U.S. patent number RE30,288 [Application Number 05/861,214] was granted by the patent office on 1980-05-27 for intrusion detection system.
This patent grant is currently assigned to Pittway Corporation. Invention is credited to Kenneth R. Hackett.
United States Patent |
RE30,288 |
Hackett |
May 27, 1980 |
Intrusion detection system
Abstract
An ultrasonic or microwave intrusion detection system uses a
transmitter having one or more transducers or antennas for
maintaining wave fields in the area to be protected and fed by one
or more transducers or antennas which detects echo reflections of
the energy from within the area. The receiver employs two mixers
energized to be mixed with the echo signals and a portion of the
energy from the transmitter, one of the mixers receiving such
transmitter energy with 90.degree. phase shift relative to the
other mixer. The outputs of the two mixers thus are in quadrature
phase relative to each other and define a rotating vector which
contains the information relating to the echo signals. The product
of these quadrature signals obtained by multiplying them together
is processed continuously to obtain target information in the
presence of substantially larger clutter return signals and other
interference. The multiplication of the two quadrature signals to
obtain cancellation of the clutter return signals and enhancement
of the target signals is achieved without further phase shift
requirements by differentiating one of the quadrature signals
before it is multiplied with the other quadrature signal. In the
preferred embodiment the differentiated quadrature signal is
normalized to enhance the dynamic range of the equipment and
eliminate the requirement for balanced channels.
Inventors: |
Hackett; Kenneth R. (Boulder,
CO) |
Assignee: |
Pittway Corporation
(Northbrook, IL)
|
Family
ID: |
27037800 |
Appl.
No.: |
05/861,214 |
Filed: |
December 16, 1977 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
Reissue of: |
455260 |
Mar 27, 1974 |
03942178 |
Mar 2, 1976 |
|
|
Current U.S.
Class: |
342/28; 340/554;
342/159; 342/194; 367/94 |
Current CPC
Class: |
G01S
13/56 (20130101); G08B 13/1618 (20130101) |
Current International
Class: |
G01S
13/00 (20060101); G01S 13/56 (20060101); G08B
13/16 (20060101); G01S 009/02 (); G08B
013/24 () |
Field of
Search: |
;340/554,560
;343/5PD |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Hubler; Malcolm F.
Attorney, Agent or Firm: Pfund; Charles E.
Claims
I claim:
1. The method of detecting moving target echos in the presence of
interference comprising the steps of radiating energy and receiving
echo return energy, hetrodyning radiated and received energy
signals to produce separate quadrature beat frequency signals,
differentiating and amplitude limiting one of said beat frequency
signals, multiplying the differentiated and limited signal and the
other beat frequency signal to obtain the four quadrant product of
the signals multiplied, and integrating said product.
2. In an echo signal intrusion alarm system having enhanced
detection capability for moving objects in the presence of
interference and clutter return echos wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising phase rotation processor means responsive to the full
information content of a predetermined bandwidth of said quadrature
output signals for producing a bipolar signal in accordance with
the product of the magnitudes of said quadrature output signals
multiplied by said beat frequency, means for integrating said
bipolar signal, and threshold means responsive to the magnitude of
the integrated bipolar signal exceeding a predetermined value for
providing an alarm signal.
3. Apparatus according to claim 2 in which said phase rotation
processor means comprises a two-phase induction motor having two
sets of orthogonally positioned stator windings and a locked rotor,
said quadrature output signals being applied to energize the
respective windings of said sets, and means coupled to said rotor
and responsive to the magnitude and sense of the torque exerted on
said rotor by the energized said windings for producing said
bipolar signal.
4. In an echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
means for multiplying each of said quadrature output signals by a
factor obtained from the time derivative of the other quadrature
output signal to obtain two four-quadrant products of the signal
and factor multiplied;
means for subtracting said products to obtain a difference
signal;
integrating means responsive to said difference signal for
cancelling opposite polarity signal components corresponding to
said interference and clutter return and accumulating an integrated
signal magnitude from signal components of the same polarity
corresponding to a given direction of motion for said moving
object; and
means responsive to a predetermined accumulated level of said
integrated signal magnitude for producing an alarm signal.
5. Apparatus according to claim 4 in which said factor is
proportional to the time derivative of said other quadrature output
signal and said four-quadrant products are proportional
respectively to the product of the amplitudes of both of said
quadrature output signals.
