Radar system and radar signal processing method

Abstract

Even with a continuous observation method using SIMO, integration loss due to beamwalks can be reduced on a small scale, improving detection performance. [Solution] In the radar system according to the embodiment, for transmission, a transmission beam is formed to cover the observation range using an array antenna having Naz × Nel (2 or more of either) sub-array elements, and for reception, the observation range is observed for a long period of time using Nrx multi-beams. As a result, if a beam walk occurs in which the target moves in the beam direction within the integration time of the slow-time axis, the Naz × Nel-Nrx signals other than the number of multi-beams Nrx of the received signal are zero-filled, the zero-filled signal is inverse FFT obtained to obtain the phase of the array element signal, a correlation calculation is performed between the reference signal, which is a type M target motion model, and the array element signal, and a search is performed to extract the maximum value of the beam obtained by slow-time axis FFT from the calculation result, and then the beam walk is corrected on the array axis to detect the target.

Description

Technical Field

[0004] , , , , , , , , ,

[0005] , , , ,

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

Background Art

[0002] In a conventional radar system, when a target is to be constantly identifiable, for the observation range, transmission forms a fan beam and reception forms a multi-beam to constantly observe the observation range, adopting a SIMO (Single Input Multiple Output) method. Since this SIMO method can constantly observe the observation space, it has the advantage of being able to observe at a high data rate. However, since the transmission is a fan beam, in order to ensure a predetermined SN (Signal to Noise), a long integration time is required. Furthermore, when the number of integration hits is large, in addition to range walk and Doppler walk, there is a problem of integration loss due to beam walk.

[0003] For this countermeasure, for example, a method by integral series maximization (velocity / acceleration correction) (see Patent Document 1) and a method of performing high-order correction on the range frequency axis (Non-Patent Document 4) have been proposed. These methods are search methods for velocity and acceleration, and reduce integration loss by maximizing the integral series. However, even with these methods, there is no method for correcting beam walk.

Prior Art Documents

Patent Documents

[0004] [[ID=​​​​​​​​​​​​​​​Pulse compression: Ouchi, 'Fundamentals of Synthetic Aperture Radar for Remote Sensing', Tokyo Denki University Press, pp. 131-149 (2003) [Non-Patent Document 2] CFAR (Constant False Alarm Rate): Yoshida, 'Revised Radar Technology', Institute of Electronics, Information and Communication Engineers, pp. 87-89 (1996) [Non-Patent Document 3] Monopulse angle measurement: Yoshida, 'Revised Radar Technology', Institute of Electronics, Information and Communication Engineers, pp. 260-264 (1996) [Non-Patent Document 4] High-order correction method: Penghui Huang,'Long-Time Coherent Integration for Weak Maneuvering Target Detection and High-Order Motion Parameter Estimation Based on Keystone Transform', IEEE Trans. On SIGNAL PROCESSING, VOL.64, NO.15, AUGUST 1, pp.4013-4026(2016) [Non-Patent Document 5] PGA (Phase gradient autofocus) method: Charles V. Jakowatz, 'Spotlight-Mode Synthetic Aperture Radar: A Signal Processing Approach', Springer, pp.251-256 (1996) [Non-Patent Document 6] DBSCAN (Density-based Spatial Clustering of Applications with Noise): Sebastian Raschka, 'Python Machine Learning Programming', Impress, pp. 319-323 (2016) [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] As described above, conventional SIMO-based radar systems that use a continuous observation method lack effective methods to address integration losses due to beam walks during long-term integration, resulting in a problem of reduced detection performance.

[0007] The objective of this embodiment is to provide a radar system and radar signal processing method that can reduce integration loss due to beamwalks and improve detection performance with a small processing scale, even in a continuous observation method using SIMO. [Means for solving the problem]

[0008] To solve the above problems, the radar system according to the embodiment forms a transmission beam to cover the observation range using an array antenna having Naz × Nel (Naz ≥ 1, Nel ≥ 1, or at least 2) sub-array elements for transmission, and forms Nrx (Nrx ≥ 1) multi-beams (SIMO: Single Input Multiple Output) for reception, in a radar system that observes the observation range for a long period of time. When a beam walk occurs in which the target moves in the beam direction within the integration time of the slow-time axis, the Naz×Nel-Nrx signals other than the number of multi-beams Nrx of the received signal are zero-filled, the zero-filled signal is inverse FFT to obtain the phase of the array element signal, M (M≧1) types of target motion models are set as reference signals and correlated with the array element signal, and a search is performed to extract the maximum value of the beam obtained by slow-time axis FFT from the correlation calculation results, and then the beam walk is corrected on the array axis to detect the target.

