阅读说明：本技术 一种基于时频特征图的目标信号检测方法及检测系统 (Target signal detection method and system based on time-frequency characteristic diagram ) 是由 张海如 汪俊 王海斌 台玉朋 于 2020-07-29 设计创作，主要内容包括：本发明属于水声信号检测与处理技术领域,具体地说,涉及一种基于时频特征图的目标信号检测方法,该方法包括：分别提取发射信号的时频特征图矩阵和接收信号的时频特征图矩阵；根据提取的发射信号的时频特征图矩阵和接收信号的时频特征图矩阵,计算二者的相似度,进而得到相似度曲线向量；根据得到的相似度曲线向量和预先设定的判断标准,检测目标信号。(The invention belongs to the technical field of underwater acoustic signal detection and processing, and particularly relates to a target signal detection method based on a time-frequency characteristic diagram, which comprises the following steps: respectively extracting a time-frequency characteristic diagram matrix of a transmitting signal and a time-frequency characteristic diagram matrix of a receiving signal; calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal, and further obtaining a similarity curve vector; and detecting a target signal according to the obtained similarity curve vector and a preset judgment standard.)

1. A target signal detection method based on a time-frequency characteristic diagram comprises the following steps:

respectively extracting a time-frequency characteristic diagram matrix of a transmitting signal and a time-frequency characteristic diagram matrix of a receiving signal;

calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal, and further obtaining a similarity curve vector;

and detecting a target signal according to the obtained similarity curve vector and a preset judgment standard.

2. The method for detecting a target signal based on a time-frequency feature map as claimed in claim 1, wherein the extracting process of the time-frequency feature map matrix of the transmission signal specifically comprises:

random frequency modulation signal is used as transmitting signal, and its discrete data sequence is S_Copy[n]：

Wherein N is equal to [1, N ∈]N is the total sampling point number of the discrete data sequence corresponding to the transmitting signal; omega₀Is the center frequency; f. of_SIs the sampling rate; b is the frequency modulation signal bandwidth; k (x) e [ -1,1]Is a 0-mean random sequence, and x is the serial number of the random sequence; x is the length of a random sequence selected for generating random frequency modulation signals; t is₀The frequency modulation time occupied by each code element in the random sequence; r 2]Is a ramp function, and is a step function u [ [ alpha ] ]]The integration result of (1);

from a discrete data sequence S of the transmitted signal_Copy[n]Establishing a time-frequency diagram matrix A of the transmitted signal_Copy；

Wherein A is_CopyCorresponds to the frequency domain information of the transmitted signal, the frequency band range of which is

Time-frequency diagram matrix A according to transmission signals_CopyExtracting time-frequency characteristic diagram matrix R of transmitting signal_Copy；

Wherein R is_CopyInitialisation to a full 0 matrix, R_CopyCorresponds to the frequency domain information of the transmitted signal, the frequency band range of which isWith a frequency point interval ofWherein N is_FeatureThe number of the target frequency band B is divided at equal intervals; r_CopyEach column of (a) corresponds to the time domain information of the transmitted signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is f_s×T₀，R_CopyTotal number of columns of

Extraction of A_CopyThe upper envelope curve of the jth column vector of (a),

The frequency value corresponding to the effective maximum value point of the jth column vector falls in the interval

3. The method for detecting a target signal based on a time-frequency feature map as claimed in claim 1, wherein the extracting process of the time-frequency feature map matrix of the received signal specifically comprises:

let the discrete data sequence corresponding to the received signal be S_Re[n]At a sampling rate f_S；

According to discrete data sequence S corresponding to received signal_Re[n]Establishing a time-frequency diagram matrix A of the received signal_Re；

Wherein A is_ReCorresponds to the frequency domain information of the received signal, the frequency band range of which is

From the time-frequency diagram matrix A of the received signal_ReExtracting the time-frequency characteristic diagram matrix R of the received signal_Re；

Wherein R is_ReInitialisation to a full 0 matrix, R_ReCorresponds to the frequency domain information of the received signal, the frequency band range of which is

Extraction of A_ReThe upper envelope curve of the jth column of vectors is detected, the amplitude value and the corresponding position of each maximum value point in the upper envelope curve of the jth column of vectors are detected, the first 2 xM is selected according to the sequence of the amplitude values of the maximum value points from large to small_CopyA larger maximum point as the effective maximum point of the jth column vector, wherein

The frequency value corresponding to the effective maximum value point of the jth column vector falls in the intervali∈[1,N_Feature]Then R is_ReJth column of (1).

