radar equipment

Abstract

To provide a technique capable of reducing errors in estimation of signal arrival directions in a radar device.SOLUTION: A radar device 100 includes: a transmitting unit 110 that transmits transmission waves from one or more transmission antennas 112; a reception unit 120 that receives reflected waves of the transmission waves reflected by an external object using a plurality of reception antennas 122; and a processing unit 130 that determines distance to the external object using a reception result by the reception unit, and determines a direction of the external object using a corrected received signal, corrected by CΓ correction by a calibration matrix. The processing unit divides a viewing angle of the reception antennas into N correction sections, where N is an integer of 2 or more, and performs CΓ correction by the calibration matrix, assuming that the calibration matrix does not depend on a reception angle in each correction section.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to a technique for reducing direction-of-arrival estimation errors in a radar device. [Background technology]

[0002] To improve the performance of millimeter-wave radar, a technology is required to reduce the estimation error of the direction of arrival of the received signal from a target. One promising method for reducing the direction of arrival estimation error is the CΓ correction technology. Here, C is a matrix that represents the mutual coupling between antennas, and Γ is a matrix that represents the gain and phase errors due to variations in antenna characteristics. The CΓ correction technology aims to reduce errors by measuring these errors in advance and correcting the received signal using a calibration matrix Q (=CΓ).

[0003] Patent Documents 1 and 2 disclose a technique called global correction, which performs correction collectively over the entire range of arrival direction angles. Global correction is a method based on the premise that the calibration matrix Q does not depend on the reception angle. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2017-211241 [Patent Document 2] Japanese Patent Application Publication No. 2019-128235 Summary of the Invention [Problem to be solved by the invention]

[0005] However, conventional techniques have had the problem that estimation errors increase as the angle from the center of the field of view increases. On the other hand, there is a method called local correction, in which the calibration matrix Q depends on the reception angle. However, local correction has the problem of large side lobes due to errors in the antenna position. These problems are also common to radar devices that use electromagnetic waves other than millimeter waves. [Means for solving the problem]

[0006] According to one embodiment of the present disclosure, there is provided a radar device (100) comprising: a transmitter (110) that transmits a transmission wave from one or more transmitting antennas (112), a receiver (120) that receives, using a plurality of receiving antennas (122), a reflected wave of the transmission wave reflected by an external object, and a processor (130) that determines the distance to the external object using a reception result from the receiver and determines the azimuth of the external object using a corrected reception signal corrected by CΓ correction using a calibration matrix, where N is an integer equal to or greater than 2, and the processor divides the viewing angle of the receiving antenna into N correction intervals, and performs the CΓ correction using the calibration matrix within each correction interval, assuming that the calibration matrix is independent of the reception angle. The N calibration matrices corresponding to the N correction intervals have different matrix element values. The processing unit sets each of the N calibration intervals used when determining the N calibration matrices corresponding to the N correction intervals to an interval that includes the corresponding correction interval and is wider than the correction interval, and determines the calibration matrix corresponding to each calibration interval using received signals obtained at a plurality of receiving angles within each calibration interval.

[0007] This radar system can improve upon the drawback of conventional global correction, where the error increases as the receiving angle deviates from the center of the field of view. In addition, because the calibration matrix is angle-independent within the correction interval, it can also improve the side lobes that occur in local correction using a calibration matrix that is dependent on the receiving angle. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a block diagram showing the configuration of a radar device. [Figure 2] FIG. 3 is an explanatory diagram showing a calibration interval and a correction interval in the first embodiment. [Figure 3] FIG. 1 is a plan view showing a measurement system for determining a calibration matrix. [Figure 4] FIG. 2 is an explanatory diagram showing the configuration of a receiving antenna of a radar device. [Figure 5] FIG. 10 is a diagram showing correction conditions with and without section division. [Figure 6] 10A and 10B are diagrams showing MUSIC correction results with and without section division. [Figure 7]FIG. 10 is an explanatory diagram showing a calibration interval and a correction interval in the second embodiment. [Figure 8] FIG. 11 is an explanatory diagram showing a calibration interval and a correction interval in the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] As shown in FIG. 1, the radar device 100 includes a transmitter 110, a receiver 120, and a processor 130. The transmitter 110 includes an oscillator 114 and a modulator 116, and transmits a transmission wave from a transmitter antenna 112. The receiver 120 receives, using multiple receiver antennas 122, waves reflected by external objects. The processor 130 determines the distance to the external object using the reception results from the receiver 120, and determines the azimuth of the external object using a corrected reception signal corrected by CΓ correction using a calibration matrix. Since the method for determining the distance to an external object is well known, a description thereof will be omitted. CΓ correction using a calibration matrix will be described later. The radar device 100 of this embodiment is a millimeter-wave radar, but the present disclosure can also be applied to radar devices that use electromagnetic waves other than millimeter waves.

