JP2012181109A - Radar device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a radar device that increases a coherent integration gain of a target moving at a low speed while maintaining a coherent integration gain of a target moving at a high speed to improve accuracy of arrival direction estimation.SOLUTION: A radar transmission unit converts a pulse compression code into a high-frequency transmission signal to transmit the high-frequency transmission signal from a transmission antenna. A radar reception unit includes: a plurality of antenna system processing sections for performing coherent integration of a correlation value between a reception signal and a transmission signal; "p"(integer) pieces of correlation matrix generation sections for generating correlation matrices based on each output of the plurality of antenna system processing sections; "p" (integer) pieces of addition sections for adding an output of at least one of the "p" pieces of correlation matrix generation sections; an output selection control section for selecting a correlation matrix generation section generating a correlation matrix that has a maximum coherent integration gain in each of the generated correlation matrices; and an output selection section for selecting one of each output of the "p" pieces of addition sections according to the correlation matrix generation section selected by the output selection control section.

Description

  The present invention relates to a radar apparatus using a pulse signal that detects a target by receiving a pulse signal of a reflected wave reflected by a target with an antenna.

  The radar device radiates radio waves from a measurement point into space, receives a pulse signal of a reflected wave reflected by the target, and measures at least one or more of the distance and direction between the measurement point and the target. In particular, in recent years, a radar apparatus capable of being detected as a target including automobiles and pedestrians by high-resolution measurement using short-wave radio waves including microwaves or millimeter waves has been developed.

  Further, the radar apparatus receives a signal in which reflected waves from a target existing at a short distance and a target existing at a long distance are mixed. In particular, a range side lobe is generated by a signal of a reflected wave from a target existing at a short distance. The range side lobe and the main lobe of the reflected wave signal from the target existing at a long distance are mixed, and the radar apparatus deteriorates the accuracy of detecting the target existing at a long distance.

  Therefore, a radar apparatus using a pulse signal that requires high-resolution measurement for a plurality of targets has an autocorrelation characteristic (hereinafter referred to as “low-range sidelobe characteristic”) that results in a low range sidelobe level. Transmission of a pulse wave or a pulse modulated wave is required.

  In addition, when a car and a pedestrian are present at the same distance from the measurement point, the radar apparatus mixed signals of reflected waves from a car and a pedestrian having different radar cross sections (RCS). Receive a signal. In general, the radar reflection cross section of a pedestrian is lower than the radar reflection cross section of an automobile.

  For this reason, the radar apparatus is required to properly receive the reflected wave signal from the automobile and the pedestrian even if the automobile and the pedestrian exist at the same distance from the measurement point. Depending on the distance or type of the target, the reflected wave signal at the reception signal level changes. The radar apparatus is required to have a reception dynamic range in which reflected wave signals having various reception signal levels can be received.

  When a conventional radar apparatus using pulse compression repeatedly transmits a pulse compression code in the transmission period Tr, the correlation values calculated by the pulse compression process are added and averaged to obtain an SNR ( Improve signal to noise ratio). Note that addition averaging includes coherent integration and non-coherent integration (also called incoherent integration).

  For example, coherent integration is possible in a period (Nc × Tp) in which the time correlation is high among the correlation values calculated by the pulse compression process. The parameter Tp is the pulse width [seconds]. The SNR can be improved by coherent integration, as shown in Equation (1). The parameter Gc is a gain by coherent integration, and is calculated according to Equation (2).

  The parameter Nc is a coherent integration number, and is set depending on the assumed maximum moving speed of the target. Therefore, as the assumed maximum moving speed of the target increases, the Doppler frequency fluctuation included in the reflected wave signal from the target increases, and the time correlation period becomes shorter. For this reason, the coherent integration number Nc is reduced, and the gain Gc by the coherent integration is reduced according to the equation (2). That is, in the formula (1), the effect of improving the SNR by coherent integration is reduced.

  On the other hand, even when non-coherent integration is used, the SNR can be improved as shown in Equation (3) by adding the amplitude of the correlation value or the received power component calculated by the pulse compression processing. The parameter Gd is a gain by non-coherent integration, and is calculated according to Equation (4).

  The parameter Nd is a non-coherent integration number. When the coherent integration number Nc and the non-coherent integration number Nd are the same, the coherent integration contributes more to gain improvement than the non-coherent integration. However, in order to obtain an ideal gain by coherent integration, it is necessary that the phase components of the received signal match within a predetermined range in the section (time) in which coherent integration is performed, and the possible range of coherent integration is limited. .

  Further, when a conventional pulse compression radar measures a target moving at a high speed from a target moving at a low speed, the coherent integration number Nc is fixedly set depending on the assumed maximum moving speed of the target. Is small. For this reason, regarding the positioning of a target that moves at a low speed, there is a problem that a gain other than the gain by coherent integration in some of the sections having high time correlation cannot be obtained.

  In relation to this problem, for example, a radar apparatus disclosed in Patent Document 1 is known. The radar apparatus includes a plurality of range gates whose range gate width is determined by a pulse width, and for each range gate, a plurality of coherent integrators, a plurality of detectors, a plurality of non-coherent integrators, And a plurality of threshold detectors.

  Furthermore, the radar apparatus performs a plurality of integration processes with different ratios of the coherent integration number and the non-coherent integration number by using the plurality of coherent integrators and the plurality of non-coherent integrators, A configuration is disclosed in which a measurement target is detected for each range by comparing a signal with a predetermined threshold value using a plurality of threshold detectors.

JP-A-5-45449

  However, when the arrival direction is estimated according to the phase difference of the reflected wave signal from the received target using the array antenna that can be configured using a plurality of antenna elements in the radar device of Patent Document 1, the following is shown. The configuration is as follows.

  The output of the multiple coherent integrators for each range gate includes the phase information of the reflected wave signal from the target, but the phase information is removed in the detector, so the output of the multiple coherent integrators for each range gate A configuration in which an arrival direction estimation unit is added can be considered as an example. However, since the arrival direction is estimated for all combinations of a plurality of coherent integrators and non-coherent integrators, the circuit scale of the radar apparatus increases.

  Another example is a configuration in which an arrival direction estimation unit is added to the subsequent stage of a plurality of threshold detectors for each range gate. Each threshold detector need not estimate the direction of arrival for signals below the threshold. However, it is difficult to determine in advance which detector of the plurality of detectors the received signal satisfies. For this reason, the output data of each of the plurality of coherent integrators is stored in a memory, and the arrival direction is estimated based on the detection result of the threshold detector. Accordingly, the required amount of memory increases, and the delay until the arrival direction estimation processing result is obtained also increases.

  Further, in the above configuration, as disclosed in the radar apparatus of Patent Document 1, when a plurality of combinations are set in a relationship in which the product of the coherent integral number and the non-coherent integral number is constant, 1) the coherent integral number is There are combinations in which the number of non-coherent integrals decreases as the number increases, or 2) combinations in which the number of non-coherent integration numbers increases as the number of coherent integration numbers decreases.

  Under such a relationship, when a reflected wave signal from a target moving at high speed is included in a certain range gate, the following phenomenon occurs, and the accuracy of estimation of the direction of arrival of the target is greatly degraded. There was a problem.

  1) A combination with a large number of coherent integrals and a small number of non-coherent integrals:

  The reflected wave from the target moving at high speed is subjected to coherent integration up to a period of low time correlation, and is smaller than the ideally obtained coherent integration gain.

  For example, when a phase rotation of 180 degrees or more is included in the coherent integration period, the coherent integration gain becomes negative. On the other hand, since the non-coherent integration number is small, the noise suppression effect is small, and a value on which the noise component is superimposed is output as the integration processing result.

  2) A combination with a small number of coherent integrals and a large number of non-coherent integrals:

  Since the number of coherent integrals is small, the gain improvement effect is small. However, since the number of noncoherent integrals is large, the noise component is sufficiently suppressed.

  Originally, the integration processing result of 2) should be selected as the optimum combination of integration numbers. However, when the SNR of the reflected wave signal from the target is low, it is more apparent that more noise components are superimposed. Since the upper integration processing output level becomes higher, the integration processing result of 1) is erroneously selected as the optimum combination of integration numbers.

  Estimating the arrival direction based on the selected combination of the coherent integration number and the non-coherent integration number has a problem that the estimation accuracy of the arrival direction of the target is greatly deteriorated due to the influence of noise components.

  FIG. 15 shows a case where a signal including reflected waves from four different moving speed targets is received by the receiving unit of the radar device for a signal that the radar device repeatedly transmits a total of 480 pulse compression codes. This is an example of the result of integration processing using a combination of different coherent integration numbers and non-coherent integration numbers for each range gate.

  In the figure, four different moving speed targets are target # 1 moving at high speed (equivalent to 80 km / h), target # 2 moving at medium speed (equivalent to 40 km / h), and moving at low speed. Target # 3 (equivalent to 20 km / h per hour) and target # 4 (equivalent to 5 km / h per hour) moving at walking speed, and the reflected wave from each target to the range gate at the arrow position in FIG. Is receiving.

  FIG. 9A shows the detection result, where the coherent integration number is 30 and the non-coherent integration number is 16.

  FIG. 5B shows the detection result, where the coherent integration number is 60 times and the non-coherent integration number is 8 times.

  FIG. 4C shows the detection result, where the coherent integration number is 120 times and the non-coherent integration number is 4 times.

  FIG. 4D shows the detection result, where the coherent integration number is 240 times and the non-coherent integration number is two times.

  FIG. 5E shows the detection result, where the coherent integration number is 480 times and the non-coherent integration number is one time.

  In FIG. 15, the horizontal axis indicates the arrival timing of the received signal and is displayed using the range gate number defined by the transmission pulse width. The vertical axis represents the output level [dB] obtained by coherent integration and non-coherent integration after pulse compression processing.

  In FIG. 15, as the coherent integration number increases, the signal of the reflected wave from the slowly moving target obtains a large coherent integration gain. On the other hand, the signal of the reflected wave from the target moving at high speed has a reduced coherent integral gain, and further attenuated as a negative coherent integral gain.

  Further, as the number of coherent integrations increases, the non-coherent integration number decreases, so that the variance of noise components increases. The probability of occurrence of the phenomenon as described above (a phenomenon in which the estimation accuracy of the direction of arrival of the target due to an error in selecting the optimum combination of integration numbers for the reflected wave signal from the target moving at high speed greatly increases) is increased.

  The present invention has been made in view of the above-described conventional circumstances, and maintains the coherent integration gain of a target moving at high speed and increases the coherent integration gain of a target moving at low speed, thereby estimating the direction of arrival using a simple configuration. An object of the present invention is to provide a radar device that improves accuracy.

  The present invention is the above-described radar device, which converts a transmission signal into a high-frequency transmission signal, transmits the high-frequency transmission signal from a transmission antenna, and a reflected wave that is the high-frequency transmission signal reflected by the target. A radar receiving unit from a plurality of receiving antennas for estimating the direction of arrival of the reflected wave from the target using the signal of the signal, wherein the radar receiving unit has different coherent integration between the correlation value of the received signal and the transmitted signal A plurality of antenna system processing units that perform a plurality of coherent integrations using a number, and a correlation matrix that is phase difference information caused by a receiving antenna arrangement based on each output of the plurality of coherent integrations of the plurality of antenna system processing units. A plurality of correlation matrix generation units to be generated and outputs of the correlation matrix generation unit obtained from the coherent integration output of the coherent integration number N On the other hand, based on the outputs of the one or more correlation matrix generation units obtained from the coherent integration output of the coherent integration number smaller than N and the output of the addition unit, An arrival direction estimation unit that estimates the arrival direction of the reflected wave.