4. The method for detecting a target signal based on a time-frequency feature map as claimed in claim 1, wherein the similarity between the extracted time-frequency feature map of the transmitted signal and the extracted time-frequency feature map of the received signal is calculated to obtain a similarity curve vector; the method specifically comprises the following steps:

time-frequency characteristic diagram matrix R of received signal_ReIn the method, the number of sampling points corresponding to the time domain interval corresponding to each column isTime-frequency characteristic diagram matrix R of transmitting signal_CopyIn (1), the number of sampling points corresponding to the time domain interval corresponding to each column is f_s×T₀；

Taking the kth column of the time-frequency characteristic diagram matrix of the received signal as the starting time, and carrying out interval f on subsequent column vectors after the kth column of the time-frequency characteristic diagram matrix of the received signal_s×T₀Down-sampling to obtain a matrix with the same column number as the time-frequency characteristic diagram matrix of the transmitted signalComputing matricesTime-frequency characteristic diagram matrix R of transmitting signal_CopyDegree of similarity d_k：

Wherein the functionRepresenting first-pair matrix

from the calculated matrixTime-frequency characteristic diagram moment of transmitted signalArray R_CopyDegree of similarity d_kCalculating the time-frequency characteristic diagram matrix R of the received signal_ReThe time-frequency characteristic diagram matrix R of each column and the transmitting signal_CopyAnd then a similarity curve vector D ═ D is obtained₁d₂… d_k]。

5. The method for detecting a target signal based on a time-frequency characteristic diagram according to claim 4, wherein the target signal is detected according to the obtained similarity curve vector and a preset judgment standard; the method specifically comprises the following steps:

extracting an upper envelope curve of the similarity curve vector D according to the obtained similarity curve vector, and detecting the amplitude value and the corresponding position of each maximum value point in the upper envelope curve;

judging whether the amplitude of each maximum value point in the upper envelope curve is larger than a preset detection threshold value psi,

if the amplitude of a certain maximum value point in the upper envelope curve is greater than a preset detection threshold psi, the maximum value point is judged as a target signal, and arrival time information of the target signal is obtained;

and if the amplitude of a certain maximum value point in the upper envelope curve is less than or equal to a preset detection threshold psi, the maximum value point is judged as an interference signal.

6. A target signal detection system based on a time-frequency characteristic diagram is characterized by comprising:

the characteristic extraction module is used for respectively extracting a time-frequency characteristic diagram matrix of the transmitting signal and a time-frequency characteristic diagram matrix of the receiving signal;

the similarity curve vector acquisition module is used for calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal so as to obtain a similarity curve vector; and

and the detection module is used for detecting the target signal according to the obtained similarity curve vector and a preset judgment standard.

Technical Field

The invention belongs to the technical field of underwater acoustic signal detection and processing, and particularly relates to a target signal detection method and a target signal detection system based on a time-frequency characteristic diagram.

Background

The random frequency modulation signal has good autocorrelation characteristic and cross-correlation characteristic, and is often applied to the fields of underwater acoustic communication, navigation positioning, active sonar detection and the like. In the underwater acoustic channel, the mutual motion between the signal transmitter and the signal receiver and the flow of the water body cause the doppler shift of the signal, and the doppler shift of the signal usually reduces the detection capability of the matched filter, and when the doppler shift of the signal exceeds a certain tolerance, the reliability of signal communication is seriously reduced, so that the signal cannot be normally communicated. Generally, if a doppler shift signal is to be subjected to wideband matched filtering, first all copy signals possibly corresponding to doppler shift are generated, then search matching is performed by using an exhaustive method or a newton method, a copy signal closest to the doppler shift signal is found, and an optimal matching result is obtained, which may be considered as mismatch correction.

At present, all the existing matched filtering methods perform constant doppler coefficient correction on all signals within a time length corresponding to a copy signal, that is, the doppler coefficient for correction is a constant. When the signal transmitter and the receiver make relative variable speed motion, the corresponding doppler coefficient is a change curve rather than a constant in the time length corresponding to the copy signal, which results in the matched filtering method based on doppler linear correction, and the signal detection performance is reduced or even fails.

