JPS646705B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an unstaggered pulse Doppler radar, and more particularly to a clutter suppression device that suppresses unnecessary signals other than a target signal in a received signal and enables detection of the target signal.

In addition to the target signal, radar received signals include unnecessary received signals (for example, reflected received signals from the earth, mountains, etc., reflected received signals from the sea surface, reflected received signals from rain clouds, etc.; hereinafter, these unnecessary received signals are These clutter signals may make it impossible to detect the target signal, or the clutter signal may be mistakenly detected as a target signal other than the target signal. For this reason, various techniques for removing or reducing clutter signals have been employed in radar devices. A typical example of these is a Moving Target Indicator (MTI).
In recent years, FFT (Fast
Doppler processors using Fourier Tvansform filters and bandpass filters (hereinafter referred to as bandpass filters) have been used (for example, RO'Donnell, C.
Muehe, M. Labitt and L. Cartledge.
“ADVANCED SIGNAL PROCESSING FOR
AIRPORT SURVEILLANCE RADARS”
EASCON-74 Record, 1974, PP71-71F).

The MTI device removes reflected reception signals from the ground, mountains, buildings, etc. (hereinafter referred to as fixed clutter signals), and also removes reflected reception signals (hereinafter referred to as targets) from moving aircraft, ships, vehicles, etc. (hereinafter referred to as targets). This device detects a target in fixed clutter by increasing the target signal power to clutter signal power (hereinafter referred to as the S/C ratio) by increasing the power of the target signal. The MTI device used in radar equipment detects targets in clutter signals other than fixed clutter signals (for example, sea surface clutter signals, rain cloud clutter signals, etc.; hereinafter referred to as moving clutter signals) and fixed clutter signals in a tangential direction to the radar equipment. This method has the disadvantage that it is almost impossible to detect targets that move rapidly or move slowly through fixed clutter.

On the other hand, Dotsupura processor uses FFT filter,
Alternatively, a configuration is adopted in which a plurality of bandpass filters are provided in place of the conventional MTI canceler (hereinafter referred to as a canceller) or in a subsequent stage of the canceller. The filter characteristics are used to separate the target signal and clutter signal based on the difference in their Doppler frequencies, and the target is detected.It has superior performance compared to the above-mentioned MTI device in the following respects. In other words, with an MTI device, it is almost impossible to detect a target signal in a moving clutter signal, but with a Doppler processor, the clutter signal and target signal are separated based on the difference in the Doppler frequency of the received signal, and the target signal is detected. In addition to the ability to detect targets in fixed clutter signals, it is also possible to detect targets in moving clutter signals, unless the Doppler frequencies of the moving clutter signals are approximately the same. Furthermore, in the case of a Doppler processor that does not have a canceller, the power of the target does not drop drastically, unlike in the case of an MTI device, even if the radial velocity of the target approaches zero due to its filter characteristics.
Figures 1a and b show the frequency response characteristics (or speed response characteristics) of the MTI device and Doppler processor.
Here is an example. In the figure, T indicates the pulse repetition period.

As mentioned above, Doppler processors compensate for the drawbacks of MTI devices in several ways, but Doppler processors also eliminate the difference between the Doppler frequency of the moving clutter signal and the Doppler frequency of the target signal to zero. As the Doppler frequency of the target signal in the fixed clutter approaches zero (for example, as the Doppler frequency of the target signal in the fixed clutter approaches zero), it becomes impossible to obtain sufficient target detection ability. In this case, in order to detect the target while taking advantage of the advantages of the Doppler processor mentioned above, it is necessary to increase the frequency resolution of the FFT filter and bandpass filter used in the Doppler processor. . This will be discussed in detail later. However, in order to increase the frequency resolution,
It becomes necessary to increase the number of FFT points or the order of the bandpass filter, which means that the hardware of the Doppler processor becomes more complex and larger.

Furthermore, by increasing the number of FFT points or the order of the bandpass filter, it is conceivable that the interval from when one filter output is obtained to when the next output is obtained becomes longer. For example, when considering N-point FFT, the FFT processing result can be obtained for each received signal for N sweeps of the radar (for example, GERALD O'LEARY,
“Nonrecursive Digital Filtering Using
Cascade Fast Fourier Transformers”IEEE
TRANS.AUDIO & ELECTRO
ACOUSTICS vol.AU；18 No.2 June 1970)
It is.