6. Apparatus according to claim 4 and including zero-slicer means
operative on each said time derivative to obtain said factor for
making said four-quadrant products proportional to the amplitude of
only one of said quadrature output signals.
7. Apparatus according to claim 4 and including zero-slicer means
operative on one of the inputs to said means for multiplying for
making said four-quadrant products proportional to the amplitude of
only one of said quadrature output signals.
8. In an echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
means for multiplying one of said quadrature output signals by a
factor obtained from the time derivative of the other quadrature
output signal to obtain the four-quadrant product of the signal and
factor multiplied;
zero-slicer means operative on said time derivative to obtain said
factor for making said four-quadrant product proportional to the
amplitude of only said one of said quadrature output signals;
integrating means responsive to said product for cancelling
opposite polarity signal components corresponding to said
interference and clutter return and accumulating an integrated
signal magnitude from signal components of the same polarity
corresponding to a given direction of motion for said moving
object; and
means responsive to a predetermined accumulated level of said
integrated signal magnitude for producing an alarm signal.
9. In an echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
a four quadrant multiplier having a bipolar product output;
a time differentiator coupled to differentiate one of said
quadrature output signals;
means for coupling the other of said quadrature output signals and
the output of said differentiator as input factors to said
multiplier;
zero-slicer means coupled to the output of said time differentiator
operative to remove amplitude information from the one of said
input factors derived from said time differentiator;
an integrator for said bipolar output product of said multiplier;
and
threshold means responsive to a predetermined value of the output
of said integrator for indicating a signal condition.
10. In a echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
a balanced modulator having signal and switching input
terminals;
means coupling one of said quadrature output signals to said signal
input terminals;
differentiating zero-slicing operational amplifier means operating
on the other of said quadrature output signals and having the
output thereof coupled to said switching input terminals;
integrating and signal limiting means operative on the output of
said balanced modulator;
a balanced bias circuit coupled to the output of said balanced
modulator; and
a threshold circuit biased from a reference point in said bias
circuit and responsive to signal excursions of either polarity
relative to said reference point for producing an output actuation
signal for said excursions which exceed a predetermined percentage
change relative to said bias point. .Iadd.
11. In an echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
means for multiplying one of said quadrature output signals by a
factor obtained from the time derivative of the other quadrature
output signal to obtain the four-quadrant product of the signal and
factor multiplied;
limiting means operative on said time derivative to obtain said
factor for making said four-quadrant product proportional to the
amplitude of only said one of said quadrature output signals;
integrating means responsive to said product for cancelling
opposite polarity signal components corresponding to said
interference and clutter return and accumulating an integrated
signal magnitude from signal components of the same polarity
corresponding to a given direction of motion for said moving
object; and
means responsive to a predetermined accumulated level of said
integrated signal magnitude for producing an alarm signal.
.Iaddend. .Iadd.
12. In an echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
a four quadrant multiplier having a bipolar product output;
a time differentiator coupled to differentiate one of said
quadrature output signals;
means for coupling the other of said quadrature output signals and
the output of said differentiator as input factors to said
multiplier;
limiting means coupled to the output of said time differentiator
operative to remove amplitude information from the one of said
input factors derived from said time differentiator;
an integrator for said bipolar output product of said multiplier;
and
threshold means responsive to a predetermined value of the output
of said integrator for indicating a signal condition. .Iaddend.
.Iadd.
13. In an echo signal intrusion alarm system having enhanced
detection capability for return echos from a moving object in the
presence of interference and clutter return wherein a signal is
transmitted into a protected zone and reflected signals are
received from said zone, said system including first and second
quadrature mixing means for mixing signals corresponding to the
transmitted and received signals to produce direction sensitive
quadrature output signals in accordance with the beat frequency
between said transmitted and received signals, the improvement
comprising:
a balanced modulator having signal and switching input
terminals;
means coupling one of said quadrature output signals to said signal
input terminals;
differentiating, limiting, operational amplifier means operating on
the other of said quadrature output signals and having the output
thereof coupled to said switching input terminals;
integrating and signal limiting means operative on the output of
said balanced modulator;
a balanced bias circuit coupled to the output of said balanced
modulator; and
a threshold circuit biased from a reference point in said bias
circuit and responsive to signal excursions of either polarity
relative to said reference point for producing an output actuation
signal for said excursions which exceed a predetermined percentage
change relative to said bias point. .Iaddend.