Under the environment of white Gaussian noise, the longer the time length corresponding to the copy signal is, the larger the time gain obtained by the matched filtering method is, and the better the target signal detection is. However, since the marine environment has time, space, and frequency-dependent characteristics, when the time length corresponding to the copy signal is greater than the time-dependent radius corresponding to the marine environment, the matched filtering method cannot obtain the entire time gain of the copy signal, and the target signal detection performance is reduced as the time length corresponding to the copy signal increases.

In active sonar detection application, the relative motion of the three parts of 'transmitter-target-receiver' can also bring nonlinear Doppler distortion of the received signal compared with the transmitted signal; because the target echo signal is weaker, in order to achieve a better detection effect, a long pulse signal needs to be transmitted, a sufficiently large time gain is expected to be obtained, and the problem that the detection performance of the matched filtering method on the target signal is deteriorated when the time length corresponding to the copy signal is greater than the time correlation radius corresponding to the marine environment can also be faced; meanwhile, the problem that the matched filtering technology based on Doppler linear correction fails due to the fact that nonlinear Doppler distortion occurs when the copy signal is transmitted in an underwater acoustic channel cannot be solved, and the robustness of the nonlinear Doppler distortion is reduced.

Disclosure of Invention

In order to solve the defects of the existing random frequency modulation signal detection method, the invention provides a target signal detection method based on a time-frequency characteristic diagram, which comprises the following steps:

respectively extracting a time-frequency characteristic diagram matrix of a transmitting signal and a time-frequency characteristic diagram matrix of a receiving signal;

calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal, and further obtaining a similarity curve vector;

and detecting a target signal according to the obtained similarity curve vector and a preset judgment standard.

As an improvement of the above technical solution, the extraction process of the time-frequency characteristic diagram matrix of the transmission signal specifically includes:

random frequency modulation signal is used as transmitting signal, and its discrete data sequence is S_Copy[n]：

Wherein N is equal to [1, N ∈]N is the total sampling point number of the discrete data sequence corresponding to the transmitting signal; omega₀Is the center frequency; f. of_SIs the sampling rate; b is the frequency modulation signal bandwidth; k (x) e [ -1,1]Is a 0-mean random sequence, and x is the serial number of the random sequence; x is the length of a random sequence selected for generating random frequency modulation signals; t is₀The frequency modulation time occupied by each code element in the random sequence; r 2]Is a ramp function, and is a step function u [ [ alpha ] ]]The integration result of (1);

from a discrete data sequence S of the transmitted signal_Copy[n]Establishing a time-frequency diagram matrix A of the transmitted signal_Copy；

Wherein A is_CopyCorresponds to the frequency domain information of the transmitted signal, the frequency band range of which isWith a frequency point interval of

Wherein N is_FFTSampling points selected for short-time Fourier transform; a. the_CopyEach column of (a) corresponds to the time domain information of the transmitted signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is f_s×T₀，A_CopyTotal number of columns of

Time-frequency diagram matrix A according to transmission signals_CopyExtracting time-frequency characteristic diagram matrix R of transmitting signal_Copy；

Wherein R is_CopyInitialisation to a full 0 matrix, R_CopyCorresponds to the frequency domain information of the transmitted signal, the frequency band range of which isWith a frequency point interval ofWherein N is_FeatureThe number of the target frequency band B is divided at equal intervals; r_CopyEach column of (a) corresponds to the time domain information of the transmitted signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is f_s×T₀，R_CopyTotal number of columns of

Extraction of A_CopyThe upper envelope curve of the jth column vector of (a),and detecting the amplitude value and the corresponding position of each maximum value point in the upper envelope curve of the jth column vector, sorting the amplitude values of the maximum value points from large to small, and selecting the first M_CopyA larger maximum point as the effective maximum point of the j-th column vector, wherein

The frequency value corresponding to the effective maximum value point of the jth column vector falls in the interval Then R is_CopyJth column of (1).

