RFPA radar coherent accumulation method and system under influence of group delay distortion

By performing distance-speed compensation and group time delay distortion compensation in the phase parameter accumulation method of radar, using phase factors based on the second-order coefficient of group time delay, the phase distortion problem of radar echo signal under group time delay distortion is solved. Improves the estimation accuracy of target motion parameters and reduces the computational complexity.

CN119024294APending Publication Date: 2024-11-26BEIJING INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410976032.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing phase comparison accumulation method cannot effectively suppress the phase distortion of radar echo signals under the case of group time delay distortion, resulting in a decrease in the accuracy of target motion parameters estimation, and traditional channel equalization algorithms are difficult to apply to radar scenarios.

Method used

A phase-parameter accumulation method for random carrier frequency refrequency agile radar under the influence of group time delay distortion is proposed. By performing distance-speed compensation and group time delay distortion compensation in phase-parameter accumulation operation, the second-order coefficient based on group time delay is used. The phase factor is constructed, using three multidimensional parameter searches to simplify computational complexity and improve estimation accuracy.

Benefits of technology

Effectively suppress group time delay distortion, improve the estimation accuracy of target motion parameters, and reduce the computational complexity by simplifying the phase expression method, and optimize the search strategy to reduce unnecessary calculation steps.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024294A_ABST
    Figure CN119024294A_ABST
Patent Text Reader

Abstract

The invention provides a phase-coherent accumulation method and system for a random carrier frequency and repetition frequency agile radar under the influence of group delay distortion, and the method not only carries out the compensation of distance and speed, but also carries out the compensation of group delay distortion during the phase-coherent accumulation. A phase factor of group delay distortion compensation is constructed based on a group delay second-order coefficient. By using the method, group delay distortion can be suppressed, and the estimation accuracy of the target motion parameters is improved. According to the method, when the group delay distortion compensation phase is constructed, the influence of the first-order and second-order coefficients of the group delay on the coherent accumulation result is analyzed, and the compensation phase expression mode is simplified based on the analysis result, so that the calculation complexity is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of radar signal processing, and in particular to a coherent accumulation method and system for a random carrier frequency re-frequency agile radar under the influence of group delay distortion. Background Art

[0002] The random carrier frequency repetition interval agile (RFPA) radar signal is anti-interference in both the frequency domain and the time domain, and has excellent electronic countermeasure performance, thus having excellent anti-interference capability and low probability of intercept. However, due to the random switching of the carrier frequency of the pulse in the broadband and the group delay fluctuation in the large synthesis bandwidth, the coherent accumulation performance will be greatly deteriorated.

[0003] Since the agile radar parameters will cause discontinuity in the phase of the echo signal, traditional coherent integration methods such as Keystone transform, Hough transform and Radon-Fourier transform will fail in the application of agile radar. In order to solve this phase discontinuity, some improved coherent integration methods have been proposed for agile radar signals and RFPA signals in recent years. However, current research has not considered the group delay fluctuation in the echo signal. For RFPA signals, the group delay fluctuation of the channel will cause phase distortion of echo signals with different carrier frequencies. Under the influence of group delay distortion, the performance of the above coherent integration methods will be significantly reduced.

[0004] In communication systems, channel equalization algorithms are used to correct the effects of group delay distortion. However, channel equalization algorithms in communication systems are designed to reduce inter-symbol interference and lower the bit error rate in data transmission. Unlike radar application scenarios, related algorithms are difficult to use in radar group delay distortion suppression. In addition, some literature has studied the group delay distortion of wideband frequency modulated continuous wave radars, pointing out that the nonlinear phase caused by group delay distortion will lead to frequency nonlinearity, thereby causing the deterioration of the radar range image. However, the effect of group delay distortion in RFPA radar echo signals is different from that of wideband frequency modulated continuous wave radar signals. The carrier frequency of RFPA radar echo signals changes randomly and rapidly, and each pulse is subject to varying degrees of phase distortion. Summary of the invention

[0005] In view of this, the present invention provides a coherent accumulation method for a random carrier frequency re-frequency agile radar under the influence of group delay distortion, which increases the phase compensation of group delay distortion, can suppress group delay distortion, and improve the estimation accuracy of target motion parameters. In addition, when constructing the group delay distortion compensation phase, the present invention analyzes the influence of the first-order and second-order coefficients of the group delay on the coherent accumulation result, and simplifies the compensation phase expression based on the analysis results, thereby reducing the calculation complexity.

[0006] In order to solve the above technical problems, the present invention is implemented as follows.

[0007] A coherent integration method for a random carrier frequency re-frequency agile radar under the influence of group delay distortion, comprising:

[0008] Coherently integrate the radar echo signals after reorganization and matched filtering; perform distance-speed compensation and group delay distortion compensation in the coherent accumulation operation;

[0009] The phase factor of the group delay distortion compensation is constructed based on the second-order coefficient of the group delay and is expressed as:

[0010]

[0011] Where Δf m is the frequency agility parameter of the pulse; α 2 (k 2 ) is the second-order coefficient of the group delay response polynomial, A 2min is the minimum search value of the group delay response parameter, is the search range of the group delay response parameter, is the search interval for the group delay response parameter.

