A micro-moving target detection method based on OFDM sounding integrated signal

Abstract

The present invention belongs to the technical field of integrated detection and communication, and specifically relates to a micro-motion target detection method based on OFDM sounding integrated signal. The present invention extracts the peak position sequence of broadband HRRP of micro-motion target, and roughly estimates the micro-motion information as prior information; then performs time-frequency analysis on narrowband echo to obtain the micro-Doppler information it carries, and performs extreme point screening on the time-frequency distribution result, and uses the dynamic programming Viterbi algorithm to search for the shortest extreme path at the cost of gradient change, and obtains the secondary estimation result after curve fitting, and performs joint feature estimation and extraction on the target micro-motion information, and more comprehensively utilizes the target micro-motion feature information carried in the echo, so as to obtain higher-precision target micro-motion parameters, and can realize parameter extraction and estimation of multiple micro-motion targets, while ensuring accuracy with smaller calculation amount.

Description

Technical Field

[0001] The present invention belongs to the technical field of detection and communication integration, and in particular relates to a micro-motion target detection method based on OFDM sounding integrated signal. Background Art

[0002] In a complex environment with multiple moving targets, the echo signal is composed of the Doppler information generated by multiple different movements. It is impossible to extract the information of small-amplitude moving targets from multiple moving targets, such as gesture targets, wing targets, and changes in warhead targets, etc., only by using the range-Doppler information. Analyzing the micro-Doppler of the target and extracting its micro-motion characteristics from the detection echo is an important way to identify the target. For a narrowband detection system, the micro-motion characteristics of the target are mainly manifested in its broadened Doppler spectrum and instantaneous micro-Doppler; for a broadband detection system, if the micro-motion amplitude of the target is smaller than the range resolution unit, the micro-motion will cause the nonlinearity of the echo phase process. If the micro-motion amplitude of the target is larger than the range resolution unit, the micro-motion will cause the target to move between range units.

[0003] Common integrated signals mainly include LFM signals and frequency step (SF) signals. Generally speaking, LFM signals have serious time-frequency coupling, which not only has an adverse effect on the analysis and extraction of micro-Doppler, but also causes range Doppler ambiguity. SF signals use the method of synthesizing bandwidth with pulse trains, which will cause the echo signal under target motion to shift the range image or even cause waveform distortion, and its micro-motion modulation effect is more complicated. Orthogonal frequency division multiplexing (OFDM) signals are large-bandwidth signals synthesized by multiple sub-band signals of different frequencies, and have a thumbtack-shaped ambiguity function. There is no time-frequency coupling problem, so they not only have broadband high-resolution capabilities, but also have narrowband micro-Doppler characteristics, which are more conducive to the detection and extraction of micro-motion targets. Therefore, finding a micro-motion parameter estimation method that can simultaneously utilize the broadband range high-resolution characteristics and narrowband Doppler characteristics of OFDM integrated signals is of great significance to improving the detection accuracy of micro-motion targets. Summary of the invention

[0004] The purpose of the present invention is to provide a joint characteristic parameter estimation method for micro-motion targets based on OFDM signals in view of the above-mentioned problems and deficiencies. In the process of using OFDM signals to detect target micro-motion, the method roughly estimates the micro-motion information by using the broadband HRRP of the micro-motion target as prior information, and then uses the narrowband micro-Doppler information to perform joint characteristic estimation; for the problem of extracting the target micro-motion curve parameters, the present invention designs a characteristic curve estimation method based on extreme value path, which uses the Viterbi algorithm to search for the shortest extreme value path at the cost of the gradient change, and then respectively performs curve fitting and parameter extraction on the extreme value paths of different targets, and finally selects the one with a higher degree of fitting in the two estimates as the target micro-motion parameter. In this way, not only can the target micro-motion information be more finely portrayed, but also the amount of calculation can be reduced at the same time, which is an efficient method for extracting and estimating the parameters of OFDM signal micro-motion targets.

[0005] The technical solution of the present invention is: a method for estimating joint characteristic parameters of a micro-motion target based on OFDM signals, comprising the following steps:

[0006] Step 1: Preprocess the original signal s(t) at the receiving end by mixing it with the transmitting carrier frequency exp{j2πf0t} to obtain the integrated signal baseband echo r(t) after down-conversion processing, and then use the polyphase filter group h n (m) Divide the signal into sub-band echo signals equal to the number of transmitted carriers N and demodulate them to obtain N sub-band echo signals r m,n (t).