The above points will be further explained below by giving specific examples.

First, target detection in a clutter signal by the Doppler processor will be explained. 4 points as filters used in Dotsupura processors
Considering FFT, consider a received signal in which the Doppler frequency of the target signal and the Doppler frequency of the clutter signal are relatively far apart. Second
The figure shows the frequency response characteristics of a 4-point FFT. 4 points
The FFT can be considered to be a combination of four bandpass filters having frequency response characteristics as shown in FIG. 2 b to e, and the combined characteristics are as shown in FIG. 2 a. T in FIG. 2 represents the pulse repetition period of the pulse device of the radar device. b is at the Doppler frequency of n/T (n=0, 1, 2...), c is at the Doppler frequency of n/T+1/4T (n=0, 1, 2...), and d is n/T+1 /2T (n=0, 1, 2...), e is n/T+3/4T (
It has a frequency response characteristic that has the maximum response at the Doppler frequency of n = 0, 1, 2...), and each filter output is called a filter bank output, and is F ₀ bank output, F ₁ bank output, and F ₂ bank output. Output, we will call it _F3 bank.

FIG. 3 shows how the target signal is extracted from the clutter signal using the FFT filter. FIG. 3a shows an A-scope waveform of a target signal submerged in a clutter signal over several sweeps, where 310 represents the target signal and 311 represents the clutter signal. Now, if the Doppler frequency of the target is 1/2T, and the clutter is a moving clutter, and its Doppler frequency is around 1/4T and fluctuates for each sweep, the clutter is a moving clutter, so the conventional
MTI can hardly detect targets.
However, in this case, each filter bank output of F ₀ , F ₁ , F ₂ , F ₃ of the 4-point FFT is b, c, d, e, respectively.
The target signal 340 is detected at the _F2 bank output d. This is Figure 2b
As can be seen from ~e, the Doppler frequency of the target signal is 1/2T, so the target signal appears only in the _F2 bank output of the 4-point FFT in Figure 3d, and does not appear in the other filter bank outputs. On the other hand, since the Doppler frequency of the clutter signal changes with each sweep, the clutter signal component appears in all filter bank outputs, but since the fluctuation is near the Doppler frequency of 1/4T, its energy most of
This is because the _F1 bank output appears in Figure 3c.

By performing four-point FFT in this way, it is possible to separate the target signal and clutter signal as different filter bank outputs, for example.
F ₃ bank output If only d in Figure 3 is taken out as the output of the Doppler processor, then the 4-point FFT
This means that clutter has been suppressed, and the target signal hidden in the clutter signal can be detected. In this way, the Doppler processor extracts the clutter signal by using the relative difference between the Doppler frequency of the clutter signal and the Doppler frequency of the target signal.
It is effective not only for fixed clutter signals like the MTI device, but also for target detection in moving clutter signals.

Next, in order to effectively detect the target when the difference between the Doppler frequency of the clutter signal and the Doppler frequency of the target signal approaches zero, it is necessary to
We will discuss the need to improve the frequency resolution of FFT or bandpass filters. Third
FIG. 4 shows how the four-point FFT filter cannot detect the target sufficiently when the Doppler frequency of the target approaches that of Clutter compared to the case shown in the figure. FIG. 4a shows the same A-scope waveform as FIG. 3a, but assume that the Doppler frequency of the target has shifted from 1/2T and is close to 1/4T. At this time, the four-point FFT F ₀ , F ₁ , F ₂ ,
The outputs of each filter bank of _F3 are shown in Fig. 4b, respectively.
c, d, e.