[0010] The processing unit 130 is a microcomputer equipped with a processor 132 and a memory 134. The memory 134 is, for example, a ROM and a RAM. Various functions of the microcomputer are realized by executing a computer program stored in the memory 134. In this example, the memory 134 corresponds to a non-transitory tangible recording medium storing the computer program. Some or all of the functions executed by the processor 132 may be realized by hardware circuits.

[0011] In this embodiment, the number of transmitting antennas 112 is one, and the radar is configured as a single-input multiple-output (SIMO) radar. However, the present disclosure is also applicable to a multiple-input multiple-output (MIMO) radar having two or more transmitting antennas 112. In the CΓ correction described below, a MIMO radar having p transmitting antennas 112 and m receiving antennas 122 can be treated as equivalent to a SIMO radar having p×m receiving antennas.

[0012] As shown in Figure 2, in this disclosure, the viewing angle of the radar device 100 is divided into N sections, and in each of the divided sections, the calibration matrix Q (=CΓ) used in the CΓ correction of the received signal is assumed to be independent of the receiving angle θ. N In FIG. 2, the correction section PQ in which the received signal is corrected using the calibration matrix Q is k and the calibration interval RQ used to find the calibration matrix Q. k Here, k is an ordinal number indicating the order of the sections, and takes a value from 1 to N. The number of sections N is 2 or more, and preferably 5 or more.

[0013] In the first embodiment, the correction section PQ k and calibration interval RQ k The magnitude of is equal and is a constant value. The minimum angle of the viewing angle is θ0, and the maximum angle is θ N Then, the kth divided correction interval PQ k The angle range of θ k-1 ~θ k Correction section PQ k Calibration matrix Q used in the CΓ correction k is this correction interval PQ k It is angle independent and has the same calibration matrix Q k is applied, where the calibration matrix Q k depends on k. Also, N calibration matrices Q khave different matrix element values from each other. Here, "having different matrix element values from each other" means that the value of at least one matrix element is different.

[0014] In this way, the viewing angle is divided into N correction intervals PQ k and divide each correction section PQ k The calibration matrix Q used in k is the correction interval PQ k By assuming that the calibration matrix Q does not change within the specified range, it is possible to improve the drawback of global correction, in which the error increases as the receiving angle deviates from the center of the field of view. k is the correction interval PQ k Since the calibration matrix Q k It can also improve side lobes that occur due to angle-dependent local correction.

[0015] As shown in Figure 3, the calibration matrix Q k The measurement system for determining the calibration matrix Q uses a rotator 200 that rotates the radar device 100 and a corner reflector 210 as a target. First, the rotator 200 is used to rotate the radar device 100 by an angle θ, thereby setting the reception angle θ of the reflected wave from the corner reflector 210, and measuring the received signal at that time. Figure 3 is a plan view, and the angle θ is the horizontal angle. Also, the direction Dn indicates the direction where the angle θ is zero, and the angle θ is measured clockwise from the direction Dn. The calibration matrix Q k When calculating the angle θ, where n is an integer equal to or greater than 2, the angle θ is set to n different values in each calibration interval, and the received signal is measured at each of these values. The n angles are called "correction points." The difference between the angles is called the "correction step" or "reception angle step width." The number n of angles θ set for each calibration interval is preferably set to be equal to or greater than the number m of receiving antennas 122.

[0016] As shown in Fig. 4, m receiving antennas 122 are arranged in a uniform linear array with a constant antenna spacing d. When j is an integer between 1 and m, a received signal X is transmitted from the j-th receiving antenna 122.j m received signals X1(t) to X(t) are obtained. m (t) is called "received data." The processing unit 130 calculates the calibration matrix Q used for CΓ correction from the received data at n different angles θ for each calibration interval. The method for calculating the calibration matrix Q will be described later.

