CN101458334B - A Motion Compensation Method for Bistatic Synthetic Aperture Radar Imaging - Google Patents

Abstract

The invention provides a motion compensation method for bistatic synthetic aperture radar imaging based on a phase tracking algorithm. In the method, firstly, by designing a set of narrow-band filters, inhibiting the echo effects of the neighboring targets and acquiring the phase position information of the echo signal of strong scattering points, phase position compensation and direction compression are carried out on the neighboring distance unit, thereby acquiring a bistatic SAR imaging result with high accuracy; secondly, extracting bistatic SAR imaging processing parameters by directly using echo data can lower the cost of a bistatic SAR system; in addition, self-focusing technology which utilizes the echo data to process the signals can compensate the effect of the rapid disturbance which is difficult to be detected by a navigation system. The method provides technical supports for the study of advanced high-resolution radar imaging technology.

Description

A kind of motion compensation process of double-base synthetic aperture radar imaging

Technical field

The invention belongs to the radar imagery technical field, it is particularly related to double-base synthetic aperture radar imaging and (is called for short: motion compensation technique Bi-SAR).

Background technology

The transmitter and receiver of double-base synthetic aperture radar splits on different platforms, transmitter transmits and shines a certain object scene, receiver receives the echoed signal of this irradiation object scene, carry out the orientation through pulse compression technique and the Doppler shift that utilizes flying platform and target relative motion to produce and compress, the distance that realizes target respectively to the orientation to high-resolution imaging.

Double-base synthetic aperture radar has a lot of outstanding advantages:

(1) can obtain the non-back scattering information of target or scene, help improving target identification performance.

(2) have that operating distance is far away, characteristics such as disguise and strong interference immunity, the hidden and viability of double-base synthetic aperture radar under war condition is stronger.

(3) double-base SAR cost expense is low, because the double-base SAR receiver do not contain high power device, it is low in energy consumption, volume is little, in light weight, is convenient to polytype aircraft and carries, and cost is lower.

(4) be used for interference imaging, obtain the 3-D view of target or scene.

Find out that from above-mentioned advantage double-base synthetic aperture radar is as a kind of means of earth observation from space, in future war, its viability is stronger, and the information of acquisition is more, can be widely used in military surveillance and obtain military information.In addition, also can be used for civil areas such as resource exploration, major disasters estimation and the earth mapping.

The receiving system of double-base SAR splits on different platforms with emission coefficient, the motion of two platforms is influenced by air flow instability etc., jolting of carrier aircraft is bigger with disturbance, the imperfect kinematic error that causes two platforms, if do not take the motion compensation measure, image quality is descended, even can not imaging.According to the document of publishing, for example: Holger Nies, Otmar Loffeld etc. has proposed to measure the exact position parameter that obtains emission, receiver based on navigation instrument in " A Solution for Bistatic MotionCompensation " literary composition, in order to motion compensation, and obtained good effect.But, there be not inertial navigation system (InertialNavigation System, be called for short: INS) and GPS (Global Position System, abbreviation: GPS) or under the not high situation of the parameters precision that provides of navigational system, said method is invalid.Moreover INS and GPS navigation instrument system also are difficult to detect the quick disturbance of carrier aircraft.In addition, use INS and GPS itself also can increase the weight of the cost of double-base SAR system.

Summary of the invention

Influence for the kinematic error of eliminating the double-base SAR imaging, improve the double-base SAR imaging precision, the present invention proposes a kind of double-base synthetic aperture radar imaging method based on the Phase Tracking algorithm, it is to utilize the double-base synthetic aperture radar echo data to obtain the method for strong scattering point target phase place in order to motion compensation, this method can obtain the phase change course of strong scattering point target signal, be applied to the motion compensation of double-base SAR imaging, can obtain the high precision imaging results; Secondly, directly utilize echo data to extract double-basis SAR imaging processing parameter, can reduce double-basis SAR system cost.In addition, the influence of the self-focusing technology of utilizing echo data the to carry out signal Processing quick disturbance that navigational system can also be difficult to detect is compensated.

Content of the present invention for convenience of description, at first make following term definition:

Define 1 Phase Tracking

To the double-base SAR echo data, adopt the narrow-band filtering technology to extract the phase place of strong scattering point echoed signal.

