U.S. patent number RE31,509 [Application Number 06/183,396] was granted by the patent office on 1984-01-24 for echo location systems.
Invention is credited to Norman S. Neidell.
United States Patent |
RE31,509 |
Neidell |
January 24, 1984 |
Echo location systems
Abstract
This invention generally relates to echo location or ranging
systems in which the individual properties of targets or reflectors
(such as range, bearing, elevation angle, relative velocity,
impedance contrast, etc.) in a field of targets within some
propagation medium are identified by the emission of signals into
the propagation medium and processing of the detected reflections
from the target field.
Inventors: |
Neidell; Norman S. (Houston,
TX) |
Family
ID: |
26879077 |
Appl.
No.: |
06/183,396 |
Filed: |
September 2, 1980 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
483202 |
Jun 26, 1974 |
|
|
|
Reissue of: |
691674 |
Jun 1, 1976 |
04114153 |
Sep 12, 1978 |
|
|
Current U.S.
Class: |
342/108; 342/378;
367/100 |
Current CPC
Class: |
G01S
3/80 (20130101); G01S 7/292 (20130101); G01S
7/41 (20130101); G01S 7/527 (20130101); G01S
13/02 (20130101); G01S 13/106 (20130101); G01S
15/582 (20130101); G01S 13/30 (20130101); G01S
13/42 (20130101); G01S 13/58 (20130101); G01S
13/878 (20130101); G01S 15/108 (20130101); G01S
13/28 (20130101) |
Current International
Class: |
G01S
15/58 (20060101); G01S 13/02 (20060101); G01S
13/00 (20060101); G01S 13/58 (20060101); G01S
15/00 (20060101); G01S 13/87 (20060101); G01S
13/28 (20060101); G01S 13/42 (20060101); G01S
13/30 (20060101); G01S 15/10 (20060101); G01S
13/10 (20060101); G01S 3/80 (20060101); G01S
3/00 (20060101); G01S 7/02 (20060101); G01S
7/41 (20060101); G01S 7/527 (20060101); G01S
7/292 (20060101); G01S 7/523 (20060101); G01S
013/58 () |
Field of
Search: |
;343/5NQ,9R,1PE,1CL
;367/90,100 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Tubbesing; T. H.
Attorney, Agent or Firm: Pravel, Gambrell, Hewitt, Kirk
& Kimball
Parent Case Text
This is a continuation of application Ser. No. 483,202, filed June
26, 1974, now abandoned.
Claims
What is claimed is:
1. A method of using a transmitter member and an arrangement of
receivers having at least one receiver member, all members being at
locations referred to a coordinate system of fixed origin and in a
medium having a known signal propagation velocity function, for
ascertaining, for a field containing one or more reflecting
targets, information about the targets' .[.relfectivity.].
.Iadd.reflectivity .Iaddend.strengths, positions and velocities as
referred to said coorinate system, comprising the steps of:
producing a signal pattern having at least two individual member
signals having preassigned time intervals, said time intervals and
the durations of the member signals being short relative to a
period over which velocity vectors describing motions of the
transmitter, receivers and targets are approximately constants;
forming said individual member signals as a weighted sum of a
design base signal pair, a pair of base signals approximately
sharing a common smooth and essentially .[.unimodel.].