As can be seen from Fig. 2 b, c, d, and e, when the Doppler frequency of the target signal deviates from 1/2T, the amplitude of the target signal in the _F2 bank output decreases, and accordingly, F ₀ , F ₁ , F The target signal component will also be included in the _3- bank output. When the Doppler frequency of the target approaches 1/4T, most of the energy of the target signal that previously appeared only in the _F2 bank is transferred to the _F1 bank.
It will appear on the bank output. On the other hand, since the clutter signal is the same as in Figure 3, most of the energy of the clutter signal is the same as in Figure 3c.
Appears on the _F1 bank output. Therefore, the S/C ratio at the _F1 bank output Fig. 4c is the received signal 4
In this case, the target signal and clutter signal cannot be separated as different filter bank outputs as shown in Figure 3 c and d, and the clutter signal It becomes impossible to detect the hidden target signal. In such cases, FFT used in Doppler processors can be used to separate the two.
Either it is necessary to increase the number of points in the filter, or it is necessary to provide a larger number of bandpass filters with narrower passband widths than before. By doing this, even if the Doppler frequency of the target approaches 1/2T to 1/4T, the energy of the target signal will reach 1/2T before it appears at the filter bank output where most of the clutter signal components appear. It appears as the output of a filter bank that has a maximum response characteristic at a certain frequency between 1/4T. Therefore, since the filter bank output where the energy of the clutter signal is concentrated and the filter bank output where the energy of the target signal is concentrated are different,
As in d, the clutter signal and target signal can be separated, and in the case of 4-point FFT, targets with undetectable Doppler frequencies can also be detected. This situation will be explained using FIGS. 5 and 6.

Figure 5 shows the frequency characteristic of 8 compared to the 4-point FFT.
This shows the frequency response characteristics of point FFT.
The 8-point FFT can be thought of as a combination of 8 bandpass filters with the characteristics shown in Figure 5 a to h, and can be considered to have twice the frequency resolution compared to the 4-point FFT. . T in FIG. 5 represents the pulse repetition period as in FIG. 2. At this time, Fig. 5 a shows l/T (l=0, 1, 2, ...), and b shows l/T+1/8T (l=0, 1, 2, ...)
, c is l/T+1/4T (l=0, 1, 2,...), d is l/T+3/8T (l=0, 1, 2,...), e is l/T
+ 1/2T (l=0, 1, 2,...), f is l/T+5/8T (l=0, 1, 2,...), g is l/T+3/4T (l= 0, 1 . Similarly, we will call them F ₀ bank output, F ₁ bank output, F ₂ bank output, F ₃ bank output, F ₄ bank output, F ₅ bank output, F ₆ bank output, and F ₇ bank output.

In this case, as in the case of Figure 3, if we consider the case where the Doppler frequency of the target is 1/2T and the Doppler frequency of the clutter is around 1/4T, the score is 8 points.
In the case of FFT, the target signal appears only at the _F4 bank output, and most of the clutter signal's energy appears at the _F2 bank output. As in the case of Figure 4, the clutter signal remains the same, but the target Doppler frequency deviates from 1/2T to 1/4T.
Consider the case where the value is close to . When the Doppler frequency of the target deviates from 1/2T, the amplitude of the target signal in the F4 bank output decreases, as in the case of 4 _- point FFT, and the target signal component comes to be included in all frequency bank outputs. When the Doppler frequency of the target approaches 1/4T and reaches 3/8T, the target signal will appear only at the output of this _F3 bank because there is a filter bank with characteristics that gives the maximum response at this frequency in the 8-point FFT. .

On the other hand, most of the energy of the clutter signal is at the frequency that gives the maximum response in the 4-point FFT.
Because it appears in the same F ₂ bank output as F ₁ bank,
In the case of 8-point FFT, the clutter and target signals can be separated as _F2 bank outputs. That is, by using a filter with higher frequency resolution, such as an 8-point FFT, instead of a 4-point FFT, it becomes possible to detect targets that were not possible with the 4-point FFT. Figure 6 shows the output of each filter bank in this state,
A is the input signal, which is the same as the target signal in FIGS. 3a and 4a, except that the Doppler frequency is changed. Figures 6 b to i are F ₀ respectively.
Bank output, F ₁ bank output, F ₂ bank output, F ₃
Bank output, F ₄ bank output, F ₅ bank output, F ₆
Bank output, F ₇ represents bank output, 640,
650 are the separated clutter signal and target signal, respectively. When the Doppler frequency of the target becomes even smaller than 3/8T and approaches 1/4T, it becomes difficult to separate the clutter from the target, as explained in the example of the 4-point FFT, and a filter with even higher frequency resolution becomes necessary. .