Definition 2 mixes strong signal

Transmitter transmits and shines a certain ground scene, and through the ground scene reflection, radar receiver receives the target echo signal pp (i) in the scene, and wherein, target echo signal pp (i) has comprised the echoed signal pp of strong scattering point target _MAX(i), then claim the echoed signal pp (i) that receives to be the strong signal of mixing, i=1,2 ... L, L are expressed as the length of strong scattering point reflection echoed signal in the synthetic aperture.

Define 3 bistatic distances and

If the transmitter and receiver platform is with speed linear uniform motion separately, with the orientation time be parameter, the position vector of transmitter

Position vector with the receiver platform

Can be expressed as

${\stackrel{\→}{\mathrm{\ξ}}}_R\left(t\right)={\stackrel{\→}{i}x}_R\left(t\right)+{\stackrel{\→}{j}y}_R\left(t\right)+{\stackrel{\→}{k}z}_R\left(t\right)={\stackrel{\→}{v}}_{R}t+{\stackrel{\→}{\mathrm{\ξ}}}_{R0}---\left(1\right)$

${\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_T\left(t\right)={\stackrel{\&RightArrow;}{i}x}_T\left(t\right)+{\stackrel{\&RightArrow;}{j}y}_T\left(t\right)+{\stackrel{\&RightArrow;}{k}z}_T\left(t\right)={\stackrel{\&RightArrow;}{v}}_{T}t+{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{<figure-callout\; id="0"\; label="T"\; filenames="S200710050856XE00011.png,S200710050856XE00021.png"\; state="\{\{state\}\}">T</figure-callout>}0}$

Wherein, x, y, the corresponding unit vector of z axle difference in the three-dimensional system of coordinate

T represents the orientation time, i.e. the slow time. With

It is respectively the velocity of transmitter and receiver.

With

It is orientation zero position vector of transmitter and receiver constantly.

The position vector of the target of setting up an office is

Then this point target to the vector of transmitter and receiver is respectively

${\stackrel{\&RightArrow;}{r}}_R(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)={\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_R\left(t\right)-{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P---\left(2\right)$

${\stackrel{\&RightArrow;}{r}}_T(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)={\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_T\left(t\right)-{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P$

Transmitter to the distance of point target is Receiver is to the distance of point target

Be

$R_R(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)=\left|{\stackrel{\&RightArrow;}{r}}_R(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\right|=|{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_R\left(t\right)-{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P|---\left(3\right)$

$R_T(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)=\left|{\stackrel{\&RightArrow;}{r}}_T(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\right|=|{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_T\left(t\right)-{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P|$

Then bistatic distance and

For:

$R_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)=R_R(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)+R_T(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)---\left(4\right)$

Define 4 narrow band filter groups

In order to extract the actual phase of strong scattering point echoed signal, designed one group of narrow band filter, its mathematic(al) representation is as follows:

F(f，f _i)＝x(f-f _i)；

F (f, f _i) be the function of narrow band filter, f _iBe centre frequency, f ₀Width for narrow band filter.

Define 5 phase unwrappings around processing

If θ (j) is for twining phase place, φ (j) is for separating the winding phase place, j=1, and 2,3 ... J.

Solutions of path integration is twined algorithm

φ(1)＝θ(1) φ(j+1)＝θ(m)+Δ(m)

Can follow the neighbouring relations of sampling point according to above-mentioned formula, the pointwise extrapolation solves phase fuzzy problem.

Define 6 quadratic fit smoothing techniques

The quadratic fit smoothing technique is at first carried out quafric curve to the level and smooth data of needs and is fitted, and promptly seeks the φ of family of functions _Fit=a ₁X ²+ a ₂X+a ₃In and one group of parameter (a of square error minimum between the level and smooth data of needs ₁, a ₂, a ₃); Utilize (a that obtains then ₁, a ₂, a ₃) substitution φ _Fit=a ₁X ²+ a ₂X+a ₃Data after obtaining smoothly.

Define 7 range migrations

In the SAR imaging system, relative motion between the target of radar antenna territory, caused target in the irradiation wave beam, to change and surpassed a distance explanation unit, made the echo of same target to be distributed in different range gates, Here it is range migration with the oblique distance of radar.