[0009] With the above configuration, beamwalks during long-term integration are corrected using the array axes, making it possible to correct the target without detection that identifies the target on the AZ and EL axes, thus enabling high signal-to-noise ratio (SNR) target detection.

[0010] Furthermore, the radar system according to the above embodiment also uses the array signal after beamwalk correction to form a monopulse beam and measure the angle of the target.

[0011] With the above configuration, a monopulse beam is formed using the beamwalk-corrected array signal, enabling high signal-to-noise ratio and high-precision angle measurement.

[0012] Furthermore, the radar system according to the above embodiment further performs cluster analysis for each beam using the coordinates of the beams that have been detected, extracts the beam range in which a target candidate exists based on the results of the cluster analysis, constrains the target candidate range by the target candidate range, constrains the M(M≧1) type target motion model by the target candidate range, performs an inverse FFT on the Naz×Nel-Nrx signal other than the multi-beam number Nrx with zeros filled, sets the M(M≧1) type target motion model constrained by the target candidate range as a reference signal for the phase of the array element signal obtained thereby, performs a correlation calculation between the reference signal and the array element signal, and then detects the target by a search method that extracts the maximum value of the beam after slow-time axis FFT.

[0013] According to the above configuration, a target motion model is set for the range in which the target exists in order to correct beamwalks during long-term integration. This makes it possible to limit the search range and correct with a small processing scale, enabling target detection with a high signal-to-noise ratio. [Brief explanation of the drawing]

[0014] [Figure 1] Figure 1 is a block diagram showing the configuration of the transmission and reception systems of a radar system according to the first embodiment. [Figure 2] Figure 2 is a conceptual diagram showing the transmit and receive beam shapes relative to the observation range of the radar system shown in Figure 1. [Figure 3] Figure 3 is a flowchart showing the target identification process of the radar system shown in Figure 1. [Figure 4] Figure 4 is a conceptual diagram showing the beamwalk correction process in the radar system shown in Figure 1. [Figure 5]FIG. 5 is a diagram for explaining a method of performing beam walk correction with respect to the AZ axis in the radar system shown in FIG. 1. [Figure 6] FIG. 6 is a diagram showing a state of data extraction for performing provisional detection and correction processing in the radar system shown in FIG. 1. [Figure 7] FIG. 7 is a diagram for explaining a method of performing beam walk correction with respect to the AZ axis and the EL axis in the radar system shown in FIG. 1. [Figure 8] FIG. 8 is a block diagram showing the configuration of a transmission system and a reception system of a radar system according to the second embodiment. [Figure 9] FIG. 9 is a diagram showing a method of angle measurement using monopulse beams of Σ, ΔAZ, and ΔEL in the radar system shown in FIG. 8. [Figure 10] FIG. 10 is a diagram showing a coordinate system for explaining angle measurement processing in the beam direction in the radar system shown in FIG. 8. [Figure 11] FIG. - is a block diagram showing the configuration of a transmission system and a reception system of a radar system according to the third embodiment. [Figure 12] FIG. 12 is a flowchart showing target detection processing in which a countermeasure against a target with low SN for beam walk correction is performed in the radar system shown in FIG. 11. [Figure 13] FIG. 13 is a conceptual diagram for explaining DBSCAN as an example of a cluster analysis method applied to the radar system shown in FIG. 11. [Figure 14] FIG. 14 is a conceptual diagram for explaining beam walk correction in the target detection processing shown in FIG. 12.

MODE FOR CARRYING OUT THE INVENTION

[0015] Hereinafter, embodiments will be described with reference to the drawings.

[0016] (First Embodiment) Beam Walk Correction Hereinafter, a radar system according to the first embodiment will be described with reference to FIGS. 1 to 7.

[0017] Figure 1 is a block diagram showing the configuration of a radar system according to the first embodiment, where (a) is a block diagram showing the configuration of the transmission system and (b) is a block diagram showing the configuration of the reception system.

