CN103278806B - Method for removing target detection Doppler dispersion of broadband signals based on sub-band processing - Google Patents

Abstract

The invention relates to a method for removing target detection Doppler dispersion of broadband signals based on sub-band processing, and belongs to the technical field of radar target acquisition. According to the method, a direct wave is decomposed into a plurality of sub-bands, phase-coherent accumulation is conducted on each sub-band, plural superposition is conducted on a phase-coherent accumulation result of each sub-band, and a final CAF output result is obtained. In order to make a target peak value accumulated by each sub-band be located on the same Doppler unit, different accumulation time needs to be selected according to a carrier frequency of each sub-band. Compared with a traditional Keystone transform method, a sub-band method is free of the hypothesis that the signal waveform is unchanged, and avoids the signal to noise ratio loss caused by the hypothesis that the signal waveform is unchanged. The method obviously removes the target Doppler dispersion through the sub-band method, effectively improves the phase-coherent accumulation grain, accordingly, improves detection performance of a system, and is especially suitable for being used on application occasions for detecting high-speed targets with long-time accumulation adopted.

Description

Based on the broadband signal target detection Doppler dispersion removing method of sub-band processing

Technical field

The present invention relates to a kind of broadband signal target detection Doppler dispersion removing method based on sub-band processing, belong to technical field of radar target acquisition.

Background technology

External illuminators-based radar utilizes the transmitting of all kinds of non-cooperation radiation source existed in space to carry out target detection, is a kind of special two multistatic radars.Because it has lower cost, good Electro Magnetic Compatibility and being paid attention to widely.Conventional non-cooperation radiation signal comprises fm broadcast signal, anolog TV signals, digital television signal etc.Wherein digital television signal is the desired illumination source of external illuminators-based radar.First, digital television signal just replaces anolog TV signals; Secondly, Digital Television adopts digital modulation technique, and emissive power is stablized; And the pseudo-random characteristics of digital modulation makes the ambiguity function of signal have lower secondary lobe; The broader bandwidth (7.56MHz) of digital television signal, makes external illuminators-based radar have higher range resolution simultaneously, is conducive to the target detecting low-latitude flying.

But during Digital Television external illuminators-based radar detection high-speed moving object, because digital television signal bandwidth ratio is wider, obvious Doppler dispersion effect can be there is, reduce correlative accumulation gain, and then influential system operating distance, integration time, the impact of longer Doppler dispersion was more serious.Keystone transform method hypothesis signal waveform conventional is at present constant, is applicable to pulsed radar range migration and compensates.For the digital television signal waveform of random variation, application Keystone method is eliminated dispersion and can be caused snr loss.

When digital television signal is used as target detection, the reason that there is Doppler dispersion effect is:

Radar relative bandwidth η is defined as signal bandwidth B and centre carrier frequency f _cratio, when η>=0.01, be wideband radar.Such as, digital television signal carrier frequency f _c=674MHz, bandwidth B=7.56MHz, relative bandwidth is

$\mathrm{\η}=\frac{B}{f_c}=\frac{7.56\mathrm{MHz}}{674\mathrm{MHz}}>0.01---\left(1\right)$

If v _rfor the radial motion speed of target, c is the light velocity, carrier frequency f _ccorresponding carrier wavelength is λ=c/f _c, then the Doppler frequency of target echo signal is

$f_d=\frac{{2v}_r}{\mathrm{\λ}}=\frac{{2v}_r{f}_c}{c}---\left(2\right)$

When the bandwidth B of signal is comparatively large, target radial speed v _rtime higher, the Doppler frequency f that different frequency composition is corresponding _dhave larger difference, this phenomenon is called Doppler dispersion.In order to describe the problem better, give one example: assuming that integration time T _int=0.8s, then frequency resolution is

$\mathrm{\Δf}=\frac{1}{T_{\mathrm{int}}}=1.25\mathrm{Hz}---\left(3\right)$

Listed by table 1 is the cell number of the Doppler caused due to Doppler dispersion diffusion corresponding to different target radial velocity.

The relation of table 1 target radial speed and the Doppler dispersion order of severity

Owing to there is Doppler dispersion effect, after Digital Television external illuminators-based radar utilizes cross ambiguity function (CAF) to complete correlative accumulation, target energy can be scattered in multiple doppler cells, thus causes the decline of accumulate augment.For this reason, need to take corresponding signal transacting measure, eliminate effect of dispersion to the impact of target detection.