The present invention proposes a kind of motion compensation process of the double-base synthetic aperture radar imaging based on the Phase Tracking algorithm, it is characterized in that it comprises following steps:

Step 1 determines to comprise the strong signal of mixing of strong scattering point target

The echo data that the double-base synthetic aperture radar receiver receives is a M _r* N _aComplex matrix M _Raw(m, n), wherein, m is that the distance of double-base synthetic aperture radar receiving radar is to sampled point ordinal number, M _rFor individual pulse repetition period inner receiver distance by radar to sampling number, m=1 ..., M _rN is the pulse number of receiver radar observation, N _aBe the pulse number of receiver radar observation, n=1 ..., N _aIs matched filtering along distance to carrying out the pulse pressure processing with the double-base synthetic aperture radar echo data, obtains distance compression echo data M afterwards _Comp(m, n);

From two-dimentional echo signal data matrix M _Comp(m, n) middle traversal search goes out the Position Approximate m0 of strong scattering point target signal place range gate; And, according to the Position Approximate m0 of strong scattering point target signal place range gate from two-dimentional echo signal data matrix M _Comp(m selects in n) and comprises strong scattering point target signal pp _MAX(i) the matrix m of local data _Comp(k, l), k=1 wherein, 2 ... M _m, l=1,2 ... N _a, i is a strong scattering point target aspect sampling ordinal number, M _mValue by two-dimentional echo signal data matrix range migration amount M _DDecision, M _DBig more M _mBig more, M _m＜M _rSearch local data matrix M _Comp(k, the peaked position of amplitude of amplitude maximal value l) and local data's matrix are set peaked 1/2nd and are thresholding, in local data's matrix M _Comp(k, in l) be with the maximum value position starting point in the orientation to point by point search, obtain comprising strong scattering point target signal pp _MAX(i) the strong signal ps of mixing _MAX(l), wherein, l=1,2 ... L, L is for mixing the length of strong signal;

Step 2 design narrow band filter extracts the strong point echo signal

The strong signal ps of mixing that step 1 is obtained _MAX(l) be divided into W segment signal ps _u(v), u=1 wherein, 2 ... W, $v=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\frac{L}{W},$ W is a natural number.To mix strong signal ps _MAX(l) section ps of each in _u(v) mixing strong signal ps respectively _MAX(l) go up original signal and keep, all zero setting of other each sections, formation and mix strong signal ps _MAX(l) isometric signal ps _New(u, l), the signal ps that to obtain W length altogether be L _New(u, l); With signal ps _New(u l) makes fast fourier transform respectively, and obtaining W length is L frequency-region signal SS _New(u, l).Search for frequency-region signal SS then _{New (}U, l) the pairing frequency f of amplitude maximal value _i, i=1,2 ... W.In order to f _iFor centre frequency, width are

The narrow band filter group be frequency-region signal SS to W length respectively _New(u l) carries out filtering, obtains filtered frequency-region signal PS ' _New(u, l); Again with W filtered frequency-region signal PS ' _New(u, l) stack, constitute length and be L frequency-region signal PS (u, l); (u l) carries out invert fast fourier transformation, obtains strong point echo signal pp to frequency-region signal PS again _MAX(i);

Step 3 phase unwrapping around

Step 2 is obtained strong point echo signal pp in the echo _MAX(i), adopt plural number to get multiple angle computing angle () and get strong point echo signal pp _MAX(i) phase information obtains phasing degree phase_angle; Around Unwrap () phase_angle is made phase unwrapping around processing with phase unwrapping then, just can obtain the phase history φ of this strong scattering point signal _s, Unwrap () is that phase unwrapping is around function.Again by formula ${\mathrm{\&phi;}}_s=\frac{2\mathrm{\&pi;}}{\mathrm{\&lambda;}}{\stackrel{\&RightArrow;}{R}}_s(t,\mathrm{\&xi;}),$ Calculate the geometric locus of this strong scattering point in data matrix

Wherein, λ is the orientation time for the pairing wavelength of the centre frequency that transmits, t, and ξ is the position vector of point target;

Step 4 phase error estimation and phase error

Emission under ideal conditions, during the receiving platform unaccelerated flight, strong scattering point target echoed signal phase history can be approximate with quadratic polynomial, obtains the match phase _Fit=a ₁T ²+ a ₂T+a ₃, wherein.a ₁, a ₂, a ₃Be match quadratic polynomial coefficient, t is the orientation time;

The phase history φ of the strong scattering point target echoed signal by quadratic fit smoothing technique match step 3 gained _s, obtain the match phase ' _s

Ask phase error to be: φ _ErR=φ ' _s-φ _Fit, obtain φ _Err=φ ' _s-(a ₁T ²+ a ₂T+a ₃); Then, to this strong point target place matrix m of local data _Comp(k, l) along the orientation to multiply by exp (j φ _Err), carry out the kinematic error compensation, finish the phase place self-focusing.