Description
BACKGROUND OF THE INVENTION
The detection of echo signals in CW systems where the echo return
is deeply immersed in noise by the use of quadrature detection
techniques is disclosed by Kalmus U.S. Pat. Nos. 3,432,855 and
3,733,581. In the basic embodiments therein shown, the outputs of
quadrature mixer-detectors are correlated after one of the mixer
outputs is subjected to a further 90.degree. phase shift. Since the
signals encountered in an ultrasonic intrusion detector alarm
system cover a range of about 3 octaves in the doppler detected
signal, the additional phase shifter required by Kalmus must
provide this range with relatively constant amplitude response.
Phase shifters of this type at the relatively low frequencies
encountered in an intrusion detection alarm system are expensive
and difficult to provide. Furthermore, these prior art circuits
require good balance in the two channels to maintain the signals
substantially equal, and when they are multiplied or correlated or
otherwise compared to cancel the noise components the desired
target components are proportional to the square of the signal
amplitudes since the two equal amplitude signals are multiplied
together. In systems which do not employ signal correlation to
produce an output proportional to the product of the two inputs
some form of signal summing is employed and the requirements for
channel balance to maintain the quadrature channel signals equal in
amplitude presents a more severe requirement. Systems which operate
by sampling the peak amplitude of one of the channels based on
sampling pulses derived from the quadrature timing points available
in the other channel discard a substantial portion of the useful
information and substitute the sample values with the result that
such systems can be readily jammed or desensitized with high
frequency noise components which have symmetrical frequency
spectrum when translated through the mixers. Such jamming can
exceed the dynamic range of the system with such symmetrical
signals and mask the statistically deficient sampled target
information to prevent an alarm being given.
SUMMARY OF THE INVENTION
The present invention employs quadrature detection in a CW
intrusion detection alarm system and processes the quadrature
outputs of the two mixers directly without further phase shift and
on a continuous basis, using signal processing which essentially
responds to the rotation of the phasor representing the quadrature
signals. By multiplying each of these signals by the derivative of
the other signal and combining the products a quantity directly
proportional to the torque causing phase rotation is obtained
without the requirement for a wide band constant amplitude phase
shift circuit. By normalizing these derivatives which are used as
products in obtaining this quantity the requirement for a balanced
channel system is greatly eased. Finally, since the products each
contain a continuous signal information signal packet containing
the entire target and clutter signal information products, one
complete channel can be eliminated and the alarm indication
obtained by integrating the response of a single product channel.
Thus circuit requirements and balance considerations between dual
channels are reduced to an absolute minimum while maintaining a
continuous full information content signal processing system
capable of high sensitivity and detection of target signals in the
presence of noise of much greater amplitude than the target echo
signals and which is not limited by the phase response over several
octaves of doppler frequency information nor saturated by the
squared products of amplitude in two channels limiting dynamic
amplitude range.
DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of a complete CW intrusion detector alarm
system.
FIG. 2 is a schematic representation of an electro-mechanical phase
rotation responsive device.
FIG. 3 is a phasor diagram showing the signal components involved
in phase rotation.
FIG. 4 is a block diagram of an electronic analog of the
electro-mechanical signal processor of FIG. 2.
FIG. 5 is a block diagram of a simplified phase rotation signal
processor.
FIG. 6 is a block diagram of a phase rotation signal processor
using normalized factors to obtain the signal products.
FIG. 7 is a simplified version of a normalized signal product
processor.
FIGS. 8A and 8B are wave form diagrams showing signal processing
for targets moving toward and away from the transmitter with
normalized differentiated signals.
FIG. 9 is a set of wave form diagrams showing the operation of the
system of FIG. 7 with both moving targets and interfering signal
phenomena occurring simultaneously.
FIG. 10 is a schematic wiring diagram for a simplified fourquadrant
multiplier with one factor normalized.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring now to FIG. 1, the general arrangement of a quadrature
detection CW radar system is shown as comprising an oscillator 1
which may be either ultrasonic or microwave to transmit energy from
a transducer into the space to be protected. Energy reflected from
the space and objects within the space is received at a receiver
input 2 and applied to identical signal mixers 3 and 4. The mixers
3 and 4 heterodyne the incoming received signals with a sample of
the frequency from oscillator 1 which is applied, as shown in FIG.
