Anti-interference channelized detection method based on weighted Gaussianity test

Abstract

The invention discloses an anti-interference channelization detection method based on weighted Gaussianity test, which belongs to the field of electronic reconnaissance. First, radar signals are preprocessed to generate ascending signal accumulation envelope sequences; the mean estimation value and variance estimation value of the signal accumulation envelope sequence are calculated using the moment estimation expression of Gaussian distribution; envelope probability density sequence and probability density sequence of reference Gaussian distribution are constructed to calculate the Bhattacharyya distance value; envelope cumulative distribution sequence and cumulative distribution sequence of reference Gaussian distribution are constructed to calculate the K-S distance value; weighting factors are set to calculate the test statistic; the test statistic is compared with the decision threshold to obtain the interference shielding code, shield the interference signal occupied channel, perform signal detection on the unshielded channel, and obtain the detection VP. The invention can realize the effective judgment and shielding of the interference signal occupied frequency band, greatly reduce the data processing pressure of electronic reconnaissance equipment, and improve the efficiency of radiation source signal sorting and positioning tracking.

Description

Technical Field

[0001] The invention belongs to the field of electronic reconnaissance, and in particular relates to an anti-interference channelized detection method based on weighted Gaussianity test. Background Art

[0002] In modern electronic warfare, the electromagnetic environment in which electronic reconnaissance equipment operates is complex and changeable. In order to achieve real-time monitoring of the entire scene and wide frequency band, the electronic reconnaissance receiver needs to have a large instantaneous bandwidth. At this time, the receiver can detect a large number of pulse responses, most of which come from non-Gaussian, broadband noise interference signals in the reconnaissance electromagnetic environment. The existence of these interference signals greatly increases the signal processing and data transmission pressure of electronic reconnaissance equipment, and even leads to the loss of key radiation source detection results, seriously affecting the adaptability of electronic reconnaissance equipment to complex electromagnetic environments. In this regard, based on the commonly used electronic reconnaissance receiver structure, the study of effective interference discrimination and shielding methods is of great significance for improving the detection performance of electronic reconnaissance equipment and reducing the data processing and transmission burden of the equipment.

[0003] In traditional digital receivers, in order to improve the ability to capture radiation sources in reconnaissance scenarios and broaden the real-time processing bandwidth, a signal detection process based on a channelized structure is usually adopted. The document "Design and Application of Multi-rate Filter Banks" proposes a channelization process based on a multi-phase filtering structure. After completing the sub-channel division of the intermediate frequency data using channelization processing, this method directly performs signal detection and parameter measurement on the data sequence of each sub-channel, lacking the judgment of whether the sub-channel is occupied by interference signals. There are often a variety of non-Gaussian, wide-band noise interference signals in the reconnaissance environment, which greatly increases the burden of radiation source reconnaissance and tracking processing. Therefore, it is of great significance to study the effective judgment method of the interference signal on the sub-channel occupancy and shield the signal output of the interference occupied sub-channel to adapt to the complex and changeable actual reconnaissance environment and realize the anti-interference channelized detection of electronic countermeasure equipment. Summary of the invention

[0004] The present invention proposes an anti-interference channelization detection method based on weighted Gaussianity test, which can realize effective judgment and shielding of the frequency band occupied by interference signals, greatly reduce the data processing pressure of electronic reconnaissance equipment, and improve the efficiency of radiation source signal sorting and positioning tracking; a weighted Gaussianity test method is proposed to realize accurate test of the statistical characteristics of sub-channel signals, and obtain effective judgment whether there is a strong non-Gaussian interference signal in a specific frequency band; real-time shielding of sub-channels occupied by non-Gaussian noise interference in complex electromagnetic environments is realized, the output of non-Gaussian and wide-band noise interference signals is effectively reduced, and the environmental adaptability of electronic reconnaissance equipment is improved.

[0005] The technical solution for implementing the present invention is: an anti-interference channelization detection method based on weighted Gaussianity test, comprising the following steps:

[0006] Step 1: After AD sampling, multi-phase filtering and time domain energy accumulation of the radar signal, the signal accumulation envelope sequence {A1, A2, …, A L}, L is the length of the signal accumulation envelope sequence, and the signal accumulation envelope sequence is processed in ascending order to generate an ascending signal accumulation envelope sequence, and then go to step 2.

[0007] Step 2: Use the moment estimation expression of Gaussian distribution to calculate the mean estimate and variance estimate of the signal accumulation envelope sequence, and then proceed to Step 3 and Step 4.