Through above step, can finish the motion compensation of double-base SAR imaging.

Essence of the present invention designs one group of narrow band filter exactly, suppresses the echo influence of adjacent target, obtains the phase information of strong scattering point echoed signal, and the neighbor distance unit object is carried out phase compensation and orientation compression, has improved the double-base SAR imaging precision.

Innovative point of the present invention is:

According to after the strong scattering point target to the big characteristics of echoed signal energy, automatically detect and comprise the strong signal of mixing that strong scattering is put high target in the echo data, centre frequency and required window width according to signal, design one group of narrow band filter, suppress the echo influence of adjacent target, mix the actual phase that extracts strong scattering point target signal the strong signal from this, after level and smooth, calculate the phase error phi of this strong point echo signal with the match of this strong point echo signal phase place through phase unwrapping _Err, be used for the motion compensation of double-base SAR imaging, improved imaging precision.

Ultimate principle of the present invention is:

If the radar emission linear FM signal,

pp(τ，t)＝w _t(τ；t)exp{-j2πf _ct}·exp{-jπK[τ-nT _r] ²} (7)

In the formula: τ=t-nT _rTime in the arteries and veins of expression radar transmitted pulse is apart from the time, i.e. the fast time, t is " slowly " time of radar transmitted pulse level in the expression synthetic aperture, T _rBe pulse-recurrence time, n indicating impulse number, f _cBe carrier frequency, K is a chirp rate, w _tBe the window function of transmitter antenna, relevant with antenna beam shape, the expression wave beam is to the modulation of echo amplitude.

Then any some target echo signals can be expressed as through down coversion and after omitting the constant amplitude factor in the scene:

$\mathrm{pp}(\mathrm{\&tau;},t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)=\mathrm{\&sigma;}\left({\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P\right)w_t(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)w_r(\mathrm{\&tau;};{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\exp \{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="S200710050856XE00011.png,S200710050856XE00021.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;}}{\mathrm{\&lambda;}}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\}---\left(8\right)$

$\&CenterDot;\exp \{-\mathrm{j\&pi;K}{[\mathrm{\&tau;}-\frac{1}{c}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)]}^2\}$

Wherein, point target echoed signal mark

Be illustrated in apart from time one orientation time domain, c is the light velocity, and λ is the pairing wavelength of the centre frequency that transmits;

It is the scattering coefficient of point target; w _rBe the window function of receiver antenna, relevant with antenna beam shape, represent the modulation of wave beam to echo amplitude,

For bistatic distance and.

First phase term $\exp \{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="S200710050856XE00011.png,S200710050856XE00021.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;}}{\mathrm{\&lambda;}}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\}$ Represent the orientation to modulate, second phase term to Doppler frequency $\exp \{-\mathrm{j\&pi;K}{[\mathrm{\&tau;}-\frac{1}{c}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)]}^2\}$ Representative comprises the distance of bistatic range delay to linear FM signal.If there is a strong scattering point in the echoed signal

Promptly Scattering coefficient wants high with respect to the scattering coefficient of the point of adjacent area in the scene, then the strong scattering point

Echoed signal can be expressed as:

${\mathrm{pp}}_{\mathrm{MAX}}(\mathrm{\&tau;},t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{sp}})={\mathrm{\&sigma;}}_{\mathrm{MAX}}\left({\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{sp}}\right)w_t(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{sp}})w_r(\mathrm{\&tau;};{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{sp}})exp\{-<figure-callout\; id="2"\; label="j"\; filenames="S200710050856XE00011.png,S200710050856XE00021.png"\; state="\{\{state\}\}">j</figure-callout>\frac{2\mathrm{\&pi;}}{\mathrm{\&lambda;}}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{sp}})\}---\left(9\right)$

$\&CenterDot;\exp \{-\mathrm{j\&pi;K}{[\mathrm{\&tau;}-\frac{1}{c}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_{\mathrm{sp}})]}^2\}$

The echoed signal that radar receives is the linear superposition of point target echo in the scene, can be expressed as:

$\mathrm{ss}(\mathrm{\&tau;},t)=\&Integral;\mathrm{pp}(\mathrm{\&tau;},t;\mathrm{\&xi;})d\stackrel{\&RightArrow;}{\mathrm{\&xi;}}---\left(10\right)$