[0017] The processing unit 130 calculates the received signal X using the calibration matrix Q. j (t) is corrected, and the corrected received signal is used to find the noise space vector and the ideal mode vector from eigenvalue analysis using the correlation matrix, and the received signal X is calculated using the MUSIC (Multiple Signal Classification) method. j The arrival direction of (t), i.e., the reception angle θ, is estimated. The arrival direction is also called the "azimuth of the external object." The correct reception angle at the n correction points is equal to the angle θ of the rotator 200 and is known, so the difference between the reception angle estimated by the MUSIC method and the angle θ of the rotator 200 can be calculated as the estimation error of the arrival direction.

[0018] CΓ correction using the calibration matrix Q can be performed as follows. The signal actually input to the receiving antenna 122 is not an ideal signal, but contains some kind of error. Factors that cause this error include mutual coupling between elements, variations in the characteristics (amplitude, phase) of each antenna element, element position errors, scatterers near the antenna, thermal noise, etc. Of the above factors, the main causes of error in a high SNR environment are thought to be mutual coupling between elements and variations in the characteristics of each antenna element, so the actually received signal can be modeled using the following equation. X=CΓ·X ideal ...(Formula 1) X: Actual received signal X ideal :Ideal received signal C: Matrix due to mutual coupling between elements (inter-element mutual coupling matrix) Γ: Matrix based on the characteristics of each element (element characteristic matrix)

[0019] On the other hand, the ideal mode vector is a(θ) and the actual mode vector is a m(θ) and the calibration matrix Q, the following equation holds: a m (θ)=Q·a(θ) (Equation 2)

[0020] If the receiving antenna 122 is a linear array with m antennas and antenna spacing d, the ideal mode vector a(θ) is expressed by the following equation. a(θ)=[1,exp(j2π(d / λ)·sinθ), exp(j2π(d / λ)2·sinθ),,,exp(j2π(d / λ)(m-1)·sinθ] T ...(Formula 3) where [ ] T denotes transposition.

[0021] Also, let A be the matrix consisting of the ideal mode vectors at n correction points θ1 to θn, and let A be the matrix consisting of the actual mode vectors. m Then, the following formula is established: A=[a(θ1), a(θ2) ,,, a(θn)] (Equation 4) A m =[a m (θ1), a m (θ2) ,,, a m (θn)] (Formula 5) A m =Q·A ··· (Equation 6)

[0022] In the above formula 1, X ideal =a(θ), X=a m (θ), the calibration matrix Q is equal to the CΓ correction matrix as follows: CΓ=Q (Equation 7)

[0023] There are various methods for obtaining the calibration matrix Q, but in this embodiment, the following simple method can be used. First, the average value of the received signal X is calculated from a number of snapshots of the actual received signal X, and the phase is normalized based on a certain channel, and the amplitude is normalized by the average value of m antennas, thereby obtaining the actual mode vector a mSince the matrix A consisting of ideal mode vectors is an m×n matrix, the calibration matrix Q can be found from the pseudo-inverse matrix pinv(A) of matrix A according to the following equation: Q=A m ·pinv(A) ···(Formula 8)

[0024] Instead, we use the Hermitian conjugate (complex conjugate + transpose) of matrix A. H is an n × m matrix, and the product AA H Since is an m×m matrix, it is also possible to find the calibration matrix Q according to the following equation: In this case, however, the condition n≧m must be met. Q=A m A H (AA H ) -1 ...(Formula 9)

[0025] Since the calibration matrix Q (= CΓ) is an m × m square matrix, the inverse matrix (CΓ) of the calibration matrix Q (= CΓ) -1 This inverse matrix (CΓ) can be obtained. -1 By multiplying the received signal X by the received signal X, the ideal received signal X is obtained. ideal can be obtained. X ideal =(CΓ) -1 ·X ···(Formula 10)

[0026] The correction by the above formula 10 is called "CΓ correction." In addition, the received signal X obtained by CΓ correction ideal Corrected received signal X ideal This corrected received signal X ideal From the eigenvalue analysis using the correlation matrix of the noise space vector and the ideal mode vector, it is possible to estimate the direction of arrival, that is, estimate the reception angle θ, using the MUSIC method. In addition, since the correct reception angle θ at the n correction points is known, the difference between the reception angle estimated by the MUSIC method and the angle θ of the rotator 200 can be calculated as the estimation error of the direction of arrival.