[0012] Preferably, the coherent accumulation is implemented by three-dimensional parameter search:

[0013] First search: group delay distortion compensation is not considered, only distance-speed compensation is performed, and coherent accumulation of speed-distance two-dimensional parameter search is performed to obtain the speed v corresponding to the maximum value of coherent accumulation max and distance r max ;

[0014] Second search: add group delay distortion compensation, fixed speed v max , determine the distance r max In the distance dimension neighborhood centered at , coherent accumulation of two-dimensional parameter search of distance-group delay parameters is performed to obtain the group delay parameter corresponding to the maximum value of the coherent accumulation as the group delay parameter estimation result;

[0015] The third search: using the group delay parameter estimation result, the coherent accumulation of the speed-distance two-dimensional parameter search is performed again to obtain the coherent accumulation result after group distortion compensation; each peak of the coherent accumulation result is the motion parameter estimation result of the moving target.

[0016] Preferably, the phase factor used in the distance-speed compensation is:

[0017] exp[j4πf m (r(i)-v(i)·t m) / c]·exp{-j4πf m ·v(i)[T w / 2+2(r(i)-v(i)·t m ) / c] / c}

[0018] r(i)=R min +iΔ r ,i=0,1,...,N r -1

[0019] v(j)=V min +jΔ v ,j=0,1,...,N v -1

[0020] Among them, r(i) is the distance, v(j) is the speed; R min and V min are the minimum search values ​​for distance and speed respectively; Δ r and Δ v are the search intervals for distance and speed respectively; N r and N v are the number of search units for distance and speed respectively; f m =f c +Δf m =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where fc is the center carrier frequency, Δf is the frequency agility interval, and d(m)∈{-N f / 2,...,0,…,N f / 2} is a random sequence of frequencies, N f is the number of frequency jump points; t m is the slow time, T w is the pulse width and c is the speed of light.

[0021] The present invention also provides a coherent accumulation system for a random carrier frequency re-frequency agile radar under the influence of group delay distortion, comprising: a reorganization module, a matched filtering module and a coherent accumulation module;

[0022] The recombining module and the matched filtering module sequentially recombined and matched-filtered the radar echo signals and sent them to the coherent accumulation module;

[0023] The coherent accumulation module is used to perform coherent accumulation on the reconstructed echo received after matched filtering; and to perform distance-speed compensation and group delay distortion compensation in the coherent accumulation operation;

[0024] The phase factor of the group delay distortion compensation is constructed based on the second-order coefficient of the group delay and is expressed as:

[0025]

[0026] Where Δf m is the frequency agility parameter of the pulse; α 2 (k 2 ) is the second-order coefficient of the group delay response polynomial, A 2min is the minimum search value of the group delay response parameter, is the search range of the group delay response parameter, is the search interval for the group delay response parameter.

[0027] Preferably, the coherent accumulation module includes a first search unit, a second search unit, and a third search unit;

[0028] The first search unit is responsible for the first search: without considering group delay distortion compensation, only distance-speed compensation is performed, and the coherent accumulation of speed-distance two-dimensional parameter search is performed to obtain the speed v corresponding to the maximum value of the coherent accumulation 0 and distance r 0 ;

[0029] The second search unit is responsible for the second search: adding group delay distortion compensation, with a fixed speed of v 0 , determine the distance r 0 In the distance dimension neighborhood centered at , coherent accumulation of two-dimensional parameter search of distance-group delay parameters is performed to obtain the group delay parameter corresponding to the maximum value of the coherent accumulation as the group delay parameter estimation result;

[0030] The third search unit is responsible for the third search: using the group delay parameter estimation result, coherent accumulation of the speed-distance two-dimensional parameter search is performed again to obtain the coherent accumulation result after group distortion compensation; each peak of the coherent accumulation result is the motion parameter estimation result of the moving target.

[0031] Preferably, when the coherent accumulation module performs distance-speed compensation, the distance-speed compensation phase factor used is:

[0032] exp[j4πf m (r(i)-v(i)·t m ) / c]·exp{-j4πf m ·v(i)[T w / 2+2(r(i)-v(i)·t m ) / c] / c}

[0033] r(i)=R min +iΔ r ,i=0,1,…,N r -1

[0034] v(j)=Vmin +jΔ v ,j=0,1,…,N v -1

[0035] Among them, r(i) is the distance, v(j) is the speed; R min and V min are the minimum search values ​​for distance and speed respectively; Δ r and Δ v are the search intervals for distance and speed respectively; N r and N v are the number of search units for distance and speed respectively; f m =f c +Δf m =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where fc is the center carrier frequency, Δf is the frequency agility interval, and d(m)∈{-N f / 2,…,0,…,N f / 2} is a random sequence of frequencies, N f =B s / Δf is the number of frequency jump points, where B s is the synthetic bandwidth; t m is the slow time, T w is the pulse width and c is the speed of light.