Furthermore, we will discuss the point that the hardware becomes complicated in order to increase the frequency resolution of the Doppler processor as described above. Now, as an example of a filter
Consider FFT and increase the frequency resolution by increasing the number of FFT points. The number of FFT points is M (M = 2 ^m , m = 1, 2, ...), and Figure 7 shows an example of a time-decimated FFT circuit for a real number input with this number of points (GERALD C.
O′LEARY “Nonrecursive Digital Filtering
Using Cascade Fast Fourier Transformers”
IEEE TRANSACTIONS ON AUDIO AND
ELECTROACOUSTICS vol.AU―18, No.2
June 1970). Figure 7 shows a well-known butterfly circuit connected in series, with 701, 70
2 and 703 are the 1st stage and 2nd stage, respectively, when viewed from the output side.
The butterfly circuits of the stage and m-th stage are shown. 704,
705 is a memory in the m-th stage butterfly circuit, each having a memory capacity for 2 ^m-1 sweeps. Further, 706 and 707 are adders, and 708 is a multiplier.

In the example shown in Figure 7, considering the increase in hardware when increasing the number of FFT points from N to N'(N' = 2 ^{m + 1} , m = 1, 2, ...), the number of FFT points will increase. Since the distance increases from 2 ^m to 2 ^m+1 , the number of butterfly circuits increases by one stage. The memory capacities of the two sweep memories of this added butterfly circuit are each for a 2 ^m sweep, and the increments of the adder and multiplier are 2 digits and 1 digit, respectively. In other words, by doubling the number of FFT points, the memory becomes 2 ^{m + 1} sweeps (2 x 2 ^m ), an adder,
The number of multipliers increases by 2 and 1, respectively.
We considered a real number input, but the radar received signal is generally processed as a complex number input consisting of the I and Q outputs of a phase detector, so the increment in this case is twice the memory and four times the multiplier compared to the above increment. , the adder is tripled, the memory is for 2 ^m+2 sweeps, there are 6 adders, and 4 multipliers. Now, since the increment of adders and multipliers is considered to be smaller than the increment of memory, if we focus on the increment of memory, we can see 704 or 7 in Fig. 7.
Assuming that the number of amplitude quantization is Q bits and the number of processing ranges is R, the additional memory capacity required is (R×Q×2 ^m+2 ) bits. In other words, in order to double the frequency resolution, it is necessary to increase this amount of memory, especially for m
This becomes a big problem when FFT is large, that is, when performing multi-point FFT in a relatively small space and with low power consumption.

Furthermore, in the circuit shown in Fig. 7, it takes time for N sweeps to complete the outputs of each frequency bank, but this means that the outputs of each frequency bank can only be obtained every N sweeps.
This becomes a problem when the number of FFT points increases.

In other words, if you simply perform FFT processing or digital filter processing like a conventional Doppler processor, it will have high frequency resolution only for Doppler frequencies near the Doppler frequency of Clutter, and will not be able to perform high frequency resolution sufficiently far from the Doppler frequency of Clutter. For Doppler frequencies, it is not necessary to construct a filter with relatively low frequency resolution, but rather to construct a filter with high frequency resolution over all frequency regions, so the hardware required for this purpose is large. The disadvantages are that the scale and complexity increase, and the time interval at which output data is updated becomes long.

In the Doppler processor of the present invention, increasing the frequency resolution of the filter does not make much sense for a target having a Doppler frequency that is sufficiently distant from the Doppler frequency of the clutter, and it is necessary to increase the frequency resolution. It is important to note that only the Doppler frequency near the Doppler frequency of the clutter signal is required, and the frequency resolution is high enough to detect the target signal in the vicinity only in the frequency region near the Doppler frequency of the clutter signal, and other By constructing a filter with a relatively low frequency resolution in the region of To provide a clutter suppression device that reduces the scale of the signal and further reduces the number of sweeps required to obtain an output signal for a target signal having a Doppler frequency relatively far from the Doppler frequency of the clutter signal. be.

