency hopping (FHSS) are essentially passive avoidance strategies that rely on resource scheduling in the frequency domain, time domain, or code domain. They have the following shortcomings: (1) Limited anti-interference capability, with insufficient processing gain when facing high-power, broadband, or intelligent dynamic interference; (2) Reliance on prior information, requiring known interference patterns or channel parameters, making it difficult to adapt to unknown or dynamically changing interference; (3) Low spectral efficiency, with spread spectrum or frequency hopping occupying additional bandwidth.

[0003] Blind Source Separation (BSS) technology can separate communication signals from mixed signals using only the statistical independence of the received signals, even when both the source signals and mixed channel parameters are unknown. This provides a novel "cognitive anti-interference" approach for communication anti-interference. In recent years, research has applied blind separation algorithms such as FastICA (Fast Independent Component Analysis), JADE, and SOBI (Second-Order Blind Identification) to anti-interference in spread spectrum and frequency hopping communication systems, achieving some success.

[0004] However, existing technologies still have the following shortcomings: First, there is a lack of quantitative design basis for antenna layout. The prerequisite for the solvability of blind source separation is that the columns of the mixing matrix are full rank, and its numerical stability is characterized by the condition number. In engineering, there is a lack of clear mathematical guidance on how antenna spacing affects separation performance. Usually, the spacing is chosen based on experience, such as "spacing greater than half a wavelength", which cannot guarantee the optimal separation effect.

[0005] Second, resolving the order ambiguity relies on additional overhead. The order of the blind separation output signals is arbitrary, requiring identification of which path represents the communication signal. Existing methods either transmit known pilots (increasing system overhead) or perform blind modulation identification (unreliable at low signal-to-noise ratios).

[0006] Third, a single algorithm cannot adapt to all types of interference. Different blind separation algorithms exhibit significant performance differences under various interference patterns and channel conditions. For example, SOBI performs best under swept-frequency interference and broadband noise interference, while JADE is more advantageous under narrowband interference. Existing studies all employ fixed algorithms and lack adaptive selection mechanisms. Summary of the Invention

[0007] To address the problems in existing blind source separation anti-interference technologies, such as the lack of quantitative criteria for antenna layout, the additional overhead required for sorting ambiguity elimination, and the difficulty of adapting a single algorithm to all types of interference, this invention provides an adaptive communication anti-interference method and system based on blind source separation. This method and system will achieve efficient and robust suppression of various unknown interferences and improve the separation accuracy of blind separation algorithms.

[0008] To achieve the above objectives, according to a first aspect of the present invention, an adaptive communication anti-interference method based on blind source separation is provided, the method comprising: The received signal is acquired through multiple antennas. The received signal includes a linear instantaneous mixed signal of communication signal and interference signal. The received signal is pre-identified for interference type to determine whether the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference. Based on the pre-identification results, an adaptive blind separation algorithm is selected; the selected algorithm is executed to obtain the separated output multi-channel signals; The locally stored pilot sequence is correlated with each of the separated output signals, and the communication signal branch is identified based on the correlation peak value; the locally stored pilot sequence is an existing pilot sequence or preamble in the communication signal frame; The identified communication signals are carrier synchronized and channel equalized, and the demodulated bit stream is output.

[0009] According to a second aspect of the present invention, an adaptive communication anti-interference system based on blind source separation is also provided, comprising: A multi-antenna receiving module is configured to acquire received signals through multiple antennas, the received signals including a linear transient mixed signal of communication signals and interference signals; The interference pre-identification module is configured to pre-identify the interference type of the received signal and determine that the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference. The algorithm selection and control module is configured to adaptively select a suitable blind separation algorithm based on the pre-identification result; execute the selected algorithm to obtain the separated output multi-channel signal; The communication signal identification module is configured to perform sliding correlation between the locally stored pilot sequence and each of the separated output signals, and identify the communication signal branch based on the correlation peak value; the locally stored pilot sequence is an existing pilot sequence or preamble in the communication signal frame; The carrier synchronization and channel equalization module is configured to perform carrier synchronization and channel equalization on the identified communication signals and output the demodulated bit stream.