[0008] Step 3: Construct an envelope probability density sequence based on the accumulated envelope sequence of the ascending signal, use the mean estimate and variance estimate to obtain the probability density sequence of the reference Gaussian distribution, and calculate the Bhattacharyya distance value D B , go to step 5.

[0009] Step 4: Construct an envelope cumulative distribution sequence based on the accumulated envelope sequence of the ascending signal, use the mean estimate and variance estimate to obtain the cumulative distribution sequence of the reference Gaussian distribution, calculate the KS distance value KSD, and proceed to step 5.

[0010] Step 5: Set the weighting factor and use the Bhattacharyya distance value D B And KS distance value KSD calculates the test statistic:

[0011] Set the weighting factor 0<α<1, and use the Bhattacharyya distance value D B And KS distance value KSD calculates the test statistic G w , the calculation expression is:

[0012] G w =α×KSD+(1-α)×D B ,0<α<1

[0013] Go to step 6.

[0014] Step 6: Compare the test statistic with the decision threshold to obtain the interference shielding code, shield the interference signal from occupying the channel, perform signal detection on the unshielded channel, and obtain the detection VP.

[0015] Compared with the prior art, the present invention has the following significant advantages:

[0016] 1) The invention realizes the inspection and analysis of the frequency band occupied by interference signals in complex electromagnetic environments, thereby improving the adaptability to complex electromagnetic environments.

[0017] 2) A noise floor parameter estimation method is proposed to improve the understanding of the statistical characteristics of the sub-channel data sequence, providing a basis for subsequent Gaussianity tests and effective judgment of interference-occupied sub-channels.

[0018] 3) An anti-interference channelized detection method based on weighted Gaussian test is proposed, which realizes the effective shielding of interference-occupied sub-channels, reduces the data processing pressure of electronic reconnaissance equipment, and improves the probability of capturing signals of key radiation sources. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Flowchart of the anti-interference channelized detection method based on weighted Gaussianity test

[0020] Figure 2 Schematic diagram of Bhattacharyya distance values ​​and KS distance values ​​of different sub-channels, where (a) is a schematic diagram of Bhattacharyya distance values ​​of different sub-channels, and (b) is a schematic diagram of KS distance values ​​of different sub-channels.

[0021] Figure 3 Schematic diagram of the output results of the traditional channelized detection PDW, where (a) is a schematic diagram of the amplitude characteristics of the traditional channelized detection, and (b) is a schematic diagram of the frequency characteristics of the traditional channelized detection.

[0022] Figure 4 Schematic diagram of the PDW output results after the weighted Gaussianity test, where (a) is the amplitude characteristic after the weighted Gaussianity test, and (b) is the frequency characteristic after the weighted Gaussianity test. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] The following will further introduce the specific implementation method, as well as the technical difficulties and inventive points of this invention in combination with this design example.

[0025] Combination Figure 1 ,Anti-interference channelization detection method based on weighted Gaussianity test, the steps are as follows:

[0026] Step 1: After AD sampling, multi-phase filtering and time domain energy accumulation of the radar signal, a signal accumulation envelope sequence is obtained, and the signal accumulation envelope sequence is processed in ascending order to generate an ascending signal accumulation envelope sequence to obtain a signal accumulation envelope sequence. Specifically, the following steps are included:

[0027] Step 11: After AD sampling and multi-phase filtering of the radar signal, the filtered data is obtained. k (n1M) is as follows:

[0028]

[0029] Among them, let the input time variable n1 correspond to the sampling frequency before filtering as 500MHz, the sampling rate M = 32, then the output sampling frequency time variable n2 after filtering = 500 / 32 = 15.625MHz; the number of channels G = 32, ω g (g=0,1,···,G-1) is the center frequency of each channel. The coefficient length allocated to each channel filter is 8, then h(n1) is a 1024-order low-pass FIR filter, x(n1) is the sampling signal sequence, and j represents the imaginary part.

[0030] Step 12: Modulo the filtered data of each channel to obtain the amplitude sequence Amp after filtering of each channel.

[0031] Step 13, according to the pulse width matching accumulation module, when one or more weak signals are drowned by noise or remain flat, the accumulation points that match the signal width are used. The result of noise superposition is that the noise almost cancels each other out, while the weak signal part is retained. The more data the signal is superimposed on, the higher the accumulated energy is, thus achieving the maximum signal-to-noise ratio improvement. Using a length of 1000ns, the time domain energy sliding accumulation is performed on the amplitude sequence Amp after filtering of each channel, and each channel generates a signal accumulation envelope sequence {A1, A2, ..., A L}, L is the length of the signal accumulation envelope sequence.