1, directly to mixer 4 and with a leading 90.degree. phase shift
through phase shifter 5 to mixer 3. Any well-known coupling device
may be employed to select a small sample of signal from the
oscillator 1 for use as the heterodyning signal. The outputs of the
mixers 3 and 4 are selected by band pass filters 6 and 7 to select
the difference frequency which difference frequency will be the
result of the doppler effect derived from targets reflecting energy
to the receiver input 2 which targets are in motion with respect to
the transmitter and receiver transducers. Thus the band pass of
filters 6 and 7 will normally correspond to the base band doppler
frequency shift expected from targets of interest. The portion of
FIG. 1 thus far described will readily be understood by those
skilled in the art and correspond generally with the disclosure of
the previously referenced patents to Kalmus.
In accordance with the present invention the direct outputs of the
mixers 3 and 4 with higher frequency mixer components removed by
filters 6 and 7 are processed directly by a phase rotation
processor 10 which produces an output signal representing the net
or resultant phase rotation forces produced by the concerted action
of the quadrature detected signals from mixers 3 and 4. The output
of the processor 10 is a bipolar signal 9 which is applied to an
integrator 10, the integrated output 11 of which is applied to a
threshold circuit 12 which when actuated operates an alarm 13. The
integrator 10 operates on the bipolar signal 9 and cooperates with
the preceding signal processing system to cancel the energy
components of the spectrum which are derived from clutter and other
noise components. According to Kalmus, this energy is distributed
on both sides of the transmitter frequency when observed over a
period of time as in integration while the energy containing the
doppler shift from a moving target produces a signal component on
one or the other side of the transmitter frequency and thus
integrates cumulatively to a value which will exceed the threshold
established by circuit 12 and thus actuate the alarm 13.
The outputs from the filters 6 and 7 representing the doppler
frequencies are derived from the mixers 3 and 4 and designated X
and Y as the inputs to phase rotation processor 10. In FIG. 2 these
signals are applied to the space quadrature windings of a fixed
rotor two-phase induction motor. Thus stator windings 14 and 15 in
series are energized with the X signal and windings 16 and 17 in
series are energized with the Y signal. The resultant flux caused
by current flow in these respective sets of coils is indicated by
the vectors b.sub.y and b.sub.x in FIG. 2. Rotor 18 of the motor is
constrained and the force acting on the rotor is operative to
generate a signal in a strain gauge 19 which produces an output
signal on line 21 representative of the torque on the rotor 18.
This torque is the result of the rotating vector field b having an
angular velocity .omega.. The signal on line 21 can be applied as
the input 9 to integrator 10 and will contain the full signal
output information from the mixers 3 and 4 processed for
integration without the requirement of electrical phase shift to
bring the X and Y signals into phase or phase opposition conditions
as required in the prior art.
Electro-mechanical processor such as shown in FIG. 2 have certain
advantages including inertia and other mechanical parameters for
performing useful operational functions. However, they do generally
have disadvantages and represent relatively costly components and
thus an all electronic system may sometimes be preferred. In order
to obtain the electrical analog of the electro-mechanical system
shown in FIG. 2 an analysis of the relation between the electric
signals and the forces involved will be made with reference to FIG.
3.
In FIG. 3 the phasor .rho. is rotating at an angular velocity
.omega. which can be positive or negative as indicated. The
tangential velocity v.sub.T can be resolved in terms of the
positive velocity v.sub.x, in the x direction and the positive
velocity v.sub.y in the y direction as shown. The torque, m, on
rotor 18 is related to the resultant magnetic vector field, .rho.
having components x corresponding to b.sub.x and y corresponding to
b.sub.y, and its angular velocity, .omega. as follows:
where K represents the motor transfer characteristics for the units
of measurement used and m is determined by the dynamics of signals
x and y. Rewriting equation (1) the torque can be expressed as
follows:
where .omega..rho. is the tangential velocity v.sub.T, at the tip
of phasor .rho., in FIG. 3. Thus:
v.sub.T consists of two components:
v.sub.T.sbsb.1 is the tangential velocity component due to v.sub.x
(the first derivative of x). v.sub.T.sbsb.2 is the tangential
velocity component due to v.sub.y, (the first derivative of y).
Because of the geometrical relationship shown in FIG. 3, the
following proportions exist: ##EQU1## solving for v.sub.T.sbsb.1
and v.sub.T.sbsb.2 : ##EQU2## From equation (4): ##EQU3## where x'
and y' (i.e., v.sub.x and v.sub.y) are the first derivatives of x
and y.
