JP6177446B1 - Signal processing apparatus, signal processing method, and radar apparatus - Google Patents

Abstract

Included in radar signals s1 (t), s2, aft (t) from radar signal s1 (t) output from transmission wave separation unit (4) and radar signals s2, aft (t) after imbalance correction A noise power estimator (13) for estimating the noise power σ2 that is present, a radar signal s1 (t) output from the transmission wave separator (4), radar signals s2 and aft (t) after imbalance correction, A signal power estimation unit (14) that estimates the signal power P (t) included in the radar signals s1 (t), s2, and aft (t) from the noise power σ2 estimated by the noise power estimation unit (13). ).

Description

  The present invention relates to a signal processing device, a signal processing method, and a radar device that estimate the power of a signal component and the power of a noise component contained in a radar signal.

The signal processing device disclosed in the following Non-Patent Document 1 includes an interference image between two Synthetic Aperture Radar (SAR) images imaged under the same geometric condition, and between two SAR images. The difference image is calculated.
And this signal processing apparatus detects the target which exists in two SAR images, and among the pixels which comprise the interference image, the average value of the pixel value of the pixel which comprises the target The signal power is estimated from
Moreover, noise power is estimated from the average value of the pixel values of the pixels constituting the target among the pixels constituting the difference image.

M. Villano, “SNR and noise variance estimation in polarimetric SAR data,” IEEE Geoscience and Remote Sensing Letters, vol.11, no.1, pp.278-282, Jan., 2014.

Since the conventional signal processing apparatus is configured as described above, in order to estimate the signal power, it is necessary to perform processing for detecting a target existing in two SAR images. There was a problem that the amount would increase.
If the number of pixels constituting the target among the pixels constituting the interference image is infinite, the signal power is calculated from the average value of the pixel values of the pixels constituting the target according to the law of a large number. Although the estimation can be performed according to regular statistics, an estimation error remains in the estimated value of the signal power because the number of pixels constituting the target is finite. For this reason, there has been a problem that the signal power cannot be accurately estimated.

  The present invention has been made to solve the above-described problems, and is a signal processing device and signal that can accurately estimate the signal power contained in the radar signal without performing target detection processing. An object is to obtain a processing method and a radar apparatus.

The signal processing apparatus according to the present invention obtains an interference signal between a plurality of radar signals, averages the interference signal with the number of samples of the plurality of radar signals, and power of noise components included in the plurality of radar signals. Is averaged by the number of samples, and using the real part of the complex number in the averaged interference signal and the averaged noise power, from the value of the signal component contained in at least two of the plurality of radar signals A signal power estimator for estimating one signal power that is a power to be obtained, wherein a plurality of radar signals are radar signals obtained from respective radio waves received by a plurality of antennas, and each received by a plurality of antennas The radio wave has the same geometric condition, the signal component values included in multiple radar signals are the same, and the average power of noise components included in the multiple radar signals is the same. That.

According to the present invention, the signal power estimation unit obtains an interference signal between a plurality of radar signals, averages the interference signal with the number of samples of the plurality of radar signals, and includes noise included in the plurality of radar signals. The noise power that is the power of the component is averaged by the number of samples, and the signal component contained in at least two of the plurality of radar signals using the complex real part and the averaged noise power in the averaged interference signal Since it is configured to estimate one signal power that is the power of the value of, it is possible to accurately estimate the signal power included in the radar signal without performing target detection processing. .

It is a block diagram which shows the radar apparatus by Embodiment 1 of this invention. BRIEF DESCRIPTION OF THE DRAWINGS It is a block diagram which shows the signal processing apparatus by Embodiment 1 of this invention. It is a hardware block diagram of the signal processing apparatus by Embodiment 1 of this invention. It is a hardware block diagram in case the signal processing apparatus by Embodiment 1 of this invention is comprised with a computer. It is a flowchart which shows the signal processing method which is the processing content of the signal processing apparatus by Embodiment 1 of this invention. It is a block diagram which shows the other signal processing apparatus by Embodiment 1 of this invention. It is a hardware block diagram of the other signal processing apparatus by Embodiment 1 of this invention. It is a block diagram which shows the other radar apparatus by Embodiment 1 of this invention. It is a block diagram which shows the signal processing apparatus by Embodiment 2 of this invention. It is a hardware block diagram of the signal processing apparatus by Embodiment 2 of this invention. It is a block diagram which shows the other signal processing apparatus by Embodiment 2 of this invention. It is a hardware block diagram of the other signal processing apparatus by Embodiment 2 of this invention. It is a block diagram which shows the signal processing apparatus by Embodiment 3 of this invention. It is a hardware block diagram of the signal processing apparatus by Embodiment 3 of this invention. It is a flowchart which shows the signal processing method which is the processing content of the signal processing apparatus by Embodiment 3 of this invention.

  Hereinafter, in order to describe the present invention in more detail, modes for carrying out the present invention will be described with reference to the accompanying drawings.

Embodiment 1 FIG.
1 is a block diagram showing a radar apparatus according to Embodiment 1 of the present invention.
2 is a block diagram showing a signal processing apparatus according to Embodiment 1 of the present invention, and FIG. 3 is a hardware block diagram of the signal processing apparatus according to Embodiment 1 of the present invention.
1 to 3, the receiving device 1 has an antenna configuration in which the phase centers of transmission and reception are the same, and outputs radio waves received by each antenna.
In the first embodiment, for simplification of description, the receiving apparatus 1 includes two antennas, performs MIMO (Multiple-Input Multiple-Output), and transmits signals simultaneously from the two antennas. It is assumed that signals are received simultaneously. As a result, a set of two transmission / reception phase centers overlapping each other can be made. Since the two signals obtained from the phase centers of transmission and reception have the same geometric condition, the embodiment will be described so as to be applied to a combination of two signals having the same geometric condition.
Details of MIMO are disclosed in Non-Patent Document 2 below, for example.
[Non-Patent Document 2]
Haowen Chen, Xiang Li, Weidong Jiang, and Zhaowen Zhuang “MIMO Radar Sensitivity Analysis of Antenna Position for Direction Finding,” IEEE TRANSACTIONS ON SIGNAL PROCESSING, VOL. 60, NO. 10, OCTOBER 2012
However, this is only an example, and the receiving apparatus 1 may include three or more antennas. Each antenna may be physically one antenna, but may be a phased array antenna including a plurality of element antennas.
In the first embodiment, the radio wave received by the receiving device 1 is assumed to be a pulse-shaped radio wave. However, this is only an example and is not limited to pulse-shaped radio waves.

The analog circuit 2 includes a receiver including an amplifier, a frequency converter, a demodulator, and the like. In the first embodiment, the receiving device 1 includes two antennas and includes two receivers to perform MIMO.
One receiver included in the analog circuit 2 demodulates the radio wave received by the receiver 1 and outputs a received signal that is the demodulated signal, and the other receiver demodulates the radio wave received by the receiver 1. Then, the received signal which is the demodulated signal is output.
Here, an example is shown in which the analog circuit 2 includes a receiver, but the analog circuit 2 includes two transmitters, the two transmitters modulate the transmission signal, and the modulated signal is transmitted to the radio wave. As a result, the radio wave may be radiated to the space by outputting to the receiving device 1.
The A / D converter 3 converts the received signal output from the analog circuit 2 from an analog signal to a digital signal, and outputs the digital signal.
The transmission wave separation unit 4 generates a radar signal from the digital signal output from the A / D converter 3, separates the signal multiplexed by MIMO for each antenna, and outputs the separated radar signal.

The storage device 5 is composed of, for example, a RAM (Random Access Memory), a hard disk, and the like, and holds the radar signal output from the transmission wave separation unit 4.
The signal processing device 6 is the signal processing device of FIG. 2, and performs processing for estimating noise power, which is the power of the noise component included in the radar signal output from the transmission wave separation unit 4, and the radar signal. The processing for estimating the signal power, which is the power of the signal component included in is performed.
The display device 7 is a device that displays the signal power, noise power, and radar signal estimated by the signal processing device 6 on a display.

In FIG. 1, the radar signal output from the transmission wave separation unit 4 is held in the storage device 5 and input to the signal processing device 6, and the output result of the signal processing device 6 is held in the storage device 5. The value held in 5 is input to the signal processing device 6.
Hereinafter, of the digital signals after A / D conversion, two radar signals in a range time t having the same geometric condition as a pair obtained by performing transmission waveform separation on a signal obtained by MIMO are represented by radar. Signals s ₁ (t) _, s _{2, bef} (t) are represented.
In the first embodiment, since the geometric conditions of the two antennas are the same, the desired signal components included in the radar signal s ₁ (t) and the radar signal s _{2, bef} (t) are the same signal. For example, since there is an individual difference between the respective antennas and receivers _, there is an imbalance between the radar signal s ₁ (t) and the radar signal s _{2, bef} (t). Has occurred.