[0032] Step 14: Sort the signal accumulation envelope sequence from small to large to obtain an ascending signal accumulation envelope sequence A = {A (1) ,A (2) ,…,A (L)}.

[0033] Step 2: Calculate the mean estimate and variance estimate of the signal accumulation envelope sequence using the moment estimate expression of the Gaussian distribution, which specifically includes the following steps:

[0034] Step 21: Calculate the mean estimate using the signal accumulation envelope sequence The specific calculation expression is:

[0035]

[0036] Among them, A l Accumulates the lth element of the envelope sequence for signal.

[0037] Step 22: The mean estimate obtained in step 21 Calculate the variance estimate using the signal accumulation envelope sequence The specific calculation expression is:

[0038]

[0039] Step 3: Construct the envelope probability density sequence p(x) based on the accumulated envelope sequence of the ascending signal, obtain the probability density sequence q(x) of the reference Gaussian distribution using the mean estimate and variance estimate, and calculate the Bhattacharyya distance value D B The specific steps include:

[0040] Step 31: Utilize the ascending signal to accumulate the envelope sequence A={A (1) ,A (2) ,…,A (L)}, divide the ascending signal accumulation envelope sequence into 200 sub-intervals, and calculate the envelope probability density distribution sequence p(x), where x = {x1, x2, …, x i ,…,x 200} represents the center value of each sub-interval, and i represents the sub-interval number of the ascending signal accumulation envelope sequence.

[0041] Step 31_1, calculate the maximum value A of the accumulated envelope sequence of the ascending signal max and minimum value A min , accumulate the elements of the envelope sequence of the ascending signal in the interval [A min ,A max ] is divided into 200 subintervals, i represents the subinterval number of the ascending signal accumulation envelope sequence, and the subinterval set Ω={ω1,ω2,…,ω i ,…,ω 200}, where ω i represents the i-th subinterval.

[0042] Step 31_2: Let the number of accumulated elements in each sub-interval be Num = {num1, num2, ..., num i ,…,num 200}, and initialize Num = 0.

[0043] Step 31_3, determine the ascending signal accumulation envelope sequence A = {A (1) ,A (2) ,…,A (L)}The size of the lth element belongs to the subinterval, if the subinterval it belongs to is ω i , then let num i =num i +1.

[0044] Step 31_4, repeat step 31_3 until l=L, and obtain a set P consisting of the total number of accumulated elements in each subinterval, and satisfy L is the length of the signal accumulation envelope sequence.

[0045] Step 31_5: Calculate the probability density value prob of the ith subinterval using the following formula:i :

[0046]

[0047] Get the envelope probability density sequence p(x)={prob i ,i=1,2,…,200}, where the center value of each subinterval is x={x1,x2,…,x i ,…,x 200} represents, and satisfies x i =(i-0.5)×(A max -A min ) / 200.

[0048] Step 32: Use the mean estimate and variance estimates The probability density sequence q(x) of the reference Gaussian distribution is obtained, and the calculation expression is as follows:

[0049]

[0050] Step 33: Calculate the Bhattacharyya distance value D using the envelope probability density sequence p(x) and the probability density sequence q(x) of the reference Gaussian distribution. B The specific calculation expression is:

[0051] D B =-ln(BC(p,q))

[0052] in It can be seen that when p(x) and q(x) are exactly the same, D B =0; the stronger the non-Gaussian characteristics of the sequence to be detected, the higher the D B The larger the value.

[0053] Step 4: construct the envelope cumulative distribution sequence P(A) based on the accumulated envelope sequence of the ascending signal, obtain the cumulative distribution sequence Q(A) of the reference Gaussian distribution using the mean estimate and variance estimate, and calculate the KS distance value KSD. Specifically, the following steps are included:

[0054] Step 41: Utilize the ascending signal to accumulate the envelope sequence A={A (1) ,A (2) ,…,A (L)}, construct the envelope cumulative distribution sequence P(A) = {P(A (1) ),P(A (2) ),…,P(A (l) ),…,P(A (L) )}, where the value of the lth element of the envelope cumulative distribution sequence is P(A (l) )=l / L.

[0055] Step 42: Use the mean estimate and variance estimates The cumulative distribution sequence Q(A) of the reference Gaussian distribution is obtained, and the calculation expression is as follows:

[0056]

[0057] Among them, erf(*) represents the error function, which is a monotonic function of the independent variable.

[0058] Step 43: Calculate the KS distance value KSD using the envelope cumulative distribution sequence P(A) and the reference Gaussian distribution cumulative distribution sequence Q(A). The specific calculation expression is:

[0059] KSD(P,Q)=max{|P(A)-Q(A)|}

[0060] The larger the KSD value, the worse the Gaussianity of the sequence to be detected, and vice versa.