[0018] In the radar system transmission system shown in Figure 1(a), a signal generator 11 generates a transmission seed signal, a modulator 12 generates a modulated signal from the transmission seed signal, a frequency converter 13 converts the modulated signal into a high-frequency signal, and then a pulse modulator 14 pulse modulates it to generate a radar signal, which is then transmitted into space from the transmitting antenna 15.

[0019] On the other hand, in the receiving system shown in Figure 1(b), n receiving antennas (n-channel sub-arrays) 211~21n receive the radar reflected signal, frequency converters 221~22n convert the received signal to baseband frequency, and AD converters 231~23n convert it to a digital signal.

[0020] Next, in the receiving system, slow-time axis FFT processing is performed in FFT processors 241 to 24n to obtain the signal on the range frequency-slow-time axis, clutter suppression processing is performed near DC (0 frequency) on the Doppler axis in clutter suppression processors 251 to 25n, and pulse compression processing is performed in pulse compressors 261 to 26n (see Non-Patent Literature 1) using the signal on the range frequency-slow-time axis, and pulse compression is performed by inverse range axis FFT processing.

[0021] Next, in the receiving system, the pulse-compressed signals obtained for each sub-array are input to the DBF (Digital Beam Forming) Σ beamformer 27 to form a Σ beam for target detection. The range walk and Doppler walk are corrected in the range Doppler corrector 28, and range cells exceeding the threshold are provisionally detected in the CFAR (see Non-Patent Literature 2) provisional detector 29. Then, in the DBF Σ beam corrector 30, the RD (Range-Doppler) data is rearranged for each range cell provisionally detected exceeding the threshold. Targets exceeding the threshold are extracted in the CFAR 31, the range of the extracted targets is selected in the target range selector 32, and the distance, velocity, and angle of the target are measured in the observation value outputter 33, and the observed values ​​of distance, velocity, AZ, and EL are output.

[0022] The processing operation of the radar system with the above configuration will be explained with reference to Figures 2 to 7. Here, Figure 2 is a conceptual diagram showing the transmit and receive beam shapes for the observation range of the radar system shown in Figure 1, Figure 3 is a flowchart showing the flow of target identification processing for the radar system shown in Figure 1, Figure 4 is a conceptual diagram showing the beam walk correction process for the radar system shown in Figure 1, Figure 5 is a diagram explaining the method of beam walk correction for the AZ axis in the radar system shown in Figure 1, Figure 6 is a diagram explaining the method of beam walk correction for the AZ axis and EL axis in the radar system shown in Figure 1, and Figure 7 is a diagram showing the data extraction process for provisional detection and correction processing in the radar system shown in Figure 1.

[0023] In this embodiment, as shown in Figure 2, the target identification process will be explained following the flowchart in Figure 3, for the case of SIMO where the observation range is covered by a transmitting fan beam and the reception range is covered by a multi-beam system using a pencil beam. Although it is generally a two-dimensional method using the AZ × EL axes, for simplicity, it will be explained in one dimension first, and then extended and formulated for the two-dimensional case.

[0024] First, in the signal transmission / reception processing (step S11), the transmitting system modulates the transmission type signal (11, 12), converts it to a high-frequency signal (13), then pulse modulates it to generate a radar signal (14), which is then sent out into space from the transmitting antenna 15. Meanwhile, in the receiving system, when the n receiving antennas 211~21n receive the reflected waves of the transmitted radar signal, the received signal is frequency-converted to the baseband (211~21n), converted to a digital signal (231~23n), and then, as signal processing (step S12), FFT processing is performed on the slow-time axis (241~24n), clutter near DC (0 frequency) is suppressed on the Doppler axis (251~25n), and pulse compression is performed (261~26n).

[0025] Similar to the signal processed with FFT on the fast-time axis, this obtains a signal on the range-frequency-slow-time axis through range-axis correlation. This can be formalized as equations (1) and (2), and multiplying them by their conjugates yields equation (3). Pulse compression is performed by using this range-frequency-slow-time axis signal to perform inverse FFT on the range axis.

[0026]

number

[0027] The above processing is performed for each subarray, and a detection Σ beam is formed by DBF processing between subarrays (27).

