Radar device and method for processing its radar signals

Abstract

Even when mounted on an aircraft, it suppresses clutter and accurately detects slow-moving targets. [Solution] In the radar device of the embodiment, transmission pulses of N (N>1) pulses are transmitted and received, the received signal is acquired in CPI units, L (L≧1) zeros are filled in on the slow-time axis, then an FFT is performed on the slow-time axis to convert it to the Doppler axis, the amplitude is added on the fast-time axis, the Doppler frequency that is maximum on the slow-time axis is extracted, the signal on the slow-time axis before zero filling is multiplied by the conjugate value of the phase gradient vector for the maximum Doppler frequency, then an M (M≧1) order approximation curve is calculated from the complex signal after multiplication by the conjugate value on the slow-time axis, and low-frequency signal components are suppressed by subtracting this from the original signal, then the peak of the phase gradient vector is multiplied, an FFT is performed on the slow-time axis to convert it to a Doppler axis signal, and the target is detected by CFAR or the like.

Description

Technical Field

[0005] , , , ,

[0001] This embodiment relates to a radar device and a radar signal processing method thereof.

Background Art

[0002] In a conventional radar device mounted on an aircraft or the like, since the center frequency of clutter from the ground or sea surface deviates from 0 Doppler, in a normal MTI (Moving Target Indicator, see Non-Patent Document 1), there is a problem that clutter cannot be suppressed. Also, in MTI, there is a problem that the clutter null width becomes wide and low-speed targets near the clutter cannot be detected.

[0003] If the clutter center frequency is near 0 Doppler, there is also a method (see Patent Document 1) of extracting an approximate curve on the slow-time axis to form a steep frequency filter to suppress clutter, but it does not deal with moving clutter.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Non-Patent Documents

[0005]

Non-Patent Document 1

Non-Patent Document 2

Non-Patent Document 3

Summary of the Invention

[0006] As described above, conventional aircraft-mounted radar systems have a problem in that they cannot suppress clutter with normal MTI because the center frequency of clutter from the ground or sea surface is shifted from 0 Doppler. In addition, MTI has the problem that the clutter null width becomes wide, making it impossible to detect low-speed targets in the vicinity of the clutter.

[0007] This embodiment has been made in view of the above-mentioned problems, and aims to provide a radar device and a radar signal processing method thereof that can suppress clutter and accurately detect low-speed targets even when mounted on an aircraft. [Means for solving the problem]

[0008] To solve the above problems, the radar device according to the first embodiment transmits and receives N (N>1) pulse transmission pulses and receives the received signal as CPI (Coherent Pulse). This system captures data in interval units to detect targets. In the case of chirp pulses, the data is compressed on the fast-time axis, L (L≧1) zeros are filled on the slow-time axis, then an FFT is performed on the slow-time axis to convert it to the Doppler axis, the amplitudes are added on the fast-time axis, the maximum Doppler frequency fdpeak is extracted on the slow-time axis, the signal on the slow-time axis before zero filling is multiplied by the conjugate value of the phase gradient vector peak phase_peak corresponding to the maximum Doppler frequency fdpeak, an M (M≧1) order approximation curve is calculated from the complex signal after multiplication by the conjugate value on the slow-time axis, and this is subtracted from the original signal to suppress low-frequency signal components (clutter, proximity reflection, etc.), then the peak phase_peak of the phase gradient vector is multiplied, an FFT is performed on the slow-time axis to convert it to the Doppler axis signal, and the target is detected by CFAR, etc.

[0009] In other words, according to the first embodiment, even when the clutter is not zero Doppler, the clutter Doppler frequency is extracted with high precision, and the clutter frequency on the slow-time axis is moved to zero Doppler by a corresponding phase gradient, then suppressed by subtraction using an approximation curve, and returned to the original Doppler axis. This suppresses the clutter with a steep filter, making it possible to detect even slow targets.

