JPH01250779A - Feedthrough nulling circuit - Google Patents

Abstract

PURPOSE:To improve a response speed of a loop by a method wherein a loop circuit is made to be of an open loop and weight computation is executed only with an input signal irrespective of a loop constant. CONSTITUTION:A reference signal d(R)4 is inputted to a weight computing circuit 31 and given to a multiplier 7a. Meanwhile, a transmission signal X(R)1 extracted from the transmission side is also inputted to the weight computing circuit 31 and given to another multiplier 7b, while it is given to a complex conjugate unit 32, where it is turned to be X*(R)33 and given to the multiplier 7a. It is turned to be d(K)X*(R) by the multiplier 7a and averaged by an averaging circuit 34a. On the other side, an output X*(R) of the complex conjugate unit 32 is given to the multiplier 7b, where X*(R)X(R) is taken, and further the output is averaged by another averaging circuit 34b. These two signals are removed by a divider 35, while the transmission signal X(R)1 is weighted by a weighting mixer 7, and a difference between this signal and the signal d(R)4 inputted from a reception antenna 12 is taken in an adder 5. A feedthrough can be removed from a weighted transmission signal W(R)X(R)14 by taking a difference from a reference signal d(R).

Description

Detailed Description of the Invention (Industrial Application Field) The present invention is directed to a CWW radar in which a transmission signal other than a necessary radar echo caused by coupling between a transmitting antenna and a receiving antenna is routed to the receiving receiver No. 8. The present invention relates to a radar signal processing device whose purpose is to remove feedthrough signals.

(Prior art) The purpose of CWW radar is to receive echoes from moving targets, but because it handles CW waves, transmission and reception are performed at the same time, so there is a large feed-through from the transmitting antenna to the receiving antenna. Problems with machine saturation and burnout occur. For this reason, a feed-through nulling circuit is used which applies an adaptive signal processing method that uses the received signal 48 and the transmitted signal to remove the transmitted signal 48 included in the received signal.

In the configuration shown in FIG. 2, the adaptive signal processing circuit assumes a signal obtained by multiplying the input signal x1 by a weight W2 as y3,
If the output obtained by subtracting the reference signal d4 by the subtracter 5 is the error ε6, then the relationship is expressed by equation (1).

ε=:y-d=WX-d...(1)
In formula (1), minimizing the error ε means finding the weight W that makes the input signal No. 8 X the same phase and amplitude as the reference signal d, and the weight W that satisfies this condition is:
Equation (2) is known as the Wiener-Hopf solution.

Wopt= [X'X] −' [dX”]
...(2) Wiener −Ho shown in equation (2)
It is known as Wtdrov's adaptive loop that expresses the solution of p'' in the form of sequential solutions. Expressing this in the formula, +r for W, = WK+2△S・εuo・X++
...(3) ε(K) = d (K) The ones marked with are samples when the circuit is executed using digital signal processing, and this is expressed as K. (3), (4)
The algorithm of Widrow's loop shown by the formula is written in C
When applied to the feed-through nulling circuit of W radar,
The configuration is as shown in Figure 3. The operation is such that a transmission signal from a transmitter 8 is radiated from a transmission antenna 9, the signal enters a target 10, and a reflected signal 13 is input to a reception antenna 12 as an echo. On the other hand, since the amount of coupling between the transmitting antenna and the receiving antenna is feedthrough 11, which is input to the receiving antenna 12, the reference signal d, 4) is considered to correspond to the sum of the desired signal and the transmitting signal.

Here, the operation of the feedthrough nulling circuit will be explained by comparing the algorithms (3) and (4) equations with FIG. 3.
The sum circuit 5 takes the difference between (, 4, and X, is multiplied by 1, the amplifier 15 multiplies the loop gain by 2ΔS, the integrator 16 averages it, and executes equation (3) to obtain a new weight W (!)
I'm looking for 2.