[0061] Step 5: Set the weighting factor 0<α<1 and use the Bhattacharyya distance value D B And KS distance value KSD calculates the test statistic G w , the calculation expression is:

[0062] G w =α×KSD+(1-α)×D B ,0<α<1

[0063] Test statistic G w The larger the value, the stronger the non-Gaussian characteristics of the subchannel are, and the greater the probability of the presence of interference signals.

[0064] Step 6: Compare the test statistic with the decision threshold to obtain an interference shielding code, shield the interference signal from occupying the channel, and perform signal detection on the unshielded channel, which specifically includes the following steps:

[0065] Step 61: construct a binary hypothesis test problem for interference channel determination. The corresponding expression of the binary hypothesis test problem for each subchannel is as follows:

[0066]

[0067] In the formula, A l 、s l 、j l and n l They correspond to the lth element of the signal accumulation envelope sequence, target accumulation envelope sequence, interference accumulation envelope sequence and noise accumulation envelope sequence respectively, and L is the length of the signal accumulation envelope sequence; in the hypothesis test, H0 and H1 respectively represent that the subchannel has no interference signal and contains interference signal.

[0068] Step 62: Set the decision threshold T and use the binary hypothesis test problem to accurately determine whether there is an interference signal in each sub-channel. The specific expression is as follows:

[0069]

[0070] According to the binary hypothesis test result, the interference mask code of each subchannel is obtained. When the test result is H1, the interference mask code of the subchannel is set to 1; when the test result is H0, the interference mask code of the subchannel is set to 0.

[0071] Step 63: If the interference masking code of the subchannel is 1, mask the subchannel and do not perform signal detection; if the interference masking code of the subchannel is 0, perform signal detection. Traverse all subchannels to obtain the signal detection VP after the interference signal is filtered.

[0072] According to the experimental analysis of measured data, the Bhattacharyya distance and KS distance values ​​of channels 7 and 9 are too large, indicating that there are a lot of non-Gaussian noise interference signals in the channel data. The channel can be closed (such as Figure 2 In addition, the measured data was channelized and PDW information was output. It was observed that strong interference signals existed in channels 7, 9 and 11, resulting in 1752 PDW information outputs, too many false signals, unclear amplitude characteristics, and a large amount of storage and data transmission resources were occupied, which was not conducive to the subsequent signal sorting and recognition process (as shown in Figure 1). Figure 3 As shown in Figure 2 ). By comparison, it is found that after the Gaussian test, the above interference channels are all identified and shielded, and 229 PDW information is output, providing an effective basis for subsequent signal processing (as shown in Figure 2 ). Figure 4 shown).

Claims

1. An anti-interference channelization detection method based on weighted Gaussianity test, characterized in that: The following steps are involved: Step 1: After AD sampling, multi-phase filtering and time domain energy accumulation of the radar signal, the signal accumulation envelope sequence {A1, A2, …, A L }, L is the length of the signal accumulation envelope sequence, and the signal accumulation envelope sequence is processed in ascending order to generate an ascending signal accumulation envelope sequence, and then proceed to step 2; Step 2: Calculate the mean estimate and variance estimate of the signal accumulation envelope sequence using the moment estimate expression of Gaussian distribution, and proceed to Step 3 and Step 4 at the same time; Step 3: Construct an envelope probability density sequence based on the accumulated envelope sequence of the ascending signal, use the mean estimate and variance estimate to obtain the probability density sequence of the reference Gaussian distribution, and calculate the Bhattacharyya distance value D B , go to step 5; Step 4: construct an envelope cumulative distribution sequence based on the accumulated envelope sequence of the ascending signal, obtain the cumulative distribution sequence of the reference Gaussian distribution using the mean estimate and the variance estimate, calculate the KS distance value KSD, and proceed to step 5; Step 5: Set the weighting factor and use the Bhattacharyya distance value D B And KS distance value KSD calculates the test statistic: Set the weighting factor 0<α<1, and use the Bhattacharyya distance value D B And KS distance value KSD calculates the test statistic G w , the calculation expression is: G w =α×KSD+(1-α)×D B ,0<α<1 Go to step 6; Step 6: Compare the test statistic with the decision threshold to obtain the interference shielding code, shield the interference signal from occupying the channel, perform signal detection on the unshielded channel, and obtain the detection VP.