" soon " time linear frequency modulation composition of thinking variable in the formula (10) is carried out after pulse pressure handles (matched filtering), and the output signal that obtains is:

$\mathrm{pc}(\mathrm{\&tau;},t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)=\mathrm{\&sigma;}\left({\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P\right)w_t(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)w_r(\mathrm{\&tau;};{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\exp \{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="S200710050856XE00011.png,S200710050856XE00021.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;}}{\mathrm{\&lambda;}}{R}_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)\}---\left(11\right)$

$\&CenterDot;\sin c\left[B(\mathrm{\&tau;}-\frac{R_s(t;{\stackrel{\&RightArrow;}{\mathrm{\&xi;}}}_P)}{C})\right]$

B=KT in the formula _pBe the bandwidth of the linear FM signal of radar emission, T _pIt is exomonental time width.

So whole echo data is after the distance compression, and the scene echoes signal indication is:

$\mathrm{sc}(\mathrm{\&tau;},t)=\&Integral;\mathrm{pc}(\mathrm{\&tau;},t;\stackrel{\&RightArrow;}{\mathrm{\&xi;}})d\stackrel{\&RightArrow;}{\mathrm{\&xi;}}---\left(12\right)$

Have the mixing large-signal that comprises strong scattering point target echo in the echoed signal matrix of above-mentioned formula (12), we can design one group of narrow band filter, extract strong scattering point target signal from echo data Phase information, after level and smooth, calculate the phase error phi of this strong point through phase unwrapping with the match of this strong point echo signal phase place _Err, can sub-aperture to this strong scattering point target place in adjacent some range gate along the orientation to multiply by e ^{J φ err}, carry out the kinematic error compensation, finish the phase place self-focusing.

The technical matters that the present invention solves:

Automatically detect the strong signal of mixing that comprises strong scattering point target echo, by one group of narrow band filter, suppress to close on the influence of echo, obtain the accurate estimated value of phase place of strong point target echo signal, through phase unwrapping around with the curve fit of strong point echo signal phase place, calculate phase error phi _Err, can be used for the motion compensation of double-base SAR imaging, reach the purpose that improves the double-base SAR imaging precision.

Advantage of the present invention is:

Under the not high situation of the parameters precision that does not have inertial navigation and GPS or navigational system to provide, directly utilize echo data to extract double-basis SAR imaging processing parameter, can reduce double-basis SAR system cost, and obtain the high precision imaging results.

Description of drawings:

Fig. 1 is a workflow block diagram of the present invention

Fig. 2 is the desired phase synoptic diagram of strong scattering point target

Wherein, transverse axis is represented slow time t, and the longitudinal axis is represented phase place;

Fig. 3 is the match phase place synoptic diagram of strong scattering point target,

Wherein, transverse axis is represented slow time t, and the longitudinal axis is represented phase place;

Fig. 4 is the phase error of strong scattering point target

Wherein, transverse axis is represented slow time t, and the longitudinal axis is represented phase error.

Embodiment

The present invention mainly adopts the method that the double-base SAR measured data is handled to verify, institute in steps, conclusion all on MATLAB 7.0 checking correct.Concrete implementation step is as follows:

Step 1 determines to comprise the strong signal of mixing of strong scattering point target

The echo data that the Bi-SAR receiver receives is one 2400 * 8000 complex matrix M _Raw(m, n), wherein, m be the distance of Bi-SAR receiving radar to the sampled point ordinal number, 2400 be individual pulse repetition period inner receiver distance by radar to sampling number, m=1 ..., 2400; N is the pulse number of receiver radar observation, N _aBe the pulse number of receiver radar observation, n=1 ..., 8000; Is matched filtering along distance to carrying out the pulse pressure processing with the Bi-SAR echo data, obtains distance compression echo data M afterwards _Comp(m, n).