[0036] Beneficial effects:

[0037] (1) When performing coherent accumulation, the present invention not only compensates for distance and speed, but also adds phase compensation for group delay distortion, which can suppress group delay distortion and improve the estimation accuracy of target motion parameters. In addition, when constructing the group delay distortion compensation phase, the present invention analyzes the influence of the first-order and second-order coefficients of the group delay on the coherent accumulation result, and finds that the first-order α in the actual analog filter is 1 The value of is much smaller than that used in the simulation and has limited impact on distance estimation, so the α in the filter can be ignored. 1 The nonlinear phase caused by only considering the second-order α 2 In order to reduce the computational complexity, the compensation phase expression is simplified based on the analysis results, thereby reducing the computational complexity.

[0038] (2) From the simulation analysis of the influence of group delay distortion, it can be seen that group delay mainly affects the range-dimensional image of the RFPA radar, and the resulting main lobe offset is limited. Therefore, the group delay parameter search can be performed only within the range-dimensional neighborhood of the strongest target to obtain the estimated result. Therefore, the present invention designs a more reasonable multi-dimensional search strategy to avoid unnecessary search steps and reduce the amount of calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 Schematic diagram of the group delay characteristics of the filter; (a) is the group delay characteristics of three typical filters, and (b) is the group delay characteristics of the cascaded filter;

[0040] Figure 2 is the coherent accumulation result under the influence of first-order group delay distortion; where (a) is α 1 =-5e-3(ns / MHz), (b) is α 1 = -1e-2 (ns / MHz);

[0041] Figure 3 is the coherent accumulation result under the influence of second-order group delay distortion; where (a) is α 2 =6e-4(ns / MHz 2 ), (b) is α 2 =8e-4(ns / MHz 2 );

[0042] Figure 4 An improved parameter search process of the present invention;

[0043] Figure 5 It is a flow chart of the coherent accumulation method of the RFPA radar under the influence of group delay distortion of the present invention;

[0044] Figure 6 It is a structural diagram of the coherent integration system of the RFPA radar under the influence of group delay distortion of the present invention;

[0045] Figure 7 The coherent accumulation results of the RFPA-RFT method with uniform carrier frequency distribution; (a) is the range-velocity plane; (b) is the range dimension slice of the three targets;

[0046] Figure 8 The coherent accumulation results of the algorithm of the present invention when the carrier frequency obeys uniform distribution; (a) is the distance-velocity plane; (b) is the distance dimension slice of three targets. DETAILED DESCRIPTION

[0047] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0048] Aiming at the coherent accumulation of RFPA radar echo signals under the conditions of rapid agile changes in radar parameters and group delay fluctuations within a huge synthetic bandwidth, the present invention proposes a coherent accumulation scheme for RFPA radar under the influence of group delay distortion. The basic idea is to construct a phase compensation factor of the peak conjugate phase of the matched filtering result, and simultaneously compensate for the discontinuous phase caused by the agile radar parameters and the nonlinear phase caused by the group delay distortion, thereby realizing the coherent accumulation of RFPA radar echo signals. This scheme suppresses group delay distortion and improves the estimation accuracy of target motion parameters. When constructing the group delay distortion compensation phase, the present invention analyzes the influence of the first-order and second-order coefficients of the group delay on the coherent accumulation results, and simplifies the compensation phase expression based on the analysis results, thereby reducing the computational complexity. The present invention further optimizes the search strategy to further reduce the computational complexity.

[0049] The derivation process of the coherent accumulation algorithm of the present invention is given below.

[0050] Step 1: Analysis of phase distortion caused by group delay distortion.

[0051] The group delay characteristics of the analog filter in the passband can be expressed by a polynomial:

[0052]

[0053] Where f is the frequency, α = [α 0 ,α 1 ,...,α q ,...] are the polynomial coefficients of the group delay response, α q is the qth order coefficient, f c is the center frequency of the passband.

[0054] Group delay is defined as the negative first-order differential of the phase response, which mainly describes the phase effect of a linear system on different frequency components. Unlike fixed-frequency systems, the instantaneous bandwidth of each pulse of the RFPA radar signal is smaller than the system synthesis bandwidth and occupies different parts of the system frequency band. The effect of group delay distortion on each pulse is related to the group delay near the pulse carrier frequency, rather than being determined by the group delay characteristics of the entire bandwidth. In addition, even though the group delay response has large fluctuations within the synthesis bandwidth of the RFPA radar, it is relatively flat within the instantaneous bandwidth. Each pulse will introduce phase distortion after passing through the nonlinear radar channel, and the magnitude of the phase distortion is the same as the phase response at the pulse carrier frequency. Therefore, the phase distortion of a pulse with a carrier frequency of f is given by the following formula:

[0055]

[0056] In the formula, is the phase response at the center frequency.

[0057] In RFPA radar systems, signals usually pass through multiple analog filters, and it is more reasonable to use cascade filters to simulate the actual group delay characteristics. Figure 1 Figure (a) shows the group delay characteristics of three commonly used typical filters, whose center frequency is 10 GHz and bandwidth is 400 MHz. Figure 1 Figure (b) in FIG. 1 is the group delay characteristic of the filter formed by cascading the above filters, which includes three Butterworth filters, two Chebyshev II filters and one Chebyshev I filter.