Substituting (7) into (3):
the .rho. term disappears. The torque is epxressed in terms of x
and y.
Since the torque which is a measure of the net phase rotation
signal can be readily derived from the x and y signals and their
derivatives, a quantity proportional to torque can be derived
electrically as shown in FIG. 4. As indicated, the x and y inputs
are applied as direct inputs to respective multipliers 22 and 23
and these input quantities after differentiation in differentiators
24 and 25 are applied as the factor input to the multiplier for the
opposite signal as indicated. Thus the output of multiplier 22 is
##EQU4## or xy' and the output of multiplier 23 is ##EQU5## or yx'.
Taking the difference of these two quantities in subtractor 26
gives the desired output quantity xy'-yx'. It should be noted that
the outputs of the multipliers 22 and 23 are continuously present
for both moving target and clutter interference signal returns. As
stated by Kalmus, the clutter energy is distributed on both sides
of the transmitter frequency and thus cancels in the output of the
multipliers after these components have been intergrated over a
sufficient period of time. The continuous frequency waves of a
moving target, however, when processed according to FIG. 4
implementation produces a constant amplitude continuous signal (DC)
the polarity of which indicates the direction of rotation and thus
whether or not the target is approaching or receding from the
transmitter. Taking the quadrature x and y signals as sine and
cosine terms at the doppler angular frequency .omega. we have:
from (8)
For constant target velocity all the factors on the right in (9)
are constant; hence, m is constant. The sign of m is a function of
the sign of .[.107..]. .Iadd..omega...Iaddend.+.omega. represents a
counter-clockwise phase roation. .omega. represents a clockwise
rotation.
Referring to FIG. 5 a simplified version of the electronic phase
rotation processor will be seen to constitute one-half of the
symmetrical system shown in FIG. 4. Thus the x input is applied
directly to a multiplier 28 and the y input is applied to a
differentiator 29, the output of which, y', is applied as the other
input to multiplier 28. The resulting output on line 30 is the
quantity .[.xu'.]. .Iadd.xy' .Iaddend.and corresponds precisely
with the output of multiplier 22 in FIG. 4. It will be noted that
the quadrature mixers 3 and 4 of FIG. 1 provide output doppler
signals which contain the full information content both as to real
targets and clutter or noise information, the only difference
therebetween being the phase relation. Thus each signal bears the
same relation to the other signal and the quantities xy' and
.[.xx'.]. .Iadd.yx' .Iaddend.provide redundancy as to their
information content. Accordingly, selecting one or the other of the
quantities xy' or yx' will provide the full information content but
is statistically less effective in accumulating moving target
signal energy and likewise in averaging or cancelling clutter
return energy. For sufficient period of integration, either xy' or
.[.xx'.]. .Iadd.yx' .Iaddend.will average clutter signals to zero
and accumulate a signal of increasing magnitude for a moving target
return. Thus the circuit of .[.FIG).]. .Iadd.FIG. .Iaddend.5 can be
directly substituted for the phase rotation processor 10 of FIG. 1
and the improved performance of applicant's invention obtained with
simplified and economical apparatus. There is in fact a further
advantage to the processor of FIG. 5 in that it does not require
balancing with respect to a channel purported to be identical
thereto but which may when malfunctioning introduce an unbalanced
condition.
A further improvement and simplification is shown in the modified
phase rotation processor shown in FIG. 6. It will be recalled from
the description of the previous embodiment and equations (9) and
(10) that the quantity being measured both as the useful signal and
clutter signal is proportional to the square of the amplitude of
the various signal components. In echo signal systems generally and
intrusion alarm systems in particular the dynamic range of signals
encountered is extremely large and the presence of large signals
results at times in exceeding the dynamic range of electronic
channels which include amplifiers, differentiators and multipliers
and similar components. The processing of signals which are squared
accentuates this problem and can result in a system which is
capable of handling only signals within a limited amplitude range.
Conversely, if the signal information could be processed with only
the first power of the amplitude quantities requiring operational
conversion, such as differentiation and multiplication, improved
results can be obtained in that wider dynamic signal range can be
accommodated and the system is not susceptible to being disabled by
high amplitude signals. At the same time it is important that the
improved processing be obtained with full utilization of the
formation content of the signals since the difficult discrimination
between the true signal and clutter return must nevertheless be
achieved with signals deeply buried in the clutter signals.