In order to achieve the above object, a clutter suppression device according to the present invention includes an A/D converter that A/D converts an input radar reception signal, and performs a discrete Fourier transform operation using the output of the A/D converter as input. The output of the discrete Fourier transform circuit and the A/D converter is used as an input, and the frequency that provides the maximum response is selected as a frequency that has a narrower pass band width than that of each filter bank of the discrete Fourier transform circuit. a plurality of bandpass filters capable of changing the frequency to a frequency close to the Doppler frequency of Clutter according to Doppler frequency information; and a plurality of bandpass filters that receive the output of the discrete Fourier transform circuit and include the Doppler frequency of Clutter according to the Doppler frequency information of Clutter. a video selection circuit that inhibits filter bank output and outputs other filter bank output; and a video that inputs the output of the band pass filter and the output of the video selection circuit, synthesizes the amplitudes of both, and outputs the synthesized signal. and a synthesis circuit.

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 8 shows a block diagram of a clutter suppressing device according to the present invention. In FIG. 8, 801,
802 is an I (In-phase) video and a Q (Quadrature-phase) video from a phase detector;
The embodiment of the present invention uses A/D converters 803 and 804 which input signals 801 and 802, and performs N-point disjunctive Fourier transform by inputting N sweep outputs of these A/D converters 803 and 804. N points to do
A video selection circuit 806 receives N filter bank outputs from FFT circuits 805 and 805 and selects and outputs an arbitrary filter bank output from the N filter bank outputs using a control signal 809; A control signal 809 receives a signal 807 regarding Doppler frequency information and controls the video selection circuit 806.
and bandpass filters 811, 81 described below.
A control circuit 808 that outputs a control signal 810 for changing the frequency response characteristics of 2,813,814, and A/D converters 803 and 80.
4 output and control signal 810 as input, N
bandpass filters 811, 812, 813, and 814 that have narrower passband widths than the point FFT circuit 805 and can change the frequency that gives the maximum response according to the control signal 810;
It includes a video synthesis circuit 815 which receives the output of the video selection circuit 806, performs timing adjustment and synthesis between the respective inputs, and outputs a signal 816 in which a clutter signal is suppressed. Here, it is assumed that the circuit already explained in FIG. 7 is used as the N-point FFT circuit, and the one shown in FIG. 9 is used as the bandpass filter.

I and Q videos 801 and 802 from the phase detector containing the target signal and clutter signal are A/D
After being converted into digital signals by converters 803 and 804, the signals are sent to N-point FFT circuit 805 and bandpass filters 811, 812, 813, and 814. An N-point FFT circuit 805 receives the outputs of the A/D converters 803 and 804 over N sweeps, performs N-point disjunctive Fourier transform for each range bin, and outputs the resultant output. Here, the time unit quantized by the A/D converter 803 or 804 is called a range bin. In the N-point FFT circuit 805, after the signals for N-1 sweeps are input from the A/D converters 803 and 804, one filter bank output is processed by N-point FFT every time a signal for one sweep is input.
The signal is output from the circuit 805, and after a time period of N sweeps, N filter bank outputs are obtained. The video selection circuit 806 outputs only the filter bank outputs with fewer clutter signal components from among the N filter bank outputs thus input, and sends them to the video synthesis circuit 815. At this time, the selection of the filter bank output in the video selection circuit 806 is performed by the control signal 809. In the bandpass filters 811, 812, 813, 814,
The outputs of A/D converters 803 and 804 are received over M sweeps (M>N), and a signal having a frequency component determined by each frequency response characteristic is output for each M sweep and sent to a video synthesis circuit 815.