[0028] Next, range walk and Doppler walk corrections are performed (28) (steps S14 to S31). This process can be performed, for example, by maximizing the integral sequence (velocity and acceleration correction) (see Patent Document 1) or by performing higher-order corrections on the range frequency axis of equation (2) (see Non-Patent Document 4), and is a method that maximizes the integral sequence by searching for velocity and acceleration. Furthermore, in order to perform Doppler walk correction, the autofocus method for synthetic aperture processing, PGA (Phase Gradient Autofocus: see Non-Patent Document 5 and Patent Document 2), can be used.

[0029] This allows us to calculate a correction coefficient for the slow-time axis corresponding to range Doppler walk, and then apply the same correction coefficient to the sub-array signal using range Doppler correction (28) to perform the correction.

[0030] For the Σ beam, range cells are extracted by CFAR preliminary detection (29), beamwalk correction is performed on those range cells by DBFΣ beam correction (30) (steps S32, S33), and the observed values ​​of the target detection results are output (step S34).

[0031] The effect of this beamwalk correction is shown in Figure 4. For clarity, Figure 4(a) shows the range walk, Doppler walk, and beamwalk without correction; Figure 4(b) shows the range walk and Doppler walk with correction, and the beamwalk without correction; and Figure 4(c) shows the range walk, Doppler walk, and beamwalk with correction. First, as shown in Figure 4(b), the range walk and Doppler walk are corrected from the state in Figure 4(a) to tentatively detect the target and extract the range cell. Next, the beamwalk is corrected for that range cell, as shown in Figure 4(c). This makes the beam angle a constant value.

[0032] Next, we will discuss the beamwalk correction method for the one-dimensional case of the AZ axis (EL axis).

[0033] First, the target motion model can be expressed by equation (4).

[0034]

number

[0035] For simplicity, equation (4) uses up to the third order coefficient, but for more complex motion models, higher orders may be used. Alternatively, other functions may be used to represent the motion model. The initial target angle Ang0 can be determined by the angle of the provisionally detected beam, etc.

[0036] Ang is a two-dimensional data set with angle axis × slow-time axis ts for each set of c1, c2, c3, t1, t2, t3. Using this, the correction signal can be expressed as equation (5).

[0037]

number

[0038]

number

[0039] By performing an inverse FFT on this along the angular axis, we can obtain the array signal of an N element.

[0040]

number

[0041]

number

[0042] Next, we perform an FFT on the slow-time axis.

[0043]

number

[0044] As shown in Figure 3, this process is carried out by varying each parameter (c1, c2, c3, t1, t2, t3) within a predetermined range, and the parameter (c1, c2, c3, t1, t2, t3) that maximizes sig(rng, fs) is selected.

[0045] This explanatory diagram is shown in Figure 5. The diagram explains the method for correcting with respect to the AZ axis. After identifying the range cell by preliminary detection, the input signal changes along the AZ axis for each slow-time axis, and therefore generally takes the form of a curve as shown in Figure 5(a). To correct this, it is converted to the position axis (AZ position axis) of the array element (subarray), as shown in Figures 5(b1) to (bn). On the array position axis, as shown in the search signals in Figures 5(c1) to (cn), the AZ axis signal for each cell on the slow-time axis is converted into a phase gradient across the entire array position axis. Therefore, correction can be performed on the AZ position axis without having to identify the AZ angle for correction, which has the advantage of eliminating the need for detection to identify the AZ angle.

[0046] By modifying the parameters and correcting the reference signal using equation (7), and then performing correlation processing on the input signal and the reference signal, as shown in Figure 5(d), the phase gradient of the AZ position axis can be corrected for each cell on the slow-time axis, and each slow-time axis can be converted into a straight line. Performing an FFT (AZ axis beamforming) on ​​this on the AZ axis corrects the beamwalk as shown in Figure 5(e). After performing the above higher-order correction processing on the reference signal, performing an FFT on the slow-time axis allows for target detection with a high signal-to-noise ratio, as shown in Figure 5(f).

[0047] Next, we formulate a method for simultaneously processing two dimensions when the angular axes are two-dimensional (AZ, EL). The target motion model when the target moves in two dimensions along the AZ and EL axes can be expressed by the following equation.

[0048]

number

[0049] For simplicity, we have used up to the third order coefficient, but for more complex motion models, higher orders may be used. Alternatively, other functions may be used to represent the motion model. The initial target angle Ang0 can be determined by the angle of the provisionally detected beam, etc.