The systems of FIGS. 6 and 7 achieve the operation and advantages
just described by processing signals as previously described for
the systems of FIGS. 4 and 5 respectively but with the amplitude
information removed from the differentiated signal component. Thus
in FIG. 6 multiplier 35 produces the product of input .[.signatl.].
.Iadd.signal .Iaddend.x and a quantity Py' which is the quantity y'
having the polarity of the y' signal but with normalized or unit
amplitude constant throughout the period during which y' remains at
each polarity. multiplier 36 produces a product yPx' by multiplying
the quantities y applied thereto and Px' derived from x' by
removing the amplitude information therefrom. Thus the system of
FIG. 6 differentiates the x input in differentiator 32 and passes
x' to a zero slicer 34 which produces the signal .[.Py'.].
.Iadd.Px' .Iaddend.having the same polarity as .[.y'. The outputs
of amplitude .]. .Iadd.x' but constant amplitude.Iaddend..
Similarly the y input signal is differentiated in a differentiator
31 and the signal y' is applied to a zero slicer 33 to produce the
unit amplitude bipolar signal Py' having the same polarity as y'.
The outputs of the multipliers 35 and 36 are applied to a
subtractor 37 to produce on output line 38 the signal xPy'-yPx'.
The system of FIG. 7 is identical to one-half of the FIG. 6 system
and applies the x input to a multiplier 39 while the y input is
differentiated in differentiator 41 to apply y' to a zero slicer 40
which applies Py' to the multiplier 39. The output line 42 thus has
the quantity xPy' thereon.
Referring now to FIGS. 8A and 8B typical target wave forms are
displayed for the operation of this system with multiplication of
those waves shown for the normalized or constant value derivatives
employed as one factor in obtaining the product. Thus lines (a) and
(b) show the cosine and sine signals obtained from the quadrature
detectors 3 and 4 of FIG. 1. As is apparent from Kalmus the
relative phase of these signals changes between 90.degree. lead and
lag as the target motion changes direction. The waveforms shown on
lines (c) and (d) represent x' and y' differentiated values of x
and y respectively. As can be seen by observing the instantaneous
values thus obtained, the products xy' and yx' when multiplied
through a four-quadrant multiplier produce continuous signal
components of opposite polarity. When both of these components are
used by application to subtractor 26 or 37, all components are
additive to accumulate the maximum signal response from the moving
target returned components. Thus output 38 for the system of FIG. 6
is indicated in line (i). This signal on line (i) is the result of
direct addition of the signals indicated at line (g) and (h) which
represent xPy' and -yPx'. In the single channel systems the signal
output 42 is as indicated on line (g). In all of the responses (g),
(h) and (i), the product obtained by signal multiplication with the
normalized derivative shown in lines (e) and (f) are obtained. Thus
no products proportional to the amplitude squared are depicted in
FIG. 8.
It will be noted that the normalized derivatives Px' and Py' shown
on lines (e) and (f) of FIG. 8 precisely correspond to the half
cycles of the derivatives x' and y' from which they were derived.
Thus the polarity information of the derivatives is preserved while
its amplitude information is removed in obtaining the quantities
Px' and Py'. Since all of the amplitude information is available in
the x and y input signals no essential information is lost in
obtaining the products xPy' and yPx'. On the contrary, the
advantages of processing the amplitude of the signal as the first
power thereof while maintaining the polarity as obtained with four
quadrant multiplication is achieved by the simple expedient of
differentiating the x and y waves and limiting their amplitude
value. Although the limiting function is shown as it would appear
with infinite gain amplification, "soft" limiting may be employed
to provide some dynamic response to the product xy' for low level
signals with the zero slicers providing a hard limiting
polarity-only signal for larger magnitude echo returns.
Referring now to FIG. 9 a representation of the operation of the
system of FIG. 7 in the presence of both moving targets and
interfering signal phenomena will be described. In FIGS. 9(a) and
(b) the dotted curve represents the substantially sinusoidal signal
component produced by a good target reflector moving at
approximately constant velocity. The clutter signal components
which are uniformly distributed in the spectrum on opposite sides
of the transmitter frequency will result in components which are
distributed relative to the signal return from the moving target.