[0058] A second-order polynomial is used to fit the group delay, ignoring the slight effect of higher-order group delay terms. The polynomial fit curve of the cascaded filter is plotted on Figure 1 In Figure (b), the zero-order coefficient of the polynomial is 15.4ns. The first-order and second-order coefficients are -1.88e-3ns / MHz and 3.27e-4ns / MHz respectively. 2 , they produce group delay changes of 0.752ns and 13.08ns respectively in the 400MHz passband, which shows that the fluctuation of group delay is mainly caused by the second-order group delay. Figure 1 As can be seen in Figure (b), the error between the fitting curve and the actual response is very small and can be ignored. Therefore, it is feasible to use a second-order polynomial to fit the in-band group delay.

[0059] It is worth noting that the zero-order group delay coefficient only adds a fixed delay to the signal, which will cause a range error on the range image. In practical applications, this range error is usually included in the radar's range zero value and can be easily eliminated by prior calibration. Therefore, in the subsequent discussion, only the non-constant component of the group delay characteristic will be considered.

[0060] Each pulse of the RFPA signal has a random carrier frequency and pulse repetition interval (PRI). The transmitted signal can be expressed as

[0061]

[0062] Where M is the number of pulses. w is the pulse width. m =mT R +T m is the random start time of the transmitted pulse, i.e. the slow time of the conventional radar. R is the average PRI of the radar, T m is the PRI agility of the mth pulse, which is in [0,T ρ ] is randomly distributed within the ρ <T Ris the agility range of the radar PRI. m =f c +Δf m =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where f c is the center carrier frequency, Δf m is the frequency agility parameter of the pulse, Δf is the frequency agility interval, d(m)∈{-N f / 2,...,0,...,N f / 2} is a random sequence of frequencies, N f =B s / Δf is the number of frequency jump points, where B s is the synthesis bandwidth. 0 is the initial phase. μ(t) is the unit rectangular function defined as:

[0063]

[0064] Assuming there are P moving targets, the RFPA received signal affected by group delay distortion can be expressed as

[0065]

[0066] Among them A p 、r p and v p are the complex backscatter coefficient, distance and speed of the pth target at time t. m,p (t) is the echo signal corresponding to the mth pulse of the pth target. c is the speed of light. is the phase distortion caused by group delay distortion, which can be expressed as

[0067]

[0068] Step 2: Signal expression of radar echo after baseband reconstructing.

[0069] RFPA radar has excellent range-free detection capability due to its agile carrier frequency. By processing echo reassembly, the pulse echoes received from multiple PRIs are spliced ​​into a reassembled echo, and the radar's detection range can be greatly increased. In this paper, the received signal will be divided into M reassembled echoes after echo reassembly. If the maximum detection range of the radar is defined as R max , the corresponding distance fuzzy number is Therefore, assuming that the mth reconstructed echo contains the reflected echoes of L transmitted pulses of the same target, and among these L echoes, the echo number of the mth transmitted pulse is K. Then after the carrier frequency of the mth transmitted pulse is down-converted, the mth baseband reconstructed echo can be written as

[0070]

[0071] In the formula

[0072]

[0073] For l+m-K,p (t) is a down-converted pulse signal. D,l+m-K,p =2f l+m-K v p / c is the Doppler frequency of the pth target in the (l+mK)th pulse. Generally speaking, the target speed is much smaller than the speed of light, that is, |v p |<<c, in T W The width change of the pulse envelope can be ignored during this time. Define τ l+m-K,p =2(r p -v p t l+m-K ) / c is the delay of the pth target in the (l+mK)th pulse, then equation (1.8) can be simplified to:

[0074]

[0075] After matched filtering, equation (1.9) can be obtained:

[0076]

[0077] In the formula,

[0078] h m (t) = μ[(-tt m ) / T w ] (1.11)

[0079] is the impulse response of the matched filter,

[0080]

[0081] is the center moment of the two rectangular pulses,

[0082]

[0083] is the overlap ratio of the two rectangular pulses.

[0084] Step 3: Expression of matched filtering results of reconstructed echo.

[0085] It can be seen from equation (1.7) that the mth reconstructed echo contains multiple pulse echo components from all P targets and different frequencies. Here, the pth target is selected as the desired target and analyzed. At this time, the matched filtering result of the reconstructed echo can be written as

[0086]

[0087] y 1 (t,m) is the matched filtering result of the expected echo signal of the pth target to the mth transmitted pulse. It can be seen from the formula that when t is the same as the delay of the echo component, that is, t = τ m,p , the energy of the expected echo pulse is effectively accumulated, and the matched filtering result y m,p (t,p) obtains the maximum value.

[0088] y 2 (t,m) is the matched filtering result of the reflected echo components of the pth target to other transmitted pulses. Due to the mismatch between the carrier frequency and the down-conversion frequency, the envelope amplitude of these matched filtering results is smaller than that of y 1 (t,m) is very small.

[0089] y 3 (t,m) is the matched filtering result of other targets on the reflected echo component of the mth transmitted pulse. Among them, the neighboring targets of the pth target will have a certain impact on the matched filtering result of the pth target, and the impact of other targets is limited due to their long distance. If there is a strong target at a close distance of the pth target, the matched filtering result of the pth target will be submerged in the proximal sidelobes of these strong targets.

