JPH0369207B2 - - Google Patents

Description

[Detailed Description of the Invention] The present invention provides a sidelobe canceller (hereinafter referred to as SLC).
This relates to the improvement of the correlation loop gain of (abbreviated as).

A conventional SLC had a configuration as shown in Figure 1. In the figure, u and x are the carrier waves of the input signals from the main antenna and the auxiliary antenna, respectively.
It represents a signal converted to an intermediate frequency (IF) of w ₁ and w ₂ (w ₁ >w ₂ ) (hereinafter, u and x are called the main signal and reference signal, respectively). 1 is a subtracter, 2 and 5 are mixers, 3 is a narrow band pass filter with a center frequency of w ₁ -w ₂ , and 4 is an amplifier with a gain of g. In this configuration, there is a loop from subtracter 1 through mixer 2 and mixer 5 to subtracter 1 again. This is called a correlation loop.

Next, the operation of this device will be explained. Since the auxiliary antenna is omnidirectional and its gain is about the same as the gain of the side lobe of the main antenna, the power of the signal u is greater than the power of the signal x for radio waves incident from the main lobe of the main antenna. On the other hand, for radio waves incident from the side lobes of the main antenna, the power of the reference signal x is greater than that of the main signal u. This device uses this to estimate components in the main signal u that have a correlation with the reference signal x as a signal y obtained by multiplying the reference signal x by the control weight signal w, and estimates this from the main signal u. It has a structure in which the signal z obtained by subtracting the signal y is used as an output signal, and only the influence of the radio waves incident from the main lobe of the main antenna remains. The output signal z from the SLC is input to mixer 2 and fed back. If the output signals of the mixer 2 are v and v, and the signal after passing through the narrow band filter 3 and the amplifier 4 is w, then w becomes a signal with a center frequency w ₁ -w ₂ . Mixer 5 multiplies signal w and auxiliary antenna input signal x and outputs signal y. The above process can be expressed mathematically as follows. (All signal components shall be expressed as complex numbers.) First, the relationship between the main antenna input signal u and the output signal z is z=u−y (1). Next, if the narrowband filter 3 is a first-order lag filter with a time constant τ ₀ , then the following differential equation holds true with respect to the control load signal w, which is the output signal of the amplifier 4.

τ ₀ dw/dt+w=gzx ^* ...(2) where x ^* represents the conjugate complex number of x. Also,
Regarding mixer 5, y=wx...(3). If (1), (2), and (3) are used here, τ ₀ dw/dt+(1+g|x| ² )w=gux ^* ...(4). This is a first-order linear differential equation for the control load signal w. In equation (4), if the value of |x| ² and the envelope signal component of ux ^* change more slowly than τ ₀ and are approximately constant values, equation (4) has the following solution.

However, in formula (5), w ₀ is the initial value of w when t=0, and T ₀ and α are T ₀ =τ ₀ /1+g|x| ² ...(6) α=gux ^* The value is ^{/1+g|x|2...(7} ). In equation (5), the control load signal w
The value of will approach the value α as t becomes sufficiently large. Figure 2 shows the absolute value of the control load signal w |w|
It shows how converges to the value |α|. The time constant at this time is T ₀ . On the other hand, as shown in the figure, the output signal z is z=u−wx ...(8), so the signal suppression ratio (CR) by the sidelobe canceller is CR=<(u/z) ² >=<(u/ u−wx) ² > ...(9) If defined as (where <> indicates the time average) t
When becomes sufficiently large, α in equation (7) is substituted for w in equation (9), and CR=(1+g|x| ² ) ² ...(10). As is clear from equation (10), the signal suppression ratio CR
depends on the value of reference signal power |x| ² and increases as reference signal power |x| ² increases. This relationship is shown in FIG.

In this way, the convergence time constant T ₀ of the control load voltage w and the signal suppression ratio CR due to SLC are determined by the amplifier gain g
and the reference signal power |x| ² , which is the product g|x| ² . Especially when the reference signal power |x| ² is a small value, the gain g of the amplifier 4 must be set large.
The convergence time constant T ₀ of the control load voltage w becomes large. However, when the gain g is large, the loop gain becomes too large when the reference signal power |x| ² is large, and oscillation occurs due to the nonlinear elements of the circuit. In the conventional SLC circuit, as described above, the convergence speed of the control load voltage w is the reference signal power |x| ²
It has the disadvantage that it changes depending on the value of .