The combination of these distributed signals with the sinusoidal
return from the moving target will produce typically an x signal as
shown in line (a) as the solid curve. A similar signal displaced
90.degree. in phase is obtained as the y signal from the other
mixer and after differentiation this signal would appear relative
to the x signal as shown in line (b) of FIG. 9. By zero slicing or
limiting the signal on line (b) the normalized or constant
amplitude of line (c) is obtained having the same polarity
information as contained in the y' signal of line (b). The
multiplication of lines (a) and (c) produces the signal indicated
on line (d). As can be seen the four-quadrant multiplication has
rectified target signal components to be of a single (positive)
polarity with the clutter component signals distributed above and
below the dotted rectified sinusoidal waveform. During integration,
of course, the plus and minus clutter error relative to the dotted
line curve will cancel and the true target signal is recovered
despite the presence of the clutter return signals.
FIG. 10 is a schematic wiring diagram for a solid state or
integrated circuit implementation of a simplified system generally
corresponding to that described and shown by FIG. 7. In this
implementation the four-quadrant multiplier has one linear channel
for faithfully introducing the amplitude function of one signal
channel while the other factor is processed as polarity information
only as described in FIG. 7 and thus the channel is simplified and
combined as part of the slicing or limiting function. Referring to
FIG. 10 the circuit will be seen to operate from a DC supply such
as 12 volts applied to terminal 51 and ground 52. Across this
voltage supply a voltage divider is connected consisting of the
series connection of resistors 53 50, 54 and the intermediate
balancing resistor combination 55. This voltage divider provides
appropriate operating potentials for the circuit as hereinafter
described.
The polarity-only factor for the multiplication is obtained by
applying the y input signal corresponding to that shown in FIG. 7
to a terminal 56 which is capacitor couplied to the negative input
of an operational amplifier 57 which has its positive input
referenced on line 58 to the junction between voltage divider
resistors 53 and 50. A pair of back-to-back diodes 59 is connected
from the output to the negative input of op amp 57 and the
combination with the coupling capacitor thus operates to
differentiate the y input signal on terminal 56 and switch the
output signal between a maximum positive and negative value as the
polarity of the differentiated wave varies above and below the
reference potential supplied as an input on line 58. Thus the
output on line 61 corresponds to the Py' signal of the
differentiated but polarity-only y input signal.
The linear factor for the multiplication is introduced on a
terminal 62 where it is applied to a balanced series cascade of
transistors comprising switching pairs Q.sub.5, Q.sub.6, and
Q.sub.7, Q.sub.8, which pairs have their emitters joined and
respectively connected to the collectors of transistors Q.sub.9,
Q.sub.10, which in turn have their emitters bridged by a resistor
63 an connected respectively to the collectors of transistors
Q.sub.11, Q.sub.12, the emitters of which are returned through 500
ohms resistors 64 and 65 to ground. The collectors of Q.sub.5 and
Q.sub.7 are joined and connected through a 4.7 Kohm resistor 66 and
a balancing potentiometer 67 to the positive supply. The collectors
of Q.sub.6 and Q.sub.8 are joined and returned through a 4.7 Kohm
resistor 68 and the balancing potentiometer 67 to the positive
supply. The joined collector pairs of Q.sub.5, Q.sub.7, and
Q.sub.6, Q.sub.8, are bridged by a capacitor 69 and a pair of
back-to-back diodes 71. The output of the multiplier is derived
from these same joined collector pairs on lines 72, 73.
The bases of Q.sub.5 and Q.sub.8 are joined and connected to the
reference potential of line 58. The bases of Q.sub.6, Q.sub.7, are
joined and connected to the polarity switching factor for the
multiplication on lines 61. Thus the transistor pairs Q.sub.5,
Q.sub.6, and Q.sub.7, Q.sub.8, switch the polarity of their output
on the joined collectors connected to output lines 72, 73 in
accordance with the polarity of the signal on line 61 referenced to
the potential on line 58 and the polarity of the x input on line 62
relative to the biase in Q.sub.9, Q.sub.10, to provide four
quadrant product polarity from these two factors.
The transistors Q.sub.9, Q.sub.10 provide a current balance for the
multiplier system and an appropriate bias level for the
introduction of the linear x input signal from line 62. For this
purpose the base of Q.sub.9 is connected through a 10 Kohm resistor
75 to a mid-point on the balancing resistor grouping 55 while the
base of Q.sub.10 is connected to the adjustable tap on a
potentiometer 74. The base of Q.sub.9 is connected through coupling
capacitor 76 to the input terminal 62.