[0010] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: (1) The adaptive communication anti-interference method based on blind source separation provided by this invention is applied to the communication signal receiver. Utilizing blind source separation technology, it can adaptively select the blind separation algorithm according to the type of interference. It can significantly reduce the bit error rate under single-tone, multi-tone, frequency sweep, and broadband noise interference. The processing delay is in the millisecond range, making it suitable for cognitive anti-interference communication systems with high real-time requirements. It can achieve efficient and robust suppression of various unknown interferences. Furthermore, by using the existing pilot sequence in the communication signal frame for sliding correlation peak detection, it is possible to identify communication signal branches with zero additional overhead.

[0011] (2) The adaptive communication anti-interference method based on blind source separation provided by this invention optimizes the antenna spacing by minimizing the condition number of the mixing matrix, and selects the antenna spacing that minimizes the condition number as the optimal antenna spacing for the layout of multiple antennas. This ensures the optimal numerical stability of the mixing matrix and improves the separation accuracy of the blind separation algorithm. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart illustrating an adaptive communication anti-interference method based on blind source separation provided in this application embodiment; Figure 2 A structural block diagram of an adaptive communication anti-interference system based on blind source separation is provided in an embodiment of this application; Figure 3 A flowchart illustrating the steps of acquiring received signals through multiple antennas in an adaptive communication anti-interference method based on blind source separation provided in this application embodiment; Figure 4 A flowchart illustrating the antenna layout optimization steps in an adaptive communication anti-interference method based on blind source separation, provided in an embodiment of this application; Figure 5 A flowchart illustrating the interference type pre-identification step in an adaptive communication anti-interference method based on blind source separation provided in this application embodiment; Figure 6 A flowchart illustrating the blind separation algorithm steps for adaptive selection and adaptation in an adaptive communication anti-interference method based on blind source separation provided in this application embodiment; Figure 7 A flowchart illustrating the step of identifying communication signal branches based on relevant peak values ​​in an adaptive communication anti-interference method based on blind source separation provided in this application embodiment; Figure 8 This application provides an analysis of the bit error rate under a fixed interference-to-signal ratio of 10 during frequency sweeping interference. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0015] The terms "first," "second," "third," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0016] like Figure 1 As shown, an adaptive communication anti-interference method based on blind source separation is provided. This method is executed by the communication signal receiver and includes the following steps: Step 101: Acquire the received signal through multiple antennas. The received signal includes a linear instantaneous mixed signal of communication signal and interference signal.

[0017] Preferably, the multiple antennas are arranged with an optimal antenna spacing, which is determined according to the criterion of minimizing the condition number of the hybrid matrix.

[0018] Specifically, by acquiring received signals through multiple antennas and optimizing the antenna spacing with the goal of minimizing the condition number of the mixing matrix, the antenna spacing that minimizes this condition number is selected as the optimal antenna spacing for the multi-antenna layout. This ensures optimal numerical stability of the mixing matrix and improves the separation accuracy of the blind separation algorithm.

[0019] Step 102: Perform interference type pre-identification on the received signal to determine that the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference.

[0020] Step 103: Based on the pre-identification results, adaptively select the appropriate blind separation algorithm; execute the selected algorithm to obtain the separated output multi-channel signals.

[0021] Blind separation algorithms include FastICA, JADE, or SOBI algorithms.

[0022] Step 104: Perform sliding correlation between the locally stored pilot sequence and each of the separated output signals, and identify the communication signal branch based on the correlation peak value.

[0023] The locally stored pilot sequence is an existing pilot sequence or preamble in the communication signal frame.

[0024] Step 105: Perform carrier synchronization and channel equalization on the identified communication signal, and output the demodulated bit stream.

[0025] The aforementioned adaptive communication anti-interference method based on blind source separation, applied to the communication signal receiver, utilizes blind source separation technology to adaptively select the blind separation algorithm according to the interference type. It significantly reduces the bit error rate under single-tone, multi-tone, frequency-sweeping, and broadband noise interference, making it suitable for cognitive anti-interference communication systems and capable of efficiently and robustly suppressing various unknown interferences. Furthermore, by using the existing pilot sequence in the communication signal frame for sliding correlation peak detection, it is possible to identify communication signal branches with zero additional overhead.