As an improvement of the above technical solution, the process of extracting the time-frequency characteristic diagram matrix of the received signal specifically includes:

let the discrete data sequence corresponding to the received signal be S_Re[n]At a sampling rate f_S；

According to discrete data sequence S corresponding to received signal_Re[n]Establishing a time-frequency diagram matrix A of the received signal_Re；

Wherein A is_ReCorresponds to the frequency domain information of the received signal, the frequency band range of which is

With a frequency point interval ofWherein N is_FFTSampling points selected for short-time Fourier transform; a. the_ReEach column of (a) corresponds to the time domain information of the received signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is

According to the receiving messageTime-frequency diagram matrix A of numbers_ReExtracting the time-frequency characteristic diagram matrix R of the received signal_Re；

Wherein R is_ReInitialisation to a full 0 matrix, R_ReCorresponds to the frequency domain information of the received signal, the frequency band range of which is

With a frequency point interval ofWherein N is_FeatureThe number of the target frequency band B is divided at equal intervals; r_ReEach column of (a) corresponds to the time domain information of the received signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is

Extraction of A_ReThe upper envelope curve of the jth column of vectors is detected, the amplitude value and the corresponding position of each maximum value point in the upper envelope curve of the jth column of vectors are detected, the first 2 xM is selected according to the sequence of the amplitude values of the maximum value points from large to small_CopyA larger maximum point as the effective maximum point of the jth column vector, wherein

The frequency value corresponding to the effective maximum value point of the jth column vector falls in the interval

Then R is_ReJth column of (1).

As one improvement of the above technical solution, the similarity between the time-frequency characteristic graph of the transmission signal and the time-frequency characteristic graph of the reception signal is calculated according to the extracted time-frequency characteristic graph of the transmission signal and the time-frequency characteristic graph of the reception signal, so as to obtain a similarity curve vector; the method specifically comprises the following steps:

time-frequency characteristic diagram matrix R of received signal_ReIn the method, the number of sampling points corresponding to the time domain interval corresponding to each column isTime-frequency characteristic diagram matrix R of transmitting signal_CopyIn (1), the number of sampling points corresponding to the time domain interval corresponding to each column is f_s×T₀；

Taking the kth column of the time-frequency characteristic diagram matrix of the received signal as the starting time, and carrying out interval f on subsequent column vectors after the kth column of the time-frequency characteristic diagram matrix of the received signal_s×T₀Down-sampling to obtain a matrix with the same column number as the time-frequency characteristic diagram matrix of the transmitted signalComputing matricesTime-frequency characteristic diagram matrix R of transmitting signal_CopyDegree of similarity d_k：

Wherein the functionRepresenting first-pair matrixAnd R_CopyAnd operation is carried out element by element to obtain a matrixReacquiring matrix

The number of medium element values is 1; function(s)

Representing an acquisition matrix R_CopyThe number of medium element values is 1;

from the calculated matrix

Time-frequency characteristic diagram matrix R of transmitting signal_CopyDegree of similarity d_kCalculating the time-frequency characteristic diagram matrix R of the received signal_ReThe time-frequency characteristic diagram matrix R of each column and the transmitting signal_CopyAnd then a similarity curve vector D ═ D is obtained₁d₂…d_k]。

As one improvement of the above technical solution, the target signal is detected according to the obtained similarity curve vector and a preset judgment standard; the method specifically comprises the following steps:

extracting an upper envelope curve of the similarity curve vector D according to the obtained similarity curve vector, and detecting the amplitude value and the corresponding position of each maximum value point in the upper envelope curve;

judging whether the amplitude of each maximum value point in the upper envelope curve is larger than a preset detection threshold value psi,

if the amplitude of a certain maximum value point in the upper envelope curve is greater than a preset detection threshold psi, the maximum value point is judged as a target signal, and arrival time information of the target signal is obtained;

and if the amplitude of a certain maximum value point in the upper envelope curve is less than or equal to a preset detection threshold psi, the maximum value point is judged as an interference signal.

The invention also provides a target signal detection system based on the time-frequency characteristic diagram, which comprises the following components:

the characteristic extraction module is used for respectively extracting a time-frequency characteristic diagram matrix of the transmitting signal and a time-frequency characteristic diagram matrix of the receiving signal;

the similarity curve vector acquisition module is used for calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal so as to obtain a similarity curve vector; and

and the detection module is used for detecting the target signal according to the obtained similarity curve vector and a preset judgment standard.

Compared with the prior art, the invention has the beneficial effects that:

the method of the invention has stronger capability of resisting nonlinear Doppler distortion. When the pulse width of the transmitted random frequency modulation signal is larger than the corresponding time correlation radius of the marine environment or the transmitted signal in the received signal has large-scale nonlinear Doppler frequency shift distortion, the existing matched filtering signal detection method is not applicable any more, and the method can still obtain higher signal time gain when being used for detecting the transmitted signal, thereby obviously improving the detection capability of the random frequency modulation signal.