Summary of the invention

The object of the invention is for effectively eliminating based on the Doppler dispersion in the external illuminators-based radar target detection of broadband signal, improve correlative accumulation gain and systems axiol-ogy performance, propose a kind of broadband signal target detection Doppler dispersion removing method based on sub-band processing, main subband cross ambiguity function (CAF) that adopts realizes Doppler dispersion elimination.

The object of the invention is to be achieved through the following technical solutions.

Based on a broadband signal target detection Doppler dispersion removing method for sub-band processing, specifically comprise the steps:

Step one, sub-band division

The broadband direct-path signal received by antenna, by sub-band division bank of filters, is decomposed into K subband:

r _k(n)=r(n)*h _k(n) (4)

Wherein, r _kthe signal of a n kth subband that () obtains through sub-band division for direct-path signal vector r (n); h _kn () is a kth Subband Analysis Filter;

The mathematical description of described sub-band division bank of filters is:

h _k(n)=h(n)e ^-j2πkn/K(5)

K=0 in formula, 1,2 ... K-1, K are number of sub-bands; H (n) is the prototype filter of Subband Analysis Filter.

Step 2, conjugate complex is taken advantage of

Direct-path signal in each subband that step one is obtained, the wideband echoes signal received with another antenna respectively does conjugate complex and takes advantage of, and a kth subband conjugate complex takes advantage of formula to be:

$x_k(\mathrm{\τ},n)=e(n+\mathrm{\τ})r_k^{*}\left(n\right)---\left(6\right)$

Wherein, e (n) is target echo signal, and τ is time delay.

Step 3, takes out under signal

For reducing correlative accumulation operand, each road signal obtained after being taken advantage of by step 2 conjugate complex carries out anti-aliasing low-pass filtering, M doubly lower extraction successively.X _kn taking out rear gained signal under () is:

$y_k(\mathrm{\τ},n)=\underset{l=-\∞}{\overset{\∞}{\mathrm{\Σ}}}g\left(l\right)x_k(\mathrm{\τ},\mathrm{Mn}-l)---\left(7\right)$

Wherein M takes out multiple under being; G (l) is ideal low-pass filter, and its DTFT form is

Step 4, the correlative accumulation time adjusts

Respectively the adjustment of correlative accumulation time is carried out to each way band signal after taking out under step 3, makes the corresponding different wave length λ of different sub-band k( _k) target peak that obtains is in identical doppler cells.Concrete grammar is:

Step 4.1, tries to achieve counting of the correlative accumulation of each way band signal

Doppler frequency is f _dtarget, residing doppler cells position is:

N _{Dop_bin}=f _d/Δf=f _dT _int(9)

Wherein, the frequency resolution Δ f=1/T of target detection _int, T _intfor the correlative accumulation time.

According to the Doppler frequency formula of target echo signal wherein v _rfor the radial motion speed of target, c is the light velocity, carrier frequency f _ccorresponding carrier wavelength is λ=c/f _c, obtain target place N further _{dop_bin}for:

N _{Dop_bin}=2v _rf _cT _int/c (10)

For ensureing doppler cells position, target place N in each subband _{dop_bin}identical, the correlative accumulation time T of each subband _intbe proportional to 1/f _c, therefore correlative accumulation count for:

Wherein f _sfor sampling rate, f _{c_k}the carrier frequency of a kth subband, f _centerit is the centre frequency of whole digital television signal carrier wave.

Step 4.2, according to counting of the correlative accumulation of each way band signal, intercepts each way band signal of corresponding length of counting, as the subband signal participating in subband cross ambiguity function and calculate.

Wherein, the initial interception position of every way band signal is identical.

Step 5, calculates each subband CAF

According to each subband correlative accumulation time that step 4 adjustment obtains, the correlative accumulation result adopting CAF to calculate each subband is:

χ _k(τ,f _d)=DFT{y _k(τ,n)} (12)

Step 6, subband correlative accumulation is comprehensive

The correlative accumulation resulting complex values of each subband gained is added, has been eliminated the object detection results of Doppler dispersion:

${\mathrm{\χ}}_{\mathrm{\Σ}}(\mathrm{\τ},f_d)=\underset{k=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{\χ}}_k(\mathrm{\τ},f_d)---\left(13\right)$

So far, broadband signal target detection is completed.