[0010] Furthermore, the radar device according to the second embodiment, in addition to the configuration of the first embodiment, generates a Σ sequence from the received signal on the slow-time axis with N cells left as they are, and a Δ sequence by dividing the N cells in the middle and inverting the sign of one side. After filling each with L (L≧1) zeros, it performs an FFT to convert to the Doppler axis, calculates the Σ signal and Δ signal by adding the amplitudes on the fast-time axis, and extracts the Doppler frequency fdpeak in which the absolute value of Δ / Σ is minimized on the slow-time axis among the cells in which the Σ signal exceeds a predetermined amplitude threshold. Then, the signal on the slow-time axis before zero-padding is multiplied by the conjugate value of the phase gradient vector peak phase_peak for the smallest Doppler frequency fdpeak. After that, an M (M≧1) order approximation curve is calculated from the complex signal on the slow-time axis and subtracted from the original signal to suppress low-frequency signal components (clutter, proximity reflections, etc.). After multiplying by the phase gradient vector peak phase_peak, an FFT is performed on the slow-time axis to obtain the Doppler axis signal, and the target is detected using CFAR or similar.

[0011] In other words, according to the second embodiment, even when the clutter is not zero Doppler, by using Σ and Δ signals, the clutter-Doppler frequency can be extracted with higher accuracy than when using Σ alone. Then, by shifting the clutter frequency on the slow-time axis to zero Doppler using the corresponding phase gradient, it is suppressed by subtracting it using an approximation curve and returned to the original Doppler axis. This suppresses the clutter with a steep filter, making it possible to detect even slow targets.

[0012] As described above, the radar device according to this embodiment can detect slow-moving targets by suppressing clutter even when the clutter's center frequency is other than 0 Doppler, by calculating an M-order (M≧1) approximation curve from the complex signal on the slow-time axis and subtracting it from the original signal. [Brief explanation of the drawing]

[0013] [Figure 1] Figure 1 is a block diagram showing the configuration of a radar device according to the first embodiment. [Figure 2] Figure 2 is a flowchart showing the flow of the target detection process after signal transmission and reception in the first embodiment. [Figure 3] Figure 3 is a waveform diagram showing the timing of the transmission pulses for each PRI generated in the transmission system in the first embodiment. [Figure 4] Figure 4 shows two-dimensional data of fast-time and slow-time between PRIs for each CPI data range cell based on PRI data of the transmitted pulse, in the first embodiment. [Figure 5] Figure 5 is a waveform diagram showing how the center frequency of the clutter is extracted by increasing the resolution of the signal for each range cell converted to the Doppler axis in the first embodiment. [Figure 6] Figure 6 is a waveform diagram illustrating the process of suppressing extracted clutter and detecting targets in the first embodiment. [Figure 7] Figure 7 is a waveform diagram showing the process of detecting targets near the clutter by suppressing moving clutter with a steep filter in the first embodiment. [Figure 8] Figure 8 is a block diagram showing the configuration of a radar device according to the second embodiment. [Figure 9] Figure 9 is a flowchart showing the flow of the target detection process after signal transmission and reception in the second embodiment. [Figure 10] Figure 10 is a waveform diagram showing how the center frequency of the clutter is extracted by increasing the resolution of the signal for each range cell converted to the Doppler axis in the second embodiment.

Best Mode for Carrying Out the Invention

[0014] Hereinafter, embodiments will be described with reference to the drawings. In the description of each embodiment, the same parts are denoted by the same reference numerals and redundant descriptions are omitted.

[0015] (First Embodiment) Referring to FIGS. 1 to 7, a radar device according to the first embodiment will be described. FIG. 1 is a block diagram showing the configuration of the radar device, FIG. 2 is a flowchart showing the flow of target detection processing after signal transmission and reception, FIG. 3 is a waveform diagram showing the timing of transmission pulses for each PRI generated in the transmission system, FIG. 4 is a diagram showing two-dimensional data of fast-time between PRIs and slow-time for each CPI data range cell according to the PRI data of the transmission pulse, FIG. 5 is a waveform diagram showing a state of extracting the center frequency of clutter by high-resolution of signals for each range cell converted to the Doppler axis in the first embodiment, FIG. 6 is a waveform diagram for explaining a process of suppressing the extracted clutter and detecting a target, and FIG. 7 is a waveform diagram showing a process of suppressing a moving clutter with a steep filter and detecting a target near the clutter.