[0090] y 4 (t,m) is the matched filtering result of the reflected echo components of other targets on other transmitted pulses. Since the transmitted pulses are located in different PRIs, the impact of these undesired signal components is mainly caused by targets that are far away from the pth target. If the reflection intensity of the pth target is weak, its matched filtering result will also be submerged by the far side lobes of these targets that are far away from the pth target.

[0091] Step 4: Distance-speed-group delay distortion compensation factor analysis.

[0092] What needs to be paid attention to is the coherent integration algorithm of the RFPA radar echo signal. Consider the scenario where the difference in the echo intensity of each target is relatively small. In this scenario, after matched filtering, the problem of the near sidelobe and far sidelobe of the large target blocking the small target is relatively minor, and it is not the main problem in the coherent integration. In addition, in the multi-pulse coherent accumulation process, the echo components of different targets and different frequencies are still inconsistent in phase after phase compensation, and will be further suppressed after accumulation. Therefore, in order to simplify the subsequent analysis, these undesired reflected echo components can be ignored. 1 (t,m) is the matched filtering result y MF Next, let t = τ m,p Substitute y MF(t,m), then the peak value of the matched filtering result of the echo of the pth target in the mth reconstructed echo can be expressed as

[0093]

[0094] The phase term in the formula can be divided into three parts. 1 is the phase generated by the delay of the nth target to the mth pulse echo, ψ 2 is the phase of the intrapulse Doppler, ψ 3 It is the phase distortion caused by group delay distortion.

[0095] In RFPA radar, the carrier frequencies of different pulses change randomly, and the phase terms caused by agile radar parameters, target motion parameters and nonlinear phase distortion will also change accordingly, destroying the phase continuity between echo signals.

[0096] In order to coherently accumulate the energy of different RFPA echo pulses, it is necessary to compensate for the discontinuous phase and group delay distortion of each pulse at the same time to ensure the phase consistency of the matched filtering result. For the pth target, the phase compensation factor of the mth baseband reconstructed echo can be expressed as

[0097]

[0098] The first and second terms in formula (1.16) are used to compensate for the phase generated by the target motion parameters (range and speed), and the target motion parameters and radar agility parameters (carrier frequency and PRI) are coupled together. The third term is the compensation factor for group delay distortion.

[0099] Step 5: Construct the coherent accumulation formula including the compensation factor.

[0100] Based on the analysis of the phase compensation factor in step 4, the phase compensation factor requires the motion parameters of the target and the channel group delay parameters, and these parameters are unknown. The coherent accumulation result of the RFPA signal affected by the group delay distortion can be obtained by multi-dimensional parameter search. Its expression is:

[0101]

[0102] Where H(m,r,v,α) is the phase compensation factor, r, v and α represent the search distance, speed and group delay coefficient respectively.

[0103] Since the matched filtering result of the received signal contains the echoes of all P targets, in the coherent accumulation process of multidimensional search, if the search parameters match the motion parameters and group delay parameters of any target, the energy of the target will be effectively accumulated to form a peak. In addition, the echo pulse of each target is affected by the same group delay of the channel. The group delay estimation result of the channel can be obtained based on the coherent processing result of the strongest target in the echo, and then the group delay effect of different pulse echoes can be compensated in the coherent processing process.

[0104] Therefore, when using formula (1.17) to realize the coherent accumulation of RFPA signals, if the search parameters are consistent with the motion parameters of the strongest target in the detection area and the group delay parameters of the channel, a global peak will appear in the multidimensional space, and the estimated value of the group delay parameter can be obtained from the global peak. Next, on the distance-speed two-dimensional plane corresponding to the estimated value of the group delay parameter, the motion parameters of other targets can be obtained by searching for local peaks until the motion parameter estimation results of all targets are obtained.

[0105] In addition, the discrete form of formula (1.17) can be obtained:

[0106]

[0107] Among them, r(i), v(q) and α q (k q ) represent the discrete search units of distance, speed and qth-order group delay coefficient respectively, and their expressions are:

[0108]

[0109] In the formula, Δ r =c / 2B s , Δ v =c / (2f c MT R )and are the search intervals for distance, speed and group delay respectively, MT R is the coherent accumulation time. min 、V min and A nmin Indicates the minimum search value. N r 、N v and are the number of search units for distance, speed and group delay in multidimensional space respectively.

[0110] The above formula (1.18) is the result of the coherent accumulation after compensation. However, when using this formula for search and coherent accumulation, many parameters need to be searched, which has a high computational complexity and leads to low efficiency. Therefore, it is necessary to reduce its computational complexity. The present invention proposes two ideas to simplify the search: one is to reduce the number of search parameters (step 6), and the other is to improve the parameter search strategy (step 7).

[0111] Step 6: Simplify the coherent accumulation formula - through experimental research, determine to retain the second-order group delay coefficient, while ensuring the effect, while reducing the number of search parameters.