Drawings

FIG. 1 is a flow chart of a method for detecting a target signal based on a time-frequency feature map according to the present invention;

FIG. 2 is a schematic diagram of a time-frequency diagram matrix of a transmitted signal according to an embodiment of a method for detecting a target signal based on a time-frequency feature diagram;

FIG. 3 is a schematic diagram of a time-frequency characteristic diagram matrix of a transmitted signal according to an embodiment of a method for detecting a target signal based on a time-frequency characteristic diagram according to the present invention;

FIG. 4 is a schematic diagram of a time-frequency diagram matrix of a received signal according to an embodiment of a method for detecting a target signal based on a time-frequency feature diagram;

FIG. 5 is a schematic diagram of a time-frequency feature map matrix of a received signal according to an embodiment of a method for detecting a target signal based on a time-frequency feature map according to the present invention;

FIG. 6 is a schematic diagram of a similarity curve and an upper envelope curve of the similarity curve according to an embodiment of a method for detecting a target signal based on a time-frequency feature diagram;

fig. 7 is a schematic diagram of pulse pressure curves of a transmitted signal and a received signal according to an embodiment of a target signal detection method based on a time-frequency characteristic diagram.

Detailed Description

The invention will now be further described with reference to the accompanying drawings.

The invention provides a target signal detection method based on a time-frequency characteristic diagram, which has stronger robustness against nonlinear Doppler distortion; when the pulse width of the transmitted random frequency modulation signal is larger than the corresponding time correlation radius of the marine environment, the existing matched filtering signal detection method is not applicable any more, the transmitted signal can still be detected by adopting the method of the invention, and the detection capability of the random frequency modulation signal can be obviously improved.

As shown in fig. 1, the method includes:

respectively extracting a time-frequency characteristic diagram matrix of a transmitting signal and a time-frequency characteristic diagram matrix of a receiving signal;

specifically, the specific process of extracting the video feature map matrix of the transmitted signal is as follows:

random frequency modulation signal is used as transmitting signal, and its discrete data sequence is S_Copy[n]：

Wherein N is equal to [1, N ∈]N is the total sampling point number of the discrete data sequence corresponding to the transmitting signal; omega₀Is the center frequency; f. of_SIs the sampling rate; b is the frequency modulation signal bandwidth; k (x) e [ -1,1]Is a 0-mean random sequence, and x is the serial number of the random sequence; x is the length of a random sequence selected for generating random frequency modulation signals; t is₀The frequency modulation time occupied by each code element in the random sequence; r 2]Is a ramp function, and is a step function u [ [ alpha ] ]]The integration result of (1);

from a discrete data sequence S of the transmitted signal_Copy[n]Establishing a time-frequency diagram matrix A of the transmitted signal_Copy；

Wherein A is_CopyCorresponds to the frequency domain information of the transmitted signal, the frequency band range of which isWith a frequency point interval of

Wherein N is_FFTSampling points selected for short-time Fourier transform; a. the_CopyEach column of (a) corresponds to the time domain information of the transmitted signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is f_s×T₀，A_CopyTotal number of columns of

Time-frequency diagram matrix A according to transmission signals_CopyExtracting time-frequency characteristic diagram matrix R of transmitting signal_Copy；

Wherein R is_CopyInitialisation to a full 0 matrix, R_CopyCorresponds to the frequency domain information of the transmitted signal, the frequency band range of which isWith a frequency point interval of

Wherein N is_FeatureThe number of the target frequency band B is divided at equal intervals; r_CopyEach column of (a) corresponds to the time domain information of the transmitted signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is f_s×T₀，R_CopyTotal number of columns of

Extraction of A_CopyThe upper envelope curve of the jth column vector of (a),

and detecting the amplitude value and the corresponding position of each maximum value point in the upper envelope curve of the jth column vector, sorting the amplitude values of the maximum value points from large to small, and selecting the first M_CopyA larger maximum point as the effective maximum point of the jth column vector

The frequency value corresponding to the effective maximum value point of the jth column vector falls in the interval Then R is_CopyJth column of (1).