2. The anti-interference channelization detection method based on weighted Gaussianity test according to claim 1 is characterized in that: In step 2, the moment estimation expression of Gaussian distribution is used to calculate the signal accumulation envelope sequence {A1, A2, …, A L }, L is the length of the signal accumulation envelope sequence, specifically including the following steps: Step 21: Calculate the mean estimate using the signal accumulation envelope sequence The specific calculation expression is: Among them, A l The lth element of the signal accumulation envelope sequence; Step 22: The mean estimate obtained in step 21 Calculate the variance estimate using the signal accumulation envelope sequence The specific calculation expression is:

3. The anti-interference channelization detection method based on weighted Gaussianity test according to claim 2 is characterized in that: In step 3, the envelope probability density sequence p(x) is constructed based on the accumulated envelope sequence of the ascending signal, and the probability density sequence q(x) of the reference Gaussian distribution is obtained using the mean estimate and variance estimate, and the Bhattacharyya distance value D is calculated. B , specifically including the following steps: Step 31: Utilize the ascending signal to accumulate the envelope sequence A={A (1) ,A (2) ,…,A (L) }, divide the ascending signal accumulation envelope sequence into 200 subintervals, and calculate the envelope probability density sequence p(x), where x = {x1, x2, …, x i ,…,x 200 } represents the center value of each subinterval, i represents the subinterval number of the ascending signal accumulation envelope sequence; Step 32: Use the mean estimate and variance estimates The probability density sequence q(x) of the reference Gaussian distribution is obtained, and the calculation expression is as follows: Step 33: Calculate the Bhattacharyya distance value D using the envelope probability density sequence p(x) and the probability density sequence q(x) of the reference Gaussian distribution. B , the specific calculation expression is: D B =-ln(BC(p,q)) in It can be seen that when p(x) and q(x) are exactly the same, D B =0; the stronger the non-Gaussian characteristics of the sequence to be detected, the higher the D B The larger the value.

4. The anti-interference channelization detection method based on weighted Gaussianity test according to claim 3 is characterized in that: In step 4, the envelope cumulative distribution sequence P(A) is constructed according to the ascending signal accumulation envelope sequence, the cumulative distribution sequence Q(A) of the reference Gaussian distribution is obtained by using the mean estimate and the variance estimate, and the KS distance value KSD is calculated, which specifically includes the following steps: Step 41: Utilize the ascending signal to accumulate the envelope sequence A={A (1) ,A (2) ,…,A (L) }, construct the envelope cumulative distribution sequence P(A) = {P(A (1) ),P(A (2) ),…,P(A (l) ),…,P(A (L) )}, where the value of the lth element of the envelope cumulative distribution sequence is P(A (l) ) = l / L; Step 42: Use the mean estimate and variance estimates The cumulative distribution sequence Q(A) of the reference Gaussian distribution is obtained, and the calculation expression is as follows: Among them, erf(*) represents the error function, which is a monotonic function of the independent variable; Step 43: Calculate the KS distance value KSD using the envelope cumulative distribution sequence P(A) and the reference Gaussian distribution cumulative distribution sequence Q(A). The specific calculation expression is: KSD(P,Q)=max{|P(A)-Q(A)|} The larger the KSD value, the worse the Gaussianity of the sequence to be detected, and vice versa.

5. The anti-interference channelization detection method based on weighted Gaussianity test according to claim 4 is characterized in that: In step 6, the test statistic is compared with the decision threshold to obtain an interference shielding code, shield the interference signal from occupying the channel, perform signal detection on the unshielded channel, and obtain a detection VP, which specifically includes the following steps: Step 61: construct a binary hypothesis test problem for interference channel determination. The corresponding expression of the binary hypothesis test problem for each subchannel is as follows: In the formula, A l 、s l 、j l and n l They correspond to the lth element of the signal accumulation envelope sequence, the target accumulation envelope sequence, the interference accumulation envelope sequence and the noise accumulation envelope sequence, respectively, and L is the length of the signal accumulation envelope sequence; in the hypothesis test, H0 and H1 correspond to the subchannel without interference signal and with interference signal respectively; Step 62: Set the decision threshold T and use the binary hypothesis test problem to accurately determine whether there is an interference signal in each sub-channel. The specific expression is as follows: According to the binary hypothesis test result, the interference mask code of each subchannel is obtained; when the test result is H1, the interference mask code of the subchannel is set to 1; when the test result is H0, the interference mask code of the subchannel is set to 0; Step 63: If the interference masking code of the subchannel is 1, mask the subchannel and do not perform signal detection; if the interference masking code of the subchannel is 0, perform signal detection; traverse all subchannels to obtain the signal detection VP after the interference signal is filtered.

Images