[0026] In one embodiment, such as Figure 2 As shown, an adaptive communication anti-interference system based on blind source separation is also provided, including the following modules: A multi-antenna receiving module is configured to acquire received signals through multiple antennas, the received signals including a linear transient mixed signal of communication signals and interference signals; The interference pre-identification module is configured to pre-identify the interference type of the received signal and determine that the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference. The algorithm selection and control module is configured to adaptively select a suitable blind separation algorithm based on the pre-identification result. The blind separation algorithm includes FastICA, JADE, or SOBI algorithms. The selected algorithm is executed to obtain the separated output multi-channel signals. The communication signal identification module is configured to perform sliding correlation between the locally stored pilot sequence and each of the separated output signals, and identify the communication signal branch based on the correlation peak value; the locally stored pilot sequence is an existing pilot sequence or preamble in the communication signal frame; The carrier synchronization and channel equalization module is configured to perform carrier synchronization and channel equalization on the identified communication signals. The demodulation module is configured to demodulate the signal after carrier synchronization and channel equalization processing and output the demodulated bit stream.

[0027] like Figure 2 As shown, the above-mentioned adaptive communication anti-interference system based on blind source separation also includes a preprocessing module, which is configured to remove the mean from the received signal to obtain a zero-mean signal; to whiten the zero-mean signal to obtain a whitened observation signal; and to use the whitened observation signal as the input of the selected algorithm.

[0028] In one specific embodiment, an adaptive communication anti-interference method based on blind source separation includes the following steps: Step S1: Antenna layout optimization (offline / initialization phase) Figure 3 This is a schematic diagram illustrating the steps of acquiring received signals using multiple antennas. Figure 4 A flowchart for antenna layout optimization. Before system deployment (i.e., the communication signal receiver), i.e., during the offline or initialization phase, different antenna spacings are obtained through electromagnetic simulation or actual measurement. The following is a sample of the mixture matrix. Mixture matrix elements Indicates the first The source signal (including communication signals and interference signals) to the first The complex gain of the root receiving antenna, For the number of receiving antennas, The number of source signals.

[0029] Calculate the condition number for each sample: , for Find the inverse or pseudo-inverse matrix of the matrix and calculate the average condition number: The optimal antenna spacing (also known as optimal layout) is determined by minimizing the condition number of the hybrid matrix, i.e., selecting the antennas that minimize the average condition number. Minimum antenna spacing As the optimal layout, the condition number minimization criterion of the mixture matrix ensures the best numerical stability of the mixture matrix, thereby improving the separation accuracy of the blind separation algorithm.

[0030] Step S2: Multi-antenna reception of mixed signals The received signal is acquired by using at least two receiving antennas arranged in an optimal layout. The received signal comprises a linear transient mixture of communication and interference signals. Specifically, the mathematical model of the received signal can be a linear transient mixture model, expressed as:

[0031] in, The source signal vector (containing one communication signal and several interference signals). It is an additive white Gaussian noise vector.

[0032] Figure 3 The case with two receiving antennas is shown. At this point, the hybrid matrix... The received signal of antenna 1 is The received signal of antenna 2 is .

[0033] Step S3: Interference type pre-identification Figure 5 This is a flowchart illustrating the interference type pre-identification step for the received signal provided in an embodiment of this application. The flowchart shows two paths: frequency domain energy distribution analysis and time domain autocorrelation analysis. Figure 5 As shown, pre-identification of the received signal is performed to determine the current type of interference, including two pre-identification methods: Method A (based on frequency domain energy distribution): Perform a Fast Fourier Transform on the received signal to obtain the frequency domain amplitude spectrum. Calculate its kurtosis Single-tone / multi-tone interference has its frequency domain energy concentrated at a few frequencies, with a high kurtosis (typically >10); broadband noise interference has a flat frequency domain energy distribution, with a kurtosis close to 0; swept-frequency interference has a kurtosis between the two. A first threshold is set. Second threshold , ,like If it is, then it is judged as single-tone / multi-tone interference; if If it is positive, it is judged as broadband noise interference; otherwise, it is judged as frequency sweep interference.

[0034] Method B (based on time-domain autocorrelation): Calculate the autocorrelation function of the received signal. .in, For time variables, For time delay variables, In order to receive signals, for conjugate, This represents the mathematical expectation operation. Due to the periodic frequency variation, the autocorrelation function of swept-frequency interference exhibits a distinct periodic peak, with the peak interval equal to the swept-frequency period; the autocorrelation function of broadband noise interference... The frequency decays rapidly to zero; the autocorrelation function of single-tone / multi-tone interference exhibits a persistent non-zero value. Peak detection can distinguish frequency sweep interference from other interferences.