(Problem M to be Solved by the Invention) Due to this circuit configuration, the circuit is a closed loop, and the response speed of the loop is determined by the loop constant 2ΔS, so there is a limit to the convergence time of the loop. Furthermore, since the convergence speed of the loop is determined by the input signal level and the loop constant, there is a drawback that the calculation speed is easily influenced by external conditions.

(Means for Solving the Problems) In order to eliminate these drawbacks, the present invention directly solves the W 1ner -Hop f solution shown in equation (2), and creates a circuit configuration expressed in hardware from the obtained optimal weights. An open loop is provided, and examples will be explained in detail below with reference to the drawings.

(Example) When equation (2) is directly solved, the circuit configuration that is expressed entirely in hardware is shown in Figure 1. In this figure, elements or signals having the same functions as those in FIG. 2 are represented by the same symbols.

To explain the operation of this circuit, the reference signal d+a+
4 is input to the weight calculation circuit 31 and added to the multiplication N7a.

On the other hand, the transmitted signal X(R)1 extracted from the transmitting side is also input to the weight calculation circuit 31, and is added to another multiplier 7b as well as the complex conjugator 32, where it becomes X''(R)33 and the multiplier Join 7a.

The multiplier 7a yields d(K)X''(R), and the average circuit 34
a, and a discrete cross-correlation between the transmitted signal, which is the numerator of equation (5), and reference No. 48 is taken. On the other hand, complex conjugator 3
2's output X'' (R) is added to the multiplier 7b where
``(R)X(R) is taken and averaged by another averaging circuit 34b, and discrete autocorrelation of the transmitted signal, which becomes the denominator of equation (5), is performed.

Both signals can be divided by the divider 35, and the weight W (R)2
As for the weighting, the difference between the transmit signal
(R) X (R) 14 Lt, standard (No. 8 d<r
The signal has the same phase and amplitude as the feedthrough included in r>4, and the feedthrough can be removed by taking the difference from d(R).

In this way, it can be seen that the weight calculation circuit itself does not include a closed loop, and the output side of the summation circuit 5 has no relation to the weight calculation circuit 31 and does not form a closed loop.

All the components of this example were hardware, but
From an economical point of view, it is possible to construct the system using software to a considerable extent, and the present invention is not limited by this.

Although the present invention has been described as being applied to a CW radar, the idea can also be applied to preventing interference in wireless communication devices.

(Effects of the Invention) As described above, the loop circuit is an open loop, and weight calculation can be performed using only input signals regardless of the loop constant. Therefore, by increasing the sampling speed, the response speed of the loop can be improved. Furthermore, since equation (5) does not include a loop constant in the weight calculation, it has the advantage that the convergence speed of the weight calculation can be obtained regardless of the input level.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an embodiment of the present invention, in which it is applied to a 0W radar, Fig. 2 is a conceptual diagram of an adaptive signal processing circuit, and Fig. 3 shows a conventional feedthrough nulling circuit in CW mode.
Example applied to radar. 1... Input signal X (R1, 2... Weight W, R,, 4
...Reference signal d (rll, 5...sum circuit, 7...
・Multiplier, 8...Transmitter, 9...Transmission antenna, 1
0...Target (TGT), 11...Feed through, 12...Receiving antenna, 13---Reflected signal,
14... x, - w, 31... Weight calculation circuit, 32... Complex conjugator, 33... Summation circuit, 34... X6. 1.35...divider. Patent applicant Japan Radio Co., Ltd.

Claims (1)

[Claims] In the feed-through nulling circuit that removes signals that go directly from the CW radar transmitting side to the receiving side and reflected signals from fixed objects, the transmitted signal X_( partially extracted from the transmitting side)
_R_) and the reference signal d_(_R_) input to the receiving side.
Wiener by dividing by the sample mean value of the discrete autocorrelation between _(_R_) and its complex conjugate
- Obtain the optimal weight Wopt of the direct solution of the Hopf solution, and obtain the transmitted signal X_(_R_) from the reference signal d_(_R_).
What is claimed is: 1. A feed-through nulling circuit configured to obtain a desired radar reflected signal by subtracting a value multiplied by an optimal weight, and characterized in that the circuit configuration is an open loop.