The imbalance correction unit 11 of the signal processing device 6 is realized by, for example, a semiconductor integrated circuit on which a CPU (Central Processing Unit) is mounted, or an imbalance correction processing circuit 21 on which a one-chip microcomputer or the like is mounted. Yes, a process of correcting the imbalance of a plurality of radar signals output from the transmission wave separation unit 4 is performed. In the first embodiment, the transmitting wave separator 4 two radar signals _s 1 _{(t), s 2,} since _bef (t) is output, the radar signal _{s 2} for the radar signal _s 1 _{(t), The} imbalance of _bef (t) is corrected, and the radar signal s _{2, aft} (t) after the imbalance correction is output to the power estimation unit 12.
The radar signals s ₁ (t), s _{2, bef} (t), s _{2, aft} (t) are complex data.

The power estimation unit 12 includes a noise power estimation unit 13 and a signal power estimation unit 14.
The noise power estimation unit 13 is realized by, for example, a semiconductor integrated circuit on which a CPU is mounted or a noise power estimation processing circuit 22 on which a one-chip microcomputer or the like is mounted, and is output from the transmission wave separation unit 4. From the radar signal s ₁ (t) and the radar signal s _{2, aft} (t) after imbalance correction, the power of the noise component contained in the radar signal s ₁ (t), s _{2, aft} (t) A process of estimating a certain noise power σ ² is performed. The noise power σ ² is real number data.

The signal power estimation unit 14 is realized by, for example, a semiconductor integrated circuit in which a CPU is mounted, or a signal power estimation processing circuit 23 in which a one-chip microcomputer or the like is mounted, and is output from the transmission wave separation unit 4. Radar signal s ₁ (t), radar signal s _{2, aft} (t) after imbalance correction _, and noise power σ ² estimated by noise power estimation unit 13, radar signal s ₁ (t), Processing for estimating the signal power P (t), which is the power of the signal component included in s _{2, aft} (t), is performed. The signal power P (t) is real number data.

In the example of FIG. 2, it is assumed that each of the imbalance correction unit 11, the noise power estimation unit 13, and the signal power estimation unit 14 that are components of the signal processing device 6 is configured by dedicated hardware. However, the signal processing device 6 may be a computer.
FIG. 4 is a hardware configuration diagram when the signal processing device 6 according to the first embodiment of the present invention is configured by a computer.
When the signal processing device 6 is configured by a computer, a program describing the processing contents of the imbalance correction unit 11, the noise power estimation unit 13, and the signal power estimation unit 14 is stored in the memory 31 of the computer, and the processor of the computer 32 may execute the program stored in the memory 31.
FIG. 5 is a flowchart showing a signal processing method which is a processing content of the signal processing device 6 according to the first embodiment of the present invention.

Next, the operation will be described.
The receiving device 1 has two antennas having an antenna configuration in which the geometric conditions of the received radio waves are the same. After the pulse-shaped radio waves are radiated from the two antennas of the own device to the space, A MIMO radar that receives a pulse-shaped radio wave reflected back from an existing target is realized.
The receiving device 1 outputs the radio wave received by one antenna to the analog circuit 2 and outputs the radio wave received by the other antenna to the analog circuit 2.

When the analog circuit 2 receives the radio wave received by the one antenna from the receiving device 1, the analog circuit 2 demodulates the radio wave and outputs the received signal as the demodulated signal to the A / D converter 3. When the radio wave received by the other antenna is received, the radio wave is demodulated and a received signal as a demodulated signal is output to the A / D converter 3.
When the A / D converter 3 receives a reception signal related to a radio wave received by one antenna from the analog circuit 2, the A / D converter 3 converts the reception signal from an analog signal to a digital signal and converts the digital signal into a transmission wave separation unit 4. Output to.
In addition, when the A / D converter 3 receives a reception signal related to the radio wave received by the other antenna from the analog circuit 2, the A / D converter 3 converts the reception signal from an analog signal to a digital signal and separates the digital signal from a transmission wave. Output to part 4.
When receiving a digital signal from the A / D converter 3, the transmission wave separation unit 4 generates a radar signal from the digital signal, and separates the signal multiplexed by MIMO for each antenna.
The transmission wave separation unit 4 outputs the radar signal s ₁ (t) and the radar signal s _{2, bef} (t) to the storage device 5 and the signal processing device 6 as the separated radar signals. Note that an imbalance occurs between the radar signal s ₁ (t) and the radar signal s _{2, bef} (t).

Imbalance correction unit 11 of the signal processing device 6, the radar signal _s 1 from the transmitting wave separation section 4 _{(t), s 2,} when receiving the _bef (t), the radar signal _{s 2} with respect to the radar signal _s 1 (t) _{, Bef} (t) is corrected, and the imbalance-corrected radar signal s _{2, aft} (t) is output to the power estimation unit 12 (step ST1 in FIG. 5).
As imbalance correction processing, in addition to registration processing for aligning waveforms in radar signal s ₁ (t) and radar signal s _{2, bef} (t), amplitude imbalance correction processing and phase imbalance correction There is processing.
Hereinafter, the imbalance correction process by the imbalance correction unit 11 will be described in detail.

(1) Registration Processing The imbalance correction unit 11 takes the cross-correlation between the radar signal s ₁ (t) and the radar signals s _{2 and bef} (t), and thereby the radar signal s ₁ (t) and the radar signal s _{2. , Bef} (t) is calculated. Since the process for obtaining the cross-correlation between two signals is a known technique, a detailed description thereof will be omitted.
When the imbalance correction unit 11 calculates the position shift amount, the imbalance correction unit 11 resamples the radar signals s _{2 and bef} (t) so that the position shift amount becomes zero, thereby _obtaining the radar signal s ₁ (t) and Position alignment of the waveforms in the radar signals s _{2, bef} (t) is performed.
For example, if the registration processing is represented by reg (), the radar signal s _{2, aft} (t) after the registration processing can be expressed as reg (s _{2, bef} (t)).

式（１）において、Ｎａｌｌはレーダ信号ｓ_１（ｔ），ｓ_{２，ｂｅｆ}（ｔ）のサンプル数であり、Σ_Ｎａｌｌは全サンプル数分の加算を示している。
インバランス補正部１１は、振幅のインバランス補正量ａｍｐを算出すると、そのインバランス補正量ａｍｐをレジストレーション処理後のレーダ信号ｒｅｇ（ｓ_{２，ｂｅｆ}（ｔ））に乗算することで、振幅のインバランスを補正する。(2) Amplitude Imbalance Correction Processing The imbalance correction unit 11 calculates an amplitude imbalance correction amount amp as shown in the following equation (1).

In Expression (1), Nall is the number of samples of the radar signals s ₁ (t), s _{2, and bef} (t), and Σ _Nall indicates the addition for the total number of samples.
After calculating the amplitude imbalance correction amount amp, the imbalance correction unit 11 multiplies the radar signal reg (s _{2, bef} (t)) after the registration processing by the imbalance correction amount amp, thereby calculating the amplitude. Correct the imbalance.

式（２）において、ａｒｇは偏角を示す記号である。
インバランス補正部１１は、位相のインバランス量ｐｈａｓｅを算出すると、指数関数であるｅｘｐ（）にインバランス量ｐｈａｓｅを代入し、ｅｘｐ（ｊ・ｐｈａｓｅ）をレジストレーション処理後のレーダ信号ｒｅｇ（ｓ_{２，ｂｅｆ}（ｔ））に乗算することで、位相のインバランスを補正する。ここで、ｊは虚数単位である。(3) Phase Imbalance Correction Processing The imbalance correction unit 11 calculates a phase imbalance amount phase as shown in the following equation (2).

In Expression (2), arg is a symbol indicating a declination.
After calculating the phase imbalance amount phase, the imbalance correction unit 11 substitutes the imbalance amount phase into exp () that is an exponential function, and uses exp (j · phase) as a radar signal reg (s) after registration processing. _{2, bef} (t)) is multiplied to correct the phase imbalance. Here, j is an imaginary unit.

Therefore, the imbalance-corrected radar signal s _{2, aft} (t) output from the imbalance correction unit 11 is expressed by the following equation (3).