  According to the present invention, it is possible to maintain the estimation accuracy of the direction of arrival of a target moving at high speed, and to improve the coherent integration gain of the target moving at low speed using a simple configuration.

The block diagram which shows simply the internal structure of the radar apparatus of 1st Embodiment The block diagram which shows the internal structure of the radar apparatus of 1st Embodiment in detail Explanatory drawing which shows the relationship between the transmission interval and transmission period of a high frequency transmission signal The block diagram which shows the other internal structure of a transmission signal generation part in detail Explanatory drawing which shows the relationship between the transmission section of a high frequency transmission signal, a transmission period, and a measurement range Block diagram showing in detail another internal configuration of the signal processing unit Explanatory drawing which shows the simulation result which estimated the arrival direction of the target based on the output of a certain correlation matrix production | generation part, (a) The simulation result whose coherent integration number is 30 times and non-coherent integration number is 120 times, ( b) Simulation result with 60 coherent integrals and 60 non-coherent integrals, (c) 120 coherent integrals, 30 non-coherent integrals, non-coherent integrals and non-coherent integrals Simulation results under the condition that the products of coherent integral numbers are constant An explanatory diagram showing a simulation result in which the arrival direction of the target is estimated based on an output of a certain correlation matrix generation unit, (a) a simulation result in which the coherent integration number is 240 times and the non-coherent integration number is 15 times, ( b) Simulation result under the condition that the coherent integral number is 480, the non-coherent integral number is 7, and the product of the coherent integral number and the non-coherent integral number is constant. Explanatory drawing which shows the simulation result which estimated the arrival direction of the target based on the output of the addition part selected by the output selection part An explanatory diagram showing a simulation result indicating an accumulated probability distribution indicating an error in the estimation accuracy of the arrival direction with respect to the position of each target, (a) an accumulated probability distribution indicating an error in the estimation accuracy of the arrival direction with respect to the target # 1, and (b). Cumulative probability distribution showing error in direction of arrival estimation accuracy for target # 2 An explanatory diagram showing a simulation result indicating an accumulated probability distribution indicating an error in the estimation accuracy of the arrival direction with respect to each target position, (a) a cumulative probability distribution indicating an error in the estimation accuracy of the arrival direction with respect to the target # 3, and (b). Cumulative probability distribution showing error in direction of arrival estimation accuracy for target # 4 The block diagram which shows in detail the internal structure of the radar apparatus of the modification 1 of 1st Embodiment. The block diagram which shows in detail the internal structure of the radar apparatus of the modification 2 of 1st Embodiment. The block diagram which shows in detail the internal structure of the radar apparatus of the modification 3 of 1st Embodiment. A signal including reflected waves from four different moving speed targets is received by the radar device for the signal repeatedly transmitted by the radar device for a total of 480 pulse compression codes. After the pulse compression processing, each range gate is received. , An example of a result of integration processing using a combination of different coherent integration numbers and non-coherent integration numbers, (a) a detection result in which the coherent integration number is 30 and the non-coherent integration number is 16, ) Detection result with 60 coherent integrals and 8 noncoherent integrals, (c) Detection with 120 coherent integrals and 4 noncoherent integrals, (d) Coherent As a result of detection that the integration number is 240 times and the non-coherent integration number is 2, (e) the coherent integration number is 480 times. There, the detection result noncoherent integration number is one

  Before describing each embodiment of the radar apparatus according to the present invention, a pulse compression process and a complementary code will be briefly described below as technical contents which are the premise of each embodiment described later.

(Pulse compression processing)
First, the pulse compression process will be described. For example, pulse compression that transmits a high-frequency radar transmission signal using a pulse compression code including at least one of a Barker code, an M-sequence code, and a complementary code as the pulse wave or pulse modulated wave having the low-range sidelobe characteristics described above Radar is known.

  Pulse compression is a signal after the radar device receives a reflected wave by pulse-modulating or phase-modulating a plurality of pulse signals using the above-described pulse compression code, and using an equivalently wide pulse width signal. In the processing, the received signal is demodulated, and the correlation value is calculated by converting (compressing) the original signal with a narrow pulse width by obtaining the correlation with the pulse compression code used for transmission. According to the pulse compression, the received power is increased equivalently, the detection distance of the target is increased, and the distance estimation accuracy with respect to the detection distance is further improved.

(Complementary code)
Next, complementary codes will be described. The complementary code is a code using a plurality of complementary code sequences (a _n , b _n ), for example, two pairs. Complementary code in each autocorrelation calculation result of one of the complementary code sequences a _n and the other of the complementary code sequence b _n, by the addition of the auto-correlation calculation result to match the delay time tau [sec], the range side lobe It has the property of becoming zero. The parameter n is n = 1, 2,. The parameter L indicates the code sequence length or simply the code length.

A method for generating a complementary code is disclosed, for example, in Reference Non-Patent Document 1 below.
(Reference Non-Patent Document 1) BUDISIN, S. Z, “NEW COMPLEMENTARY PAIRS OF SEQUENCES”, Electron. Lett. , 26, (13), pp. 881-883 (1990)

Complementary code sequences _(a n, _{b n)} of the autocorrelation value calculation result of one of the complementary code sequence _{a n} is calculated according to Equation (5). Autocorrelation value calculation result of the other complementary code sequence b _n is calculated according to Equation (6). The parameter R indicates the autocorrelation value calculation result. However, when n> L or n <1, the complementary code sequences a _n and b _n are set to zero (that is, when n> L or n <1, a _n = 0 and b _n = 0). The asterisk * indicates a complex conjugate operator.

Equation (5) autocorrelation value calculation result of the computed complementary code sequence a _n R _{aa (τ)} in accordance with the delay time (or shift time) tau peak occurs when is zero, except the delay time tau is zero Then there is a range side lobe. Similarly, the autocorrelation value calculation result R _bb (τ) of the complementary code sequence b _n calculated according to Equation (6) has a peak when the delay time τ is zero, and when the delay time τ is not zero, Range side lobes are present.

The sum of these autocorrelation value calculation results (R _aa (τ), R _bb (τ)) has a peak when the delay time τ is zero, and there is a range sidelobe when the delay time τ is not zero. It becomes zero without. Hereinafter, a peak that occurs when the delay time τ is zero is referred to as a “main lobe”. This relationship is shown in Equation (7).

  In the complementary code, the peak sidelobe level can be reduced with a shorter code length from the above-described autocorrelation characteristics. For this reason, with a complementary code using a short code length, the reception dynamic range can be reduced even when a signal in which reflected waves from a target at a short distance and a target at a long distance are mixed is received.

Further, by using a Barker code or an M-sequence code having a code length L as a complementary code, the peak side lobe ratio is given by 20 log ₁₀ (1 / L) [dB]. For this reason, the complementary code can obtain excellent low-range sidelobe characteristics by increasing the code length L.

(Embodiments of the present invention)
Next, embodiments of the present invention will be described with reference to the drawings.

  In the following description, a radar apparatus according to the present invention includes a plurality of receiving antennas that receive a reflected wave signal from a target. For example, a configuration having four receiving antennas (array antennas) is shown as an example, but the present invention is not limited to this. The four reception antennas may be four reception antenna elements.

(First embodiment)
The configuration and operation of the radar apparatus 1 according to the first embodiment will be described with reference to FIGS. FIG. 1 is a block diagram schematically showing the internal configuration of the radar apparatus 1 according to the first embodiment. FIG. 2 is a block diagram showing in detail the internal configuration of the radar apparatus 1 according to the first embodiment. FIG. 3 is an explanatory diagram showing the relationship between the transmission interval Tw of the high-frequency transmission signal and the transmission cycle Tr. FIG. 4 is a block diagram showing in detail another internal configuration of the transmission signal generation unit 4.

  FIG. 5 is an explanatory diagram showing the relationship among the transmission interval Tw, the transmission cycle Tr, and the measurement range of the high-frequency transmission signal. FIG. 6 is a block diagram showing in detail another internal configuration of the signal processing unit.

  The radar apparatus 1 transmits (emits) the high-frequency transmission signal generated by the radar transmission unit 2 from the transmission antenna AN1. The radar apparatus 1 receives reflected wave signals, which are high-frequency transmission signals reflected by the target, for example, at four receiving antennas AN2 to AN2-4 (not shown, the same applies hereinafter) as shown in FIG. The radar apparatus 1 detects the presence or absence of a target by signal processing of signals received by the receiving antennas AN2 to AN2-4.

  The target is an object to be detected by the radar apparatus 1 and includes, for example, a car or a person, and the same applies to each of the following embodiments.

  First, the configuration of each part of the radar apparatus 1 will be briefly described.

  As shown in FIG. 2, the radar apparatus 1 includes a radar transmitter 2 and a radar receiver 3. The radar transmission unit 2 includes a transmission signal generation unit 4 and a transmission RF unit 5 connected to the transmission antenna AN1. The radar transmitter 2 and the radar receiver 3 are connected to a reference signal oscillator Lo, and a signal is supplied from the reference signal oscillator Lo so that the processing of the radar transmitter 2 and the radar receiver 3 is synchronized.

  The radar receiver 3 includes D antenna system processors 11-1 to 11-D, p correlation matrix generators 21-1 to 21-p, p adders 22-1 to 22-p, and outputs. A selection control unit 23 and an output selection unit 24 are included. Parameter D and parameter p are integers of 2 or more. Each antenna system processing unit has the same configuration, and in the following description, the antenna system processing unit 11-1 will be described as an example.

  The antenna system processing unit 11-1 includes at least a reception RF unit 12, a correlation value calculation unit 19, and p coherent integration units 20-1-1-1 to 20-1-p connected to the reception antenna AN2.

(Radar transmitter)
Next, the configuration of each part of the radar transmitter 2 will be described in detail. Hereinafter, the configuration of each part of the radar transmitter 2 will be described with reference to FIG.

  As shown in FIG. 3, the radar transmitter 2 includes a transmission signal generator 4 and a transmission RF unit 5 to which a transmission antenna AN1 is connected.

  The transmission signal generation unit 4 includes a code generation unit 6, a modulation unit 7, and an LPF (Low Pass Filter) 8. In FIG. 3, the transmission signal generation unit 4 is configured to include the LPF 8, but the LPF 8 may be configured in the radar transmission unit 2 independently of the transmission signal generation unit 4.

  The transmission RF unit 5 includes a frequency conversion unit 9 and an amplifier 10.

  Next, the operation of each part of the radar transmitter 2 will be described in detail.

  The transmission signal generation unit 4 generates a signal obtained by multiplying the reference signal by a predetermined factor based on the reference signal generated by the reference signal oscillator Lo. Each unit of the transmission signal generation unit 4 operates based on the generated signal.

Transmission signal generating unit 4, by the modulation code sequence a _n of the code length L, the transmission signal r baseband shown in equation (8) (k, M) ( also referred to as pulse compression code) to periodically generate. Here, the parameter n = 1, · · ·, L, and the parameter L represents the code length of the code sequence _{a n.} The parameter j is an imaginary unit that satisfies j ² = −1.