[0050] Here, Figure 6 shows the data extraction process for provisional detection and correction processing. As shown in Figure 6(a), with the number of beams set to Naz (AZ axis) × Nel (EL axis) (Nrx = Naz × Nel), the Doppler axis data of the provisional detection range is extracted, and the Doppler axis is subjected to an inverse FFT. As a result, as shown in Figure 6(b), data for the angle axis (AZ, EL) × slow-time axis is obtained. This is used as the subarray signal Sin(rng, naz, nel, ts) for the following processing.

[0051] Ang is a three-dimensional data set with angle axis (AZ,EL) × slow-time axis ts for each set of c1, c2, c3, t1, t2, t3. Using this, the correction signal can be expressed by the following equation.

[0052]

number

[0053] To use this, for Nrx (Nrx>1) 2D (AZ,EL) beam signals, the NAZ×NEL-Nrx signal is zero-filled (2D) to generate NAZ×NEL signals, and an inverse FFT is performed in 2D to obtain the NAZ×NEL element (subarray) signal.

[0054]

number

[0055] By performing an inverse FFT on this with respect to the two-dimensional angular axis, we can obtain the array signal (position axis signal) of the NAZ×NEL elements.

[0056]

number

[0057]

number

[0058] Next, we perform an FFT on the slow-time axis.

[0059]

number

[0060] As shown in Figure 3, this process is performed by varying each parameter (c1, c2, c3, t1, t2, t3) within a predetermined range, and the parameter (c1, c2, c3, t1, t2, t3) that maximizes sig(rng, fs) (where rng and fs are the provisionally detected range cell and Doppler cell) is selected.

[0061] The input signal is generally a curved surface along the AZ and EL angle axes for each of the slow-time axes, as shown in Figure 7(a). To correct this, it is converted to the position axes (AZ and EL position axes) of the array elements (subarray), as shown in Figures 7(b1) to (bn). On the array position axes, as shown in Figures 6(c1) to (cn), the angle axis signal for each cell on the slow-time axis is converted to a phase gradient across the entire array element axis. This has the advantage of allowing correction on the position axes (two-dimensional AZ and EL axes) without having to specify the angle for correction, thus eliminating the need for detection to identify angle cells.

[0062] By correcting the reference signal using equation (14) while changing the parameters, the phase gradient can be corrected on the array position axis (2D) for each cell on the slow-time axis, as shown in Figure 7(d), resulting in a straight plane for each slow-time axis. Performing a 2D FFT (AZ axis and EL axis beamforming) on ​​this plane on the AZ axis and EL axis corrects the beamwalk, as shown in Figure 7(e). After the above higher-order correction processing of the reference signal, an FFT is performed on the slow-time axis. This allows for target detection with a high signal-to-noise ratio, as shown in Figure 7(f).

[0063] As described above, according to this embodiment, since beamwalks during long-term integration are corrected using the array axes, correction can be performed without detection that identifies the correction target using the AZ and EL axes, thereby enabling target detection with a high signal-to-noise ratio.

[0064] (Second Embodiment) Beamwalk Corrected Monopulse Angle Measurement In the first embodiment, a method for detecting targets with a high signal-to-noise ratio (SNR) by beamwalk correction was described. In the second embodiment, the monopulse angle measurement process after beamwalk correction is described with reference to Figures 8 to 10. Here, the case of two-dimensional (AZ,EL) angle measurement is generally described.

[0065] Figure 8 is a block diagram showing the configuration of the transmission and reception systems of the radar system according to the second embodiment; Figure 9 is a diagram showing the method of measuring angles using monopulse beams of Σ, ΔAZ, and ΔEL in the radar system shown in Figure 8; and Figure 10 is a diagram showing the coordinate system for explaining the beam direction angle measurement process in the radar system shown in Figure 8. In Figure 8, the same parts as in Figure 1 are denoted by the same reference numerals, and the different parts will be explained here.

[0066] The transmission system shown in Figure 8(a) is the same as that in Figure 1(a), while the reception system shown in Figure 8(b) has the same configuration as that shown in Figure 1(b) with the addition of a ΔAZΔEL beamformer 36 and an AZ / EL monopulse angle detector 37.

[0067] In other words, in the second embodiment, the array signal after beamwalk correction according to equation (15) is used to perform ΔAZΔEL beam formation and AZ / EL monopulse angle measurement. In this case, the correction coefficients include the range Doppler correction coefficient generated by the Σ beam range Doppler corrector 28 and the beam correction coefficient generated by the DBF Σ beam corrector 30. For both, the same coefficients used for correction on the Σ beam can be used in the range Doppler correctors 341~34n and beam correctors 351~35n for each array signal.