Power estimation unit 12, the radar signal _s 1 receives a (t) from the transmission wave separating section 4, when receiving the radar signal _{s 2} after imbalance correction from imbalance correction unit _{11, aft} (t), the radar signal _{s 1 (t), s 2,} for each sample of the _aft (t), the radar signal _{s 1 (t), s 2} , estimated _aft and signal power P contained in (t) (t) and a noise power sigma ² To do.
Here, the radar signals s ₁ (t), s _{2 and aft} (t) can be expressed as the following equations (4) and (5).
s ₁ (t) = s (t) + n ₁ (t) (4)
s _{2, aft} (t) = s (t) + n ₂ (t) (5)

In equations (4) and (5), s (t) is a common component included in both radar signals s ₁ (t), s _{2 and aft} (t), and this common component is the target component. This corresponds to a desired signal which is a signal reflected by.
In the first embodiment, since the phase center of transmission / reception in which two signals are observed overlaps and the geometric condition is the same, the desired signal included in the radar signal s ₁ (t) and the radar signal s _{2 and aft} (t) coincide with the desired signal and become a common component.
n ₁ (t) is a noise component included in the radar signal s ₁ (t) at the range time t, and n ₂ (t) is the radar signal s _{2, aft} (t) at the range time t. This is a noise component included in.
At this time, the average power E (| n ₁ (t) | ² ) of the noise component n ₁ (t) included in the radar signal s ₁ (t) and the radar signal s _{2, aft} ( The average power E (| n ₂ (t) | ² ) of the noise component n ₂ (t) included in t) is the same as shown in the following formula (6).
E (| n ₁ (t) | ² ) = E (| n ₂ (t) | ² ) = σ ² (6)

Hereinafter, the estimation process of the noise power σ ² by the noise power estimation unit 13 of the power estimation unit 12 will be specifically described.
Noise power estimation unit 13, as shown in the following formula (7), the radar signal _s 1 _{(t), s 2,} for each sample of the _aft (t), the radar signal _s 1 in the sample and (t) Then, a difference signal Δs _1-2 (t) between the radar signal s _{2 and aft} (t) after the imbalance correction is calculated.
Δs _1-2 (t) = s ₁ (t) −s _{2, aft} (t) (7)

At this time, the difference signal Δs _1-2 (t) is expressed by the following equation (8) from the equations (4) and (5).
Δs _1-2 (t) = n ₁ (t) −n ₂ (t) (8)
Therefore, E (| s ₁ (t) −s _{2, aft} (t) | ² ), which is an expected value of the power of the difference signal Δs _1-2 (t), is expressed as the following equation (9). The

この実施の形態１では、レーダ信号ｓ_１（ｔ）と、インバランス補正後のレーダ信号ｓ_{２，ａｆｔ}（ｔ）から雑音電力σ^２を推定するものを示したが、各々の受信機内の雑音レベルを把握していれば、その雑音レベルから雑音電力σ^２を推定するようにしてもよい。As described above, when the number of samples of the radar signals s ₁ (t), s _{2, aft} (t) is Null, the noise power estimation unit 13 calculates the difference signal Δs ₁₋ as shown in the following equation (10). _A square value of ₂ (t) is calculated, and an estimated value σ ² hat of the noise power σ ² is calculated by dividing the sum of the square values of each sample by 2Nall (step ST2 in FIG. 5). In the formula (10), the symbol “^” representing the estimated value is attached above the character “σ”. However, in the text of the specification, “^” above the character because of the electronic application. Since the symbol cannot be attached, it is expressed as “σ ² hat”.

In the first embodiment, the noise power σ ² is estimated from the radar signal s ₁ (t) and the imbalance-corrected radar signal s _{2, aft} (t). However, the noise in each receiver is shown. If the level is known, the noise power σ ² may be estimated from the noise level.

Hereinafter, the estimation process of the signal power P (t) by the signal power estimation unit 14 of the power estimation unit 12 will be specifically described.
As a method for estimating the signal power P (t), a method for estimating the signal power P (t) using the interference signal s _fero (t), and a signal power P (t) using the addition average signal s _add (t). A method for estimating t) is conceivable.

First, a method for estimating the signal power P (t) using the interference signal s _fero (t) will be described.
First, as shown in the following equation (11), the signal power estimation unit 14 is an interference signal s _fero (between the radar signal s ₁ (t) and the radar signal s _{2, aft} (t) after imbalance correction. t) is calculated.
s _fero (t) = s ₁ (t) s _{2, aft} (t) ^* (11)
In the formula (11), * is a symbol indicating the complex conjugate of s _{2, aft} (t).

また、干渉信号ｓ_ｆｅｒｏ（ｔ）を有限のサンプル数Ｎで平均化を行うと、下記の式（１３）のようになる。

Here, the interference signal s _fero (t) can be developed as shown in the following expression (12) from the expressions (4) and (5).

Further, when the interference signal s _fero (t) is averaged with a finite number of samples N, the following equation (13) is obtained.

Assuming that the spatial correlation between N samples is “1” and Σ _N | s (t) | ² / N = P (t), the interference signal | Σ _N s averaged with a finite number of samples _N The mean square error (2ε _f ) ² between _fero (t) | / N and the signal power P (t) is expressed by the following equation (14).

From Equation (6), the power obtained by averaging the noise components n ₁ (t) and n ₂ (t) with a finite number of samples N is expressed as Equation (15) below.
E (Σ _N | n ₁ (t) | ² / N)
= E (Σ _N | n ₂ (t) | ² / N)
_{= E (Σ N | n 1} (t) n 2 (t) * | 2 / N)
_{= E (2Σ N Re [n} 1 (t) n 2 (t) *] / N)
= Σ ²
(15)
Thus, a finite number N of samples averaged interference signal _{| Σ N s fero (t)} | average error 2 [epsilon] _f included in the / N is as shown in the following equation (16).

The average error included in the real part Re [Σ _N s _fero (t)] / N of the complex number in the interference signal averaged with a finite number of samples N is 0 because the imaginary component is 0. It becomes half of the average error 2ε _f included in Σ _N s _fero (t) | / N.
Thus, the average error included in the real part _{Re [Σ N s fero (t} )] / N complex in the interference signal is given by epsilon _f, the average number of radar signals used in estimating the signal power is finite Therefore, the statistical lower limit value of the average error ε _f remaining in the estimated signal power is expressed by the following equation (17).

このように、干渉信号ｓ_ｆｅｒｏ（ｔ）を用いて、信号電力Ｐ（ｔ）を推定する方法によれば、Σ_Ｎ ｜ｎ_１（ｔ）ｎ_２（ｔ）^＊｜の項の影響による誤差を軽減して、信号電力Ｐ（ｔ）を推定することができる。Therefore, the signal power estimation unit 14 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 13 to estimate the signal power P (t) as in the following equation (18). A value P (t) hat is calculated (step ST3 in FIG. 5).

Thus, according to the method of estimating the signal power P (t) using the interference signal s _fero (t), the error due to the influence of the term Σ _N | n ₁ (t) n ₂ (t) ^* | And the signal power P (t) can be estimated.

Incidentally, if the polarization angle of the interference signal _{s Fero} (t) in a desired signal does not become zero, as shown in the following equation (19), two times a correction term for the absolute value of the interference signal _{s Fero} (t) Thus, the estimated value P (t) hat of the signal power P (t) may be calculated.

In the first embodiment, an example in which the interference signal s _fero (t) of two radar signals is calculated is shown. However, when an interference signal is obtained from L (L ≧ 3) radar signals, _L a combination of C _{2 ways,} it is possible to calculate the L _{C 2} pieces of interfering signals.
Thus, _L averaging signal Σ _N Σ _{p ≠ q} ¹ of C ₂ pieces of the interference signal ^{<p <L 1 <q <} p | s p s q * | / a (N _L C _2), the signal power P (t ) And the mean square error (2ε _g ) ² are expressed by the following equation (20). However, the 2 <p, _{s p} is assumed to be given as a radar signal.

Therefore, the complex real part Re [Σ _N Σ _{p ≠ q} ^{1 <p <L 1 <q <p} | s _p (t) s _q (t) ^* |] / (N _L C ₂ ) in the interference signal mean errors, since the imaginary component is zero, a half of the average error 2 [epsilon] _g.
Therefore, the real part Re [Σ _N Σ _{p ≠ q} ^{1 <p <L 1 <q <p} | s _p (t) s _q (t) ^* |] / (N _L C ₂ ) in the interference signal The included average error is given by ε _g , and since the average number of radar signals used when estimating the signal power is finite, the statistical lower limit of the average error ε _g remaining in the estimated signal power is The following equation (21) is obtained.

Therefore, the signal power estimation unit 14 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 13 to estimate the signal power P (t) as in the following equation (22). The value P (t) hat is calculated.