  The baseband transmission signal r (k, M) shown in Expression (8) indicates a transmission signal at a discrete time k in the Mth transmission cycle Tr, and an in-phase signal component Ir (k, M), This is indicated by the addition result with the orthogonal signal component Qr (k, M) multiplied by the imaginary unit j.

The transmission signal generated by the transmission signal generator 4, as shown in FIG. 3, for example, the transmission interval Tw [sec] of each transmission period Tr, to the code sequence a _n of the code length L, 1 single There are No [number] samples per pulse code. Therefore, Nr (= No × L) samples are included in the transmission interval Tw. Further, in the non-transmission section (Tr−Tw) [seconds] of each transmission cycle Tr, there are Nu [number] samples as baseband transmission signals. The parameter k is a discrete time.

Code generator 6, for each transmission period Tr, to generate a transmission code for pulse compression code sequence a _n of the code length L. For example, the transmission code is preferably a code including any one of the Barker code sequence and the M-sequence code in addition to the above-described code sequence constituting the pair of complementary codes.

Code generator 6 outputs a transmission code of the generated code sequence a _n to the modulator 7. Hereinafter, the transmission code of the code sequence a _n, are described as conveniently transmitted symbols a _n.

Incidentally, the code generating unit 6, in the transmission period Tr, the transmission code when generating a pair of complementary code as a _n, using two transmission period (2Tr), a pair alternately every transmission cycle code Pn, Each Qn is generated.

That is, in the M-th transmission period (Tr), transmits the code Pn as a pulse compression code _a n (M), as followed by the (M + 1) th transmission period (Tr) in the pulse compression code _a n (M + 1) , Qn is transmitted. Thereafter, in the (M + 2) th and subsequent transmission cycles, the repetition codes Pn and Qn are generated in the same manner with the Mth transmission cycle and the (M + 1) th two transmission cycles as one unit.

Modulation unit 7 inputs the transmission code a _n output by the code generator 6. Modulation unit 7, the pulse modulation of the transmitted symbols a _n input, generates a transmission signal r baseband shown in equation (8) (k, M) . The pulse modulation is amplitude modulation, ASK (Amplitude Shift Keying) or phase modulation (PSK (Phase Shift Keying)), and the modulation unit 7 generates a transmission signal r (k) generated via the LPF 8. , M), a transmission signal r (k, M) that is equal to or smaller than a preset limit band is output to the transmission RF unit 5.

  The transmission RF unit 5 generates a signal obtained by multiplying the reference signal by a predetermined multiple based on the reference signal generated by the reference signal oscillator Lo. The transmission RF unit 5 operates based on the generated signal.

  The frequency conversion unit 9 receives the transmission signal r (k, M) generated by the transmission signal generation unit 4 and up-converts the input baseband transmission signal r (k, M) to thereby generate a carrier frequency. A high-frequency transmission signal in a band is generated. The frequency conversion unit 9 outputs the generated high-frequency transmission signal to the amplifier 10.

  The amplifier 10 receives the output high-frequency transmission signal, amplifies the level of the input high-frequency transmission signal to a predetermined level, and outputs it to the transmission antenna AN1. This amplified high-frequency transmission signal is transmitted by radiation to the space via the transmission antenna AN1.

  The transmission antenna AN1 transmits the high-frequency transmission signal output from the transmission RF unit 5 by radiating it to the space. As shown in FIG. 4, the high-frequency transmission signal is transmitted during the transmission interval Tw in the transmission cycle Tr, and is not transmitted during the non-transmission interval (Tr−Tw).

  In addition, the reference signal generated by the reference signal oscillator Lo is transmitted to the transmission RF unit 5 and the reception RF units 12 to 12-4 (not shown, the same applies hereinafter) of the antenna system processing units 11-1 to 11-4. A signal multiplied by a predetermined factor is supplied in common. Thereby, the operation | movement which each transmission RF part 5 and reception RF part 12-12-4 synchronized was performed.

Note that without providing the code generating unit 6 described above to the transmission signal generator 4, as shown in FIG. 4, provided with a transmission code storage unit CM for previously storing a transmission code a _n generated by the transmission signal generator 4 Also good. Furthermore, in order to generate a complementary code by the transmission signal generation unit 4, it is preferable to store a pair of complementary codes, for example, a pair of codes Pn and Qn alternately for each transmission cycle in the transmission code storage unit CM. .

  Note that the transmission code storage unit CM shown in FIG. 4 is not limited to the first embodiment, and can be similarly applied to each embodiment described later. As shown in FIG. 4, the transmission signal generation unit 4 includes a transmission code storage unit CM, a transmission code control unit CT, a modulation unit 7, and an LPF 8.

4, Send code controller CT is based on a reference signal output by the reference signal oscillator Lo with the multiplied signal to the predetermined number of times, each transmission period Tr, which constitutes the transmission code a _{n (or} complementary code The code Pn and the transmission code Qn) are read cyclically from the transmission code storage unit CM and output to the modulation unit 7. Since the subsequent operations output to the modulation unit 7 are the same as those of the modulation unit 7 and the LPF 8 described above, description of the operations is omitted.

(Radar receiver)
Next, the configuration of each part of the radar receiver 3 will be described in detail.

  The radar receiving unit 3 includes a plurality of, for example, four antenna system processing units, and one receiving antenna is connected to each antenna system processing unit, and constitutes an array antenna including four receiving antennas. Hereinafter, the configuration of each part of the radar receiver 3 will be described with reference to FIG.

  In the radar receiver 3 shown in FIG. 2, the parameter D indicating the number of antenna system processors is 4, and the parameter P indicating the number of coherent integrators, correlation matrix generators, and adders is 3. In each of the following embodiments including the first embodiment, the number of coherent integration units, correlation matrix generation units, and addition units is the same.

  As shown in FIG. 2, the radar receiving unit 3 includes four antenna system processing units 11-1 to 11-4 and first to third correlation matrices provided corresponding to the number of receiving antennas constituting the array antenna. The configuration includes generation units 21-1 to 21-3, first to third addition units 22-1 to 22-3, an output selection control unit 23, an output selection unit 24, and an arrival direction estimation unit 25.

  In the following description, since the configuration and operation of each part of the four antenna system processing units 11-1 to 11-4 are the same, the antenna system processing unit 11-1 will be described as an example, and each embodiment described later will be described. The same applies to. Furthermore, regarding the reference numerals of the respective parts of the antenna system processing unit 11-1 shown in FIG. 2, for example, the reception RF unit 12 is simply described instead of the reception RF unit 12-1.

  The same applies to the description of the reference numerals of the respective units of the respective antenna system processing units shown in FIGS. In addition, in the description of each part of the antenna system processing unit other than the first antenna system processing unit as necessary, for example, the reception RF unit 12-2 is described.

  The antenna system processing unit 11-1 includes a reception RF unit 12 and a signal processing unit 13 to which the reception antenna AN2 is connected. The reception RF unit 12 includes an amplifier 14, a frequency conversion unit 15, and a quadrature detection unit 16. The signal processing unit 13 includes A / D conversion units 17 and 18, a correlation value calculation unit 19, and first to third coherent integration units 20-1 to 20-3. The signal processing unit 13 periodically calculates each transmission cycle Tr as a signal processing section.

  Next, the operation of each part of the radar receiver 3 will be described in detail.

  The reception antenna AN2 receives a reflected wave signal obtained by reflecting the high-frequency transmission signal transmitted from the radar transmitter 2 by the target. The reception signal received by the reception antenna AN2 is output to the reception RF unit 12.

  Similarly to the transmission RF unit 5, the reception RF unit 12 generates a signal obtained by multiplying the reference signal by a predetermined number based on the reference signal generated by the reference signal oscillator Lo. The reception RF unit 12 operates based on the generated signal.

  The amplifier 14 receives a high-frequency band reception signal received by the reception antenna AN2, amplifies the level of the input reception signal, and outputs the amplified signal to the frequency converter 15.

  The frequency conversion unit 15 receives the high frequency band reception signal output from the amplifier 14, down-converts the input high frequency band reception signal to baseband, and outputs the down-converted reception signal to the quadrature detection unit 16. To do.

  The quadrature detection unit 16 performs baseband detection on the baseband received signal output from the frequency conversion unit 15, thereby forming a baseband using an in-phase signal and a quadrate signal. The received signal is generated. The quadrature detection unit 16 outputs the in-phase signal of the generated reception signals to the A / D conversion unit 17 and outputs the quadrature signal to the A / D conversion unit 18.

  The A / D conversion unit 17 converts the in-phase signal of analog data into digital data by sampling the baseband in-phase signal output from the quadrature detection unit 16 at each discrete time k. The A / D conversion unit 17 outputs the in-phase signal component of the digital data converted at each discrete time k to the correlation value calculation unit 19 as a discrete sample value.

  The A / D converter 17 converts the baseband in-phase signal output from the quadrature detector 16 into one pulse width (pulse time) Tp (= Tw / L) of the transmission signal generated by the transmission signal generator 4. ) Is sampled at a rate of Ns [pieces]. Accordingly, the sampling rate of the A / D conversion unit 17 is Ns / Tp, and the number of oversamples per pulse is Ns [pieces].

  Similarly, the A / D conversion unit 18 converts the quadrature signal of analog data into digital data by sampling the baseband quadrature signal output from the quadrature detection unit 16 at each discrete time k. The A / D converter 18 outputs the orthogonal signal component of the digital data converted at each discrete time k to the correlation value calculator 19 as a discrete sample value.

  The A / D converter 18 converts the baseband quadrature signal output from the quadrature detector 16 into one pulse width (pulse time) Tp (= Tw / L) of the transmission signal generated by the transmission signal generator 4. Sampling is performed at a ratio of Ns [pieces]. Therefore, the sampling rate of the A / D conversion unit 18 is Ns / Tp, and the number of oversamples per pulse is Ns [pieces].

  The received signal at the discrete time k in the Mth transmission cycle Tr converted by the A / D converters 17 and 18 is the in-phase signal component I (k, M) and the quadrature signal component Q (k , M), and expressed as a complex signal x (k, M) in equation (9). The same applies to the following embodiments. Here, j is an imaginary unit.

  The parameter k indicates the timing when sampling is performed by the A / D converters 17 and 18, and the discrete time k = 1 indicates the start time of each transmission cycle Tr. The discrete time k = Ns × (Nr / No) indicates the end point of the transmission section Tw in each transmission cycle Tr.

  Further, the discrete time k = (Nr + Nu) × (Ns / No) indicates the time immediately before the end of each transmission cycle Tr. That is, the discrete time k is based on the start timing of the radar transmission cycle (Tr) as the reference (k = 1). The A / D converters 17 and 18 periodically measure until a discrete time k (= (Nr + Nu) × Ns / No) that is a sample point before the radar transmission cycle Tr ends. That is, it is a discrete time that satisfies k = 1 to Ns (Nr + Nu) / No. The range of the discrete time k is the same in each embodiment described later.