[0007] Step 2: For N sub-band echo signals r m,n (t) Perform inverse fast Fourier transform processing within the pulse to synthesize a large bandwidth to obtain the HRRP of the echo signal; determine the migration of the target between distance units: if the HRRP peak position sequence is approximately a straight line, extract the distance unit Y(m) where the peak is located for time-frequency analysis to obtain the micro-Doppler curve of the target; if the HRRP peak position changes significantly over time under the influence of micro-motion, extract the target motion trajectory curve k(m) after eliminating the translation, fit the obtained curve, and obtain a preliminary estimate of the target micro-motion parameter γ1.

[0008] Step 3: For any sub-band signal r m,n (t) Perform time-frequency analysis to obtain the time-frequency distribution matrix F(n,k), perform extreme value detection on the matrix, and obtain the extreme value point matrix F ext (n,k), passing the noise threshold η g Screen the extreme points and set the values ​​below the noise threshold η g The extreme points of the effective signal are filtered out, and the set of extreme points is F' ext(n,k), using the estimated value γ1 obtained by the wideband high-resolution characteristics of the signal as prior information, the dynamic programming algorithm is used to find the shortest path sequence l(n) within the time 1≤n≤N at the cost of gradient change, and the micro-motion characteristic curve is fitted to the result to obtain the micro-motion estimate γ2 based on the narrowband micro-Doppler characteristics.

[0009] Step 4: Compare the two estimation results γ1 and γ2 obtained through different feature information, and select the estimation results with higher fitting degree of different target curves to describe the micro-motions of different targets respectively.

[0010] The present invention extracts the peak position sequence of broadband HRRP of micro-motion target to roughly estimate the micro-motion information as prior information; then performs time-frequency analysis on narrowband echo to obtain the micro-Doppler information it carries, screens the extreme points of the time-frequency distribution results, uses the dynamic programming Viterbi algorithm to search for the shortest extreme path at the cost of gradient change, obtains the secondary estimation result after curve fitting, and performs joint feature estimation and extraction on the target micro-motion information to ensure the estimation accuracy of target micro-motion parameters and reduce the amount of calculation.

[0011] The beneficial effects of the present invention are that the target micro-motion feature information carried in the echo is more comprehensively utilized to obtain target micro-motion parameters with higher precision; when the micro-motion feature of one aspect fails, the micro-motion feature of another aspect can still be used to extract the target micro-motion feature; and by using a dynamic programming algorithm to perform extreme value path search, parameter extraction and estimation of multiple micro-motion targets can be achieved, and the accuracy can be guaranteed while having a lower amount of calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a flow chart of the present invention;

[0013] Figure 2 It is the flow chart of the characteristic curve estimation algorithm based on extreme value path;

[0014] Figure 3 is the graph of target HRRP changing over time;

[0015] Figure 4 It is the result of different target motion trajectories and their curve fitting;

[0016] Figure 5 This is the result diagram of narrowband echo time-frequency analysis;

[0017] Figure 6 Yes Figure 5 Result graph of extreme value detection;

[0018] Figure 7 yes Figure 6 Passing noise threshold η g Result graph of extreme point screening;

[0019] Figure 8 This is the result diagram of extreme value path selection and curve fitting for different micro-motion targets; DETAILED DESCRIPTION

[0020] The present invention is described in detail below with reference to the accompanying drawings and simulation examples to demonstrate the practicability of the present invention.

[0021] As attached Figure 1 As shown, through the OFDM signal-based micro-motion target joint feature parameter estimation method of the present invention, the broadband HRRP information and narrowband micro-Doppler information of the micro-motion target can be used simultaneously to perform joint feature estimation on the micro-motion target, and the target micro-motion information can be more finely characterized. The specific implementation steps are as follows:

[0022] Step 1: Preprocess the original signal s(t) at the receiving end, that is, mix it with the transmission carrier frequency exp{j2πf0t} to obtain the integrated signal baseband echo r(t) after down-conversion processing, where f0 is the transmission carrier frequency; then use the polyphase filter group h n (m) Divide the signal into sub-band echo signals equal to the number of transmitted carriers N and demodulate them to obtain N sub-band echo signals r m,n (t), its expression is:

[0023]

[0024] Among them, r m,n (t) is the signal of the nth subcarrier in the mth pulse after passing through the polyphase filter; ρ represents the backscatter coefficient of the target scattering center; A m,n is the amplitude attenuation and phase shift caused by the echo signal gain of the nth subcarrier in the mth pulse; △f is the interval between adjacent carriers; f n =f0+n△f is the carrier frequency of the nth subcarrier; the target delay is t=2R d (t) / c; where the radial relative distance between the scattering center of the target and the integrated system is:

[0025]

[0026] Where R0 is the radial distance between the target micromotion center and the integrated system at the initial moment; vt is the translation generated by the radial relative velocity v between the target and the integrated system, D is the radial amplitude of the target micromotion, the micromotion angular frequency is ω, and the initial phase is

[0027] Step 2: For N sub-band echo signals r m,n (t) Perform inverse fast Fourier transform processing within the pulse to synthesize a large bandwidth to obtain the HRRP of the mth pulse echo:

[0028]

[0029] Where c is the propagation speed of electromagnetic waves, T r is the pulse repetition period, R m is the radial relative distance between the target and the system within the mth pulse,

[0030] Determine the migration of the target between range units: If the HRRP peak position sequence is approximately a straight line, align it for translational velocity compensation, extract the range unit Y(m) where the peak is located for time-frequency analysis, and obtain the micro-Doppler curve of the target, where Y(m) is:

[0031]

[0032] If the HRRP peak position changes significantly over time under the influence of micromotion, the target motion trajectory curve k(m) after eliminating translation is extracted as:

[0033]

[0034] Among them, △R is the distance resolution of the system. The obtained target broadband micro-motion characteristic curve is fitted to obtain the preliminary estimate of the target micro-motion parameter γ1.

[0035] Step 3: For any subband signal r m,n (t) Perform time-frequency analysis to obtain the time-frequency distribution result as a matrix F(n,k). Perform extreme value detection on the matrix to obtain the extreme value point matrix F ext (n,k), that is:

[0036]

[0037] In the formula, arg{·} is the quantity operation; peak{·} is the extreme point calculation function, n is the time index, and k is the frequency index. Suppose the echo signal noise ratio is SNR, then the noise threshold η of the signal can be obtained under the condition of this signal noise ratio. g for:

[0038]

[0039] Passing noise threshold η g Screen the extreme points and set the values ​​below the noise threshold η g The extreme points of the signal are filtered out to obtain the set of effective signal extreme points F' ext (n,k) is:

[0040] F' ext (n,k)=arg{F ext (n,k)>ηg}

[0041] The estimated value γ1 obtained by using the high-resolution feature of the signal broadband is used as prior information, and the shortest path sequence l(n) within the time 1≤n≤N is found through a dynamic programming algorithm at the cost of gradient change:

[0042] l(n)=V iterbi {F' ext (n,k)}

[0043] Where V iterbi {·} represents the Viterbi calculation. According to the period of the path gradient change, l(n) is further curve-fitted to obtain a time-frequency curve of the micro-motion component. Clear the extreme points near the fitting curve, that is, n k ∈[n k+i ,n k-i ], and a new extreme point matrix is ​​obtained; when the number of extracted extreme value paths is consistent with the number of micro-motion targets, the micro-motion characteristic curve extraction ends; otherwise, continue to find the next shortest path. The flowchart of the characteristic curve estimation method based on extreme value paths is shown in the attached figure. Figure 2 The micro-motion characteristic curve is fitted to the extreme path result, and the micro-motion estimation value γ2 based on the narrow-band micro-Doppler characteristics is obtained.

[0044] Step 4: Compare the two estimation results γ1 and γ2 obtained through different feature information, and select the estimation results with higher fitting degree of different target curves to describe the micro-motions of different targets respectively.

[0045] Simulation Example

[0046] According to the above method, Figure 1 The simulation parameters are set as follows: the center carrier frequency of the transmitted signal is 220 GHz, the bandwidth is 1.6 GHz, the number of subcarriers is 128, and the target micro-motion model is R i (t) = R 0i +D i sin(ω i t+φ i ), the initial positions of the two point targets were 5m and 5.5m, the translational velocities were 3m / s and 0m / s, the micro-motion amplitudes were 0.1m and 0.4m, the micro-motion angular frequencies were 14rad / s and 7rad / s, and the initial phases were 0° and 60°;

[0047] Figure 2 The characteristic curve estimation method based on the extreme value path described in step 3 is given;

[0048] Figure 3The change of HRRP of the target over time is given. It can be seen that the smaller the amplitude of the target micromotion, the more the motion trajectory is close to a straight line; the more the number of distance units the target micromotion crosses, the more obvious the motion trajectory changes;

[0049] Figure 4 The fitting results of broadband characteristic curve of micro-motion target after eliminating translational velocity are given. It can be found that the target with larger micro-motion amplitude has smoother motion trajectory and higher degree of curve fitting.