Here, band pass filters 811, 812, 8
The operation of No. 13,814 is as follows. FIG. 9 shows the configuration of the bandpass filter in this embodiment, which corresponds to a specific filter bank in the disjunctive Fourier transform. 9th
The algorithm of the circuit shown in the figure is as follows: where the input complex number signal is x(nT), n=0, 1, 2,..., T: sampling period, and the output signal is y(nT) (real number), y(nT)= _M 〓 ⁿ⁼⁰ x(nT)・W ^nk W ^nkT = exp [−j2πnk/M] (weighting coefficient) k=0, 1, 2,... M: An integer larger than N. In FIG. 9, multiplier 903,
904, 905, 906, and adders 911, 9
12, the input I, Q video 901, 90
2 is multiplied by weighting coefficients 909 and 910. Adders 913 and 914 execute addition in the above equation. The addition result is stored in the sweep memory 91
5,916, and processing is performed for the required number of range bins in this way. The outputs of the sweep memories 915 and 916 undergo calculation of the amplitude of the complex signal in an amplitude calculation circuit 917, and are outputted through a gate 918. Here, 907 is a control signal for changing the frequency characteristics, and 90
8 is a weighting coefficient generation circuit that receives this control signal and generates a predetermined weighting coefficient, and a gate signal outputted every M sweep is a signal for outputting the calculation result only when the calculation of the above formula is completely performed. It is. Bandpass filter 811, 812, 8
13 and 814 are different from 909 and 910 in FIG. 9 depending on their frequency characteristics.

In the video synthesis circuit 815, the video selection circuit 80
6 output and bandpass filters 811, 812,
813 and 814, and combine the outputs of each filter bank sent from the video selection circuit 806 so that the signal is updated every N sweeps. , 813, 814 are combined and output. As a result, the output 816 of the video synthesis circuit 815 is
This is a signal that is updated every sweep. Selection of filter bank output in video selection circuit 806 and bandpass filters 811, 812, 813,
The frequency response characteristics of 814 are changed by outputs 809 and 810 of control circuit 808. The input 807 of the control circuit 808 is, for example, a signal artificially set from a control panel or the like, or a signal provided from a device (or circuit) that detects the Doppler frequency of the clutter signal. For example, the frequency corresponding to the filter bank with the highest clutter level is set as the Doppler frequency of the clutter signal. 809 is a gate signal for passing or cutting off the signal from the N-point FFT circuit 805 in the video selection circuit 806, and 810 is a gate signal for passing or cutting off the signal from the N-point FFT circuit 805, for example.
1,812,813,814 are signals that control the coefficients of a transversal filter.

Now, the operation of the embodiment when the Doppler frequency of the clutter signal is close to 0 is as follows. First, information regarding the Doppler frequency of the clutter signal is sent to the control circuit 808, and from this circuit 808, for example, among the outputs of the N-point FFT circuit 805,
A signal is sent to video selection circuit 806 to block each filter output bank of F ₀ , F ₁ , and F ₂ .
At this time, the frequency response characteristic seen from the output of the video selection circuit 806 is as shown in FIG. 10a, for example. However, it is assumed that the Doppler frequency of the clutter at this time is near 0. Figure 10 1010,10
11 and 1012 are frequency response characteristics of the blocked filter bank among the frequency characteristics of the N-point FFT, which are indicated by broken lines. On the other hand, control circuit 8
The band pass filters 811, 812, 813, 814 are controlled by the control signal 810 from 08.
are 1020, 1021, 1 in Figure 10b, respectively.
It is set to have response characteristics such as 022 and 1023. Therefore, the bandpass filter 81
When 1,812, 813, and 814 are viewed comprehensively, the frequency response characteristics are as shown in FIG. 10b.
Since the video synthesis circuit 815 in FIG. 8 synthesizes the input signals that have passed through the filters in FIG. 10, the frequency response characteristic seen at the output side of the video synthesis circuit 815 becomes as shown in FIG. 10c. As a result, in this example, clutter such as ground clutter where the Doppler frequency is almost zero is sufficiently removed, and the first
It is possible to detect targets up to a Doppler frequency of approximately f ₂ as shown in Figure 0b.

The scale of the hardware according to this embodiment is compared with that shown in FIG. 5 as follows. As already mentioned, from the FFT of N points (N=2 ^m )
If you increase the FFT of N′ points (N′=2 ^m+2 ) and the number of points, the memory increment is R×Q×2 ^m+2
(Q: number of amplitude quantization, R: number of processing range bins), but as shown in Figures 6 and 8, FFT
If you install 4 bandpass filters instead of increasing the number of points, the memory increase will be 4 x 2 x R.
×Q, which is 2 ³ /2 ^m+2 = 1/2 ^m-1 compared to the conventional case
and decreases, and as m increases, the degree of decrease also increases.