Next, a method for estimating the signal power P (t) using the addition average signal s _add (t) will be described.
First, as shown in the following equation (23), the signal power estimation unit 14 averages the radar signal s ₁ (t) and the imbalance-corrected radar signal s _{2, aft} (t), that is, An addition average signal s _add (t) that is an average value of the radar signal s ₁ (t) and the radar signal s _{2, aft} (t) is calculated.
s _add (t) = (s ₁ (t) + s _{2, aft} (t)) / 2 (23)

また、加算平均信号ｓ_ａｄｄ（ｔ）の２乗値を有限のサンプル数Ｎで平均化を行うと、下記の式（２５）のようになる。

Here, the square value of the addition average signal s _add (t) can be developed as in the following Expression (24).

When the square value of the addition average signal s _add (t) is averaged with a finite number of samples N, the following equation (25) is obtained.

Assuming that the spatial correlation between N samples is “1”, Σ _N | s (t) | ² / N = P (t).
The finite averaged sum average signal by the number of samples N | and a / N, the mean square error between the signal power P (t) and ^{_{ε a 2, | | Σ N}} s add (t) Σ N s add (t) | / and N, calculating the mean square error epsilon _b ² and P (t) + σ hat ^2/2, for the average number of radar signals used in estimating the signal power is limited, signal The statistical lower limit value of the mean square error ε _b ² remaining in the estimated power value is expressed by the following equation (26).

Therefore, the signal power estimation unit 14 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 13 to estimate the signal power P (t) as shown in the following equation (28). The value P (t) hat is calculated.

In the first embodiment, two is shown an example that calculates averaged signal s _{the add} a _(t) of the radar signal, L (L ≧ 3) pieces averaging signals from the radar signal s _{the add} ( when obtaining _{t), | Σ N s add} (t) | and / N, signal power the mean square error between P (t) and ^{_{ε c 2, | Σ N s}} add (t) | and / N, P When the mean square error ε _d ² with (t) + σ hat ² / L is calculated, the average number of radar signals used when estimating the signal power is finite, so the mean square remaining in the estimated value of the signal power The statistical lower limit value of the error ε _d ² is expressed by the following equation (29).

Therefore, the signal power estimation unit 14 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 13 to estimate the signal power P (t) as in the following equation (31). The value P (t) hat is calculated.

The estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 13 and the estimated value P (t) hat of the signal power P (t) calculated by the signal power estimation unit 14 are the storage device 5. Saved in.
The display device 7 displays the estimated value σ ² hat of the noise power σ ² and the estimated value P (t) hat of the signal power P (t) stored in the storage device 5 on the display. Further, the radar signals s ₁ (t), s _{2, bef} (t) or s _{2, aft} (t) stored in the storage device 5 are displayed on the display.
Note that the estimated value P (t) hat of the signal power P (t) calculated by the signal power estimation unit 14 is used as auxiliary data when the radar apparatus performs target detection processing, tracking processing, and the like, for example. be able to.

As apparent from the above, according to the first embodiment, the radar signal s ₁ (t) output from the transmission wave separation unit 4 and the radar signal s _{2, aft} (t) after imbalance correction are used to calculate the radar signal. A noise power estimation unit 13 that estimates the noise power σ ² included in the signals s ₁ (t), s _{2, aft} (t), and a radar signal s ₁ (t) output from the transmission wave separation unit 4 The radar signal s _{2, aft} (t) after imbalance correction and the noise power σ ² estimated by the noise power estimation unit 13 are included in the radar signal s ₁ (t), s _{2, aft} (t). Since the signal power estimation unit 14 for estimating the signal power P (t) is provided, the signal power and the noise power included in the radar signal with high accuracy without performing the target detection process. There is an effect that can be estimated.

In the first embodiment, the noise power estimation unit 13 estimates the noise power σ ² and the signal power estimation unit 14 estimates the signal power P (t), but as shown in FIG. The signal-to-noise power ratio is calculated from the estimated value σ ² hat of the noise power σ ² calculated by the power estimating unit 13 and the estimated value P (t) hat of the signal power P (t) calculated by the signal power estimating unit 14. The signal-to-noise power ratio calculation unit 15 for calculating (Signal-to-Noise Ratio: SNR) and the square root of the estimated value P (t) hat of the signal power P (t) calculated by the signal power estimation unit 14 Thus, the estimated value P (t) hat is converted into the amplitude a (t) of the signal, the phase θ (t) is the phase of the radar signal s ₁ (t), and the amplitude is the converted amplitude a (t). A signal generator 16 for generating a signal It may be.

By including the signal-to-noise power ratio calculation unit 15, for example, if the display device 7 displays the SNR calculated by the signal-to-noise power ratio calculation unit 15, the SNR can be monitored.
By including the signal generation unit 16, the signal generated by the signal generation unit 16 can be used for, for example, target detection processing or tracking processing.
As shown in FIG. 7, the signal-to-noise power ratio calculation unit 15 is, for example, a semiconductor integrated circuit in which a CPU is mounted, or a signal-to-noise power ratio calculation processing circuit 24 in which a one-chip microcomputer is mounted. Can be realized.
Further, as shown in FIG. 7, the signal generation unit 16 can be realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or a signal generation processing circuit 25 on which a one-chip microcomputer or the like is mounted.

In the first embodiment, the signal processing device 6 is mounted in one radar device. However, as shown in FIG. 8, the radar device is divided into two devices. The signal processing device 6 may be mounted in the device.
FIG. 8 is a block diagram showing another radar apparatus according to Embodiment 1 of the present invention. In FIG. 8, the same reference numerals as those in FIG.
The radar apparatus of FIG. 8 includes a first apparatus 41 and a second apparatus 42, and the first apparatus 41 transmits a digital signal stored in the storage device 5 to the second apparatus 42. It has.
The second device 42 includes a communication device 52 that receives the digital signal transmitted from the first device 41, and radar signals s ₁ (t), s _{2, bef} (t (t) from the digital signal received by the communication device 52. ) And a storage device 53 for storing the radar signals s ₁ (t), s _{2 and bef} (t).
The second device 42 includes the signal processing device 6 shown in FIG. 2 or 6 and the display device 7.
Communication between the communication device 51 and the communication device 52 may be wired communication or wireless communication.
Thus, even when the radar device is divided into two devices, the same effect as that of the radar device of FIG. 1 can be obtained.

Embodiment 2. FIG.
In the first embodiment, the radar signals s ₁ (t), s _{2, bef} (t) given from the transmission wave separation unit 4 to the signal processing device 6 are one-dimensional signals. A two-dimensional signal may be given from the D converter 3 to the signal processing device 6.
When the radar signal is a two-dimensional signal, for example, image data of a SAR image can be considered as the radar signal.

Hereinafter, in the second embodiment, radar signals output from the A / D converter 3 to the signal processing device 6 are SAR image data s ₁ (x, y), s _{2, bef} (x, y). It explains as being. In the image data, for example, x indicates the position of the azimuth bin, and y indicates the position of the range bin.
Here, two SAR images are obtained from pulse-shaped radio waves received by two antennas in the receiving apparatus 1 that are arranged so as to be shifted in the orbit direction after pulse-shaped radio waves are radiated from one transmission apparatus. It is an image obtained by performing image reproduction. Therefore, if it is assumed that registration for performing image alignment is performed, the two images are images captured under the same geometric condition.

The second embodiment is different from the first embodiment in that the transmission wave separation unit 4 is not mounted and the A / D converter 3 outputs a two-dimensional signal. The configuration of the radar apparatus is assumed to be that of FIG. 1 or FIG. 8 as in the first embodiment except that the transmission wave separation unit 4 is not mounted.
However, in the second embodiment, it is assumed that the analog circuit 2 has an image reproduction function for reproducing an SAR image from a pulse-shaped radio wave.
Also in the second embodiment, since there is an individual difference between the respective antennas and receivers, there is a difference between the image data s ₁ (x, y) of the SAR image and the image data s _{2, bef} (x, y). Has an imbalance.
When the radar apparatus adopts MIMO as in the first embodiment and uses a pulse Doppler radar, two range Doppler maps obtained by the pulse Doppler radar, which are the same geometric condition, are used in the second embodiment. It may be handled as two-dimensional data.

FIG. 9 is a block diagram showing a signal processing apparatus according to Embodiment 2 of the present invention, and FIG. 10 is a hardware block diagram of the signal processing apparatus according to Embodiment 2 of the present invention.
9 and FIG. 10, the same reference numerals as those in FIG.
The imbalance correction unit 61 of the signal processing device 6 is realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or an imbalance correction processing circuit 71 on which a one-chip microcomputer or the like is mounted. The imbalance of the image data s _{2, bef} (x, y) with respect to the image data s ₁ (x, y) of the SAR image output from the converter 3 is corrected, and the image data s _{2, bef} ( A process of outputting x, y) to the power estimation unit 62 is performed.
Note that the image data s ₁ (x, y), s _{2, bef} (x, y), s _{2, aft} (x, y) is complex number data.