[0016] First, in FIG. 1, in the transmission system, a transmission signal generator 11 generates a transmission signal such as a modulation pulse (FIG. 2), converts it into an analog signal by a DA converter 12, converts it into a high-frequency (RF) signal by a frequency converter 13, amplifies the power by a high-output amplifier 14, and transmits it by an antenna 16 via a circulator 15.

[0017] In the receiving system, reflected signals from targets, etc., are captured by the antenna 16, transmitted and received are separated by the circulator 15, noise is reduced and the signals are amplified by the low-noise amplifier 17, then the frequency is converted to baseband by the frequency converter 18, and the signal is converted to a digital signal by the AD converter 19. Received data (PRI data, PRI: Pulse Repetition Interval) of N (N≧1) hit transmitted pulses is acquired in CPI data units by the CPI (Coherent Pulse Interval) input 20, and the following target detection processing is performed.

[0018] First, the fast-time axis pulse compression unit 21 pulse-compresses the signal on the fast-time axis, the slow-time axis zero-padding unit 22 converts it to a signal on the slow-time axis and zero-paddings it, and the slow-time axis FFT unit 23 performs FFT processing on the received signal (data) on the slow-time axis. Next, the range axis amplitude addition unit 24 adds the amplitudes on the range axis, the clutter center frequency extraction unit 25 extracts the clutter center frequency, the phase gradient calculation unit 26 calculates the phase gradient of the received signal along the extracted clutter, the inverse phase gradient multiplication unit 27 multiplies the calculated phase gradient by the inverse phase gradient (conjugate phase), the approximation curve calculation unit 28 calculates an approximation curve of the inverse phase gradient multiplication result, the approximation curve subtraction unit 29 subtracts the calculated approximation curve from the original received signal, and then the phase gradient multiplication unit 30 finds the phase gradient of the received signal and multiplies it. Subsequently, the slow-time axis FFT unit 31 returns to the original Doppler axis, CFAR 32 performs target detection using CFAR, and the distance measurement / angle measurement unit 33 generates and outputs three-dimensional information of the target.

[0019] In the radar system with the above configuration, as shown in Figure 2, the received data of N (N≧1) hit transmitted pulses is acquired in CPI data units (step S11), the fast-time axis signal is pulse-compressed (step S12), converted to a slow-time axis signal and zero-padding is performed (step S13), and the slow-time axis received signal (data) is subjected to FFT processing (step S14). Subsequently, amplitude summation is performed on the range axis (step S15), the center frequency of the clutter is extracted (step S16), the phase gradient of the received signal is calculated along the extracted clutter (step S17), the inverse phase gradient of the calculated phase gradient is multiplied with the pulse-compressed signal (conjugate phase) (step S18), an approximation curve of the inverse phase gradient multiplication result is calculated (step S19), the calculated approximation curve is subtracted from the original received signal (step S20), and the phase gradient of the received signal is found and multiplied (step S21). Subsequently, the image is returned to the original Doppler axis using a slow-time axis FFT (step S22). The processes described in steps S19 to S22 are performed on all range cells in the observation region (steps S23, S24), target detection is performed using CFAR, and three-dimensional information of the target is generated and output (step S25).

[0020] In the target detection process described above, the CPI data is two-dimensional data consisting of fast-time data for each range cell and slow-time data between PRIs, as shown in Figure 4. The slow-time axis signal for each range cell is shown in Figure 5(a). Depending on the range cell, this signal includes the target signal and clutter components. In the case of ground clutter, especially in ground-based radar, the clutter component is a low-frequency component near the Doppler frequency of 0. However, in the case of aircraft-mounted radar, the relative velocity is different from 0 m / s, so the center velocity of the clutter is different from 0 m / s, and the Doppler frequency deviates from 0. If the relative velocity can be obtained, that velocity can be used, but in this embodiment, the relative velocity is extracted from the transmitted and received signals.

[0021] First, if the transmission is a chirp pulse, pulse compression is performed on the fast-time axis of the CPI data to increase the signal-to-noise ratio (SNR) (see Non-Patent Literature 2). Next, in order to extract the clutter center frequency with high accuracy, as shown in Figure 5(b), the CPI data is converted to the slow-time axis, zero-padding is performed, and then an FFT is performed to convert it to the Doppler axis, thereby achieving pseudo-high resolution of the range-Doppler signal.