.Iadd.unimodal .Iaddend.amplitude spectrum occupying a contiguous
band of frequencies, said base signals being in mutual
quadrature;
propagating said signal pattern with defined polarization
character;
developing return signal patterns by reflection from the target
field to each receiver;
processing reflected return signal patterns from each receiver by
cross-correlating replicas of these return signal patterns with a
detection base signal pair, thereby producing a pair of correlation
component functions, said detection base signals having properties
analogous to said design base signals, but with an amplitude
spectrum overlapping that of the design base signal pair by an
interval of frequency greater than any Doppler shift attributable
to propagation toward and refelection from moving targets, the
difference in phase angle between counterpart design and detection
base signals at any common frequency component being described, in
good approximation, mathematically by a constant and a term linear
with frequency;
forming for each received return signal pattern a correlation
amplitude function, formed from term-by-term sums of the absolute
values of the correlation component functions raised to a like
power not less than one, said sums then raised to a like power
greater than zero but not greater than one;
identifying from the significant maxima of the correlation
amplitude function return signal patterns corresponding to
individual target reflections;
estimating for each detected target at each receiver a relative
velocity component using the known signal propagation velocity
function and the extension or compression of the initially
preassigned time intervals between member signals in the return
signal pattern for the target at the particular receiver, as
determined from significant maxima of the correlation amplitude
function;
estimating for each detected target at each receiver initial
relative range information using an effective signal propagation
velocity developed from the target's relative velocity estimate and
known medium velocity function along with the elapsed signal travel
time for onset of the return signal pattern identifying a
target;
calculating final relative range information by introducing timing
corrections to the elapsed signal travel time, such corrections
representing linear elements of phase distortions arising from
propagation, such corrections being dependent upon empirical tests
using targets of known parameters, these corrections being
catalogued according to initial range estimates, the conversion to
final range information using the effective signal propagation
velocity with the corrected elapsed time;
referring all estimates of target ranges and velocities to the
fixed coordinate system; and
characterizing each member of the target field according to
reflectivity, magnitudes being estimated from magnitudes of
particular target correlation amplitude functions, accounting for
losses associated with propagation and other observable and
controllable signal amplitude modifications which are obtainable
from empirical tests using targets of known parameters, and
reflectivity polarities being determined from return signal pattern
polarities as compared with the initial outgoing signal
polarities.
2. The method according to claim 1 wherein more than one
transmitter is used concurrently, and
producing outgoing signal patterns from each transmitter separable
by a combination of any distinction according to polarization
character and the frequency filtering produced by said
cross-correlating.
3. The method according to claim 2 wherein for at least one
received return signal train, positional information in the form of
one angle for each target is encoded, and differentiating
reflecting targets additionally by using amplitude spectral
modifications of the member signals of the received return signal
train;
employing for transmitters and receivers related to said encoding,
impulse-like design and detection base signals, individual base
signals having the further property that the phase angles for all
significant frequencies are described mathematically, in good
approximation, by a constant and a term linear in frequency;
encoding the angular positional information for each target as a
phase modification, which is alike for each member signal of the
signal train to be received, by interposing a phase adjustment
element into the signal propagation path such that targets
distinctive in their angular positional information will have
distinctive phase modifications, any phase modification being
described, in good approximation mathematically, by a constant and
a term linear in frequency for all significant frequencies, the
angular positional information being functionally related to the
constant phase modifications by single valued functions determined
by empirical tests using targets of known parameters, the linear
phase modification with frequency being determined and catalogued
according to the constant phase modification as a timing
correction;
forming an initial constant phase modification estimate for
individual member signals of the received return signal train and
the correlation components, wherein member signals are identified
at time origins corresponding to the times of the significant
maxima of the correlation amplitude function;
compensating the initial constant phase modification for the
initial constant value of the design base signal, for the phase
difference between the design and detection base signals, and for
any constant phase distortion approximating effects of propagation,
said distortion due to propagation being determined by empirical
tests using targets of known parameters and catalogued according to
initial target range estimates;
calculating final relative range information, including an
additional timing correction to the elapsed signal travel time,
such corrections representing linear elements of phase distortion
arising from said phase adjustment element, said distortions being
determined from the catalogued values corresponding to the
compensated initial constant phase modification estimate;
forming the angular positional information from the compensated
initial constant phase modification estimates using said
single-valued functions;
incorporating angular positional information with other estimated
target positional and velocity information in the fixed coordinate
system; and
differentiating reflecting targets by using amplitude spectra of
returning member signals for each target, adjusting said spectra
for known modifications resulting from signal propagation and
accompanying phase adjustment element effects, determined from
empirical tests using targets of known parameters.
4. The method according to claim 3 wherein the initial constant
phase modification estimate is formed from calculations of phase
spectra for individual member signals of the received return signal
train and the correlation components.
5. The method according to claim 3 wherein initial constant phase
modification estimates for individual member signals are derived
from the arctangent functions of ratios formed from values of the
correlation components at times corresponding to time origins for
member signals appertaining to reflection from a particular target,
any pair of said values being appropriate to forming a ratio when
their corresponding member signals are in quadrature; and
developing additional ratios from values of said reflected return
signal pattern at times corresponding to member signal origins
appertaining to a particular target, whose numerator is derived
from a member signal in quadrature with the denominator.