  The correlation value calculation unit 19 receives the discrete sample values x (k, M) output by the A / D conversion units 17 and 18. In order to establish synchronization with the operation of the transmission signal generation unit 4, the correlation value calculation unit 19 multiplies the reference signal by a predetermined multiple on the basis of the reference signal generated in the reference signal oscillator Lo, similarly to the transmission signal generation unit 4. A signal multiplied by is generated. In FIG. 2, the input of the reference signal to the correlation value calculation unit 19 is omitted.

The correlation computing unit 19, based on a signal multiplying the reference signal to a predetermined multiple, discrete time depending on k, the pulse compression code a _n of the code length L to be transmitted in the M-th transmission period Tr _(M) Are generated periodically. Here, parameters n = 1,..., L. The parameter L is a code length.

Correlation computing unit 19 computes an input discrete sample values x (k, M), the correlation value AC (k, M) of the pulse compression code _a n (M) a. The correlation value AC (k, M) indicates a correlation value at the discrete time k in the Mth transmission cycle Tr.

  Specifically, the correlation value calculation unit 19 calculates the correlation value AC (k, M) according to Equation (10) in each transmission cycle Tr shown in FIG. 5, that is, at discrete times k = 1 to (Nr + Nu) / No. Calculate. The correlation value calculator 19 outputs the correlation values AC (k, M) calculated according to Equation (10) to the first to third coherent integrators 20-1 to 20-3, respectively.

  The first stage of FIG. 5 shows the transmission timing of the high-frequency transmission signal, and the second stage of FIG. 5 shows the reception timing of the reflected wave signal. The reflected wave signal is a wave in which a high-frequency transmission signal transmitted during the transmission section Tw is reflected by the target.

  The correlation value calculation unit 19 calculates at discrete times k = 1 to Ns (Nr + Nu) / No. Note that the measurement range (range k) is further narrowed, for example, from k = Ns (L + 1) to {(Ns (Nr + Nu) / No) −NsL} depending on the range of the target to be measured by the radar apparatus 1. You may be limited.

  As shown in the second stage of FIG. 5, the discrete time k = Ns (L + 1) indicates the next discrete time after the end time of the transmission interval Tw. Further, the discrete time k = Ns (L + 1) is a reception start time at which the reflected wave signal starts after a delay time τ1 from the discrete time k = 0. This delay time τ1 is expressed by Equation (11).

  As shown in the second stage of FIG. 5, the discrete time k = {(Ns (Nr + Nu) / No) −NsL} corresponds to the time before the transmission interval Tw (= Tp × L) from the end time of the transmission cycle Tr. To do. Further, the discrete time k = {(Ns (Nr + Nu) / No) −NsL} is the reception start time when the reflected wave signal is started with a delay time τ2 from the discrete time k = 0. This delay time τ2 is expressed by Equation (12).

  Therefore, the correlation value calculation unit 19 has a correlation value AC (k, M) shown in the mathematical formula (10) described above at least in the range of the discrete time k = Ns (L + 1) to {(Ns (Nr + Nu) / No) −NsL}. May be calculated. Thereby, the radar apparatus 1 can reduce the calculation amount of the correlation value calculation unit 19. That is, the radar apparatus 1 can reduce the power consumption based on the reduction of the calculation amount by the signal processing unit 13. The same applies to the other antenna system processing units 11-2 to 11-4.

  Further, the radar apparatus 1 does not measure in the transmission section Tw of the high-frequency transmission signal, and even if the high-frequency transmission signal directly wraps around the radar receiver 3, it can measure without the influence of the wraparound. By limiting the measurement range (range of discrete time k), first to third coherent integrators 20-1 to 20-3, first to third correlation matrix generators 21-1 to 21-3, and first, which will be described later, are provided. The operations of the third addition units 22-1 to 22-3, the output selection control unit 23, the output selection unit 24, and the arrival direction estimation unit 25 are also limited to the same measurement range.

The first coherent integrator 20-1 receives the correlation value AC (k, M) output from the correlation value calculator 19. The first coherent integration unit 20-1 performs a plurality of (N ₁ times) transmission cycles Tr based on the correlation value AC (k, M) calculated for each discrete time k in the Mth transmission cycle Tr. Coherent integration with an integration number N _{1 is performed} over a period (N ₁ × Tr).

The parameter N ₁ is the number of times of coherent integration by the first coherent integrator 20-1. The first coherent integrator 20-1 performs coherent integration with the parameter p = 1 in Equation (13). The parameter m is a natural number. The parameter q is a delay time.

That is, the first coherent integrator 20-1 performs the (N ₁ ××) from the correlation value AC (k, N ₁ (m−1) +1) in the {N ₁ (m−1) +1} th transmission cycle Tr. m) The correlation value CI ₁ (k, m) is calculated with the timing of the discrete time k aligned using the correlation value AC (k, N ₁ × m) in the _first transmission cycle Tr as a unit. The first coherent integrator 20-1 outputs the calculated correlation value CI ₁ (k, m) to the first correlation matrix generator 21-1. Here, m is a natural number.

The second coherent integrator 20-2 receives the correlation value AC (k, M) output from the correlation value calculator 19. The second coherent integration unit 20-2 uses a plurality of (N ₂ times) transmission cycles Tr based on the correlation value AC (k, M) calculated for each discrete time k in the Mth transmission cycle Tr. Coherent integration with an integration number N _{2 is performed} over a period (N ₂ × Tr).

Parameter _{N 2} is the number of integrations of the coherent integration by the second coherent integrator 20-2. The second coherent integrator 20-2 performs coherent integration with the parameter p = 2 in Expression (13).

That is, the second coherent integrator 20-2 calculates the (N ₂ ××) from the correlation value AC (k, N ₂ (m−1) +1) in the {N ₂ (m−1) +1} th transmission cycle Tr. m) Correlation value CI ₂ (k, m) is calculated by aligning the timings of discrete time k with the correlation value AC (k, N ₂ × m) in the first transmission cycle Tr as a unit. The second coherent integrator 20-2 outputs the calculated correlation value CI ₂ (k, m) to the second correlation matrix generator 21-2. Here, m is a natural number.

The third coherent integrator 20-3 receives the correlation value AC (k, M) output from the correlation value calculator 19. The third coherent integration unit 20-3 uses a plurality of (N ₃ times) transmission cycles Tr based on the correlation value AC (k, M) calculated for each discrete time k in the Mth transmission cycle Tr. Coherent integration with an integration number N _{3 is performed} over a period (N ₃ × Tr).

Parameter _{N 3} is an integral number of the coherent integration by the third coherent integrator 20-3. The third coherent integrator 20-3 performs coherent integration with the parameter p = 3 in Equation (13). The parameter m is a natural number.

That is, the third coherent integrator 20-3 performs the (N ₃ ××) from the correlation value AC (k, N ₃ (m−1) +1) in the {N ₃ (m−1) +1} th transmission cycle Tr. m) The correlation value CI ₃ (k, m) is calculated with the timing of the discrete time k aligned using the correlation value AC (k, N ₃ × m) in the first transmission cycle Tr as a unit. The third coherent integrator 20-3 outputs the calculated correlation value CI ₃ (k, m) to the third correlation matrix generator 21-3. Here, m is a natural number.

The number of integrations N _{1 to} N ₃ in the _first to third coherent integrators 20-1 to 20-3 satisfies Expression (14). Further, the number of integrations N _{1 to} N ₃ corresponds to the signal of each reflected wave from each target moving at high speed, medium speed, and low speed. Further, each of the integration times N _{1 to} N ₃ is set so that a coherent integration gain corresponding to a section with a high time correlation is obtained in order to improve the SNR when receiving the signal of each reflected wave from each target. Has been.

As a result, the radar apparatus 1 converts the reflected wave signal from the target moving at any one of the high speed, the medium speed, and the low speed into P different coherent integration numbers (N _{p in a} section having a high time correlation. Coherent integration is performed using a coherent integration unit having times (p = 1 to P, where P = 3).

  The radar apparatus 1 improves the SNR in receiving the signal of each reflected wave according to the speed of the moving target by the coherent integration described above.

  Furthermore, the radar apparatus 1 improves the distance performance related to the estimation of the target arrival distance.

  Each of the coherent integrators 20-1 to 20-3 may individually perform coherent integration, but a result of coherent integration with a small number of integrals may be used in a coherent integration calculation process with a large number of integrals. Thereby, the radar apparatus 1 can reduce the coherent integration number in the signal processing unit 13, and can reduce the circuit scale of the signal processing unit 13.

  For example, as shown in FIG. 6, in the signal processing unit 13, the integration result of the first coherent integration unit 20-1 is input to the second coherent integration unit 20-2, and the integration result of the second coherent integration unit 20-2. Is input to the third coherent integrator 20-3.

  When the integration number of each of the coherent integrators 20-1 to 20-3 satisfies Expression (15), the p-th coherent integrator performs coherent integration using the output of the (p-1) -th coherent integrator. The parameters α and β are natural numbers of 2 or more.

If the (p-1) the ratio of the number of integration _{N p} of the p coherent integrator for integrating the number _{N p-1} of the coherent integrator _(N p _{/ N p-1)} is not a natural number, the first p coherent integrator, the greatest common divisor of N _p and N _p-1 as a unit, to coherent integration using the output of the (p-1) coherent integrator. However, between the parameters _{N p} and a parameter _{N p-1,} equation (16) is satisfied. Thereby, the radar apparatus 1 can reduce the circuit scale of the signal processing unit 13.

The first correlation matrix generator 21-1 outputs the correlation values CI ₁ ¹ (k, m),..., CI ₁ ^D (output from the first coherent integrators of the antenna system processors 11 to 11-4 ( k, m). The first correlation matrix generation unit 21-1 receives the reflected wave signal from the target based on the input correlation values CI ₁ ¹ (k, m),..., CI ₁ ^D (k, m). In order to detect a phase difference between them, a correlation matrix H ₁ (k, m) is generated at each discrete time k.

The correlation matrix H ₁ (k, m) is generated according to Equation (17) where the parameter p = 1. In Equation (17), the superscript H is an operator representing complex conjugate transpose. In each embodiment, the correlation matrix is phase difference information that is generated based on the outputs of the plurality of coherent integration units of the plurality of antenna system processing units and is caused by the arrangement of the reception antennas.

Further, the first correlation matrix generation unit 21-1, a period of _{N f} times the transmission period Tr _{(N f} × Tr), was obtained for each period of the transmission period Tr of _{one N (N 1} × Tr) Each correlation matrix is averaged according to Equation (18) where the parameter p = 1.

Parameter _{D 1} represents the number of the correlation matrix that is averaged by the first correlation matrix generation unit 21-1, satisfies the formula (19) is a parameter p = 1. The first correlation matrix generation unit 21-1 outputs the averaged correlation matrix to the first to third addition units 22-1 to 22-3 and the output selection control unit 23.

The second correlation matrix generation unit 21-2 outputs the correlation values CI ₂ ¹ (k, m),..., CI ₂ ^D (output from the second coherent integration units of the antenna system processing units 11 to 11-4. k, m). The second correlation matrix generation unit 21-2 receives a reflected wave signal from the target based on each input correlation value CI ₂ ¹ (k, m),..., CI ₂ ^D (k, m). In order to detect a phase difference between them, a correlation matrix H ₂ (k, m) is generated at each discrete time k.

The correlation matrix H ₂ (k, m) is generated according to Equation (17) where the parameter p = 2.