[0022]

number

[0023] Using this range-Doppler signal RDmatΣfft, amplitude summing is performed along the range axis to improve the signal-to-noise ratio (SNR). Next, the Doppler frequency fdpeak, which maximizes the slow-time axis, is extracted. From this, the phase gradient of the slow-time axis corresponding to fdpeak can be calculated.

[0024]

number

[0025] Next, the clutter frequency is shifted to 0 Doppler by multiplying the original slow-time axis range-Doppler signal RDmat, before zero-padding, by the conjugate of the phase gradient (inverse phase gradient).

[0026]

number

[0027] Next, the slow-time axis signal (complex signal) for each range cell is extracted, and a low-order approximation curve is calculated as shown in Figure 6(b). This approximation curve can be determined by determining the coefficients using, for example, the least squares method with the polynomial approximation formula shown in the following equation.

[0028]

number

[0029] In this case, increasing the order M can widen the null width of the clutter, but it will also suppress the target signal. Therefore, it is necessary to set the order to a low level so as not to suppress the intended target signal.

[0030] Therefore, by subtracting the approximation calculated in equation (5) from the original signal, the clutter of the low-frequency component is suppressed, as shown in Figure 6(c). This is repeated for each range cell (fast-time axis cell) to generate the CPI range-Doppler signal RDmatcal. By multiplying this CPI range-Doppler signal RDmatcal by the phase gradient matrix and performing a slow-time axis FFT, a signal is obtained that suppresses clutter and returns to the original Doppler frequency, as shown in Figure 6(d).

[0031]

number

[0032] Using this range-Doppler signal RDmatcal, target detection processing is performed using CFAR (Non-Patent Document 3), etc. The detected range-Doppler cells are then processed for distance and angle measurement and output as target information.

[0033] As a result of the above process, even when a target is located near moving clutter, as shown in Figure 7(a), the low-frequency component suppression (approximation curve subtraction) process suppresses the moving clutter with a steep filter, as shown in Figure 7(b), making it possible to detect targets near the clutter.

[0034] As described above, according to the radar device of the second embodiment, even when the clutter is not zero Doppler, the clutter Doppler frequency is extracted with high precision, the clutter frequency on the slow-time axis is moved to zero Doppler by a corresponding phase gradient, then suppressed by subtraction using an approximation curve, and returned to the original Doppler axis, thereby suppressing the clutter with a steep filter and enabling detection of even low-speed targets.

[0035] (Second embodiment) A radar device according to a second embodiment will be described with reference to Figures 8 to 10. In the first embodiment, a method for extracting the clutter center frequency from the maximum value of a Σ signal with high resolution on the slow-time axis was described. In this embodiment, a method for extracting the clutter center frequency, which greatly affects clutter suppression performance, with even greater precision is described.

[0036] Figure 8 is a block diagram showing the configuration of the radar device according to the second embodiment, Figure 9 is a flowchart showing the flow of target detection processing after signal transmission and reception in the second embodiment, and Figure 10 is a waveform diagram showing how the center frequency of clutter is extracted by increasing the resolution of the signal for each range cell converted to the Doppler axis in the second embodiment. Note that in Figures 7 and 8, the same parts as in Figures 1 and 2 are denoted by the same reference numerals, and the process up to pulse compression of the transmitted and received data on the fast-time axis of the CPI data is the same as in the first embodiment and is therefore omitted.

[0037] A key feature of the radar system according to this embodiment, as shown in Figure 8, is that in the receiving system, the slow-time axis Δ signal generation unit 34 generates a slow-time axis Δ signal from the output of the fast-time axis pulse compression unit 21, the slow-time axis zero-filling unit 35 zeros out unnecessary parts, the low-time axis FFT unit 36 ​​performs FFT processing, the range axis amplitude summing unit 39 performs range axis amplitude summing, and the data is sent to the Δ / Σ calculation unit 40. Meanwhile, from the range axis amplitude summing result of the Σ signal obtained by the range axis amplitude summing unit 24, the clutter center frequency candidate extraction unit 41 extracts candidate center frequencies for moving clutter, the Δ / Σ calculation unit 40 extracts the Doppler cell with the minimum Δ / Σ from among the candidates, the data is sent to the clutter center frequency extraction unit 25 to extract the clutter center frequency for the extracted Doppler cell, and the process thereafter is the same as in the first embodiment.