6. The method of claim 3 wherein for at least one received return
signal train, positional information in the form of at least one
angle for each target is encoded, such that for each target
different phase encodings are imparted for the same angular
information in different member signals,
interposing for each transmitter and receiver related to said
encoding a second phase adjustment element into the signal
propagation path, changing at least one phase adjustment element
for at least one member signal of the train to be received the
total number of said changes being no less than one less than the
number of positional information angles desired for each
target;
designating phase encodings having independence such that the
compensated constant phase modifications estimated from individual
member signals and the known functional relationships between the
phase adjustment element encodings and the desired angles develop a
set of equations having as unknowns the target positional
information, the number of equations being at least equal to the
number of unknowns thereby allowing their solutions, and
applying timing corrections to individual member signals to
compensate also for linear phase effects of phase adjustment
elements, such corrections now modifying also final relative
velocity component estimates as well as relative range
information.
7. The method according to claim 6 wherein for at least one
received return signal train, constant phase distortions
approximating the properties of each reflecting target are
estimated,
developing at least one set of equations for the desired angular
positional information for each target having at least one more
equation than unknowns and including as an additional unknown
variable the constant phase distortion introduced by the target,
phase encodings having been selected to allow at least as many
independent equations as the total of unknowns, and
incorporating phase distortion properties of each target as an
additional target identifier.
8. The method according to claim 6 wherein for at least one
received return signal train a constant phase distortion
approximating uncompensated properties of the propagation medium is
estimated, and
developing at least one set of equations for the desired angular
positional information for each target having at least one more
equation than unknowns and including as an additional unknown
variable the constant phase distortion approximating properties of
the propagation medium, phase encodings having been selected to
allow at least as many independent equations as the total of
unknowns. .Iadd. 9. The method according to claim 1 wherein for at
least one received return signal train, positional information in
the form of one angle for each target is encoded, and
differentiating reflecting targets additionally by using amplitude
spectral modifications of the member signals of the received return
signal train;
employing for transmitters and receivers related to said encoding,
impulse-like design and detection base signals, individual base
signals having the further property that the phase angles for all
significant freqencies are described mathematically, in good
approximation, by a constant and a term linear in frequency;
encoding the angular positional information for each target as a
phase modification, which is alike for each member signal of the
signal train to be received, by interposing a phase adjustment
element into the signal propagation path such that targets
distinctive in their angular positional information will have
distinctive phase modifications, any phase modification being
described, in good approximation mathematically, by a constant and
a term linear in frequency for all significant frequencies, the
angular positional information being functionally related to the
constant phase modifications by single value functions determined
by empirical tests using targets of known parameters, the linear
phase modification with frequency being determined and catalogued
according to the constant phase modification as a timing
correction;
forming an initial constant phase modification estimate for
individual member signals of the received return signal train and
the correlation components, wherein member signals are identified
at time origins corresponding to the times of the significant
maxima of the correlation amplitude function;
compensating the initial constant phase modification for the
initial constant value of the design base signal, for the phase
difference between the design and detection base signals, and for
any constant phase distortion approximating effects of propagation,
said distortion due to propagation being determined by empirical
tests using targets of known parameters and catalogued according to
initial target range estimates;
calculating final relative range information, including an
additional timing correction to the elapsed signal travel time,
such corrections representing linear elements of phase distortion
arising from said phase adjustment element, said distortions being
determined from the catalogues values corresponding to the
compensated initial constant phase modification estimate;
forming the angular positional information from the compensated
initial constant phase modification estimates using said
single-valued functions;
incorporating angular positional information with other estimated
target positional and velocity information in the fixed coordinate
system; and
differentiating reflecting targets by using amplitude spectra of
returning member signal for each target, adjusting said spectra for
known modifications resulting from signal propagation and
accompanying phase adjustment element effects determined from
empirical tests using targets
of known parameters. .Iaddend..Iadd. 10. The method according to
claim 9 wherein said step of forming an initial constant phase
modification estimate comprises forming such initial constant phase
modification from calculations of phase spectra for individual
member signals of the received return signal train and the
correlation components. .Iaddend..Iadd. 11. The method according to
claim 9 wherein said step of forming an initial constant phase
modification comprises forming estimates for individual member
signals derived from the arctangent functions of ratios formed from
values of the correlation components at times corresponding to time
origins for member signals appertaining to reflection from a
particular target, any pair of said values being appropriate to
forming a ratio when their corresponding member signals are in
quadrature; and
developing additional ratios from values of said reflected return
signal pattern, at times corresonding to member signal origins
appertaining to a particular target, whose numerator is derived
from a member signal in quadrature with the denominator.