[0068] The ΔAZΔEL beamformer 36 uses the corrected array signal to form monopulse beams of Σ, ΔAZ, and ΔEL by summing and subtracting the 2×2=4 division signals of the AZ axis and EL axis. The AZ / EL monopulse angle meter 37 uses the monopulse beams of Σ, ΔAZ, and ΔEL to calculate the error voltage ε shown by the following equation using the Σ beam and Δ beam, as shown in Figure 9(a), and measures the angle using the error voltage table with characteristics shown in Figure 9(b) that was acquired in advance.

[0069]

number

[0070] From the angle measurements and distance measurements using the range cell, the three-dimensional coordinates (X, Y, Z) of the target shown in Figure 10 can be calculated using the following formula.

[0071]

number

[0072] As described above, according to this embodiment, since a monopulse beam is formed using a beamwalk-corrected array signal, it is possible to perform angle measurement with a high signal-to-noise ratio and high accuracy.

[0073] (Third Embodiment) Target Low S / N Countermeasures for Beamwalk Correction In the first and second embodiments, a method for beamwalk correction using all multi-beams (Nrx beams) was described. In this case, the initial angle of the search range is the angle that was tentatively detected. However, if the target signal-to-noise ratio (SN) is low, observations may span across multiple beams, widening the search angle range and potentially increasing the processing load. In the third embodiment, with reference to Figures 11 to 14, a method for addressing low SN targets in beamwalk correction is described.

[0074] Figure 11 is a block diagram showing the configuration of the transmission and reception systems of a radar system according to the third embodiment; Figure 12 is a flowchart showing the target detection process with beamwalk correction target low signal-to-noise ratio countermeasures in the radar system shown in Figure 11; Figure 13 is a conceptual diagram to explain DBSCAN (Density-based Spatial Clustering of Applications with Noise) as an example of a cluster analysis method applied to the radar system shown in Figure 11; and Figure 14 is a conceptual diagram to explain beamwalk correction in the target detection process shown in Figure 12. In Figure 11, the same parts as in Figure 8 are denoted by the same reference numerals, and the different parts will be explained here.

[0075] The transmission system shown in Figure 11(a) is the same as that in Figure 8(a).

[0076] The receiving system shown in Figure 11(b) has a configuration that adds a cluster analyzer 38 and a beam search range extractor 39 to the configuration shown in Figure 8(b). Specifically, as shown in Figure 12, after signal transmission and reception (step S41), a Σ beam is formed (step S42), the range Doppler walk of the Σ beam is corrected (step S43), and the target is provisionally detected by CFAR provisional detection (step S44). Next, the provisionally detected cells are placed on the AZ-EL axis (step S45), cluster analysis is performed (step S46), the beam search range is extracted based on the cluster analysis results (step S47), beam walk correction is performed in the beam search range (step S48), and the process of detecting the target by CFAR detection is performed (step S49). The above processes from steps S45 to S19 are executed for each range cell (steps S50, S51), the observed values ​​of the detected target are output (step S52), and the series of processes is completed.

[0077] In other words, in this embodiment, provisional detection is performed using range-Doppler data for each beam, and cluster analysis is performed for each provisional detection range cell using the provisional detection data of the AZ-EL beam as input, in order to extract the beam containing the target using the coordinates (AZ, EL) of the multi-beam where the target or false detection occurred.

[0078] Cluster analysis is a method for grouping input values ​​according to predetermined conditions. As a cluster analysis method, for example, DBSCAN (Non-Patent Literature 6) shown in Figure 13 is used. This method analyzes clusters of noise points by specifying a gate (radius ε) centered on a core point and the number of observed points (border points) within the gate, MinPts. According to this method, near the beam where the target exists, there may be multiple detected beams within the gate radius, making grouping possible.