[0035] Step S4: Adaptive blind separation algorithm selection Figure 6 This is a flowchart illustrating the adaptive selection and adaptation steps of the blind separation algorithm provided in this application embodiment. The flowchart shows the decision tree for selecting FastICA, JADE, or SOBI based on the interference type label. Specifically, based on the pre-identification result of step S3, the optimal blind separation algorithm is automatically selected according to the following rules: Table 1

[0036] Step S5: Preprocessing and Blind Separation Preprocess the received signal: First, remove the mean. Then, whitening is performed. Whitening matrix. By analyzing the sampling covariance matrix Eigenvalue decomposition yields:

[0037] in, The eigenvector matrix, It is an eigenvalue diagonal matrix. Sampling time.

[0038] The signal after whitening is ,satisfy , It is an identity matrix.

[0039] Then execute the blind separation algorithm selected in step S4: (1) FastICA algorithm: The separation vector is updated using the complex field fixed-point iterative formula:

[0040] in, This is the whitened observed signal vector. This is the separation vector for the current iteration. The updated separation vector, for The conjugate transpose of . for conjugate, This represents the mathematical expectation operation. The derivative of a nonlinear function. for right The derivative of This represents the projected power of the separating vector.

[0041] Normalization is performed after each iteration: Continue until convergence. After convergence, the separated outputs are obtained. ,in This is the unmixing matrix.

[0042] (2) JADE algorithm: Estimate the fourth-order cumulant tensor of the whitened signal, construct a set of cumulant matrices, and solve the unitary matrix by joint approximate diagonalization. The unmixing matrix is ​​obtained. .

[0043] (3) SOBI algorithm: Select a set of time delays (Typical value K=10, time delay from 1 sampling interval to 10 sampling intervals), estimate the covariance matrix for each time delay. By jointly diagonalizing the unitary matrix Make all Approximate diagonalization yields the unmixing matrix. .

[0044] Step S6: Communication signal identification based on pilot correlation peaks Figure 7 This is a flowchart illustrating the steps for identifying communication signal branches based on correlation peak values ​​in an embodiment of this application. The flowchart shows the two separated output signals being correlated with a locally stored pilot sequence (hereinafter referred to as the local pilot sequence), and the communication signal branch is determined by comparing the peak values. Multiple signals are output via blind separation. There is sorting fuzziness, where This refers to the number of output signal channels. This embodiment utilizes the existing pilot sequence (or preamble) in the communication signal frame. Identification can be performed without additional overhead.

[0045] Specific steps: Each signal output by blind separation (Discrete sampling form), and local pilot sequence Perform sliding related tasks:

[0046] in, For the starting sampling point index of the sliding correlation, The length of the pilot sequence. This is the sample index for the local pilot sequence. The first in the local pilot sequence One sample point, for conjugate, For the separated output of the first road signal, This represents the total number of received signal sampling points (i.e., the total length of the signal samples) used to perform the sliding correlation operation.

[0047] Calculate the maximum correlation peak value for each signal: .

[0048] Determine the communication signal branch The remaining branches are interference signals or noise.

[0049] Step S7: Carrier Synchronization and Channel Equalization For the identified communication signals Carrier synchronization is performed. Due to amplitude and phase ambiguity in blind separation, the communication signal can be modeled as follows:

[0050] in, For unknown magnitude, For residual frequency offset, For the sake of bias, To send symbols, For symbol period, It is additive Gaussian noise.

[0051] Finally, channel equalization is performed. If the channel exhibits flat fading (satisfying the linear instantaneous mixing model assumption of this invention), complex equalization is unnecessary. If inter-symbol interference caused by multipath propagation exists, a linear transverse equalizer or a decision feedback equalizer can be used, updating the tap coefficients using the minimum mean square error criterion. The equalized signal is then fed into a demodulator, outputting the final bitstream.

[0052] The adaptive communication anti-interference method based on blind source separation provided in this embodiment has the following beneficial effects: 1. Pioneering quantitative design principle for antenna layout: Optimize antenna spacing with the goal of minimizing the condition number of the hybrid matrix, providing a mathematical basis for engineering implementation. Compared with the empirical "spacing greater than half wavelength", it can improve separation accuracy by more than 10% (simulation verification).