.Iaddend..Iadd. 12. The method of claim 9 wherein for at least one
received return signal train, positional information in the form of
at least one angle for each target is encoded, such that for each
target different phase encodings are imparted for the same angular
information is different member signals,
interposing for each transmitter and receiver related to said
encoding a second phase adjustment element into the signal
propagation path, changing at least one phase adjustment element
for at least one member signal of the train to be received, the
total number of said changes being no less than one less than the
number of positional information angles desired for each
target;
designating phase encodings having independence such that the
compensated constant phase modifications estimated from individual
member signals and the known functional relationships between the
phase adjustment element encodings and the desired angles develop a
set of equations having as unknowns the target positional
information, the number of equations being at least equal to the
number of unknowns thereby allowing thier solution, and
applying timing corrections to individual member signals to
compensate also for linear phase effects of phase adjustment
elements, such corrections now modifying alos final relative
velocity component estimates as well as relative range information.
.Iaddend..Iadd. 13. The method according to claim 12 wherein for at
least one received return signal train, constant phase distortions
approximating the properties of each reflecting target are
estimated.
developing at least one set of equations for the desired angular
positional information for each target having at least one more
equation than unknowns and including as an additional unknown
variable the constant phase distortion introduced by the target,
phase encodings having been selected to allow at least as many
independent equations as the total of unknowns, and
incorporating phase distortion properties of each target as an
additional
target identifier. .Iaddend..Iadd. 14. The method according to
claim 12 wherein for at least one received return signal train a
constant phase distortion approximating uncompensated properties of
the propagation medium is estimated, and
developing at least one set of equations for the desired angular
positional information for each target having at least one moe
equation than unknowns and including as an additional unknown
variable the constant phase distortion approximating properties of
the propagation medium, phase encodings having been selected to
allow at least as many independent equations as the total of
unknowns. .Iaddend..Iadd. 15. A method of determining the velocity
of a signal reflecting object positioned in a medium of known
signal-propagation characteristics, relative to a signal
transmitter and a signal receiver, comprising the steps of:
(a) transmitting at two distinct instants of time at least two
signals having a pedetermined time interval therebetween, each
signal having a continuous amplitude spectrum between low and high
frequency limits, for reflection of the signals by the object;
(b) receiving the reflected signals;
(c) measuring only the time interval between the received signals;
and
(d) determining the relative velocity of the object based only upon
the measured time interval between the received signals and the
predetermined time interval between the transmitted signals.
.Iaddend. .Iadd. 16. The method of claim 15, wherein said step of
transmitting the signals comprises transmitting signals having an
essentially short time duration. .Iaddend..Iadd. 17. The method of
claim 16 wherein said step of transmitting the signals includes
transmitting the signals in phase quadrature relative to each
other. .Iaddend..Iadd. 18. The method of claim 16 wherein said
steps of transmitting the signals comprises transmitting at least
one signal which is transformable to a symmetric signal relative to
a time reference by adding a constant phase angle to the phase at
each frequency. .Iaddend..Iadd. 19. The method of claim 16 wherein
said step of transmitting the signal comprises transmitting pulse
signals, each of which is transformable to a symmetric signal
relative to a time reference by adding a constant phase angle to
the phase at each frequency. .Iaddend..Iadd. 20. The method of
claim 15, wherein the two distinct instants of time are
predetermined, and further including the steps of:
(a) measuring the arrival time of at least one reflected
signal,
(b) determining the signal transit time from the measured arrival
time and the predetermined time instant of its transmission;
(c) correcting the transit time as a function of the determined
velocity; and
(d) determining the distance of said object relative to the
transmitter and the receiver as a function of the corrected transit
time and the
propagation characteristics of the medium. .Iaddend..Iadd. 21. A
method of determining the direction of a signal reflecting object
relative to a signal transmitter and a signal receiver, comprising
the steps of:
(a) transmitting a signal from the transmitter, such signal having
a continuous amplitude spectrum between low and high frequency
limits, for reflection from the object;
(b) receiving the reflected signal with the receiver;
(c) changing the phase of the transmitted signal in dependence upon
the direction of its travel while propagating between the
transmitter and the receiver, such phase change characterized other
than by a simple time delay;
(d) measuring the phase of the received reflected signal; and
(e) determining the direction of the object based on the measured
phase. .Iaddend..Iadd. 22. The method of claim 21 wherein said step
of transmitting a signal comprises transmitting a signal having a
phase change comprising a constant for each direction independent
of frequency. .Iaddend. .Iadd. 23. The method of claim 22 wherein
said step of changing the phase of the signal comprises:
propagating the signal through a medium having a dimension which
varies in dependence upon the direction of travel of the signal.