[0112] Based on the influence of group delay parameters, the computational complexity can be reduced by reducing the number of group delay parameters to be searched. 1 (k 1 ),α 2 (k 2 ),…,α q (k q ),…, take some factors. In order to estimate the influence of group delay coefficients of different orders, the group delay distortion is not compensated during the coherent accumulation period, and the phase compensation factor is

[0113]

[0114] The influence of different order group delay coefficients is estimated through simulation. The RFPA parameters in the simulation are shown in Table I. The initial distance and radial velocity of the target are set to 1000m and 700m / s respectively, and the signal-to-noise ratio is -10dB. Figure 1 The fitting coefficients of the cascade filter in Figure (b) are used to set the values ​​of different coefficients.

[0115] Table I RFPA radar simulation parameters

[0116]

[0117] The coherent accumulation results under the influence of first-order and second-order group delay distortion are as follows: Figure 2 , Figure 3 As shown in the figure, in order to facilitate the comparison of performance differences, the simulation results are normalized using the coherent integrated power without group delay distortion. As can be seen from the figure, the speed estimation results are very accurate, but the quality of the range image is affected. 1 The impact of Figure 2 The main lobe of the mid-range image becomes wider, and the widening is more obvious on the far side, making the main lobe asymmetric. Figure 3 In the distance estimation result, the parameter α 2 The main lobe not only has a slight deviation, but also has serious leakage to the far side, the main lobe is significantly widened, and the distance image is also seriously asymmetric.

[0118] Cascade filter α 1 The value of is much smaller than that used in the simulation and has limited impact on distance estimation, so the α in the filter can be ignored. 1 The nonlinear phase caused by this can reduce the computational complexity. In summary, the coherent accumulation algorithm can be simplified as follows:

[0119]

[0120] Then, the group delay distortion compensation factor is:

[0121]

[0122] The phase compensation factor of the mth baseband reconstructed echo can be expressed as

[0123]

[0124] Step 7: Coherent accumulation of multi-dimensional search.

[0125] The present invention adopts a more reasonable multi-dimensional search strategy to avoid unnecessary search steps and reduce the amount of calculation. From the simulation analysis of the influence of group delay distortion, it can be seen that group delay mainly affects the distance dimension image of RFPA radar, and the main lobe offset generated is limited. Therefore, the group delay parameter search can be performed only in the distance dimension neighborhood of the strongest target to obtain the estimated result. The search algorithm is a three-dimensional multi-dimensional parameter search, and the specific implementation steps are as follows: Figure 4 As shown:

[0126] Step 701: Ignore group delay distortion compensation, that is, set the group delay compensation parameter to 0, perform distance-speed compensation only, perform coherent accumulation of speed-distance two-dimensional parameter search, and obtain the speed and distance corresponding to the maximum value of coherent accumulation, recorded as v max and r max .

[0127] Step 702: Add group delay distortion compensation, with a fixed speed of v max , determine the distance r max In the distance dimension neighborhood as the center, coherent accumulation of two-dimensional parameter search of distance-group delay parameters is performed to obtain the group delay parameter corresponding to the maximum value of coherent accumulation as the group delay parameter estimation result.

[0128] Step 703: using the group delay parameter estimation result, perform coherent accumulation of the speed-distance two-dimensional parameter search again to obtain the coherent accumulation result after group distortion compensation; each peak of the coherent accumulation result is the motion parameter estimation result of the moving target.

[0129] Based on the above analysis, the present invention provides a coherent accumulation method for a random carrier frequency repetition frequency agile radar under the influence of group delay distortion. Figure 5As shown, the following steps are included:

[0130] Step 1: Recombining and matching filtering the radar echo signal;

[0131] Step 2: Coherently accumulate the matched filtering results; perform distance-speed compensation and group delay distortion compensation in the coherent accumulation operation:

[0132] Here, the phase factor of group delay distortion compensation is constructed based on the second-order coefficient of group delay and is expressed as:

[0133]

[0134] Where Δf m is the frequency agility parameter of the pulse; α 2 (k 2 ) is the second-order coefficient of the group delay response polynomial, A 2min is the minimum search value of the group delay response parameter, is the search range of the group delay response parameter, is the search interval for the group delay response parameter.

[0135] The phase factor used for distance-speed compensation is:

[0136] exp[j4πf m (r(i)-v(i)·t m ) / c]·exp{-j4πf m ·v(i)[T w / 2+2(r(i)-v(i)·t m ) / c] / c}

[0137] r(i)=R min +iΔ r ,i=0,1,...,N r -1

[0138] v(j)=V min +jΔ v ,j=0,1,...,N v -1

[0139] Among them, r(i) is the distance, v(j) is the speed; R min and V min are the minimum search values ​​for distance and speed respectively; Δ r and Δ v are the search intervals for distance and speed respectively; N r and N v are the number of search units for distance and speed respectively; f m =f c +Δfm =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where fc is the center carrier frequency, Δf is the frequency agility interval, and d(m)∈{-N f / 2,...,0,...,N f / 2} is a random sequence of frequencies, N f =B s / Δf is the number of frequency jump points, where B s is the synthetic bandwidth; t m is the slow time, T w is the pulse width and c is the speed of light.

[0140] When performing a coherent accumulation search based on the above compensation factor, the above-mentioned three-dimensional multi-dimensional parameter search can be used for implementation.