The power estimation unit 62 includes a noise power estimation unit 63 and a signal power estimation unit 64.
The noise power estimation unit 63 is realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or a noise power estimation processing circuit 72 on which a one-chip microcomputer or the like is mounted, and is output from the A / D converter 3. image data _s 1 (x, y) of the SAR image and the image data _{s 2} after imbalance _correction, aft (x, y) from the image data _s 1 SAR images (x, _{y), s 2, aft} A process of estimating the noise power σ ² that is the power of the noise component included in (x, y) is performed. The noise power σ ² is real number data.

The signal power estimation unit 64 is realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or a signal power estimation processing circuit 73 on which a one-chip microcomputer or the like is mounted, and is output from the A / D converter 3. From the image data s ₁ (x, y) of the generated SAR image, the image data s _{2, aft} (x, y) after imbalance correction _, and the noise power σ ² estimated by the noise power estimation unit 63, Processing for estimating the signal power P (x, y), which is the power of the signal component included in the image data s ₁ (x, y), s _{2, aft} (x, y) of the SAR image, is performed. The signal power P (x, y) is real number data.

In the example of FIG. 10, it is assumed that each of the imbalance correction unit 61, the noise power estimation unit 63, and the signal power estimation unit 64, which are components of the signal processing device 6, is configured with dedicated hardware. However, the signal processing device 6 may be a computer.
When the signal processing device 6 is composed of a computer, a program describing the processing contents of the imbalance correction unit 61, noise power estimation unit 63, and signal power estimation unit 64 is stored in the memory 31 of the computer shown in FIG. The processor 32 of the computer shown in FIG. 4 may execute the program stored in the memory 31.

Next, the operation will be described.
When the imbalance correction unit 61 of the signal processing device 6 receives the image data s ₁ (x, y), s _{2, bef} (x, y) of the SAR image from the A / D converter 3, the image data s ₁ The imbalance of the image data s _{2, bef} (x, y) with respect to (x, y) is corrected, and the image data s _{2, aft} (x, y) after the imbalance correction is output to the power estimation unit 62.
As imbalance correction processing, in addition to registration processing for performing alignment between image data s ₁ (x, y) of the SAR image and image data s _{2, bef} (x, y), amplitude imbalance correction processing, There is a phase imbalance correction process.
Hereinafter, the imbalance correction processing by the imbalance correction unit 61 will be described in detail.

(1) Registration Processing The imbalance correction unit 61 _obtains the cross correlation between the image data s ₁ (x, y) and the image data s _{2, bef} (x, y), thereby obtaining the image data s ₁ (x, y). ) And the image data s _{2, bef} (x, y). Since the process for obtaining the cross-correlation between two signals is a known technique, a detailed description thereof will be omitted.
When the imbalance correction unit 61 calculates the position shift amount, the imbalance correction unit 61 resamples the image data s _{2, bef} (x, y) so that the position shift amount becomes zero, whereby the image data s ₁ (x , Y) and the image data s _{2, bef} (x, y) are aligned.
For example, if the registration process is expressed by reg (), the image data s _{2, aft} (x, y) after the registration process is expressed as reg (s _{2, bef} (x, y)). be able to.

式（３２）において、Ｎａｌｌは画像データｓ_１（ｘ，ｙ），ｓ_{２，ｂｅｆ}（ｘ，ｙ）のピクセル数であり、Σ_Ｎａｌｌは全ピクセル数分の加算を示している。
インバランス補正部６１は、振幅のインバランス補正量ａｍｐを算出すると、そのインバランス補正量ａｍｐをレジストレーション処理後の画像データｒｅｇ（ｓ_{２，ｂｅｆ}（ｘ，ｙ））に乗算することで、振幅のインバランスを補正する。(2) Amplitude Imbalance Correction Processing The imbalance correction unit 61 calculates an amplitude imbalance correction amount amp as shown in the following equation (32).

In Expression (32), Nall is the number of pixels of the image data s ₁ (x, y), s _{2, bef} (x, y), and Σ _Nall indicates the addition for the total number of pixels.
After calculating the amplitude imbalance correction amount amp, the imbalance correction unit 61 multiplies the image data reg (s _{2, bef} (x, y)) after registration processing by the imbalance correction amount amp. Correct the amplitude imbalance.

インバランス補正部６１は、位相のインバランス量ｐｈａｓｅを算出すると、指数関数であるｅｘｐ（）にインバランス量ｐｈａｓｅを代入し、ｅｘｐ（ｊ・ｐｈａｓｅ）をレジストレーション処理後の画像データｒｅｇ（ｓ_{２，ｂｅｆ}（ｘ，ｙ））に乗算することで、位相のインバランスを補正する。(3) Phase Imbalance Correction Processing The imbalance correction unit 61 calculates a phase imbalance amount phase as shown in the following equation (33).

After calculating the phase imbalance amount phase, the imbalance correction unit 61 substitutes the imbalance amount phase into exp () that is an exponential function, and sets exp (j · phase) to the image data reg (s) after registration processing. _{2, bef} (x, y)) is multiplied to correct the phase imbalance.

Accordingly, the imbalance-corrected image data s _{2, aft} (x, y) output from the imbalance correction unit 61 is expressed by the following equation (34).

The power estimation unit 62 receives the image data s ₁ (x, y) of the SAR image from the A / D converter 3, and the image data s _{2, aft} (x, y) after imbalance correction from the imbalance correction unit 61. When receiving the image data _{s 1 (x, y),} s 2, aft (x, y) for each pixel of the image data _s 1 _{(x, y),} contained in the _{s 2, aft (x, y} ) Signal power P (x, y) and noise power σ ² are estimated.
Here, the image data s ₁ (x, y), s _{2, aft} (x, y) can be expressed as the following equations (35) and (36).
s ₁ (x, y) = s (x, y) + n ₁ (x, y) (35)
s _{2, aft} (x, y) = s (x, y) + n ₂ (x, y) (36)

In the expressions (35) and (36), s (x, y) is a common component included in both of the image data s ₁ (x, y), s _{2, and aft} (x, y). The common component corresponds to a desired signal that is a signal reflected by the target.
In the second embodiment, since the geometrical conditions of the two antennas are the same, the desired signal included in the image data s ₁ (x, y) and the image data s _{2, aft} (x, y ) And the desired signal contained in the same component.
n ₁ (x, y) is Ajimasubin x contained in the image data _s 1 (x, y) of the SAR image, a noise component in the range bin _{y, n 2 (x, y} ) is the post-imbalance correction This is a noise component in the azimuth bin x and the range bin y included in the image data s _{2, aft} (x, y).
At this time, the average power E (| n ₁ (x, y) | ² ) of the noise component n ₁ (x, y) included in the image data s ₁ (x, y) and the image after imbalance correction The average power E (| n ₂ (x, y) | ² ) of the noise component n ₂ (x, y) included in the data s _{2, aft} (x, y) is expressed by the following equation (37). Identical as shown.
E (| n ₁ (x, y) | ² ) = E (| n ₂ (x, y) | ² ) = σ ² (37)

Hereinafter, the estimation process of the noise power σ ² by the noise power estimation unit 63 of the power estimation unit 62 will be specifically described.
The noise power estimation unit 63, for each pixel of the image data s ₁ (x, y), s _{2, aft} (x, y), as shown in the following equation (38), the image data s _{1 at} that pixel. A difference image Δs _1-2 (x, y) between (x, y) and the image data s _{2, aft} (x, y) after imbalance correction is calculated.
Δs _1-2 (x, y) = s ₁ (x, y) −s _{2, aft} (x, y) (38)

At this time, from the expressions (35) and (36), the difference image Δs _1-2 (x, y) is expressed as the following expression (39).
Δs _1-2 (x, y) = n ₁ (x, y) −n ₂ (x, y) (39)
Therefore, E (| s ₁ (x, y) −s _{2, aft} (x, y) | ² ), which is an expected value of the power of the difference image Δs _1-2 (x, y), is expressed by the following equation (40 ).