Beneficial effect

The present invention proposes a kind of Doppler dispersion removing method based on sub-band approach being applicable to the target detection of Digital Television external illuminators-based radar, and the method structure is simple, is easy to realize, successful.There is not the constant hypothesis of signal waveform in sub-band approach, avoids the therefore snr loss brought.

Accompanying drawing explanation

Fig. 1 is based on the Doppler dispersion elimination algorithm process flow diagram in the target detection of digital television signal;

Fig. 2 is under different correlative accumulation time conditions, and based on the CAF result of Keystone conversion, wherein, (a) ~ (g) is respectively the CAF result based on Keystone conversion of integration time 0.4s, 0.8s, 1.2s, 1.6s, 2.0s, 2.4s, 2.8s;

Fig. 3 is under different correlative accumulation time conditions, and based on the CAF result of sub-band processing, wherein, (a) ~ (g) is respectively the CAF result based on sub-band processing of integration time 0.4s, 0.8s, 1.2s, 1.6s, 2.0s, 2.4s, 2.8s;

Fig. 4 is under condition different integration time, based on the CAF signal to noise ratio (S/N ratio) comparison diagram of sub-band processing and Keystone conversion.

Embodiment

Below in conjunction with drawings and Examples, the present invention will be further described.

Embodiment

The present embodiment utilizes the means of experiment to verify validity of the present invention.

Experiment parameter: certain digital television broadcasting tower transmitted signal bandwidth is 7.56MHz, the centre frequency of signal is 674MHz, and radar signal processor baseband sampling rate is 9MHz.Test target is airliner, and flight radial velocity is about 99m/s.Integration time gets 0.4s, 0.8s, 1.2s, 1.6s, 2.0s, 2.4s, 2.8s respectively, contrasts the CAF result based on sub-band processing and Keystone conversion.

Method step as shown in Figure 1, is specially in the present embodiment:

1) direct-path signal is passed through analysis filterbank, be decomposed into K subband; In the present embodiment, K=32.For different broadband signals, the value of K is different.K is larger, and the effect that dispersion is eliminated is better, but operand is also larger.Therefore the value of K is compromise between dispersion eradicating efficacy and operand chooses.Traditionally, subband gets the integral number power of 2.For digital television signal, K gets 32 or 64 proper, and 32 is most suitable.

2) do conjugate complex respectively to the direct-path signal of each subband and echoed signal to take advantage of;

3) for reducing correlative accumulation operand, finish the lower extraction that conjugate complex Cheng Ge road signal does 1800 times respectively, after extracting, data transfer rate is 5kHz, and positive negative frequency range-2.5 ~ 2.5kHz, radial velocity coverage is v _r=f _dλ/2=± 2.5kHz × 0.4451m/2=556.375m/s.

4) the correlative accumulation time adjustment, correlative accumulation count for:

$N_{\mathrm{point}}=\frac{f_s{T}_{\mathrm{int}}{f}_{\mathrm{center}}}{{\mathrm{Mf}}_{c\_k}}=\frac{9\mathrm{MHz}\×674\mathrm{MHz}}{1800}(T_{\mathrm{int}}/f_{c\_k})=3.37\×{10}^{12}\×(T_{\mathrm{int}}/f_{c\_k})---\left(14\right)$

Wherein T _intget 0.4s, 0.8s, 1.2s, 1.6s, 2.0s, 2.4s, 2.8s respectively, f _{c_k}value is the centre frequency of 32 subbands be evenly distributed between 670.22MHz to 677.78MHz.

5) each subband data is DFT on different delay τ, calculates the CAF of 32 subbands.

6) the CAF result of 32 subbands plural number is added, namely obtains final CAF Output rusults.

Fig. 3 gives (integration time, scope was from 0.4s to 2.8s) under condition different integration time, the CAF result figure of sub-band approach.In order to contrast the effect after Doppler dispersion elimination, Fig. 2 is same measured data, under the same terms, adopts the result of Keystone conversion.

Under different integration time condition, based on the CAF of sub-band processing and Keystone conversion signal to noise ratio (S/N ratio) contrast as shown in Figure 4.Can find out, along with the increase of integration time, the signal to noise ratio (S/N ratio) advantage of sub-band processing CAF starts to manifest, for this embodiment, and the about high 1.5dB of signal to noise ratio (S/N ratio).If target velocity is higher and integration time is longer, then Doppler dispersion phenomenon is more serious, thus the advantage of sub-band processing CAF will be more obvious.