[0027] As shown in Fig. 5, in the example and comparative example, a radar with 12 receiving antennas m was used, and the angle θ of the rotator 200 was changed in steps of θ0 / 30 within the range between a lower limit value -θ0 and an upper limit value θ0, to acquire 61 pieces of received data. In the comparative example, the entire range of field of view from -θ0 to θ0 was treated as one correction interval, and CΓ correction of the received signal was performed using one calibration matrix Q. In the example, the range of field of view from -θ0 to θ0 was equally divided into five correction intervals, and a different calibration matrix Q was used in each correction interval to perform CΓ correction of the received signal.

[0028] FIG. 6 shows the results of the estimation error of the reception angle using the MUSIC method for the comparative example and the example shown in FIG. 5. The horizontal axis represents the reception angle θ normalized by its maximum value θ0, and the vertical axis represents the reception angle estimation error Δθ normalized by θ0. If the true value of the reception angle is θ and the estimated value of the reception angle using the MUSIC method is θ^, the reception angle estimation error Δθ is Δθ = θ^ - θ. The results of the comparative example without section division are shown by a dotted line, and the results of the example with section division are shown by a solid line. In the example, the reception angle estimation error Δθ is reduced compared to the comparative example, and it can be seen that the accuracy of the direction estimation is improved. In particular, the effect of reducing the error Δθ in the example is remarkable near both ends of the field of view angle.

[0029] As described above, according to this embodiment, the field of view of the receiving antenna is divided into N correction intervals, and within each correction interval, CΓ correction is performed using the calibration matrix Q assuming that the calibration matrix Q does not depend on the receiving angle θ. This improves the drawback of conventional global correction, in which errors increase as the receiving angle deviates from the center of the field of view. Furthermore, because the calibration matrix Q does not depend on the angle within the correction interval, it is also possible to improve the side lobes that occur in local correction using a calibration matrix that depends on the receiving angle.

[0030] As shown in FIG. 7, in the second embodiment, N calibration matrices Q1 to Q N N calibration intervals RQ1 to RQ N Each of these is divided into the corresponding correction intervals PQ1 to PQ Nand correction interval PQ1 to PQ N However, the correction step (step width of the reception angle) remains constant, as in the first embodiment. N is the calibration interval RQ1~RQ N The correction intervals PQ1 to PQ are determined using received signals obtained at multiple correction points within the N correction intervals PQ1 to PQ N The magnitude of is constant. Calibration section RQ1~RQ N Each of these is divided into the corresponding correction intervals PQ1 to PQ N If the interval is wider than N Since the number of correction points used in determining the angle of reception can be increased, the estimation error of the angle of reception can be further reduced.

[0031] In the second embodiment shown in FIG. 7, N calibration intervals RQ1 to RQ N The magnitude of each calibration section RQ1 to RQ N In this way, it is possible to further reduce the estimation error of the reception angle. Specifically, for example, in N calibration periods RQ1 to RQ N The magnitude of the angle may be increased closer to the center of the field of view and decreased closer to both ends of the field of view. In other words, the magnitude of the angle may be increased in a calibration interval where the reception angle is closer to 0. This makes it possible to further reduce the estimation error of the reception angle.

[0032] As shown in FIG. 8, in the third embodiment, N correction sections PQ1 to PQ N The size of each correction section PQ1 to PQ N In this way, it is possible to further reduce the estimation error of the reception angle. Specifically, for example, N correction sections PQ1 to PQ N The magnitude of the error may be increased closer to the center of the viewing angle and decreased closer to both ends of the viewing angle. In other words, the closer the correction section is to 0 receiving angle, the larger the magnitude of the error may be. This makes it possible to further reduce estimation errors in the receiving angle.