Further, the second correlation matrix generation unit 21-2, a period of _{N f} times the transmission period Tr _{(N f} × Tr), was obtained for each period of the transmission period Tr of _{N 2} times _{(N 2} × Tr) Each correlation matrix is averaged according to Equation (18) where the parameter p = 2.

Parameter _{D 2} represents the number of the correlation matrix that is averaged by the second correlation matrix generation unit 21-2, satisfies the formula (19) is a parameter p = 2. The second correlation matrix generation unit 21-2 outputs the addition-averaged correlation matrix to the second and third addition units 22-2 and 22-3 and the output selection control unit 23.

The third correlation matrix generation unit 21-3 outputs the correlation values CI ₃ ¹ (k, m),..., CI ₃ ^D (output from the third coherent integration units of the antenna system processing units 11 to 11-4. k, m). The third correlation matrix generation unit 21-3 receives the reflected wave signal from the target based on the input correlation values CI ₃ ¹ (k, m),..., CI ₃ ^D (k, m). In order to detect a phase difference between them, a correlation matrix H ₃ (k, m) is generated at each discrete time k.

The correlation matrix H ₃ (k, m) is generated according to Equation (17) where the parameter p = 3.

Furthermore, third correlation matrix generation unit 21-3, a period of _{N f} times the transmission period Tr _{(N f} × Tr), was obtained for each period of the transmission period Tr of _{N 3} times _{(N 3} × Tr) Each correlation matrix is averaged according to Equation (18) with parameter p = 3.

Parameter _{D 3} represents the number of the correlation matrix that is averaged by the third correlation matrix generation unit 21-3, satisfies the formula (18) is a parameter p = 3. The third correlation matrix generation unit 21-3 outputs the correlation matrix obtained by the averaging process to the third addition unit 22-3 and the output selection control unit 23.

The parameter N _f is preferably the least common multiple of the parameters N ₁ , N ₂ , N ₃ or an integer multiple thereof, and D ₁ , D ₂ , D ₃ are preferably integer values, but the equation (19) is not an integer value. In this case, for example, a decimal point value truncation process is added and combined into an integer value, so that it can be similarly applied.

  Each correlation matrix generation unit uses Equation (20) instead of Equation (17) and is received by the reception antenna of any one of the plurality of antenna system processing units 11 to 11-4. A correlation vector may be generated using the phase of the signal as a reference phase.

  In Equation (20), the superscript asterisk (*) represents a complex conjugate operator. Thereby, the radar apparatus 1 can reduce the amount of calculation of each correlation matrix production | generation part, and can detect the phase difference between the receiving antennas of the signal of the reflected wave from a target easily.

The first addition unit 22-1 receives the correlation matrix B ₁ (k) generated according to Equation (18) where the parameter p = 1. The first addition unit 22-1 adds the input correlation matrix B ₁ (k) according to Equation (21) in which the parameter p = 1. In addition, in order to add a correlation matrix, the 1st addition part 22-1 may multiply and add the weighting coefficient proportional to the magnitude | size of a correlation output. The first addition unit 22-1 outputs the added correlation matrix A ₁ (k) to the output selection unit 24.

The second addition unit 22-2 receives the correlation matrices B ₁ (k) and B ₂ (k) generated according to Equation (18) where the parameters p = 1 and p = 2. The second addition unit 22-2 adds the input correlation matrices B ₁ (k) and B ₂ (k) according to Equation (21) in which the parameter p = 2. Note that the second adding unit 22-2 may add by multiplying a weighting coefficient proportional to the magnitude of the correlation output in order to add the correlation matrix. The second addition unit 22-2 outputs the added correlation matrix A ₂ (k) to the output selection unit 24.

The third adding unit 22-3 generates correlation matrices B ₁ (k), B ₂ (k), and B ₃ (k) generated according to Equation (18) with parameters p = 1, p = 2, and p = 3. Enter. The third addition unit 22-3 adds the input correlation matrices B ₁ (k), B ₂ (k), and B ₃ (k) according to Expression (21) in which the parameter p = 3. Note that the third adder 22-3 may add by multiplying a weighting coefficient proportional to the magnitude of the correlation output in order to add the correlation matrix. The third addition unit 22-3 outputs the added correlation matrix A ₃ (k) to the output selection unit 24.

The output selection control unit 23 inputs correlation matrices B ₁ (k), B ₂ (k), and B ₃ (k) generated according to Equation (18) with parameters p = 1, p = 2, and p = 3. To do. Based on each input correlation matrix, the output selection control unit 23 selects a correlation matrix generation unit that generates a correlation matrix having the maximum coherent integration gain at each discrete time k out of the correlation matrices.

  Specifically, the output selection control unit 23 generates a correlation matrix that maximizes the sum of the diagonal components of the correlation matrix corresponding to the average received power component after coherent integration, based on each input correlation matrix. The index of the correlation matrix generation unit is selected for each discrete time k. That is, the output selection control unit 23 selects the index select_index (k) of the correlation matrix generation unit according to Expression (22).

In Equation (22), diag [B _p (k)] is an operator that calculates the sum of the diagonal components of the correlation matrix B _p (k). The output selection control unit 23 outputs the selected index select_index (k) to the output selection unit 24 at each discrete time k.

  The output selection control unit 23 generates a correlation matrix having the maximum coherent integration gain for each discrete time k according to Equation (22) using the outputs of the coherent integration units of all D antenna system processing units. A correlation matrix generator was selected. Note that the output selection control unit 23 uses the output of each coherent integration unit of one or a part of the antenna system processing units instead of D, and the coherent integration gain is maximized at each discrete time k according to Equation (22). You may select the correlation matrix production | generation part which produced | generated these correlation matrices.

Note that the output selection control unit 23 uses the function coeff (x _p ) based on the value of the ratio shown in Equation (23) as a coefficient in the sum of the diagonal components of the correlation matrix in diag [B _p (k)]. Of the multiplied values (coeff (x _p ) × diag [B _p (k)]), the maximum index select_index (k) of the correlation matrix generation unit may be selected according to the equation (24). Note that the function coeff (x _p ) needs to satisfy the formula (25) or the formula (26), for example.

In the output selection control unit 23, the process of multiplying the function coeff (x _p ) as a coefficient results in a situation where the coherent integration gain by coherent integration exceeds the increase in the variance value of the noise component.

  For this reason, the radar apparatus 1 selects an adder output corresponding to an index select_index (k) having a larger number of coherent integrations when ideally coherent integration is performed.

  Thereby, the radar apparatus 1 can reduce target selection errors such as detecting a target moving at high speed as a target moving at low speed.

The output selection control unit 23 does not estimate the arrival direction of the target at the discrete time k when the diag [B _p (k)] at the discrete time k does not satisfy the predetermined level. May be controlled. Thereby, the radar apparatus 1 does not require redundant calculation at the discrete time k when no target is detected, and the processing delay in the radar receiver 3 can be reduced.

The output selection unit 24 outputs the index select_index (k) output by the output selection control unit 23 and the correlation matrix A ₁ output by the first to third addition units 22-1 to 22-3 at each discrete time k. Input (k) to A ₃ (k).

The output selection unit 24 outputs the output A _{select_index (k)} (k) of the addition unit corresponding to the correlation matrix generation unit of the index select_index (k) for each discrete time k based on the input index select_index (k). select. The output selection unit 24 outputs the output A _{select_index (k)} (k) of the addition unit selected at each discrete time k to the arrival direction estimation unit 25.

For example, since the index select_index (k) selected by the output selection control unit 23 is 1, the output selection unit 24 corresponds to the first correlation matrix generation unit 21-1 and the first addition unit 22-1 corresponding to one-to-one. The correlation matrix A ₁ (k) is selected.

For example, since the index select_index (k) selected by the output selection control unit 23 is 2, the output selection unit 24 includes a second correlation matrix generation unit 21-2 and a second addition unit 22-2 corresponding one-to-one. The correlation matrix A ₂ (k) that is the output of is selected.

For example, since the index select_index (k) selected by the output selection control unit 23 is 3, the output selection unit 24 corresponds to the third correlation matrix generation unit 21-3 and the third addition unit 22-3 corresponding one-to-one. The correlation matrix A ₃ (k) that is the output of is selected.

The arrival direction estimation unit 25 receives the correlation matrix A _{select_index (k)} (k) output by the output selection unit 24 at each discrete time k. The arrival direction estimation unit 25 estimates the arrival direction of the target based on the correlation matrix A _{select_index (k)} (k) input at every discrete time k.

  Note that the direction of arrival estimation to the target by the direction-of-arrival estimation unit 25 is a known technique, and can be realized using, for example, an estimation method using an array antenna disclosed in Reference Non-Patent Document 2 below. It is.

  (Reference Non-Patent Document 2) JAMES A. Cadzow, “Direction of Arrival Estimating Signal Subspace Modeling”, IEEE, Vol. 64-79 (1992)

(Simulation result of radar apparatus according to the present invention)
FIG. 7 is an explanatory diagram illustrating a simulation result in which the arrival direction of the target is estimated based on the output of a certain correlation matrix generation unit. FIG. 5A shows the simulation result when the coherent integration number is 30 and the non-coherent integration number is 120. FIG. 5B shows a simulation result in which the coherent integration number is 60 times and the non-coherent integration number is 60 times. FIG. 4C shows the simulation result under the condition that the coherent integration number is 120, the non-coherent integration number is 30, and the product of the coherent integration number and the non-coherent integration number is constant.

  FIG. 8 is an explanatory diagram illustrating a simulation result in which the arrival direction of the target is estimated based on the output of a certain correlation matrix generation unit. FIG. 9A shows the simulation result when the coherent integration number is 240 times and the non-coherent integration number is 15 times. FIG. 4B shows the simulation result under the condition that the coherent integral number is 480, the non-coherent integral number is 7, and the product of the coherent integral number and the non-coherent integral number is constant.

  FIG. 9 is an explanatory diagram illustrating a simulation result in which the arrival direction of the target is estimated based on the output of the addition unit selected by the output selection unit 24.

7 to 9, the horizontal axis represents the distance r from the radar apparatus 1 to the target converted according to Equation (27) based on the discrete time k, and the vertical axis represents the maximum received power obtained by the arrival direction estimation unit 25. The relative received power value [dB] in the level direction is shown. The constant value C represents the speed of light≈3 × 10 ⁸ [m / s], the parameter L represents the code length, and the parameter Ns represents the number of oversamples for one pulse width.

  In the simulation results shown in FIG. 7 to FIG. 9, the targets are target # 1 moving at high speed, target # 2 moving at medium speed, target # 3 moving at low speed, and target moving at walking speed. # 4 is assumed. The parameters for each target are as follows.

  The moving speed of the target # 1 is 80 [km / h], the position of the target # 1 is 37 [m] from the radar apparatus 1, and the estimated arrival direction of the target # 1 is 20 [degree].

  The moving speed of the target # 2 is 40 [km / h], the position of the target # 2 is the location of the radar apparatus 1 to 38 [m], and the estimated arrival direction of the target # 2 is 10 [degree].

  The moving speed of the target # 3 is 20 [km / h], the position of the target # 3 is the location of the radar apparatus 1 to 39 [m], and the estimated arrival direction of the target # 3 is 0 [degree].