[0053] 2. Zero-overhead sorting ambiguity elimination: Communication signal identification is completed using the existing pilot sequence or preamble in the communication signal frame without sending any additional signaling, and the correlation peak detection still has a recognition accuracy of nearly 100% even at low signal-to-noise ratios (SNR≥0dB).

[0054] 3. Interference type adaptive algorithm selection: For the first time, three blind separation algorithms, FastICA, JADE and SOBI, are integrated into the same system. They are automatically switched through a lightweight pre-identification step (computation amount less than one FFT), and the overall bit error rate performance is better than any fixed algorithm.

[0055] 4. Comprehensive Interference Type Coverage: Verified through 500 Monte Carlo simulations, this invention can reduce the bit error rate (BER) from 10 under four typical suppression interference types: single-tone, multi-tone, frequency sweep, and broadband noise. 1 The above has been reduced to 10. 3 For the following (ISR ≤ 20dB, SNR ≥ 10dB), the performance is improved by more than two orders of magnitude. Figure 8 The bit error rate results are for a fixed interference-to-signal ratio (ISR) of 10 under frequency sweeping interference.

[0056] 5. Adaptability to complex channel environments: The present invention maintains stable anti-interference performance in both Gaussian white noise channels and Ricean fading channels (K factor = 10dB), with the performance degradation of the SOBI algorithm in Ricean channels being less than 0.5 orders of magnitude.

[0057] In one specific embodiment, a real-time anti-interference receiver is used to execute the above-described adaptive communication anti-interference method based on blind source separation in the following scenarios.

[0058] The scene parameters are as follows: Channel: Additive White Gaussian Noise (AWGN) channel; Modulation: QPSK-OFDM, FFT points 64, cyclic prefix 16, 200 symbols per frame, sampling rate 1MHz; Interference type: frequency sweep interference, frequency sweep range 0.2-0.8kHz; Signal-to-noise ratio (SNR): SNR = 10 dB; Interference-to-Signal Ratio (ISR): ISR = 10 dB; Monte Carlo trials: 500.

[0059] Algorithm selection: Due to the high real-time requirements of the system, and the fact that FastICA and SOBI perform similarly under frequency sweep interference, the FastICA algorithm was chosen.

[0060] The performance results are as follows: The bit error rate (BER) of the received signal without anti-interference processing is approximately 1.18 × 10⁻⁶. 1 (Frequency sweeping interference, SNR=10dB, ISR=10dB). However, after adopting the adaptive communication anti-interference method based on the FastICA algorithm of this invention, the bit error rate was reduced to 5.89×10⁻⁶. 3 Therefore, by employing the method provided by this invention, the bit error rate (BER) can be reduced from 10. 1 The above has been reduced to 10. 3 The performance has improved by approximately two orders of magnitude.

[0061] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0062] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0063] The above description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of embodiments of this disclosure upon considering the specification and practicing the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

[0064] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0065] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An adaptive communication anti-interference method based on blind source separation, applied to a communication signal receiver, characterized in that, include: The received signal is acquired by multiple antennas, and the received signal includes a linear instantaneous mixed signal of communication signal and interference signal; The received signal is pre-identified for interference type to determine whether the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference. Based on the pre-identification results, an adaptive blind separation algorithm is selected. Execute the selected algorithm to obtain the separated output multiple signals; The locally stored pilot sequence is correlated with each of the separated output signals, and the communication signal branch is identified based on the correlation peak value; the locally stored pilot sequence is an existing pilot sequence or preamble in the communication signal frame; The identified communication signals are carrier synchronized and channel equalized, and the demodulated bit stream is output.

2. The adaptive communication anti-interference method based on blind source separation as described in claim 1, characterized in that, The multiple antennas are arranged according to the optimal antenna spacing, which is determined based on the criterion of minimizing the condition number of the hybrid matrix.

3. The adaptive communication anti-interference method based on blind source separation as described in claim 2, characterized in that, The condition number minimization criterion for the mixture matrix is: During the offline or initialization phase, hybrid matrix samples with different antenna spacings are obtained through electromagnetic simulation or actual measurement, and the average condition number is calculated. in, Antenna spacing, It is a mixed matrix. , For the number of receiving antennas, The number of source signals, elements Indicates the first The source signal to the first The complex gain of the root receiving antenna, for The inverse matrix or pseudo-inverse matrix, This represents the mathematical expectation operation; The antenna spacing that minimizes the average condition number is selected as the optimal antenna spacing.