[0038] As explained in the flowchart shown in Figure 9, a slow-time axis Δ signal is generated from the fast-time axis pulse compression output (step S26), unnecessary parts are zero-filled (step S27), FFT processing is performed (step S28), and amplitude summing of the range axis is performed (step S29). Meanwhile, candidate center frequencies of the moving clutter are extracted from the amplitude summing result of the range axis of the Σ signal (step S31), and the Doppler cell with the minimum Δ / Σ is extracted from among the candidates (step S30), and the process moves to clutter center frequency extraction processing (step S16), and thereafter processing is carried out in the same manner as in the first embodiment.

[0039] In the radar system with the above configuration, a delta signal is generated using the slow-time axis signal in order to extract the clutter center frequency with high accuracy. For this purpose, the N cells of the slow-time axis are divided into two, and the polarity of one of them is reversed (the sign of one side is set to -).

[0040]

number

[0041] Figure 10(a) shows the slow-time axis signals for each range cell. Depending on the range cell, these signals include the target signal and clutter components. Therefore, as shown in Figure 10(b), a pseudo-high-resolution Δ signal is generated by zero-padding the slow-time axis data and then performing an FFT.

[0042]

number

[0043] By using this Δ range-Doppler signal RDmatΔ_fft and performing amplitude summing along the range axis (RDmatΔ_fftsum), the signal-to-noise ratio (SNR) is improved. On the other hand, using the methods of equations (1) and (2) of the first embodiment, a Σ range-Doppler signal RDmatΣ_fft can be obtained, and further, a signal obtained by summing the amplitudes along the range axis (RDmatΣ_fftsum) can be obtained.

[0044] First, we extract Doppler cells where RDmatΔ_fftsum exceeds the amplitude threshold, and within those cells, we extract the Doppler cell that minimizes the following equation, and define it as the clutter center frequency (clutter peak) fdpeak.

[0045]

number

[0046] By using this highly accurate clutter center frequency fdpeak and performing the processing described in equations (3) to (6), clutter can be suppressed.

[0047] As described above, according to the radar device of the second embodiment, even when the clutter is not zero Doppler, by using Σ and Δ signals, the clutter-Doppler frequency can be extracted with higher accuracy than when Σ is used alone. Then, by shifting the clutter frequency on the slow-time axis to zero Doppler using a corresponding phase gradient, it is suppressed by subtracting it using an approximation curve and returned to the original Doppler axis. This suppresses the clutter with a steep filter, making it possible to detect even low-speed targets.

[0048] It should be noted that the present invention is not limited to the above embodiments, and the components can be modified and implemented in practice without departing from the spirit of the invention. Furthermore, various inventions can be formed by appropriately combining the multiple components disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiments. Moreover, components from different embodiments may be appropriately combined. [Explanation of symbols]

[0049] 11...Transmit signal generator, 12...DA converter, 13...Frequency converter, 14...High-power amplifier, 15...Circulator, 16...Antenna, 17...Low-noise amplifier, 18...Frequency converter, 19...AD converter, 20...CPI input, 21...Fast-time axis pulse compression unit, 22...Slow-time axis zero-filling unit, 23...Slow-time axis FFT unit, 24...Range axis amplitude summing unit, 25...Clutter center frequency extraction unit, 26 ...Phase gradient calculation section, 27...Inverse phase gradient multiplication section, 28...Approximate curve calculation section, 29...Approximate curve subtraction section, 30...Phase gradient multiplication section, 31...Slow-time axis FFT section, 32...CFAR, 33...Distance measurement / angle measurement section, 34... slow-time axis Δ signal generation section, 35... slow-time axis zero filling section, 36... low-time axis FFT section, 39... range axis amplitude addition section, 40... Δ / Σ calculation section, 41... clutter center frequency candidate extraction section.