The specific process of extracting the video characteristic diagram matrix of the received signal is as follows:

let the discrete data sequence corresponding to the received signal be S_Re[n]At a sampling rate f_S；

According to discrete data sequence S corresponding to received signal_Re[n]Establishing a time-frequency diagram matrix A of the received signal_Re；

Wherein A is_ReCorresponds to the frequency domain information of the received signal, the frequency band range of which is

With a frequency point interval ofWherein N is_FFTSampling points selected for short-time Fourier transform; a. the_ReEach column of (a) corresponds to the time domain information of the received signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is

From the time-frequency diagram matrix A of the received signal_ReExtracting the time-frequency characteristic diagram matrix R of the received signal_Re；

Wherein R is_ReInitialisation to a full 0 matrix, R_ReCorresponds to the frequency domain information of the received signal, the frequency band range of which isWith a frequency point interval of

Wherein N is_FeatureThe number of the target frequency band B is divided at equal intervals; r_ReEach column of (a) corresponds to the time domain information of the received signal, the time domain interval of which depends on the FM time T occupied by each symbol₀The number of sampling points corresponding to the time domain interval is

Extraction of A_ReThe upper envelope curve of the jth column of vectors is detected, the amplitude value and the corresponding position of each maximum value point in the upper envelope curve of the jth column of vectors are detected, the first 2 xM is selected according to the sequence of the amplitude values of the maximum value points from large to small_CopyA larger maximum point as the effective maximum point of the jth column vector

The frequency value corresponding to the effective maximum value point of the jth column vector falls in the interval Then R is_ReJth column of (1).

Calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal, and further obtaining a similarity curve vector;

specifically, the number of rows of the time-frequency characteristic diagram matrix of the received signal is the same as that of the time-frequency characteristic diagram matrix of the transmitted signal, and the subband intervals corresponding to the same rows are also the same, that is, each row of the time-frequency characteristic diagram matrix of the received signal and each row of the time-frequency characteristic diagram matrix of the transmitted signal have the same dimension, which is beneficial to performing similarity discrimination on the time-frequency characteristics of the received signal and the transmitted signal;

time-frequency characteristic diagram matrix R of received signal_ReIn the method, the number of sampling points corresponding to the time domain interval corresponding to each column isTime-frequency characteristic diagram matrix R of transmitting signal_CopyIn (1), the number of sampling points corresponding to the time domain interval corresponding to each column is f_s×T₀；

Taking the kth column of the time-frequency characteristic diagram matrix of the received signal as the starting time, and carrying out the following column vectors after the kth +1 column of the time-frequency characteristic diagram matrix of the received signal according to the interval f_s×T₀Down-sampling to obtain a matrix with the same column number as the time-frequency characteristic diagram matrix of the transmitted signalComputing matrices

Time-frequency characteristic diagram matrix R of transmitting signal_CopyDegree of similarity d_k：

Wherein the functionRepresenting first-pair matrixAnd R_CopyAnd operation is carried out element by element to obtain a matrixReacquiring matrixThe number of medium element values is 1; function(s)

Representing an acquisition matrix R_CopyThe number of medium element values is 1;

from the calculated matrixTime-frequency characteristic diagram matrix R of transmitting signal_CopyDegree of similarity d_kCalculating the time-frequency characteristic diagram matrix R of the received signal_ReThe time-frequency characteristic diagram matrix R of each column and the transmitting signal_CopyAnd then a similarity curve vector D ═ D is obtained₁d₂…d_k]。

And detecting a target signal according to the obtained similarity curve vector and a preset judgment standard.

Specifically, according to the obtained similarity curve vector, extracting an upper envelope curve of the similarity curve vector D, and detecting the amplitude and the corresponding position of each maximum value point in the upper envelope curve;

judging whether the amplitude of each maximum value point in the upper envelope curve is larger than a preset detection threshold value psi,

if the amplitude of a certain maximum value point in the upper envelope curve is greater than a preset detection threshold psi, the maximum value point is judged as a target signal, and arrival time information of the target signal is obtained;

and if the amplitude of a certain maximum value point in the upper envelope curve is less than or equal to a preset detection threshold psi, the maximum value point is judged as an interference signal.

The invention also provides a target signal detection system based on the time-frequency characteristic diagram, which comprises the following components:

the characteristic extraction module is used for respectively extracting a time-frequency characteristic diagram matrix of the transmitting signal and a time-frequency characteristic diagram matrix of the receiving signal;

the similarity curve vector acquisition module is used for calculating the similarity of the time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal according to the extracted time-frequency characteristic diagram matrix of the transmitting signal and the time-frequency characteristic diagram matrix of the receiving signal so as to obtain a similarity curve vector; and

and the detection module is used for detecting the target signal according to the obtained similarity curve vector and a preset judgment standard.