[0050] Figure 5 The time-frequency analysis results of narrowband echo are given, and it can be found that the echo contains two micro-motion targets and there is the influence of cross terms;

[0051] Figure 6 Given the Figure 5 The result diagram of extreme value detection of time-frequency distribution shown in the figure;

[0052] Figure 7 Given Figure 6 Filter out noise below the noise threshold η g The result diagram after the extreme point shows two characteristic curves of micro-motion targets;

[0053] Figure 8 The extreme path selection results and curve fitting results of different micro-motion targets are given, and the micro-motion estimation value γ2 based on narrow-band micro-Doppler characteristics can be obtained. The two estimation results γ1 and γ2 obtained with different feature information are compared, and the estimation results with higher curve fitting degree are selected to describe the micro-motion of different targets respectively.

[0054] In summary, judging from the processing results, the method provided by the present invention has practical value.

Claims

1. A method for detecting micro-motion targets based on OFDM sounding integrated signals, characterized in that: The following steps are involved: S1. Preprocess the original signal s(t) at the receiving end, that is, mix it with the transmission carrier frequency exp{j2πf0t} to obtain the integrated signal baseband echo r(t) after down-conversion processing, where f0 is the transmission carrier frequency; then use the polyphase filter group h n (m) Divide the signal into sub-band echo signals equal to the number of transmitted carriers N and demodulate them to obtain N sub-band echo signals r m,n (t), its expression is: Where ρ represents the backscattering coefficient of the target scattering center; A m,n is the amplitude attenuation and phase shift caused by the echo signal gain of the nth subcarrier in the mth pulse; △f is the interval between adjacent carriers; f n =f0+n△f is the carrier frequency of the nth subcarrier; the target delay is τ=2R d (t) / c; where the radial relative distance between the scattering center of the target and the integrated system is: Where R0 is the radial distance between the target micromotion center and the integrated system at the initial moment; vt is the translation generated by the radial relative velocity v between the target and the integrated system, D is the radial amplitude of the target micromotion, the micromotion angular frequency is ω, and the initial phase is S2, for N sub-band echo signals r m,n (t) Perform inverse fast Fourier transform processing within the pulse to synthesize a broadband echo signal, and obtain the HRRP of the mth pulse echo as: Where c is the propagation speed of electromagnetic waves, T r is the pulse repetition period, R m is the radial relative distance between the target and the system within the mth pulse, Determine the migration of the target between range units: If the HRRP peak position sequence is approximately a straight line, align it for translational velocity compensation, extract the range unit Y(m) where the peak is located for time-frequency analysis, and obtain the micro-Doppler curve of the target, where Y(m) is: If the HRRP peak position changes significantly over time under the influence of micromotion, the target motion trajectory curve k(m) after eliminating translation is extracted as: Among them, △R is the distance resolution of the system, and the obtained target broadband micro-motion characteristic curve is fitted to obtain the preliminary estimate of the target micro-motion parameter γ1; S3, for any sub-band signal r m,n (t) Perform time-frequency analysis to obtain the time-frequency distribution result as a matrix F(h,k). Perform extreme value detection on the matrix to obtain the extreme value point matrix F ext (h,k), that is: Where arg{·} is the quantity operation; peak{·} is the extreme point calculation function, h is the time index, and k is the frequency index; let the echo signal noise ratio be SNR, then the noise threshold η of the signal is obtained under the condition of this signal noise ratio g for: Passing noise threshold η g Screen the extreme points and set the values ​​below the noise threshold η g The extreme points of the signal are filtered out to obtain the set of effective signal extreme points F' ext (h,k) is: F' ext (h,k)=arg{F ext (h,k)>η g } The estimated value γ1 obtained by using the signal broadband high-resolution feature is used as prior information, and the shortest path sequence l(h) within the time 1≤h≤H is found through a dynamic programming algorithm at the cost of gradient change: l(h)=V iterbi {F' ext (h,k)} Where V iterbi {·} represents the Viterbi calculation. According to the period of path gradient change, l(h) is curve fitted to obtain a time-frequency curve of the micro-motion component. The extreme point h near the fitting curve k Clear, that is, h k ∈[h k+i ,h k-i ], and a new extreme point matrix is ​​obtained; when the number of the extracted extreme value paths is consistent with the number of micro-motion targets, the micro-motion characteristic curve extraction ends; otherwise, continue to find the next shortest path; fit the micro-motion characteristic curve of the extreme value path results to obtain the micro-motion estimation value γ2 based on the narrow-band micro-Doppler characteristics; S4. Compare the two estimation results γ1 and γ2 obtained through different feature information, and select the estimation results with higher fitting degree of different target curves to describe the micro-motions of different targets respectively.

Images