Claims

1. In a radar system that transmits and receives N (N>1) pulses at predetermined intervals and acquires the received signal in units of CPI (Coherent Pulse Interval) to detect a target, A means for zero-padding L (L≧1) cells in a Σ sequence signal, with N cells left as they are, from the received signal in CPI units, along the slow-time axis, A means for performing an FFT on the slow-time axis of the zero-filled Σ sequence signal and converting it to the Doppler axis, Means for amplitude summing the Σ sequence signals converted to the Doppler axis on the fast-time axis, A means for determining the clutter center frequency by extracting the Doppler frequency that is maximum on the slow-time axis from the result of amplitude summing of the aforementioned Σ sequence of signals, A means for multiplying the signal of the Σ sequence on the slow-time axis before zero-padding by the conjugate value of the peak of the phase gradient vector with respect to the maximum Doppler frequency, A means for calculating an M-order (M≧1) approximation curve from the complex signal obtained by multiplying the conjugate value of the slow-time axis, and for suppressing low-frequency signal components by subtracting it from the original signal, A means for multiplying the suppression result of the signal component by the peak of the phase gradient vector, A means for performing an FFT on the slow-time axis and converting it into a signal on the Doppler axis, means for detecting a target from the received signal converted to the Doppler axis, A radar device equipped with the following.

2. The radar device according to claim 1, further comprising means for compressing the received signal along the fast-time axis when the transmitted pulse is a chirp pulse, and for zero-padding the compressed received signal L (L≧1) times along the slow-time axis.

3. Furthermore, the means for generating the Σ sequence signal with the N cells left as they are, and the Δ sequence signal by dividing the N cells in half at the center and inverting the sign of one side, from the received signal on the slow-time axis, Means for zero-padding L (L≧1) signals of the Δ sequence along the slow-time axis, A means for performing an FFT on the slow-time axis to convert the zero-filled Δ sequence signal into a Doppler axis, Means for amplitude summing the Δ-sequence signals converted to the Doppler axis on the fast-time axis, A means for determining the clutter center frequency by extracting the Doppler frequency in which the absolute value of Δ / Σ is minimized on the slow-time axis, from the amplitude sum of the Σ sequence signal and the Δ sequence signal, in the cells where the Σ sequence signal exceeds a predetermined amplitude threshold. A means for multiplying the signal on the slow-time axis before zero-padding the Δ sequence signal by the conjugate value of the peak of the phase gradient vector for the minimum Doppler frequency, Equipped with, From the complex signal on the slow-time axis, an M-order (M≧1) approximation curve is calculated and subtracted from the original signal to suppress low-frequency signal components. After multiplying by the phase gradient vector, an FFT is performed on the slow-time axis to obtain the Doppler axis signal and detect the target. The radar device according to claim 1.

4. In a radar signal processing method for a radar device that transmits and receives N (N>1) pulses at predetermined intervals and acquires the received signals in units of CPI (Coherent Pulse Interval) to detect a target, From the received signal in CPI units, the Σ sequence signal, with N cells left as is, is padded with L (L≧1) zeros along the slow-time axis. The zero-filled Σ sequence signal is subjected to FFT along the slow-time axis and converted to the Doppler axis. The Σ sequence signal, converted to the Doppler axis, is amplitude-summed along the fast-time axis. From the results of amplitude summing of the aforementioned Σ sequence signals, the Doppler frequency that is maximum on the slow-time axis is extracted to determine the clutter center frequency. The signal of the Σ sequence on the slow-time axis before zero-padding is multiplied by the conjugate value of the peak of the phase gradient vector with respect to the maximum Doppler frequency. From the complex signal obtained by multiplying the conjugate value of the slow-time axis, an M-order (M≧1) approximation curve is calculated and subtracted from the original signal to suppress low-frequency signal components. The suppression result of the signal component is multiplied by the phase gradient vector. The signal is converted to a Doppler axis signal by performing an FFT on the aforementioned slow-time axis. The target is detected from the received signal converted to the Doppler axis. A method for processing radar signals from a radar system.

Images