  The moving speed of the target # 4 is 5 [km / h], the position of the target # 4 is the location of the radar apparatus 1 to 40 [m], and the estimated arrival direction of the target # 4 is −10 [degree].

  7 and FIG. 8, for example, in FIG. 7A and FIG. 7B where the number of non-coherent integrations is large, the noise components at discrete times where there is no target are relatively suppressed.

  However, in FIGS. 7A and 7B, since the coherent integration number is small, when the reception SNR of the reception signal of the reflected wave from the target is low, a sufficient signal level may not be obtained even if the coherent integration gain is obtained. .

  On the other hand, in FIG. 7C and FIG. 8, since the non-coherent integration number is small, the noise component at the discrete time when the target does not exist is not sufficiently suppressed. Furthermore, since the number of coherent integrations is large, for example, the coherent integration is performed until a period with a short time correlation in addition to a period with a high time correlation of target # 1 moving at high speed.

  As a result, the relative received power value of target # 1 and the like moving at high speed is buried in the noise component. However, since the noise component is not sufficiently suppressed, the reception SNR of the reception signal of the reflected wave from the target is low. When the number of coherent integrations is large, the reception level may increase, and the optimum coherent integration number may be selected incorrectly.

  In the simulation result of FIG. 9, the output selection unit 24 selects one of the outputs of the addition units 22-1 to 22-3 according to the output of the output selection control unit 23 at each discrete time k.

  As shown in FIG. 9, the noise component is sufficiently suppressed, and the relative peak level from the reception level of the noise component at the discrete time when no target exists to the reception power value of each target is shown in FIG. A result almost equal to the peak level of the optimum coherent integration number is obtained according to the moving speed of each target in FIG.

  For example, the moving speed of the target # 4 is 5 [km / h], and FIG. 8B shows the highest coherent gain, but the peak value from the maximum noise power component at the discrete time when the target does not exist is 14 dB. On the other hand, the peak level at the target # 4 in FIG.

  FIG. 10 is an explanatory diagram showing a simulation result showing a cumulative probability distribution of estimation errors in the arrival direction with respect to the positions where the targets exist. FIG. 6A shows a cumulative probability distribution of estimation errors in the direction of arrival for target # 1. FIG. 5B shows the cumulative probability distribution of the estimation error of the arrival direction for the target # 2.

  FIG. 11 is an explanatory diagram showing a simulation result showing a cumulative probability distribution of the estimation error of the arrival direction with respect to the presence position of each target. FIG. 6A shows a cumulative probability distribution of estimation errors in the direction of arrival for target # 3. FIG. 5B shows the cumulative probability distribution of the estimation error of the arrival direction with respect to the target # 4.

  FIG. 10 and FIG. 11 show the correlation value calculation at the discrete time k corresponding to the positions (37 [m], 38 [m], 39 [m], 40 [m]) of the targets # 1 to # 4. The estimation accuracy of the result of estimating the direction of arrival based on the output of the unit 19 is shown.

  The solid lines in FIG. 10 and FIG. 11 are cumulative probability distributions indicating estimation errors when the radar apparatus 1 according to the present invention has 500 trials. The horizontal axis indicates the direction estimation error with respect to the arrival direction true value of each reflected wave signal from each target # 1 to # 4, the vertical axis indicates the cumulative probability, and the horizontal axis arrival direction estimation error among 500 trials. The probability of being below is plotted. However, the resolution at the time of arrival direction estimation was set to 0.1 degree.

  The dotted lines in FIGS. 10 and 11 show the relationship between the optimum coherent integration number and the non-coherent integration number for the four different targets in the conventional radar apparatus shown in Patent Document 1, as shown below. It is a direction estimation result in the ideal combination set to.

  The coherent integration number for target # 1 is 30 and the non-coherent integration number for target # 1 is 120. These times were set as an optimal combination with respect to the moving speed of the target # 1.

  The coherent integration number for target # 2 is 60 times, and the non-coherent integration number for target # 2 is 60 times. These times were set as an optimal combination with respect to the moving speed of the target # 2.

  The coherent integration number for target # 3 is 120 times, and the non-coherent integration number for target # 3 is 30 times. These times were set as an optimal combination with respect to the moving speed of the target # 3.

  The coherent integration number for target # 4 is 480 times, and the non-coherent integration number for target # 4 is seven times. These times were set as an optimal combination with respect to the moving speed of the target # 4.

  The following can be said about the estimation accuracy for targets with different moving speeds.

  Specifically, in the target (# 1, # 2) with high moving speed, the radar apparatus 1 according to the present invention has the highest coherent integration gain with respect to the moving speed of the target in the conventional radar apparatus. A characteristic comparable to the setting characteristic can be obtained. For example, the error cumulative probability 90% value is 1 ° for the target # 1 and 1.2 ° for the target # 2.

  In the target (# 3, # 4) having a low moving speed, the radar apparatus 1 according to the present invention has a coherent integration number that is higher than the characteristic of setting the highest coherent integration gain with respect to the moving speed of the target in the conventional radar apparatus. Also, improved characteristics of about 0.1 ° can be obtained.

  This improvement is due to the addition of correlation matrices obtained by different coherent integration numbers in the adder, thereby improving the SNR of the arrival angle information included in the correlation matrix used in the arrival direction estimator and improving the direction estimation accuracy. This is because.

  10 and 11, it can be seen that according to the radar apparatus 1 of the present invention, high coherent gain and non-coherent gain can be obtained regardless of the moving speed of the target.

As described above, in the radar apparatus 1, the total addition number of the correlation matrix in the correlation matrix generation unit and the addition unit functions as non-coherent integration. That is, the total addition number N _non-coherent of the correlation matrices of the p-th correlation matrix generation unit and the p-th addition unit is expressed by Equation (28).

  For this reason, as the number of coherent integrations in the coherent integration unit increases, the number of non-coherent integrations in the correlation matrix generation unit and addition unit using the output increases, that is, the number of additions of the correlation matrix increases. The probability of selection error in the part can be suppressed.

  In addition, even if there is a selection error in the output selector, the output from the correlation example generator with a small number of coherent integrals of the high-speed target is included, so arrival by weighting proportional to the size of the diagonal component of the correlation matrix Corner information is combined to prevent significant degradation of direction estimation performance.

  In addition, the reflected wave from the low speed target has an arrival direction estimation accuracy equal to or higher than the characteristics of the coherent integration number optimized for the conventional low speed target due to the weighting effect of adding the correlation matrix a plurality of times.

  Thereby, the radar apparatus 1 can increase the coherent integration gain regardless of the moving speed of the target, and when receiving the reflected wave signal from the target, the SNR corresponding to the coherent integration number or the non-coherent integration number is obtained. It can be improved. The radar apparatus 1 also has the effect of improving the final target identification performance by suppressing the noise level (noise variance).

  Furthermore, the radar apparatus 1 suppresses the noise power component in the reflected wave signal from the target, reduces the probability of erroneously selecting the target moving at high speed as the target moving at low speed, and detects the target detection accuracy. To improve.

  Even if there is a target detection error, the radar apparatus 1 synthesizes the direction-of-arrival estimation value by weighting in proportion to the output level of the correlation matrix generation unit in the addition unit. Thereby, the radar apparatus 1 combines the arrival angle information by weighting proportional to the size of the diagonal component of the correlation matrix, and prevents a remarkable deterioration of the direction estimation performance.

  In addition, the correlation matrix for the reflected wave signal from the target moving at low speed is obtained from the correlation matrix based on the output from the coherent integrator with a small number of coherent integrals. The correlation matrix based on the output from is weighted by being added a plurality of times. The radar apparatus 1 can obtain the direction-of-arrival estimation accuracy equal to or higher than the characteristics of the coherent integration number optimized for a conventional low-speed target.

  Thereby, the radar apparatus 1 can maintain the estimation accuracy of the arrival direction of the target moving at high speed, and can improve the coherent integration gain of the target moving at low speed with a simple configuration.

(Modification 1 of the first embodiment)
In the first modification of the first embodiment, the output selection control unit controls output selection based on a phase rotation amount that is equal to or less than a predetermined value due to fluctuations in the target moving speed.

  FIG. 12 is a block diagram illustrating in detail the internal configuration of the radar apparatus 1x according to the first modification of the first embodiment. Blocks having the same configuration and operation of each part of the radar apparatus 1x and the radar apparatus 1 of the first embodiment are denoted by the same reference numerals. Hereinafter, in the description of the configuration and operation of the radar apparatus 1x, the contents of the same configuration and operation will be omitted, and contents different from the configuration and operation of the radar apparatus 1 will be described.

  In FIG. 12, a radar apparatus 1x includes a radar transmission unit 2 connected to a transmission antenna AN1, and radar system processing units 11-1 to 11-4 connected to reception antennas AN2 to AN2-4. The receiving unit 3x is included. The radar transmitter 2 and the radar receiver 3x are connected to the reference signal oscillator Lo, and a signal is supplied from the reference signal oscillator Lo, so that the processing of the radar transmitter 2 and the radar receiver 3x is synchronized. Yes.

(Radar receiver)
Next, the configuration of each part of the radar receiver 3x will be described. A plurality of, for example, four array antennas are configured in the radar receiver 3x, as in the radar device 1 of the first embodiment. Further, similarly to the radar apparatus 1 of the first embodiment, each receiving antenna constituting the array antenna is provided for each antenna system processing unit.

  The radar receiving unit 3x has a plurality of, for example, four antenna system processing units, one receiving antenna is connected to each antenna system processing unit, and constitutes an array antenna including four receiving antennas. Hereinafter, the configuration of each part of the radar receiver 3x will be described with reference to FIG. In the radar receiver 3x shown in FIG. 12, the parameter D indicating the number of antenna system processors is 4, similarly to the radar receiver 3 shown in FIG. 2, and each of the coherent integrator, the correlation matrix generator, and the adder The parameter P indicating the number is 3.

  As shown in FIG. 12, the radar receiver 3x includes four antenna system processors 11-1 to 11-4 and first to third correlation matrices provided corresponding to the number of receiving antennas constituting the array antenna. Configuration including generation units 21-1 to 21-3, first to third addition units 22-1 to 22-3, output selection control unit 23x, output selection unit 24, arrival direction estimation unit 25, and Doppler frequency detection unit 26 It is.

The Doppler frequency detection unit 26 varies the moving speed of the target at each discrete time k based on the outputs of the first coherent integration units 20-1 to 20-3 of the antenna system processing units 11-1 to 11-4. The phase rotation amount θ _d (k) resulting from is calculated according to Equation (28) where the parameter p = 1. The Doppler frequency detection unit 26 outputs the calculated phase rotation amount θ _d (k) to the output selection control unit 23x.

The Doppler frequency detection unit 26 calculates the Doppler frequency according to Expression (30) based on the phase rotation amount θ _d (k) calculated according to Expression (29). Here, angle {y} is an operator that calculates the angle (phase) component of y.

Based on the phase rotation amount θ _d (k) calculated according to Equation (29), the Doppler frequency detector 26 calculates the relative movement speed v _d (k) of the target according to Equation (30) where the parameter p = 1. Calculate. In Equation (30), the parameter λ is the wavelength of the carrier frequency for the radar transmitter 2 to transmit a high-frequency transmission signal.