[0141] Based on the above method, the present invention also provides a coherent accumulation system for random carrier frequency re-frequency agile radar under the influence of group delay distortion, such as Figure 6 As shown, the system includes: a reorganization module, a matched filtering module and a coherent accumulation module.

[0142] The recombining module and the matched filtering module sequentially recombined and matched-filtered the radar echo signals and sent them to the coherent accumulation module;

[0143] A coherent accumulation module is used to perform coherent accumulation on the reconstructed echo received after matched filtering; and to perform distance-speed compensation and group delay distortion compensation in the coherent accumulation operation;

[0144] The phase factor for group delay distortion compensation is constructed based on the second-order coefficient of group delay and is expressed as:

[0145]

[0146] Where Δf m is the frequency agility parameter of the pulse; α 2 (k 2 ) is the second-order coefficient of the group delay response polynomial, A 2min is the minimum search value of the group delay response parameter, is the search range of the group delay response parameter, is the search interval for the group delay response parameter.

[0147] When performing distance-speed compensation, the distance-speed compensation phase factor used is:

[0148] exp[j4 π f m (r(i)-v(i)·t m ) / c]·exp{-j4πfm ·v(i)[T w / 2+2(r(i)-v(i)·t m ) / c] / c}

[0149] r(i)=R min +iΔ r ,i=0,1,…,N r -1

[0150] v(j)=V min +jΔ v ,j=0,1,…,N v -1

[0151] Among them, r(i) is the distance, v(j) is the speed; R min and V min are the minimum search values ​​for distance and speed respectively; Δ r and Δ v are the search intervals for distance and speed respectively; N r and N v are the number of search units for distance and speed respectively; f m =f c +Δf m =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where fc is the center carrier frequency, Δf is the frequency agility interval, and d(m)∈{-N f / 2,...,0,...,N f / 2} is a random sequence of frequencies, N f =B s / Δf is the number of frequency jump points, where B s is the synthetic bandwidth; t m is the slow time, T w is the pulse width and c is the speed of light.

[0152] In a preferred embodiment, the coherent accumulation module performs parameter search in the coherent accumulation process by three searches, so the coherent accumulation module includes a first search unit, a second search unit, and a third search unit.

[0153] The first search unit is responsible for the first search: group delay distortion compensation is not considered, only distance-speed compensation is performed, and coherent accumulation of speed-distance two-dimensional parameter search is performed to obtain the speed v corresponding to the maximum value of coherent accumulation max and distance r max . ;

[0154] The second search unit is responsible for the second search: adding group delay distortion compensation, with a fixed speed of v max , determine the distance r maxIn the distance dimension neighborhood centered at , coherent accumulation of two-dimensional parameter search of distance-group delay parameters is performed to obtain the group delay parameter corresponding to the maximum value of the coherent accumulation as the group delay parameter estimation result;

[0155] The third search unit is responsible for the third search: using the group delay parameter estimation result, the coherent accumulation of the speed-distance two-dimensional parameter search is performed again to obtain the coherent accumulation result after group distortion compensation; each peak of the coherent accumulation result is the motion parameter estimation result of the moving target.

[0156] Here is an example.

[0157] This example analyzes the performance of the proposed coherent integration algorithm through the simulation of a moving target. The cascade filter is used to simulate the group delay characteristics of the radar channel. The search range of the group delay coefficient is [0,1e-3], and the search interval is The parameters of the three targets are set as follows: the speed, distance and signal-to-noise ratio of target 1 are v 0 =700m / s, R 0 =1000m, SNR=-10dB; the speed, distance and signal-to-noise ratio of target 2 are v 1 =700m / s, R 1 =1001.5m, SNR=-20dB; the speed, distance and signal-to-noise ratio of target 3 are v 2 =670m / s,R 2 =988m, SNR=-10dB. Other simulation parameters are the same as those in Table I.

[0158] As a comparison, the coherent accumulation results without compensating for group delay distortion are obtained using the RFPA-RFT method proposed in the literature, such as Figure 7 As shown in Figure 1. It can be seen that the group delay distortion makes it difficult to distinguish the real target from the distance estimation result. At the same time, target 2 is completely covered by the main lobe energy leakage of target 1. The coherent accumulation result obtained by the algorithm proposed by the present invention is shown in Figure 1. Figure 8 The estimated result of the second-order group delay coefficient is 3e -4 The error is 2.7e -5 , which is less than a search interval. The energy of the three targets is Figure 8 In Figure (a), the effective accumulation is three obvious peaks, and the target motion parameter estimation effect is good. In addition, the main lobe offset, main lobe widening and side lobe rise problems caused by group delay distortion have been significantly improved.

[0159] The above specific embodiments only describe the design principle of the present invention. The shapes and names of the components in the description may be different and are not limited. Therefore, those skilled in the art in the field of the present invention may modify or replace the technical solutions recorded in the above embodiments; and these modifications and replacements do not deviate from the creative purpose and technical solutions of the present invention and should all fall within the protection scope of the present invention.