[0033] In the third embodiment, the individual correction sections PQ1 to PQ N and the corresponding calibration interval RQ1~RQ N The magnitudes of the calibration intervals RQ1 to RQ2 are set to be equal to each other. The correction step (step width of the reception angle) remains constant as in the first embodiment. The second and third embodiments may be applied simultaneously. For example, in the third embodiment, N calibration intervals RQ1 to RQ2 are set to be equal to each other. N Each of these is divided into the corresponding correction intervals PQ1 to PQ N and correction interval PQ1 to PQ N In addition, the N calibration intervals RQ1 to RQ2 may be set to be wider than the N calibration intervals RQ1 to RQ3. N The size of each calibration section RQ1~RQ N At the same time, N correction sections PQ1 to PQ N The size of each correction section PQ1~PQ N The angle of reception may be changed depending on the position of the antenna. These configurations also make it possible to further reduce the estimation error of the angle of reception.

[0034] Furthermore, in the first to third embodiments described above, if the correction step (step width of the reception angle) is made small, it is possible to reduce the estimation error of the reception angle. N In each of the calibration sections RQ1 to RQ N For example, the correction step may be larger for calibration intervals closer to the center of the field of view and smaller for calibration intervals closer to both ends of the field of view. This makes it possible to further reduce the estimation error of the reception angle.

[0035] The processing units and methods described herein may be implemented by a special-purpose computer configured with a processor and memory programmed to perform one or more functions embodied in a computer program. Alternatively, the processing units and methods described herein may be implemented by a special-purpose computer configured with a processor configured with one or more dedicated hardware logic circuits. Alternatively, the processing units and methods described herein may be implemented by one or more special-purpose computers configured with a processor and memory programmed to perform one or more functions in combination with a processor configured with one or more hardware logic circuits. Furthermore, the computer program may be stored as instructions executed by a computer on a computer-readable non-transitory storage medium.

[0036] The present disclosure is not limited to the above-described embodiments and their modifications, and can be embodied in various forms without departing from the spirit of the present disclosure. Furthermore, the various characteristic configurations described above can be adopted in any combination as long as they are not inconsistent with each other. [Explanation of symbols]

[0037] 100 radar device, 110 transmitting unit, 112 transmitting antenna, 114 oscillator, 116 modulator, 120 receiving unit, 122 receiving antenna, 130 processing unit, 132 processor, 134 memory, 200 rotator, 210 corner reflector

Claims

1. A radar device (100), a transmitting unit (110) that transmits a transmission wave from one or more transmitting antennas (112); a receiving unit (120) that receives a reflected wave of the transmitted wave reflected by an external object using a plurality of receiving antennas (122); a processing unit (130) that determines the distance to the external object using the reception result by the receiving unit, and determines the direction of the external object using a corrected reception signal corrected by CΓ correction using a calibration matrix; Equipped with the processing unit divides the viewing angle of the receiving antenna into N correction intervals, where N is an integer equal to or greater than 2, and performs the CΓ correction using the calibration matrix within each correction interval, assuming that the calibration matrix is independent of the receiving angle; The N calibration matrices corresponding to the N correction intervals have different matrix element values, The processing unit setting each of the N calibration intervals used when determining the N calibration matrices corresponding to the N correction intervals to an interval that includes the corresponding correction interval and is wider than the correction interval; The radar system determines the calibration matrix corresponding to each calibration interval using received signals obtained at a plurality of reception angles across each calibration interval.

2. The radar device according to claim 1, The processing unit changes the sizes of the N calibration intervals used when determining the N calibration matrices corresponding to the N correction intervals, depending on the position of each calibration interval.

3. The radar device according to claim 2, The processing unit increases the size of the N calibration sections as they approach the center of the field of view angle and decreases the size as they approach both ends of the field of view angle.

4. The radar device according to any one of claims 1 to 3, The processing unit changes the size of the N correction intervals depending on the position of each correction interval.

5. The radar device according to claim 4, The processing unit increases the size of the N correction sections as they approach the center of the field of view angle and decreases the size as they approach both ends of the field of view angle.

6. The radar device according to any one of claims 1 to 5, the processing unit changes a step width of the reception angle set to receive the reception signal at different reception angles in each of the N calibration intervals used when determining the N calibration matrices corresponding to the N correction intervals, according to the position of each calibration interval.

7. The radar device according to claim 6, The processing unit increases the step width of the reception angle for calibration intervals closer to the center of the field of view and decreases the step width for calibration intervals closer to both ends of the field of view.

Images