The Doppler frequency detection unit 26 may further calculate the Doppler frequency f _d (k) according to Expression (31) based on the relative movement speed v _d (k) calculated according to Expression (30).

In FIG. 12, the Doppler frequency detection unit 26 uses the outputs of the coherent integration units of all D antenna system processing units in accordance with Equation (29) to obtain the phase rotation amount θ _d (k ) Was calculated. Note that the Doppler frequency detection unit 26 uses the output of each coherent integration unit of one or a part of the antenna system processing units instead of D, according to Equation (28), and the amount of phase rotation θ _{d for} each discrete time k. (K) may be calculated.

The output selection control unit 23x inputs the phase rotation amount θ _d (k) output by the Doppler frequency detection unit 26 at each discrete time k.

The output selection control unit 23x selects the outputs of the plurality of P addition units based on the input phase rotation amount θ _d (k).

Specifically, the output selection control unit 23x generates a correlation matrix that generates a maximum correlation matrix when the phase rotation amount θ _d (k) in the coherent integration period (Tr × N _p ) is equal to or less than a predetermined value TH. Is selected at each discrete time k. The predetermined value TH is preferably in the range of 45 ° to 90 °.

  The output selection control unit 23x outputs the index p of the selected correlation matrix generation unit to the output selection unit 24 as an index select_index (k). Subsequent operations are the same as those in the first embodiment.

  As described above, the radar apparatus 1x according to the first modification of the first embodiment calculates the phase rotation amount due to fluctuations in the moving speed of the target for each discrete time k, and the phase rotation amount is maximum when the phase rotation amount is equal to or less than a predetermined value. A correlation matrix generation unit is selected. Therefore, according to the radar apparatus 1x, compared with the first embodiment, it is possible to reduce the probability of erroneously selecting a target moving at high speed as a target moving at low speed, and improve target detection accuracy.

(Modification 2 of the first embodiment)
In the second modification of the first embodiment, the output selection control unit generates a correlation matrix in which the diagonal component of the addition result of each correlation matrix generated based on each coherent integration result of different coherent integration numbers is maximized. The correlation matrix generation unit selected is selected.

  FIG. 13 is a block diagram illustrating in detail the internal configuration of the radar apparatus 1y according to the second modification of the first embodiment. Blocks having the same configuration and operation of each part of the radar apparatus 1y and the radar apparatus 1 of the first embodiment are denoted by the same reference numerals. Hereinafter, in the description of the configuration and operation of the radar apparatus 1y, the contents of the same configuration and operation will be omitted, and contents different from the configuration and operation of the radar apparatus 1 will be described.

  In FIG. 13, the radar apparatus 1y includes a radar transmission unit 2 connected to the transmission antenna AN1, and antenna system processing units 11-1 to 11-4 connected to the reception antennas AN2 to AN2-4. The receiving unit 3y is included. The radar transmitter 2 and the radar receiver 3y are connected to the reference signal oscillator Lo, and a signal is supplied from the reference signal oscillator Lo so that the processing of the radar transmitter 2 and the radar receiver 3y is synchronized. Yes.

(Radar receiver)
Next, the configuration of each part of the radar receiver 3y will be described. In the radar receiver 3y, a plurality of, for example, four array antennas are configured as in the radar device 1 of the first embodiment. Further, similarly to the radar apparatus 1 of the first embodiment, each receiving antenna constituting the array antenna is provided for each antenna system processing unit.

  The radar receiving unit 3y includes a plurality of, for example, four antenna system processing units, and one receiving antenna is connected to each antenna system processing unit, and constitutes an array antenna including four receiving antennas. Hereinafter, the configuration of each part of the radar receiver 3y will be described with reference to FIG. As in the radar receiver 3 shown in FIG. 2, the radar receiver 3y shown in FIG. 13 has four parameters D indicating the number of antenna system processors, and the number of coherent integrators, correlation matrix generators, and adders. The parameter P indicating 3 is 3.

  As shown in FIG. 13, the radar receiving unit 3y includes four antenna system processing units 11-1 to 11-4 and first to third correlation matrices provided corresponding to the number of receiving antennas constituting the array antenna. The configuration includes generation units 21-1 to 21-3, first to third addition units 22 y-1 to 22 y-3, an output selection control unit 23 y, an output selection unit 24, and an arrival direction estimation unit 25.

The first addition unit 22y-1 inputs the correlation matrix B ₁ (k) generated according to Equation (18) where the parameter p = 1. The first addition unit 22y-1 adds the input correlation matrix B ₁ (k) according to the equation (21) in which the parameter p = 1. Note that the first addition unit 22y-1 may add by multiplying a weighting coefficient proportional to the magnitude of the correlation output in order to add the correlation matrix. The first addition unit 22y-1 outputs the added correlation matrix A ₁ (k) to the output selection control unit 23y and the output selection unit 24.

The second addition unit 22y-2 inputs the correlation matrices B ₁ (k) and B ₂ (k) generated according to Equation (18) with parameters p = 1 and p = 2. The second addition unit 22y-2 adds the input correlation matrices B ₁ (k) and B ₂ (k) according to Expression (21) in which the parameter p = 2. Note that the second addition unit 22y-2 may add by multiplying a weighting coefficient proportional to the magnitude of the correlation output in order to add the correlation matrix. The second addition unit 22y-2 outputs the added correlation matrix A ₂ (k) to the output selection control unit 23y and the output selection unit 24.

The third adding unit 22y-3 generates correlation matrices B ₁ (k), B ₂ (k), and B ₃ (k) generated according to Equation (18) with parameters p = 1, p = 2, and p = 3. Enter. The third addition unit 22y-3 adds the input correlation matrices B ₁ (k), B ₂ (k), and B ₃ (k) according to Equation (21) in which the parameter p = 3. Note that the third adding unit 22y-3 may add by multiplying a weighting coefficient proportional to the magnitude of the correlation output in order to add the correlation matrix. The third addition unit 22y-3 outputs the added correlation matrix A ₃ (k) to the output selection control unit 3y and the output selection unit 24.

The output selection control unit 23y inputs the correlation matrices A ₁ (k), A ₂ (k), and A ₃ (k) output by the first to third addition units 22y-1 to 22y-3. The output selection control unit 23y selects a correlation matrix generation unit that generates a correlation matrix having the maximum coherent integration gain at each discrete time k out of the correlation matrices based on the input correlation matrices.

  Specifically, the output selection control unit 23y generates a correlation matrix that maximizes the sum of the diagonal components of the correlation matrix corresponding to the average received power component after coherent integration, based on each input correlation matrix. The index of the correlation matrix generation unit is selected for each discrete time k. That is, the output selection control unit 23 selects the index select_index (k) of the correlation matrix generation unit according to Equation (32).

In Equation (32), diag [A _p (k)] is an operator that calculates the sum of the diagonal components of the correlation matrix A _p (k). The output selection control unit 23y outputs the selected index to the output selection unit 24 at each discrete time k. Subsequent operations are the same as those in the first embodiment.

Note that the output selection control unit 23y uses the function coeff (x _p ) based on the ratio value shown in Equation (23) as the sum of the diagonal components of the correlation matrix to diag [A _p (k)]. Of the multiplied coeff (x _p ) × diag [A _p (k)], the maximum index select_index (k) of the correlation matrix generation unit may be selected according to the equation (33). Note that the function coeff (x _p ) needs to satisfy the formula (25) or the formula (26), for example.

In the output selection control unit 23, the process of multiplying the function coeff (x _p ) as a coefficient results in a situation where the coherent integration gain by coherent integration exceeds the increase in the variance value of the noise component.

  For this reason, the radar apparatus 1y selects an adder output corresponding to an index select_index (k) having a larger number of coherent integrations when ideally coherent integration is performed.

  Thereby, the radar apparatus 1 can reduce target selection errors such as detecting a target moving at high speed as a target moving at low speed.

(Modification 3 of the first embodiment)
In the third modification of the first embodiment, the output selection control unit generates a coherent integration result that maximizes the coherent integration gain based on the output of the coherent integration unit before generating the correlation matrix in the correlation matrix generation unit. Select the computed coherent integrator.

  FIG. 14 is a block diagram illustrating in detail the internal configuration of the radar apparatus 1z according to the third modification of the first embodiment. Blocks having the same configuration and operation of each part of the radar apparatus 1z and the radar apparatus 1 of the first embodiment are denoted by the same reference numerals. Hereinafter, in the description of the configuration and operation of the radar apparatus 1z, the contents of the same configuration and operation will be omitted, and contents different from the configuration and operation of the radar apparatus 1 will be described.

  In FIG. 14, the radar apparatus 1z includes a radar transmitter 2 connected to the transmission antenna AN1, and radar system processors 11z-1 to 11z-4 connected to the reception antennas AN2 to AN2-4. It is the structure containing the receiving part 3z. The radar transmitter 2 and the radar receiver 3z are connected to the reference signal oscillator Lo, and a signal is supplied from the reference signal oscillator Lo so that the processing of the radar transmitter 2 and the radar receiver 3z is synchronized. Yes.

(Radar receiver)
Next, the configuration of each part of the radar receiver 3z will be described. A plurality of, for example, four array antennas are configured in the radar receiver 3z, as in the radar apparatus 1 of the first embodiment. Further, similarly to the radar apparatus 1 of the first embodiment, each receiving antenna constituting the array antenna is provided for each antenna system processing unit.

  The radar receiving unit 3z has a plurality of, for example, four antenna system processing units, and one receiving antenna is connected to each antenna system processing unit, and constitutes an array antenna including four receiving antennas. Hereinafter, the configuration of each part of the radar receiver 3z will be described with reference to FIG. As in the radar receiver 3 shown in FIG. 2, the radar receiver 3z shown in FIG. 14 has a parameter D indicating the number of antenna system processors, and each of the coherent integrator, the correlation matrix generator, and the adder The parameter P indicating the number is 3.

  As shown in FIG. 14, the radar receiving unit 3z includes four antenna system processing units 11z-1 to 11z-4 and first to third correlation matrices provided corresponding to the number of receiving antennas constituting the array antenna. It is the structure containing the production | generation parts 21-1 to 21-3, the 1st-3rd addition parts 22-1 to 22-3, the output selection control part 23z, the output selection part 24, and the arrival direction estimation part 25.

  The antenna system processing unit 11z-1 includes a reception RF unit 12 to which the reception antenna AN2 is connected and a signal processing unit 13z. The signal processor 13z includes A / D converters 17 and 18, a correlation value calculator 19, and first to third coherent integrators 20z-1 to 20z-3. The signal processing unit 13z periodically calculates each transmission cycle Tr as a signal processing interval.

The first coherent integrator 20z-1 inputs the correlation value AC (k, M) output from the correlation value calculator 19. The first coherent integration unit 20z-1 uses a plurality of (N ₁ times) transmission cycles Tr based on the correlation value AC (k, M) calculated for each discrete time k in the Mth transmission cycle Tr. Coherent integration with an integration number N _{1 is performed} over a period (N ₁ × Tr).

Parameter _{N 1} is the number of integrations of the coherent integration by the first coherent integrator 20z-1. The first coherent integrator 20z-1 performs coherent integration with the parameter p = 1 in Equation (13). The parameter m is a natural number.