Claims

1. A coherent integration method for random carrier frequency re-frequency agile radar under the influence of group delay distortion, characterized in that: include: Coherently accumulate the radar echo signals after reorganization and matched filtering; Perform distance-velocity compensation and group delay distortion compensation in coherent integration operation; The phase factor of the group delay distortion compensation is constructed based on the second-order coefficient of the group delay and is expressed as: Where Δf m is the frequency agility parameter of the pulse; α2(k2) is the second-order coefficient of the group delay response polynomial, A 2min is the minimum search value of the group delay response parameter, is the search range of the group delay response parameter, The search interval for the group delay response parameter.

2. The method according to claim 1, characterized in that The coherent accumulation is implemented by three multi-dimensional parameter searches: First search: group delay distortion compensation is not considered, only distance-speed compensation is performed, and coherent accumulation of speed-distance two-dimensional parameter search is performed to obtain the speed v corresponding to the maximum value of coherent accumulation max and distance r max ; Second search: add group delay distortion compensation, fixed speed v max , determine the distance r max In the distance dimension neighborhood centered at , coherent accumulation of two-dimensional parameter search of distance-group delay parameters is performed to obtain the group delay parameter corresponding to the maximum value of the coherent accumulation as the group delay parameter estimation result; The third search: using the group delay parameter estimation result, the coherent accumulation of the speed-distance two-dimensional parameter search is performed again to obtain the coherent accumulation result after group distortion compensation; The peaks of the coherent accumulation results are the motion parameter estimation results of the moving target.

3. The method according to claim 1 or 2, characterized in that The phase factor used in the distance-speed compensation is: exp[j4πf m (r(i)-v(i)·t m ) / c]·exp{-j4πf m ·v(i)[T w / 2+2(r(i)-v(i)·t m ) / c] / c} r(i)=R min +iΔ r ,i=0,1,...,N r -1 v(j)=V min +jΔ v ,j=0,1,...,N v -1 Among them, r(i) is the distance, v(j) is the speed; R min and V min are the minimum search values ​​for distance and speed respectively; Δ r and Δ v are the search intervals for distance and speed respectively; N r and N v are the number of search units for distance and speed respectively; f m =f c +Δf m =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where f c is the center carrier frequency, Δf is the frequency agility interval, d(m)∈{-N f / 2,...,0,...,N f / 2} is a random sequence of frequencies, N f is the number of frequency jump points; t m is the slow time, T w is the pulse width and c is the speed of light.

4. A coherent integration system for random carrier frequency re-frequency agile radar under the influence of group delay distortion, characterized in that: include: Recombination module, matched filter module and coherent accumulation module; The recombining module and the matched filtering module sequentially recombined and matched-filtered the radar echo signals and sent them to the coherent accumulation module; The coherent accumulation module is used to perform coherent accumulation on the received reconstructed echo after matched filtering; Perform distance-velocity compensation and group delay distortion compensation in coherent integration operation; The phase factor of the group delay distortion compensation is constructed based on the second-order coefficient of the group delay and is expressed as: Where Δf m is the frequency agility parameter of the pulse; α2(k2) is the second-order coefficient of the group delay response polynomial, A 2min is the minimum search value of the group delay response parameter, is the search range of the group delay response parameter, is the search interval for the group delay response parameter.

5. The system according to claim 4, characterized in that The coherent accumulation module includes a first search unit, a second search unit, and a third search unit; The first search unit is responsible for the first search: without considering group delay distortion compensation, only performing distance-speed compensation, performing coherent accumulation of speed-distance two-dimensional parameter search, and obtaining the speed v0 and distance r0 corresponding to the maximum value of the coherent accumulation; The second search unit is responsible for the second search: adding group delay distortion compensation, fixing the speed as v0, determining the distance dimension neighborhood centered on the distance r0, performing coherent accumulation of the two-dimensional parameter search of the distance-group delay parameter, and obtaining the group delay parameter corresponding to the maximum value of the coherent accumulation as the group delay parameter estimation result; The third search unit is responsible for the third search: using the group delay parameter estimation result, performing coherent accumulation of the speed-distance two-dimensional parameter search again to obtain the coherent accumulation result after group distortion compensation; The peaks of the coherent accumulation results are the motion parameter estimation results of the moving target.

6. The system according to claim 4, characterized in that, When the coherent accumulation module performs distance-speed compensation, the distance-speed compensation phase factor used is: exp[j4πf m (r(i)-v(i)·t m ) / c]·exp{-j4πf m ·v(i)[T w / 2+2(r(i)-v(i)·t m ) / c] / c} r(i)=R min +iΔ r ,i=0,1,...,N r -1 v(j)=V min +jΔ v ,j=0,1,...,N v -1 Among them, r(i) is the distance, v(j) is the speed; R min and V min are the minimum search values ​​for distance and speed respectively; Δ r and Δ v are the search intervals for distance and speed respectively; N r and N v are the number of search units for distance and speed respectively; f m =f c +Δf m =f c +d(m)Δf is the random carrier frequency of the transmitted pulse, where f c is the center carrier frequency, Δf is the frequency agility interval, d(m)∈{-N f / 2,...,0,...,N f / 2} is a random sequence of frequencies, N f =B s / Δf is the number of frequency jump points, where B s is the synthetic bandwidth; t m is the slow time, T w is the pulse width and c is the speed of light.