The operating current point for the balanced cascade is provided by
transistors Q.sub.11 and Q.sub.12 which have their bases connected
together and returned to the positive supply through a resistor 77.
A current mirror comprising diode 78 and resistor 79 is connected
from the joined bases of Q.sub.11, Q.sub.12 to ground.
In the operation of the circuit of FIG. 10 as thus far described,
the tap on potentiometer 74 is adjusted for current balance through
the series cascade transistors and the tap on potentiometer 67 is
adjusted to provide a voltage balance on output lines 72, 73, such
that the voltage between lines 72 and 73 is zero whenever either of
the input factors x or y is zero. The potential between output
lines 72, 73, will thus be an amplitude analog of the linear x
input signal at terminal 72 with polarity controlled by
four-quadrant multiplication with the signal on line 61. Thus the
potential across line 72 and 73 will be positive (arbitrary
polarity) for x and y input signals in the first and third
quadrants and of negative polarity for input signals in the second
and fourth quadrants. These signals are integrated by capacitor 69
and limited by diode 71 to prevent overloading of subsequent
circuits.
The integrated xPy' signal on lines 72 and 73 will remain at
substantially zero level for absence of moving reflectors in the
energy field between the transmitting and receiving transducers and
will show a positive or negative polarity signal for an approaching
or receding target. To utilize a signal which integrates with
either positive or negative polarity preponderance, a balanced
threshold circuit comprising transistors Q.sub.13, Q.sub.14 is
used. The signal across lines 72, 73 is applied between the bases
of Q.sub.13, Q.sub.14 and a balanced output signal is developed in
the symmetrical resistor network 81 connected to the collectors of
Q.sub.13 and Q.sub.14. Test points 82 and 83 may be provided to
connect a meter for bipolar indication of approaching and receding
targets.
A threshold biased transistor Q.sub.15 is provided to operate an
alarm system connected to output lead 85 upon the occurrence of a
signal of either polarity of sufficient magnitude at the collectors
of transistors Q.sub.13, Q.sub.14. For this purpose, these
collectors are connected through diodes 86 and 87 to a point
returned to the positive potential through a resistor 88 and
connected through a diode 89 to the base of transistor Q.sub.15.
The base of Q.sub.15 is biased by a resistor 91 which is returned
to ground and bypassed by capacitor 92 which provides additional
integration. The operating point for transistor Q.sub.15 is
obtained by referencing the emitter through diode 93 to a mid-tap
on the resistor network 81 and returning the emitter through a
resistor 94 to the positive supply. Thus in the operation of the
circuit of FIG. 10 an actuating output signal will appear at output
line 85 whenever the bias of transistor Q.sub.15 is overcome by the
magnitude of a signal on the base thereof derived from either
polarity of the output from transistors Q.sub.13, Q.sub.14,
corresponding to the bipolar moving target signal inputs thereto on
lines 72, 73. As previously described, this bipolar signal on lines
72, 73 is the four-quadrant multiplication of the x and y input
signals with one channel normalized or containing polarity
information only but otherwise containing the full information
content of the signal and noise components. Furthermore, since the
threshold bias for Q.sub.15 is referenced to a point on the
balanced network 81 and direct coupled through Q.sub.13, Q.sub.14,
to the reference level of the bipolar signal on lines 72, 73, the
alarm actuation will occur for a percentage change in signal level
and relatively independent of changes in supply voltage.
It will be appreciated that the present invention by operating
continuously on the return signals provides the maximum available
statistical base over which to accumulate the like polarity signal
components contributed by returns from a true moving target.
Similarly, all of the noise and clutter components which are
distributed throughout the signal spectrum within the normal
doppler signal range relative to the transmitter frequency are
translated respectively into opposite polarity components and thus
cancel. Thus maximum discrimination between these two types of
signal returns is possible together with the other explicit and
inherent advantages of the invention as disclosed.
Many modifications of the invention will be apparent to those
skilled in the art, particularly in relation to achieving the
operational functions desired. Thus various forms of
differentiators and four-quadrant multipliers may be employed to
achieve the functions required for the operations specified herein.
The invention, accordingly, is not to be limited to the specific
embodiments disclosed but only by the scope of the appended
claims.
* * * * *