That is, the first coherent integration unit 20z-1 performs the (N ₁ × X) from the correlation value AC (k, N ₁ (m−1) +1) in the {N ₁ (m−1) +1} th transmission cycle Tr. m) Correlation value CI ₁ (k, M) is calculated with the timing of discrete time k aligned using the correlation value AC (k, N ₁ × m) in the _first transmission cycle Tr as a unit. The first coherent integration unit 20z-1 outputs the calculated correlation value CI ₁ (k, M) to the first correlation matrix generation unit 21-1 and the output selection control unit 23z.

The second coherent integrator 20z-2 receives the correlation value AC (k, M) output from the correlation value calculator 19. The second coherent integration unit 20-2 uses a plurality of (N ₂ times) transmission cycles Tr based on the correlation value AC (k, M) calculated for each discrete time k in the Mth transmission cycle Tr. Coherent integration with an integration number N _{2 is performed} over a period (N ₂ × Tr).

Parameter _{N 2} is the number of integrations of the coherent integration by the second coherent integrator 20z-2. The second coherent integrator 20z-2 performs coherent integration with the parameter p = 2 in Expression (13). The parameter m is a natural number.

That is, the second coherent integrator 20z-2 is the _{{N 2 (m-1)} +1} th correlation value AC in the transmission period _{Tr (k, N 2 (m} -1) +1) from the _{(N 2} × m) Correlation value CI ₂ (k, M) is calculated by aligning the timings of discrete time k with the correlation value AC (k, N ₂ × m) in the transmission cycle Tr as the unit. The second coherent integrator 20z-2 outputs the calculated correlation value CI ₂ (k, M) to the second correlation matrix generator 21-2 and the output selection controller 23z.

The third coherent integrator 20z-3 inputs the correlation value AC (k, M) output from the correlation value calculator 19. The third coherent integration unit 20z-3 uses a plurality of (N ₃ times) transmission cycles Tr based on the correlation value AC (k, M) calculated for each discrete time k in the Mth transmission cycle Tr. Coherent integration with an integration number N _{3 is performed} over a period (N ₃ × Tr).

Parameter _{N 3} is an integral number of the coherent integration by the third coherent integrator 20z-3. The third coherent integrator 20z-3 performs coherent integration with the parameter p = 3 in Equation (13). The parameter m is a natural number.

That is, the third coherent integrator 20z-3 is first correlation value _{AC (k, N 3 (m} -1) +1) in _{{N 3 (m-1)} +1} th transmission period Tr from the _{(N 3} × m) Correlation value CI ₃ (k, M) is calculated with the timing of discrete time k aligned using the correlation value AC (k, N ₃ × m) in the transmission cycle Tr as the unit. The third coherent integrator 20z-3 outputs the calculated correlation value CI ₃ (k, M) to the third correlation matrix generator 20z-3 and the output selection controller 23z.

  The output selection control unit 23z inputs coherent integration results that are the outputs of all the coherent integration units of all the antenna system processing units 11z-1 to 11z-4. Based on the outputs of all the coherent integration units of all the antenna system processing units 11z-1 to 11z-4, the output selection control unit 23z has a maximum coherent integration gain at each discrete time k in each correlation matrix. Select the coherent integrator that calculated the integration result.

  Specifically, the output selection control unit 23z calculates the index of the coherent integration unit that calculates the coherent integration result that maximizes the value corresponding to the average received power component after the coherent integration based on each input coherent integration result. Are selected for each discrete time k. That is, the output selection control unit 23z selects the index select_index (k) of the coherent integration unit according to Expression (34).

  The output selection control unit 23z outputs the selected index to the output selection unit 24z at each discrete time k.

The output selection unit 24z outputs the index select_index (k) output by the output selection control unit 23z and the correlation matrix A ₁ output by the first to third addition units 22-1 to 22-3 at each discrete time k. Input (k) to A ₃ (k). The output selection unit 24z selects the output A _p (k) of the addition unit corresponding to the coherent integration unit of the index for each discrete time k based on the input index select_index (k). The output selection unit 24 outputs the output A _p (k) of the addition unit selected at each discrete time k to the arrival direction estimation unit 25.

For example, since the index selected by the output selection control unit 23z is 1, the output selection unit 24z is a correlation that is an output of the first addition unit 22-1 corresponding to the first coherent integration unit 20z-1 on a one-to-one basis. Select the matrix A ₁ (k).

For example, since the index selected by the output selection control unit 23z is 2, the output selection unit 24z is a correlation that is an output of the second adder 22-2 corresponding one-to-one with the second coherent integrator 20z-2. Select the matrix A ₂ (k).

For example, since the index selected by the output selection control unit 23z is 3, the output selection unit 24z is a correlation that is an output of the third adder 22-3 corresponding to the third coherent integrator 20z-3. Select the matrix A ₃ (k).

  Subsequent operations are the same as those in the first embodiment.

  Therefore, the radar apparatus 1z according to the third modification of the first embodiment selects a coherent integration unit that calculates a coherent integration result that maximizes the coherent integration gain among the coherent integration results. Thereby, in addition to the effect of the radar apparatus 1 of the first embodiment, the radar apparatus 1z can reduce the processing amount of the radar receiving unit 3z rather than the processing amount of the radar receiving unit 3 of the radar apparatus 1.

  In addition, although the output selection control part 23z demonstrated the case where the coherent integration result which is the output of all the coherent integration parts of all the D antenna system process parts 11z-1 to 11z-4 was input, Alternatively, the selection may be made according to Expression (34) using the output of each coherent integration unit of one or a part of the antenna system processing units.

  While various embodiments have been described above with reference to the drawings, it goes without saying that the radar apparatus of the present invention is not limited to such an example. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

  INDUSTRIAL APPLICABILITY The present invention is useful as a radar apparatus that maintains the estimation accuracy of the direction of arrival of a target moving at high speed and improves the coherent integration gain of the target moving at low speed with a simple configuration.

1, 1x, 1y, 1z Radar device 2 Radar transmitter 3, 3x, 3y, 3z Radar receiver 4 Transmit signal generator 5 Transmit RF unit 6 Code generator 7 Modulator 8 LPF
9, 15 Frequency conversion unit 10, 14 Amplifier 11-1, 11-2, 11-3, 11-4, 11-D, 11z-1, 11z-2, 11z-3, 11z-4 Antenna system processing unit 12 Reception RF unit 13, 13z Signal processing unit 16 Quadrature detection unit 17, 18 A / D conversion unit 19 Correlation value calculation unit 20-1, 20-1-1, 20z-1 First coherent integration unit 20-2, 20z- 2 2nd coherent integrator 20-3, 20z-3 3rd coherent integrator 20-1-p p coherent integrator 21-1 first correlation matrix generator 21-2 second correlation matrix generator 21-3 3 correlation matrix generator 21-p p correlation matrix generators 22-1, 22y-1 first adder 22-2, 22y-2 second adder 22-3, 22y-3 third adder 22-p P-th adder 23, 23x, 23y, 23z Force selection control unit 24, 24x, 24y, 24z Output selection unit 25 Arrival direction estimation unit 26 Doppler frequency detection unit AN1 Transmission antenna AN2 Reception antenna CM Transmission code storage unit CT Transmission code control unit Lo Reference signal oscillator Tp Pulse width Tr Transmission cycle Tw Transmission interval τ1, τ2 Delay time

Claims (12)

A radar transmitter that converts a transmission signal into a high-frequency transmission signal and transmits the high-frequency transmission signal from a transmission antenna;
A radar receiving unit from a plurality of receiving antennas for estimating the arrival direction of the reflected wave from the target using a signal of the reflected wave that is the high-frequency transmission signal reflected by the target, and
The radar receiver
A plurality of antenna system processing units for performing a plurality of coherent integrations using different coherent integration numbers on a correlation value between a reception signal and the transmission signal;
A plurality of correlation matrix generation units that generate a correlation matrix that is phase difference information resulting from a receiving antenna arrangement based on each output of the plurality of coherent integrations of the plurality of antenna system processing units;
A plurality of outputs of the correlation matrix generation unit obtained from the coherent integration output having a coherent integration number smaller than N to the output of the correlation matrix generation unit obtained from the coherent integration output of the coherent integration number N The addition part of
A radar apparatus comprising: an arrival direction estimation unit configured to estimate the arrival direction of a reflected wave from the target based on an output of the addition unit. The radar apparatus according to claim 1,
An output selection control unit that selects a correlation matrix generation unit that has generated a correlation matrix having a maximum output signal level among the generated correlation matrices;
A radar device that estimates the direction of arrival using an output of an adder obtained with respect to an output of a correlation matrix generator selected by the output selection controller. The radar apparatus according to claim 1 or 2,
Each of the plurality of antenna system processing units is
A receiving antenna for receiving the reflected wave signal;
A receiving RF unit for converting the received signal into a baseband received signal;
An A / D converter that converts the converted received signal into digital data;
A correlation value calculating unit for calculating a correlation value between the converted digital data of the received signal and the pulse compression code;
A radar apparatus comprising: a plurality of coherent integrators for coherently integrating the calculated correlation values by different predetermined times. The radar apparatus according to claim 2 or 3,
A radar device in which the number of integrations of each of the plurality of coherent integration units is determined according to the moving speed of the target. The radar apparatus according to claim 2 or 3,
A radar apparatus in which a part of the plurality of coherent integrators performs coherent integration using an output of another coherent integrator. The radar apparatus according to any one of claims 2 to 4,
The output selection control unit is a radar that selects a correlation matrix generation unit that has generated a correlation matrix that maximizes the sum of diagonal components of each correlation matrix among the correlation matrices generated by the plurality of correlation matrix generation units. apparatus. The radar apparatus according to any one of claims 2 to 4,
The output selection control unit calculates a correlation matrix that is the largest among values obtained by multiplying a sum of diagonal components of each correlation matrix by a predetermined coefficient among the correlation matrices generated by the plurality of correlation matrix generation units. A radar device that selects a generated correlation matrix generation unit. The radar apparatus according to any one of claims 2 to 4,
The output selection control unit controls the output selection unit so that none of the outputs of the plurality of addition units is selected when the output signal level of each correlation matrix does not satisfy a predetermined level. . A radar device according to any one of claims 2 to 8,
The radar receiver further includes:
A Doppler frequency detector that calculates the amount of phase rotation caused by the movement of the target based on the output of any one of the coherent integrators among the coherent integrators,
The output selection control unit is a radar device that selects a correlation matrix generation unit that has generated the correlation matrix based on the calculated phase rotation amount. The radar apparatus according to any one of claims 2 to 9,
The radar receiver further includes:
A Doppler frequency detector that calculates the amount of phase rotation caused by the movement of the target based on the output of any one of the coherent integrators among the coherent integrators,
The output selection control unit is a radar device that selects a correlation matrix generation unit that generates the correlation matrix that maximizes the calculated phase rotation amount when the calculated value is less than a predetermined value. The radar apparatus according to claim 9, wherein
The output selection control unit is a radar device that selects the correlation matrix generation unit corresponding to the addition unit that outputs the addition output value of the correlation matrix having the maximum output signal level among the outputs of the plurality of addition units. . The radar apparatus according to claim 11, wherein
The output selection control unit, instead of selecting the correlation matrix generation unit, selects a coherent integration unit that calculates a coherent integration output with the maximum output signal level from among the outputs of the plurality of coherent integration units. .

Images