[0079] Therefore, in this embodiment, for example, the cluster analyzer 38 analyzes the input values ​​shown in Figure 14(a) to detect clusters as shown in Figure 14(b), and the beam search range extractor 39 extracts grouped beams (Mrx beams) of the beam search range. By extracting these Mrx beams, the search angle range in which the target moves can be determined as shown in Figure 14(c), so the range of coefficients in the target motion model of equations (4) and (10) can be limited, thereby limiting the number of searches and preventing an unintentional increase in processing scale. Furthermore, the initial angle when the detected beam spans multiple beams can be determined, for example, by calculating the centroid of the grouped Mrx beams (weighted average by the amplitude intensity of each beam). After determining the search angle range, beam walk correction can be performed as shown in Figure 14(d) using the same method as in the first and second embodiments.

[0080] As described above, according to this embodiment, when beamwalk correction is performed using all multi-beams (Nrx beams), even if the target has a low signal-to-noise ratio (SNR) and is observed across multiple beams, the cluster is analyzed to extract groupings for the beam search range and determine the search angle range. This suppresses an increase in processing load and allows for observation of the target with a high SNR using the beamwalk-corrected array signal.

[0081] 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]

[0082] 11...Signal generator, 12...Modulator, 13...Frequency converter, 14...Pulse modulator, 15...Transmitting antenna, 211-21n... Receiving antenna, 221-22n... Frequency converter, 231-23n... AD converter, 241-24n... FFT processor, 251-25n... Clutter suppressor, 261-26n... Pulse compressor, 271-27n... DBFΣ beamformer, 28... Range Doppler corrector, 29... CFAR provisional detector, 30... DBFΣ beam corrector, 31... CFAR detector, 32... Target range selector, 33... Observation value outputter, 341-34n... Range Doppler corrector, 351-35n... Beam corrector, 36... ΔAZΔEL beamformer, 37... AZ / EL monopulse angle detector, 38... Cluster analyzer, 39... Beam search range extractor.

Claims

1. In a radar system that observes an observation range for a long period of time, the transmission beam is formed to cover the observation range using an array antenna with Naz × Nel (Naz ≥ 1, Nel ≥ 1, or at least 2) sub-array elements, and the reception beam is formed with Nrx (Nrx ≥ 1) multi-beams (SIMO: Single Input Multiple Output). A means for zeroing out Naz × Nel-Nrx signals other than the multi-beam number Nrx of the reception, A means for obtaining the phase of the array element signal by performing an inverse FFT on the zero-filled signal, A means for setting M (M≧1) types of target motion models as reference signals and performing correlation calculations with the array element signals, A means for performing a search to extract the maximum value of the beam obtained by slow-time axis FFT from the results of the correlation calculation, A means for detecting a target by correcting the beamwalk, in which the target moves in the beam direction within the integration time of the slow-time axis, using the results of the search described above, on the array axis, and A radar system equipped with the following features.

2. Furthermore, the array signal after correction of the beam walk is used to form a monopulse beam and measure the angle of the target. The radar system according to claim 1.

3. Furthermore, for each beam, there is a means for analyzing clusters using the coordinates of the beams for which detection has been completed, Based on the analysis results of the aforementioned cluster, a means is used to extract the beam range in which target candidates exist, to constrain the target candidate range, and to constrain the M (M≧1) type target motion model by the aforementioned target candidate range. Equipped with, The Naz × Nel-Nrx signals other than the multi-beam number Nrx are zero-filled, and the phase of the array element signal obtained by the inverse FFT is set as a reference signal to the M (M≧1) type target motion model constrained by the target candidate range. After correlation calculation between the reference signal and the array element signal, the target is detected by a search method that extracts the maximum value of the beam obtained by slow-time axis FFT. The radar system according to claim 1.

4. In a radar signal processing method for a radar system that observes an observation range for a long period of time, the transmission uses an array antenna having Naz × Nel (Naz ≥ 1, Nel ≥ 1, or at least 2) sub-array elements to form a transmission beam that covers the observation range, and the reception uses Nrx (Nrx ≥ 1) multi-beams (SIMO: Single Input Multiple Output), The Naz × Nel-Nrx signals other than the received multi-beam number Nrx are filled with zeros. The zero-filled signal is then subjected to an inverse FFT to obtain the phase of the array element signal. A target motion model of type M (M≧1) is set as a reference signal and correlated with the array element signal. From the results of the correlation calculation, a search is performed to extract the maximum value of the beam obtained by slow-time axis FFT. Using the results of the search described above, the beamwalk, in which the target moves in the beam direction within the integration time of the slow-time axis, is corrected on the array axis to detect the target. A method for processing radar signals in a radar system.

Images