From two-dimentional echo signal data matrix M _Comp(m, n) middle traversal search goes out the Position Approximate m0=876 of strong scattering point target signal place range gate; And, according to the Position Approximate m0=876 of strong scattering point target signal place range gate from two-dimentional echo signal data matrix M _Comp(m selects in n) and comprises strong scattering point target signal pp _MAX(i) the matrix m of local data _Comp(k, l), k=1 wherein, 2 ... M _m, M _m=150; L=1,2 ... M _r, M _r=600; I is a strong scattering point target aspect sampling ordinal number, M _mValue by two-dimentional echo signal data matrix range migration amount M _DDecision, M _DBig more M _mBig more, M _m＜M _rSearch local data matrix M _Comp(k, the peaked position of amplitude of amplitude maximal value l) and local data's matrix are set peaked 1/2nd and are thresholding, in local data's matrix M _Comp(k, in l) be with the maximum value position starting point in the orientation to point by point search, obtain comprising strong scattering point target signal pp _MAX(i) the strong signal ps of mixing _MAX(l), wherein, l=1,2 ... 600, L is for mixing the length of strong signal.

Step 2 design narrow band filter extracts the strong point echo signal

The strong signal ps of mixing that step 1 is obtained _MAX(l) be divided into W=120 segment signal ps _u(v), u=1 wherein, 2 ... 120, v=1,2 ... 5.To mix strong signal ps _MAX(l) each the segment signal ps in _u(v) mixing strong signal ps respectively _MAX(l) go up original signal and keep, all zero setting of other each sections, formation and mix strong signal ps _MAX(l) isometric signal ps _Ew(u, l), the signal ps that to obtain W=120 length altogether be L=600 _New(u, l).With signal ps _New(u l) makes fast fourier transform respectively, and obtaining W=120 length is L=600 frequency-region signal SS _New(u, l).Search for frequency-region signal SS then _New(u, l) the pairing frequency f of amplitude maximal value _i, i=1,2 ... 120.In order to f _iFor centre frequency, width are $\frac{L}{W}=5$ The narrow band filter group signal SS that is L=600 to W=120 length respectively _New(u l) carries out filtering, obtains filtered frequency-region signal PS _New' (u, l).Again with W=120 filtered frequency-region signal PS _New' (u, l) stack, constitute length and be L frequency-region signal PS (u, l).Be that (u l) carries out invert fast fourier transformation, obtains strong point echo signal pp for the frequency-region signal PS of L=600 to length again _MAX(i).

Step 3 phase unwrapping around

Step 2 is obtained strong point echo signal pp in the echo _MAX(i), adopt plural number to get multiple angle computing angle () and get strong point echo signal pp _MAX(i) phase information obtains phasing degree phase_angle.Around Unwrap () phase_angle is made phase unwrapping around processing with phase unwrapping then, just can obtain the phase history φ of strong scattering point signal _s, Unwrap () is that phase unwrapping is around function.Again by formula ${\mathrm{\&phi;}}_s=\frac{2\mathrm{\&pi;}}{\mathrm{\&lambda;}}{\stackrel{\&RightArrow;}{R}}_s(t,\mathrm{\&xi;}),$ Calculate the geometric locus of strong scattering point in data matrix

Wherein, λ=0.03, λ is the orientation time for the pairing wavelength of the centre frequency that transmits, t, ξ is the position vector of point target.

Step 4 phase error estimation and phase error

Emission under ideal conditions, during the receiving platform unaccelerated flight, strong scattering point target echoed signal phase history can be approximate with quadratic polynomial, obtains the match phase _Fit=a ₁T ²+ a ₂T+a ₃, wherein.a ₁, a ₂, a ₃Be match quadratic polynomial coefficient, t is the orientation time.

The phase history φ of the strong scattering point target echoed signal by quadratic fit smoothing technique match step 3 gained _s, obtain the match phase _s'.

Ask phase error to be: φ _Err=φ ' _s-φ _Fit, obtain φ _Err=φ ' _s-(a ₁T ²+ a ₂T+a ₃); Then, to this strong point target place matrix m of local data _Comp(k, l) along the orientation to multiply by exp (j φ _Err), carry out the kinematic error compensation, finish the phase place self-focusing.

Handle through above-mentioned steps, can finish the motion compensation of double-base SAR imaging.