この実施の形態２では、ＳＡＲ画像の画像データｓ_１（ｘ，ｙ）と、インバランス補正後のレーダ信号ｓ_{２，ａｆｔ}（ｘ，ｙ）から雑音電力σ^２を推定するものを示したが、各々の受信機内の雑音レベルを把握していれば、その雑音レベルから雑音電力σ^２を推定するようにしてもよい。As described above, if the number of pixels of the image data s ₁ (x, y), s _{2, aft} (x, y) is Nall, the noise power estimation unit 63 calculates the difference as shown in the following equation (41). A square value of the image Δs _1-2 (x, y) is calculated, and an estimated value σ ² hat of the noise power σ ² is calculated by dividing the sum of the square values at each pixel by 2 Nall.

In the second embodiment, the noise power σ ² is estimated from the image data s ₁ (x, y) of the SAR image and the radar signal s _{2, aft} (x, y) after imbalance correction. If the noise level in each receiver is known, the noise power σ ² may be estimated from the noise level.

Hereinafter, the estimation process of the signal power P (x, y) by the signal power estimation unit 64 of the power estimation unit 62 will be specifically described.
As a method for estimating the signal power P (x, y), a method for estimating the signal power P (x, y) using the interference image s _fero (x, y), and an addition average image s _add (x, y). And a method for estimating the signal power P (x, y) using.

First, a method for estimating the signal power P (x, y) using the interference image s _fero (x, y) will be described.
First, as shown in the following formula (42), the signal power estimation unit 64 uses the image data s ₁ (x, y) of the SAR image and the image data s _{2, aft} (x, y) after imbalance correction. The interference image s _fero (x, y) is calculated.
s _fero (x, y) = s ₁ (x, y) s _{2, aft} (x, y) ^* (42)
In the formula (42), * is a symbol indicating the complex conjugate of s _{2, aft} (x, y).

また、干渉画像ｓ_ｆｅｒｏ（ｘ，ｙ）を有限のピクセル数Ｎで平均化を行うと、下記の式（４４）のようになる。

Here, the interference signal s _fero (x, y) can be developed as shown in the following equation (43).

Further, when the interference image s _fero (x, y) is averaged with a finite number of pixels N, the following equation (44) is obtained.

Assuming that the spatial correlation between N pixels is “1” and | s (x, y) | ² = Σ _N | s (x, y) | ² / N = P (x, y) It averaged interference image in number of pixels _{N | Σ N s fero (x} , y) | / and N, mean square error (2 [epsilon] _f) ² of the signal power P (x, y) is the following formula (45 )become that way.

From Expression (37), the power obtained by averaging the noise components n ₁ (x, y) and n ₂ (x, y) with a finite number of pixels N is expressed as the following Expression (46).
_{E (Σ N | n 1 (} x, y) | 2 / N)
= E (Σ _N | n ₂ (x, y) | ² / N)
_{= E (Σ N | n 1} (x, y) n 2 (x, y) * | 2 / N)
_{= E (2Σ N Re [n} 1 (x, y) n 2 (x, y) *] / N)
= Σ ²
(46)
Thus, it averaged interference signal with a finite number of pixels _{N | Σ N s fero (x} , y) | / average error is included in the N 2 [epsilon] _f is as the following equation (47).

The average error included in the real part _{Re [Σ N s fero (x} , y)] / N of the complex in an interference image that is averaged in a finite number of pixels N, since the imaginary component is zero, the interference signal | is half of the average contained in the / N error _{2ε f | Σ N s fero (} x, y).
Accordingly, the real part _{Re [Σ N s fero (x} , y)] of the complex in the interference image / average error contained in N is given by epsilon _f, the average number of pixels to be used for estimating the signal power is a finite For this reason, the statistical lower limit value of the average error ε _f remaining in the estimated signal power is expressed by the following equation (48).

このように、干渉画像ｓ_ｆｅｒｏ（ｘ，ｙ）を用いて、信号電力Ｐ（ｘ，ｙ）を推定する方法によれば、Σ_Ｎ ｜ｎ_１（ｘ，ｙ）ｎ_２（ｘ，ｙ）^＊｜の項の影響による誤差を軽減して、信号電力Ｐ（ｘ，ｙ）を推定することができる。Therefore, the signal power estimation unit 64 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 63 to obtain the signal power P (x, y) as shown in the following equation (49). An estimated value P (x, y) hat is calculated.

Thus, according to the method of estimating the signal power P (x, y) using the interference image s _fero (x, y), Σ _N | n ₁ (x, y) n ₂ (x, y) ^{* The} signal power P (x, y) can be estimated by reducing the error due to the influence of the term |.

Incidentally, if the polarization angle of the interference image _s Fero (x, y) desired signal within does not become zero, as shown in equation (50) below, the correction for the absolute value of the interference image _s Fero (x, y) The term may be doubled to calculate the estimated value P (x, y) hat of the signal power P (x, y).

In the second embodiment, an example in which the interference image s _fero (x, y) of two pieces of image data is calculated is shown. In the case where an interference image is obtained from L (L ≧ 3) pieces of image data. , a combination of _L C _{2 ways,} you are possible to calculate the L _{C 2} pieces of interference images.
Therefore, the average of the _L C _two interference images Σ _N Σ _{p ≠ q} ^{1 <p <L 1 <q <p} | s _p s _q * | / (N _L C ₂ ) and the signal power P (x , Y) and the mean square error (2ε _g ) ² are given by the following equation (51). However, the 2 <p, _{s p} is assumed to be given as image data.

Therefore, the real part Re [Σ _N Σ _{p ≠ q} ^{1 <p <L 1 <q <p} | s _p (x, y) s _q (x, y) ^* |] / (N _L C mean error contained in _2), since the imaginary component is zero, a half of the average error 2 [epsilon] _g.
Therefore, the real part of the complex number Re [Σ _N Σ _{p ≠ q} ^{1 <p <L 1 <q <p} | s _p (x, y) s _q (x, y) ^* |] / (N _L The average error included in C ₂ ) is given by ε _g , and since the average number of pixels used in estimating the signal power is finite, the statistical lower limit of the average error ε _g remaining in the estimated value of the signal power The value is as shown in the following equation (52).

Therefore, the signal power estimator 64 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimator 63 to obtain the signal power P (x, y) as shown in the following equation (53). An estimated value P (x, y) hat is calculated.

Next, a method for estimating the signal power P (x, y) using the addition average image s _add (x, y) will be described.
First, as shown in the following formula (54), the signal power estimation unit 64 uses the image data s ₁ (x, y) of the SAR image and the image data s _{2, aft} (x, y) after imbalance correction. average image with, i.e., calculates the image data _s 1 (x, y) and the image data _{s 2, aft (x, y} ) mean value in the form of the averaged image _s add (x, y) of the.
s _add (x, y) = (s ₁ (x, y) + s _{2, aft} (x, y)) / 2
(54)

また、加算平均画像ｓ_ａｄｄ（ｘ，ｙ）の２乗値を有限のピクセル数Ｎで平均化を行うと、下記の式（５６）のようになる。

Here, the square value of the addition average image s _add (x, y) can be developed as in the following equation (55).

Further, when the square value of the addition average image s _add (x, y) is averaged with a finite number of pixels N, the following equation (56) is obtained.

Assume that | s (x, y) | ² = Σ _N | s (x, y) | ² / N = P (x, y), where the spatial correlation between N pixels is “1”.
The averaged averaged image at a finite number of pixels N | and a / N, signal power P (x, y) and the mean square error and epsilon _a ² _{of, | Σ N s add (x} , y) | _{Σ N s add (x, y} ) | and / N, P (x, y ) when calculating the mean square error epsilon _b ² a + sigma hat ^2/2, the average number of pixels used in estimating the signal power Since it is finite, the statistical lower limit value of the mean square error ε _b ² remaining in the estimated value of the signal power is expressed by the following equation (57).

Therefore, the signal power estimation unit 64 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 63, and the signal power P (x, y) as in the following equation (59). An estimated value P (x, y) hat is calculated.

In the second embodiment, an example in which the addition average image s _add (x, y) of two pieces of image data is calculated is shown. However, the addition average image s of L (L ≧ 3) pieces of image data is shown. when obtaining an _{add (x, y), |} Σ N s add (x, y) | / N and signal power P (x, y) into a mean square error between the ^{_{ε c 2, | Σ N s}} add ( When the mean square error ε _d ² between x, y) | / N and P (x, y) + σ hat ² / L is calculated, the average number of pixels used in estimating the signal power is finite. The statistical lower limit value of the mean square error ε _d ² remaining in the estimated value of the signal power is expressed by the following equation (60).

Therefore, the signal power estimator 64 uses the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimator 63 to obtain the signal power P (x, y) as shown in the following equation (62). An estimated value P (x, y) hat is calculated.

The estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 63 and the estimated value P (x, y) hat of the signal power P (x, y) calculated by the signal power estimation unit 64 And stored in the storage device 5.
The display device 7 displays the estimated value σ ² hat of the noise power σ ² and the estimated value P (x, y) hat of the signal power P (x, y) stored in the storage device 5 on the display. Further, the radar signals s ₁ (x, y), s _{2, bef} (x, y) or s _{2, aft} (x, y) stored in the storage device 5 are displayed on the display.
Note that the estimated value P (x, y) hat of the signal power P (x, y) calculated by the signal power estimation unit 64 assists, for example, when the radar apparatus performs target detection processing, tracking processing, or the like. It can be used as data.

As is apparent from the above, according to the second embodiment, the image data s ₁ (x, y) of the SAR image output from the A / D converter 3 and the image data s _{2, aft} after imbalance correction A noise power estimation unit 63 for estimating the noise power σ ² included in the image data s ₁ (x, y), s _{2, aft} (x, y) from (x, y), and an A / D converter The image data s ₁ (x, y) of the SAR image output from 3, the image data s _{2, aft} (x, y) after imbalance correction _, and the noise power σ ² estimated by the noise power estimation unit 63 And a signal power estimation unit 64 for estimating the signal power P (x, y) included in the image data s ₁ (x, y), s _{2, aft} (x, y). Therefore, the image data of the SAR image is obtained as a two-dimensional signal from the A / D converter 3. Even when s ₁ (x, y), s _{2, bef} (x, y) is given, the signal power and noise power included in the image data are accurately estimated without performing the target detection process. There is an effect that can be.

In the second embodiment, the noise power estimation unit 63 estimates the noise power σ ² and the signal power estimation unit 64 estimates the signal power P (x, y). As shown in FIG. The estimated value σ ² hat of the noise power σ ² calculated by the noise power estimation unit 63 and the estimated value P (x, y) hat of the signal power P (x, y) calculated by the signal power estimation unit 64 The signal-to-noise power ratio (SNR) is calculated from the signal-to-noise power ratio calculation unit 65, and the signal power P (x, y) estimated value P calculated by the signal power estimation unit 64 By taking the square root of the (x, y) hat, the estimated value P (x, y) hat is converted into the signal amplitude a (x, y), and the phase θ (x, y) is the image data s ₁ ( x, y) and the amplitude is the converted amplitude a (x, y). Signal may include a signal generator 66 for generating that.

By including the signal-to-noise power ratio calculation unit 65, for example, if the display device 7 displays the SNR calculated by the signal-to-noise power ratio calculation unit 65, the SNR can be monitored.
By including the signal generation unit 66, the signal generated by the signal generation unit 66 can be used for, for example, target detection processing or tracking processing.
As shown in FIG. 12, the signal-to-noise power ratio calculation unit 65 is, for example, a semiconductor integrated circuit in which a CPU is mounted, or a signal-to-noise power ratio calculation processing circuit 74 in which a one-chip microcomputer is mounted. Can be realized.
Further, as shown in FIG. 7, the signal generation unit 76 can be realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or a signal generation processing circuit 75 on which a one-chip microcomputer or the like is mounted.

Embodiment 3 FIG.
In the first and second embodiments, the power estimation units 12 and 62 estimate the signal powers P (t), P (x, y) and the noise power σ ^2. However, the signal power P (t) Alternatively, a radar signal may be generated using P (x, y) and noise power σ ² .

13 is a block diagram showing a signal processing apparatus according to Embodiment 3 of the present invention, and FIG. 14 is a hardware block diagram of the signal processing apparatus according to Embodiment 3 of the present invention.
13 and FIG. 14, the same reference numerals as those in FIG. 2 and FIG.
The addition signal calculation unit 81 is realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or an addition signal calculation processing circuit 91 on which a one-chip microcomputer or the like is mounted, and is output from the transmission wave separation unit 4. Then, a process of calculating an addition signal s _add (t) between the radar signal s ₁ (t) and the radar signal s _{2, aft} (t) after imbalance correction by the imbalance correction unit 11 is performed.
The filter unit 82 is realized by, for example, a semiconductor integrated circuit on which a CPU is mounted, or a filter arithmetic circuit 92 on which a one-chip microcomputer or the like is mounted. The added signal s calculated by the added signal calculating unit 81 is used. _add (t), an estimated value P (t) hat of the signal power P (t) calculated by the signal power estimating unit 14, and an estimated value σ ² hat of the noise power σ ² calculated by the noise power estimating unit 13 By performing a filter operation using the above, processing for generating the radar signal s _b (t) is performed.

In the example of FIG. 13, each of the imbalance correction unit 11, the noise power estimation unit 13, the signal power estimation unit 14, the addition signal calculation unit 81, and the filter unit 82 which are components of the signal processing device 6 is dedicated hardware. Although what is comprised is assumed, the signal processing apparatus 6 may be comprised with the computer.
When the signal processing device 6 is configured by a computer, a program describing processing contents of the imbalance correction unit 11, the noise power estimation unit 13, the signal power estimation unit 14, the addition signal calculation unit 81, and the filter unit 82 is illustrated. 4 is stored in the memory 31 of the computer shown in FIG. 4, and the computer processor 32 shown in FIG.
FIG. 15 is a flowchart showing a signal processing method which is a processing content of the signal processing device 6. In FIG. 15, the same ST as in FIG. 5 indicates the same processing content.

Next, the operation will be described.
Since the processing contents other than the addition signal calculation unit 81 and the filter unit 82 are the same as those in the first embodiment, the processing contents of the addition signal calculation unit 81 and the filter unit 82 will be described here.
When the addition signal calculation unit 81 receives the radar signal s ₁ (t) from the transmission wave separation unit 4 and receives the radar signal s _{2, aft} (t) after imbalance correction from the imbalance correction unit 11, As shown in (63), an addition signal s _add (t) of the radar signal s ₁ (t) and the radar signal s _{2, aft} (t) after imbalance correction is calculated (step ST4 in FIG. 15). .
s _add (t) = s ₁ (t) + s _{2, aft} (t) (63)
Here, by adding the two radar signals, but calculates the sum signal s _{add (t),} by adding more than two radar signals, to calculate the sum signal s _{add (t)} It may be.

ウィナーフィルタの具体的な内容は、例えば、以下の非特許文献３に開示されている。
［非特許文献３］
高木幹雄，下田暢久，“画像解析ハンドブック”, 東京大学出版会, 1991The filter unit 82 performs a filtering process using a Wiener filter. The added signal s _add (t) calculated by the added signal calculating unit 81 and the signal power P (t) calculated by the signal power estimating unit 14. And the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimator 13 are used to perform a filter operation as shown in the following equation (64). Thus, the radar signal s _b (t) is generated (step ST5 in FIG. 15).

Specific contents of the Wiener filter are disclosed in Non-Patent Document 3 below, for example.
[Non-Patent Document 3]
Mikio Takagi, Akihisa Shimoda, “Image Analysis Handbook”, The University of Tokyo Press, 1991

In the third embodiment, filtering processing is performed with a Wiener filter that receives two radar signals and doubles the SNR.
Therefore, when L (L ≧ 3) radar signals are received, a filter operation as shown in the following equation (65) may be performed.

By performing the filtering process as shown in the equation (64) or the equation (65), an unnecessary signal due to noise can be statistically minimized, so that a radar signal s _b (t) with small noise is obtained. Can do.

As apparent from the above, according to the third embodiment, the radar signal s ₁ (t) output from the transmission wave separation unit 4 and the radar signal s ₂ after the imbalance correction by the imbalance correction unit 11 are performed _{. aft} the addition signal computing unit 81 that calculates a (t) and the sum signal _{s the add} (t) is provided, the filter unit 82, the sum signal _{s the add} (t) calculated by the addition signal computing unit 81, the signal power estimate A filter operation using the estimated value P (t) hat of the signal power P (t) calculated by the unit 14 and the estimated value σ ² hat of the noise power σ ² calculated by the noise power estimating unit 13. in, since it is configured to generate a radar signal s _{b (t),} an effect which can be noise to obtain a radar signal s _b being minimized statistically _(t).

In the third embodiment, the addition signal calculation unit 81 and the filter unit 82 are applied to the signal processing apparatus of FIG. 2, but the addition signal calculation unit 81 and the filter unit 82 are shown in FIGS. You may make it apply to the signal processing apparatus of FIG.
However, when applied to the signal processing device of FIG. 9 or FIG. 11, the addition signal calculation unit 81 includes the image data s ₁ (x, y) of the SAR image and s _{2, aft} (x, y after imbalance correction). ) And the addition image s _add (x, y).
The filter unit 82 also includes the added image s _add (x, y), the estimated value P (x, y) hat of the signal power P (x, y) calculated by the signal power estimation unit 64, and noise power. The radar image s _b (x, y) is generated by performing a filter operation using the estimated value σ ² hat of the noise power σ ² calculated by the estimation unit 63.