4. The adaptive communication anti-interference method based on blind source separation as described in claim 1, characterized in that, The step of pre-identifying the interference type of the received signal and determining whether the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference includes: Perform a Fast Fourier Transform on the received signal to obtain the frequency domain energy distribution, and calculate its kurtosis: in, For frequency variables, The frequency domain amplitude spectrum obtained by fast Fourier transform of the received signal. for Frequency domain energy distribution, Represents the mathematical expectation operation; If the kurtosis is greater than a first threshold, it is determined to be single-tone or multi-tone interference; if the kurtosis is less than a second threshold, it is determined to be broadband noise interference, where the second threshold is less than the first threshold; if the kurtosis is not less than the second threshold and not greater than the first threshold, it is determined to be frequency sweep interference.

5. The adaptive communication anti-interference method based on blind source separation as described in claim 1, characterized in that, The step of pre-identifying the interference type of the received signal and determining whether the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference includes: Calculate the autocorrelation function of the received signal: in, For time variables, For time delay variables, In order to receive signals, for conjugate, This represents the mathematical expectation operation; If the autocorrelation function is in If there are multiple local peaks within the range, and the time interval between adjacent peaks is equal to the preset frequency sweep period, then it is determined to be frequency sweep interference.

6. The adaptive communication anti-interference method based on blind source separation as described in claim 1, characterized in that, The blind separation algorithm includes the FastICA algorithm, the JADE algorithm, or the SOBI algorithm. The adaptive selection of the appropriate blind separation algorithm based on the pre-identification result includes: If the pre-identification result is single-tone interference or multi-tone interference, then select the JADE algorithm or the FastICA algorithm; If the pre-identification result is frequency sweeping interference or broadband noise interference, then select the SOBI algorithm; If the signal processing delay requirement of the communication signal receiver is not greater than the preset delay threshold, then the FastICA algorithm will be selected for any pre-identification result.

7. The adaptive communication anti-interference method based on blind source separation as described in claim 1, characterized in that, The step of performing sliding correlation between the locally stored pilot sequence and each of the separated output signals, and identifying the communication signal branch based on the correlation peak value, includes: The locally stored pilot sequence is then subjected to sliding correlation with each of the separated output signals: ; in, For the starting sampling point index of the sliding correlation, , The total number of received signal sampling points used to perform the sliding correlation operation. The length of the pilot sequence. This is the sample index of the pilot sequence. The first in the pilot sequence One sample point, for conjugate, For the separated output of the first road signal; Calculate the maximum correlation peak value for each signal: ; Compare the maximum correlation peak values ​​of each signal, and determine the branch corresponding to the maximum value among all the maximum correlation peak values ​​as the communication signal branch.

8. The adaptive communication anti-interference method based on blind source separation as described in claim 1, characterized in that, Also includes: The received signal is mean-reduced to obtain a zero-mean signal; The zero-mean signal is whitened to obtain the whitened observation signal; The whitened observation signal is used as the input to the selected algorithm.

9. An adaptive communication anti-interference system based on blind source separation, characterized in that, include: A multi-antenna receiving module is configured to acquire received signals through multiple antennas, the received signals including a linear transient mixed signal of communication signals and interference signals; An interference pre-identification module is configured to perform interference type pre-identification on the received signal and determine that the current interference belongs to at least one of single-tone interference, multi-tone interference, frequency sweeping interference, or broadband noise interference. The algorithm selection and control module is configured to adaptively select a suitable blind separation algorithm based on the pre-identification results; Execute the selected algorithm to obtain the separated output multiple signals; The communication signal identification module is configured to perform sliding correlation between the locally stored pilot sequence and each of the separately output signals, and identify the communication signal branch based on the correlation peak value; the locally stored pilot sequence is an existing pilot sequence or preamble in the communication signal frame; The carrier synchronization and channel equalization module is configured to perform carrier synchronization and channel equalization on the identified communication signals and output the demodulated bit stream.

10. The adaptive communication anti-interference system based on blind source separation as described in claim 9, characterized in that, It also includes a preprocessing module configured to remove the mean from the received signal to obtain a zero-mean signal; to whiten the zero-mean signal to obtain a whitened observation signal; and to use the whitened observation signal as the input to the selected algorithm.

Images