The present invention attempts to improve SLC circuits to overcome this drawback. The SLC circuit of the present invention is shown in FIG. In this figure, in FIG. 1, an integrator 6 is inserted in place of the narrow band pass filter 3, and a mixer 7 is inserted in place of the amplifier 4. In FIG. 4, 9 is a mixer, by which the instantaneous power |x| ² of the reference signal x is determined. 1
0 is a gain calculator which inputs the reference signal power |x| ² and calculates the input power p to the mixer 7, which is the gain of the correlation loop. In the gain calculator 10, 11 is a mixer, 12 is an integrator,
13 is a subtracter, and 14 is a square element.

The operation of the SLC circuit of the present invention will be explained. If the output signal of the gain calculator 10 is p, then the subtracter 1
The output signal of 3 becomes p-|x| ² p ² . Therefore, the following equation holds regarding the integrator 12.

τ ₀ dp/dt=p−|x| ² p ² ...(11) The solution to equation (11) is as follows.

p=1/4(t)...(12) However, 4(t) is the value obtained by smoothing the reference signal power |x| ² with a first-order lag filter. It is. The fact that equation (12) is a solution to equation (11) means that equation (12) can be transformed into equation (1
1)
It is clear that the identity can be obtained by substituting . Further, the equation regarding the correlation loop of the load control voltage w is as follows.

τ ₀ dw/dt=p(u−wx)x ^* (14) When this equation (14) is solved for w, the following solution is obtained.

Originally, it is preferable that the control load voltage w be set to the following value.

w=ux ^* /|x| ² ...(16) However, ^* and || ² in equation (16) represent the average values of the equations ux ^* and |x| ² , respectively. On the other hand, the numerator and denominator of equation (15) are the signal
Since it represents the output passed through a first-order lag filter for ux ^* , |x| ² , its time constant τ ₀ is
It can be considered that Equation (15) and Equation (16) are the same as long as the time is sufficient to take the average value of the signals ux ^* and |x| ² . Furthermore, in the SLC of the present invention, the time for the control load voltage w to converge is defined by a constant time constant T ₀ regardless of the value of the reference signal power |x| ² , as can be seen from equation (15).

As described above, in the SLC according to the present invention, the gain of the correlation loop is variable, and the gain calculator 10 optimally determines the loop gain according to the value of the reference signal power |x| ² . The convergence time of the voltage w is constant regardless of the value of the reference signal power |x| ² , and the value of the control load voltage w after convergence can also be an expected value without bias.

[Brief explanation of drawings]

Figure 1 is a block diagram of a conventional sidelobe canceller (SLC), and Figure 2 is a block diagram of a conventional sidelobe canceller (SLC).
Diagram 3 explaining how the absolute value of the control load voltage in SLC converges to a desired value over time.
This figure is an explanatory diagram showing the relationship between the reference signal power and the degree of signal suppression in the SLC of FIG. 1, and FIG. 4 is a block diagram of the structure of the SLC improved according to the present invention. 1, 13...subtractor, 2, 5, 7, 9, 11...
...Mixer, 3...Narrowband filter, 4...Amplifier, 6, 12...Integrator, 10...Gain calculation circuit, 14...Square element In addition, the same or corresponding parts in the figure are given the same reference numerals. It is shown as follows.

Claims (1)

[Claims] 1. In order to suppress received signals due to the influence of unnecessary radio waves such as interference waves incident from the side lobes of the main antenna, a correlation loop is provided and a reference signal for the unnecessary radio waves received from the auxiliary antenna is suppressed. A sidelobe canceller has a structure that estimates unnecessary radio wave components included in the signals received by the main antenna by taking correlation and removes them.
A circuit that performs square law detection of a reference signal for unnecessary radio waves received by an auxiliary antenna and outputs a voltage r proportional to its power, and a circuit that uses the voltage r as an input signal to generate a voltage p that is proportional to the reciprocal of its low frequency component. A gain calculation circuit that outputs a solution of the differential equation τ ₀ dp/dt=p−rp ² with τ ₀ as a time variable and τ 0 as a time constant, and A sidelobe canceller characterized by having a structure that can change gain.