  In the present invention, within the scope of the invention, any combination of the embodiments, or any modification of any component in each embodiment, or omission of any component in each embodiment is possible. .

  The signal processing apparatus, the signal processing method, and the radar apparatus according to the present invention are suitable for highly accurately estimating the power of the signal component and the power of the noise component included in the radar signal.

  DESCRIPTION OF SYMBOLS 1 Reception apparatus, 2 Analog circuit, 3 A / D converter, 4 Transmission wave separation part, 5 Storage apparatus, 6 Signal processing apparatus, 7 Display apparatus, 11 Imbalance correction part, 12 Power estimation part, 13 Noise power estimation part , 14 signal power estimation unit, 15 signal to noise power ratio calculation unit, 16 signal generation unit, 21 imbalance correction processing circuit, 22 noise power estimation processing circuit, 23 signal power estimation processing circuit, 24 signal to noise power ratio calculation processing Circuit, 25 signal generation processing circuit, 31 memory, 32 processor, 41 first device, 42 second device, 51 communication device, 52 communication device, 53 storage device, 61 imbalance correction unit, 62 power estimation unit, 63 Noise power estimation unit, 64 signal power estimation unit, 65 signal-to-noise power ratio calculation unit, 66 signal generation unit, 71 imbalance correction processing circuit, 72 Sound power estimation processing circuit, 73 signal power estimation processing circuit, 74 signal-to-noise power ratio calculation processing circuit, 75 signal generation processing circuit, 81 addition signal calculation unit, 82 filter unit, 91 addition signal calculation processing circuit, 92 filter operation circuit .

Claims (13)

An interference signal between a plurality of radar signals is obtained, the interference signal is averaged by the number of samples of the plurality of radar signals, and a noise power that is a power of a noise component included in the plurality of radar signals is calculated. Power comprising the values of signal components contained in at least two of the plurality of radar signals using the real part of the complex number in the averaged interference signal and the averaged noise power averaged by the number of samples A signal power estimation unit for estimating one signal power of
The plurality of radar signals are radar signals obtained from respective radio waves received by a plurality of antennas,
The geometrical conditions of the radio waves received by the plurality of antennas are the same, the values of the signal components included in the plurality of radar signals are the same, and the average power of the noise components included in the plurality of radar signals is A signal processing apparatus characterized by being identical. An average signal of a plurality of radar signals is obtained, the average signal is averaged by the number of samples of the plurality of radar signals, and a noise power that is a power of a noise component included in the plurality of radar signals is calculated as the number of samples. And using the averaged average signal and the averaged noise power, one signal power which is a power composed of signal component values included in at least two of the plurality of radar signals is obtained. A signal power estimation unit for estimation,
The plurality of radar signals are radar signals obtained from respective radio waves received by a plurality of antennas,
The geometrical conditions of the radio waves received by the plurality of antennas are the same, the values of the signal components included in the plurality of radar signals are the same, and the average power of the noise components included in the plurality of radar signals is A signal processing apparatus characterized by being identical. A first radar signal and a second radar signal are obtained from two radio waves received by two antennas,
An imbalance correction unit that corrects an imbalance of the second radar signal with respect to the first radar signal ;
Noise that estimates noise power, which is power of a noise component included in the first and second radar signals, from the first radar signal and the second radar signal corrected by the imbalance correction unit. A power estimation unit,
The signal power estimation unit claims, characterized in that averaging in the number of samples the interference signals between the first radar signal and a second radar signal after correction by the imbalance correction unit and The signal processing apparatus according to 1 . A first radar signal and a second radar signal are obtained from two radio waves received by two antennas,
An imbalance correction unit that corrects an imbalance of the second radar signal with respect to the first radar signal;
Noise that estimates noise power, which is power of a noise component included in the first and second radar signals, from the first radar signal and the second radar signal corrected by the imbalance correction unit. A power estimation unit,
3. The signal power estimation unit averages an average signal of the first radar signal and a second radar signal corrected by the imbalance correction unit based on the number of samples. Signal processing equipment. The noise power estimation unit obtains a difference signal between the first radar signal and the second radar signal corrected by the imbalance correction unit, and calculates an estimated value of the noise power from the difference signal. the signal processing apparatus according to claim 3 or claim 4 wherein. When the number of samples or the number of pixels of the first radar signal and the second radar signal after correction is N, the noise power estimation unit and the first radar signal at each sample or each pixel The square value of the difference signal from the corrected second radar signal is calculated, and the sum of the square values at each sample or each pixel is divided by 2N to calculate the estimated value of the noise power. The signal processing apparatus according to claim 5 . A signal-to-noise power ratio for calculating a signal-to-noise power ratio, which is a ratio of the signal power to the noise power, from the signal power estimated by the signal power estimation unit and the noise power estimated by the noise power estimation unit 5. The signal processing apparatus according to claim 3, further comprising a calculation unit. A signal generation unit that converts the signal power estimated by the signal power estimation unit into a signal amplitude, and generates a signal having a phase that is the phase of the first radar signal and an amplitude that is the converted amplitude; 5. The signal processing apparatus according to claim 3 or 4, wherein: An addition signal calculation unit for calculating an addition signal between the first radar signal and the second radar signal corrected by the imbalance correction unit ;
By performing a filter operation by a Wiener filter using the addition signal calculated by the addition signal calculation unit, the signal power estimated by the signal power estimation unit, and the noise power estimated by the noise power estimation unit. 5. The signal processing apparatus according to claim 3, further comprising a filter unit that generates a radar signal. A signal power estimation unit obtains an interference signal between a plurality of radar signals, averages the interference signal with the number of samples of the plurality of radar signals, and power of noise components included in the plurality of radar signals. A signal included in at least two of the plurality of radar signals by using the real part of the complex number in the averaged interference signal and the averaged noise power . When estimating one signal power that is a power composed of component values,
Using the radar signal obtained from each radio wave received by a plurality of antennas as the plurality of radar signals,
The geometrical conditions of the radio waves received by the plurality of antennas are the same, the values of the signal components included in the plurality of radar signals are the same, and the average power of the noise components included in the plurality of radar signals is A signal processing method characterized by being identical. The signal power estimation unit obtains an average signal of a plurality of radar signals, averages the average signal by the number of samples of the plurality of radar signals, and is a power of a noise component included in the plurality of radar signals. Noise power is averaged by the number of samples, and using the averaged average signal and the averaged noise power, a power composed of signal component values included in at least two of the plurality of radar signals. When estimating one signal power,
Using the radar signal obtained from each radio wave received by a plurality of antennas as the plurality of radar signals,
The geometrical conditions of the radio waves received by the plurality of antennas are the same, the values of the signal components included in the plurality of radar signals are the same, and the average power of the noise components included in the plurality of radar signals is A signal processing method characterized by being identical. A receiving device for receiving a plurality of radio waves;
An interference signal between a plurality of radar signals obtained from a plurality of radio waves received by the receiving device is obtained, the interference signal is averaged by the number of samples of the plurality of radar signals, and included in the plurality of radar signals Noise power , which is the power of the noise component being averaged, is averaged by the number of samples, and the real part of the complex number in the averaged interference signal and the averaged noise power are used to at least 2 of the plurality of radar signals. A signal power estimator that estimates one signal power that is a power composed of values of signal components included in the
The plurality of radar signals are radar signals obtained from radio waves received by a plurality of antennas included in the receiving device,
The geometrical conditions of the radio waves received by the plurality of antennas are the same, the values of the signal components included in the plurality of radar signals are the same, and the average power of the noise components included in the plurality of radar signals is Radar apparatus characterized by being identical. A receiving device for receiving a plurality of radio waves;
An average signal of a plurality of radar signals obtained from a plurality of radio waves received by the receiving device is obtained, the average signal is averaged by the number of samples of the plurality of radar signals, and included in the plurality of radar signals A signal included in at least two of the plurality of radar signals by averaging the noise power, which is the power of the noise component, by the number of samples and using the averaged average signal and the averaged noise power A signal power estimator for estimating one signal power, which is a power composed of component values,
The plurality of radar signals are radar signals obtained from radio waves received by a plurality of antennas included in the receiving device,
The geometrical conditions of the radio waves received by the plurality of antennas are the same, the values of the signal components included in the plurality of radar signals are the same, and the average power of the noise components included in the plurality of radar signals is Radar apparatus characterized by being identical.

Images