Can be found out by data analysis, subband correlative accumulation method of the present invention can eliminate the impact that Doppler dispersion is brought significantly, effectively improves the detection perform of system.The present invention is particularly suitable for the application scenario adopting long time integration detection high-speed target.

Above-described specific descriptions; the foregoing is only specific embodiments of the invention, the protection domain be not intended to limit the present invention, within the spirit and principles in the present invention all; any amendment of making, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (3)

1., based on the broadband signal target detection Doppler dispersion removing method of sub-band processing, it is characterized in that, specifically comprise the steps:

Step one, sub-band division

The broadband direct-path signal received by antenna, by sub-band division bank of filters, is decomposed into K subband:

r _k(n)＝r(n)*h _k(n) (4)

Wherein, r _kthe signal of a n kth subband that () obtains through sub-band division for direct-path signal vector r (n);

The mathematical description of described sub-band division bank of filters is:

h _k(n)＝h(n)e ^-j2πkn/K(5)

K=0 in formula, 1,2 ... K-1, K are number of sub-bands; h _kn () is a kth subband resolution filter; H (n) is the prototype filter of sub-band division wave filter;

Step 2, conjugate complex is taken advantage of

Direct-path signal in each subband that step one is obtained, the wideband echoes signal received with another antenna respectively does conjugate complex and takes advantage of, and a kth subband conjugate complex takes advantage of formula to be:

$x_k(\mathrm{\τ},n)=e(n+\mathrm{\τ})r_k^{*}\left(n\right)---\left(6\right)$

Wherein, e (n) is target echo signal, and τ is time delay;

Step 3, takes out under signal

The each road signal obtained after being taken advantage of by step 2 conjugate complex carries out anti-aliasing low-pass filtering, M doubly lower extraction successively; x _ktaking out rear gained signal under (τ, n) is:

$y_k(\mathrm{\τ},n)=\underset{l=-\∞}{\overset{\∞}{\mathrm{\Σ}}}g\left(l\right)x_k(\mathrm{\τ},\mathrm{Mn}-l)---\left(7\right)$

Wherein M takes out multiple under being; G (l) is ideal low-pass filter, and its DTFT form is

Step 4, the correlative accumulation time adjusts

Carry out the adjustment of correlative accumulation time respectively to each way band signal after taking out under step 3, the target peak that different sub-band k is obtained is in identical doppler cells; Concrete grammar is:

Step 4.1, tries to achieve counting of the correlative accumulation of each way band signal

Doppler frequency is f _dtarget, residing doppler cells position is:

N _{Dop_bin}＝f _d/Δf＝f _dT _int(9)

Wherein, the frequency resolution Δ f=1/T of target detection _int, T _intfor the correlative accumulation time;

According to the Doppler frequency formula of target echo signal wherein v _rfor the radial motion speed of target, c is the light velocity, f _cfor carrier frequency, λ is carrier wavelength, obtains target place N further _{dop_bin}for:

N _{Dop_bin}＝2v _rf _cT _int/c (10)

The correlative accumulation time T of each subband _intbe proportional to 1/f _c, correlative accumulation count for:

Wherein f _sfor sampling rate, f _{c_k}the carrier frequency of a kth subband, f _centerit is the centre frequency of whole digital television signal carrier wave;

Step 4.2, according to counting of the correlative accumulation of each way band signal, intercepts each way band signal of corresponding length of counting, as the subband signal participating in subband cross ambiguity function and calculate;

Wherein, the initial interception position of every way band signal is identical;

Step 5, calculates each subband CAF

According to each subband correlative accumulation time that step 4 adjustment obtains, the correlative accumulation result adopting CAF to calculate each subband is:

χ _k(τ,f _d)＝DFT{y _k(τ,n)} (12)

Step 6, subband correlative accumulation is comprehensive

The correlative accumulation resulting complex values of each subband gained is added, has been eliminated the object detection results of Doppler dispersion:

${\mathrm{\χ}}_{\mathrm{\Σ}}(\mathrm{\τ},f_d)=\underset{k=1}{\overset{M}{\mathrm{\Σ}}}{\mathrm{\χ}}_k(\mathrm{\τ},f_d)---\left(13\right)$

So far, broadband signal target detection is completed.

2. the broadband signal target detection Doppler dispersion removing method based on sub-band processing according to claim 1, it is characterized in that: for digital television signal, K gets 32 or 64.

3. the broadband signal target detection Doppler dispersion removing method based on sub-band processing according to claim 1, is characterized in that: be suitable for adopting long time integration detection high-speed target.