As explained above, the present invention uses, for example, I and Q videos from a phase detector as input, an A/D converter that performs A/D conversion, and the output of the A/D converter as input, and A disjunctive Fourier transform circuit that performs transform operations, or a plurality of digital filters with similar frequency response characteristics,
Similarly, the output of the A/D converter is input, and the output of the disjunctive Fourier transform circuit is input to a plurality of band pass filters having a function of changing the frequency response characteristics. A clutter removal device is constituted by a video selection circuit that selects and outputs an arbitrary filter bank output from the video selection circuit, and a video synthesis circuit that receives the output of the video selection circuit and the output of the bandpass filter as input, and synthesizes the amplitudes of both. By doing this, the target signal that is covered in clutter can be processed over all frequency ranges like a multi-point Fourier transform circuit.
Compared to configuring a filter with high frequency resolution, this method has the advantage of being able to detect targets with Doppler frequencies close to the Doppler frequency of the clutter (fixed clutter and moving clutter) with a small-scale hardware configuration. be.

[Brief explanation of drawings]

Figures 1a and b show the frequency response characteristics of the MTI canceller and Doppler processor, Figure 2 shows the frequency response characteristics of the 4-point FFT, and Figure 3 shows the frequency response characteristics of the 4-point FFT. Figure 4 shows how the target signals are separated as different frequency bank outputs. When the Doppler frequency of the clutter signal and the Doppler frequency of the target signal are close to each other, the target signal and the clutter signal are not separated. Figure 5 is a diagram showing the frequency response characteristics of 8-point FFT, Figure 6 is a diagram showing how targets in a clutter signal are separated as different frequency bank outputs by 8-point FFT, FIG. 7 is a typical FFT operation circuit diagram, FIG. 8 is a block diagram showing an embodiment of the present invention, and FIG. 9 is a diagram showing an example of the configuration of the bandpass filters 811 to 814 in FIG. FIG. 10 is a diagram showing an example of the frequency response characteristic of the clutter suppressor according to the present invention. 701...Final stage butterfly circuit, 702,
703... middle stage and first stage butterfly circuit, 70
4,705...Sweep memory, 706,707
... Adder, 708 ... Multiplier, 709 ... Weighting coefficient in Fourier transform, 801, 802 ...
Phase detector output, 803, 804...A/D converter, 805...N-point FFT calculation circuit, 806...
Video selection circuit, 807... Signal of Kuratsuta's Doppler frequency information, 808... Control circuit, 809... Control signal, 810... Control signal, 811, 812, 813, 81
4... Band pass filter, 815... Video synthesis circuit, 816... Output signal, 901, 902...
...I, Q video, 903,904,905,90
6... Multiplier, 907... Control signal, 9
08... Weighting coefficient generation circuit, 909, 910...
Weighting coefficient signal, 911, 912, 913, 914
... Adder, 915, 916 ... Sweep memory, 917 ... Amplitude calculation circuit, 918 ... Gate circuit, 919 ... Gate signal, 920 ... Output signal.

Claims (1)

[Claims] 1 A/D that converts the input radar reception signal to A/D
a discrete Fourier transform circuit that receives the output of the A/D converter as input and performs a discrete Fourier transform operation, and each filter of the discrete Fourier transform circuit that receives the output of the A/D converter as input; a plurality of bandpass filters having passband widths narrower than the passband width of the bank and capable of changing a frequency that provides the maximum response to a frequency close to the Doppler frequency of the clutter in accordance with Doppler frequency information of the clutter; a video selection circuit that receives the output of the discrete Fourier transform circuit and inhibits the output of a filter bank including the Doppler frequency of Clutter according to the Doppler frequency information of the discrete Fourier transform circuit, and outputs the output of another filter bank; 1. A clutter suppression circuit comprising: a video synthesis circuit that inputs the output of the video selection circuit, synthesizes the amplitudes of both, and outputs the synthesized signal.