Claims (1)

1. a kind of motion compensation method based on the bistatic synthetic aperture radar imaging of phase tracking algorithm, it is characterized in that it comprises the following steps: Step 1 Identify mixed strong signals containing strong scatter point targets The echo data received by the bistatic SAR receiver is a M _r × _N a complex matrix M _raw (m, n), where m is the range sampling point sequence number of the bistatic SAR receiving radar, and M _r is the number of sampling points in the range direction of the receiver radar within a single pulse repetition period, m=1,..., M _r ; n is the sequence number of pulses observed by the receiver radar, N _a is the number of pulses observed by the receiver radar, n=1 ,..., _Na ; the bistatic SAR echo data is processed along the range direction by pulse pressure, that is, matched filtering, and the echo data M _comp (m, n) after range compression is obtained; From the two-dimensional echo signal data matrix M _comp (m, n), search for the approximate position m0 of the range gate where the strong scattering point target signal is located; and, according to the approximate position m0 of the range gate where the strong scattering point target signal is located Select the local data matrix m _comp (k, l) containing the strong scattering point target signal pp _MAX (i) from the echo signal data matrix M _comp (m, n), where k=1, 2...M _m , l=1, 2...N _a , i is the azimuth sampling number of the strong scattering point target signal, the value of M _m is determined by the distance migration amount M _D of the two-dimensional echo signal data matrix, the larger M _D is M _m is larger, M _m < M _r ; search for the maximum amplitude value of the local data matrix M _comp (k, l) and the position of the maximum amplitude value of the local data matrix, and set half of the maximum value as the threshold. In the matrix M _comp (k, l), take the maximum position as the starting point to search point by point in the azimuth direction, and obtain the mixed strong signal ps _{MAX (l) containing the strong scattering point target signal pp MAX (} i), wherein, l=1, 2......L, L is the length of the mixed strong signal; the mixed strong signal refers to the echo signal that contains the target signal of the strong scattering point; Step 2 Design a narrowband filter to extract strong scattering point target signals The mixed strong signal ps _MAX (l) obtained in step 1 is equally divided into W segment signals ps _u (v), wherein u=1,2,...W, v=1,2,

W is a natural number; each segment of the signal ps _u (v) in the mixed strong signal ps _MAX (l) is reserved on the original signal of the mixed strong signal ps _MAX (l), and the other sections are all set to zero, forming the mixed strong signal ps Signals ps _new (u, l) equal in length to _MAX (l), and a total of W signals ps _new (u, l) whose length is L are obtained; fast Fourier transform is performed on the signals ps _new (u, l) respectively, to obtain W A length is L frequency domain signal SS _new (u, l); then search frequency domain signal SS _new (u, l) frequency f _i corresponding to the maximum amplitude, i=1, 2,...W ; With f _i as the center frequency, the width is

The narrow-band filter banks of W respectively filter W frequency-domain signals SS _new (u, l) of length to obtain filtered frequency-domain signals PS′ _new (u, l); then W filtered frequency-domain signals PS′ _new (u, l) is superimposed to form a frequency-domain signal PS(u, l) of length L; then perform inverse fast Fourier transform on the frequency-domain signal PS(u, l) to obtain a strong scattering point target signal pp _MAX (i); Step 3 Phase Unwrapping For the target signal pp _MAX (i) of the strong scattering point in the echo obtained in step 2, the complex angle operation angle( ) is used to obtain the phase information of the target signal pp _MAX (i) of the strong scattering point to obtain the phase angle phase_angle; then use the phase Unwrap Unwrap

By performing phase unwrapping on phase_angle, the phase history φ _s of the strong scattering point signal can be obtained, Unwrap is the phase unwrapping function; then by the formula

Calculate the trajectory curve of the strong scattering point in the data matrix Among them, λ is the wavelength corresponding to the center frequency of the transmitted signal, t is the azimuth time, and ξ is the position vector of the point target; Step 4 Phase Error Estimation Under ideal conditions, when the transmitting and receiving platforms fly in a straight line at a constant speed, the phase history of the echo signal of the strong scattering point target can be approximated by a quadratic polynomial, and the fitted phase φ _fit = a ₁ ·t ² +a ₂ ·t+a ₃ , where , a ₁ , a ₂ , a ₃ are coefficients of fitting quadratic polynomial, t is azimuth time; Fit the phase history φ _s of the strong scattering point target echo signal obtained in step 3 by quadratic fitting smoothing technology to obtain the fitted phase φ′ _s ; Calculate the phase error as: φ _err = φ′ _s -φ _fit , and get φ _err = φ′ _s -(a ₁ ·t ² +a ₂ ·t+a ₃ ); then, the local data matrix where the strong point target is located The azimuth direction of m _comp (k, l) is multiplied by exp(jφ _err ) to perform motion error compensation and complete phase self-focusing; Through the above steps, the motion compensation of the bistatic SAR imaging can be completed. the

Images