.Iaddend..Iadd. 24. The method of claim 21 wherein said step of
changing the phase of the signal comprises:
propagating the signal through a medium having a dimension which
varies in dependence upon the direction of travel of the signal.
.Iaddend..Iadd. 25. A method of ascertaining the range of one or
more reflecting targets in a target field referred to a coordinate
system comprising the steps of:
(a) producing a signal having a linear combination of at least a
pair of base signals such signals having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of said pair of base signals is zero;
(3) each base signal of such pair being in phase quadrature to the
other member of the pair;
(b) transmitting the signal pattern into the target field;
(c) receiving reflections of the transmitted signal pattern from
the target field; and
(d) processing the received reflections to ascertain the range of
at least one of such targets in the target field. .Iaddend..Iadd.
26. A method of ascertaining the range and velocity of one or more
reflecting targets in a target field referred to a coordinate
system comprising the steps of:
(a) producing a signal pattern having at least two signals, each
being a linear combination of at least a pair of base signals said
base signals having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smootly unimodal occupying a contiguous band of frequencies;
(2) having a finite time interval before and after which each
signal of said pair of base signals is zero;
(3) each base signal of such pair being in phase quadrature to the
other member of the pair;
(b) transmitting the signal pattern into the target field during at
least partially different time intervals for each signal of the
signal pattern;
(c) receiving reflections of the transmitted signal pattern from
the target field; and
(d) processing the received reflections to ascertain the range and
velocity of at least one of such targets in the target field.
.Iaddend. .Iadd. 27. A method of ascertaining the angular
information of one or more reflecting targets in a target field
referred to a coordinate system comprising the steps of:
(a) producing a signal being a linear combination of a pair of base
signals such signals having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of frequencies
(2) having a finite time interval before and after which each
signal of said pair of base signals is zero
(3) each base signal of such a pair being in phase quadrature to
the other member of the pair; and
(4) each of such pair of base signals being transformable to a
symmetric signal relative to a time reference by adding a constant
phase angle to the phase at each frequency;
(b) transmitting the signal pattern;
(c) introducing a phase distortion, having a component independent
of frequency, which varies in a known and single valued manner
according to desired angular information into the signal
propagation path;
(d) receiving reflections of the transmitted signal pattern from
the target field; and
(e) processing the received reflections to ascertain the angular
information for at least one of such targets in the target field.
.Iaddend..Iadd. 28. The method of claim 27, further including the
step of determining the range of at least one of the targets,
comprising the step of:
processing the received reflections to ascertain the range of at
least one
of such targets in the target field. .Iaddend..Iadd. 29. The method
of claim 27, further including the step of determining the range
and velocity of at least one of the targets, comprising the steps
of:
producing a signal pattern comprising of at least two said
signals,
transmitting the signal pattern into the target field during at
least partially different time intervals for each signal of the
signal pattern, and
processing the received reflections to ascertain the range and
velocity of at least one of such targets in the target field.
.Iaddend..Iadd. 30. The method of claim 27, wherein said step of
producing a pair of base signals comprises producing signals with
the property of:
one of each such pair of base signals being odd about a central
coordinate value in its finite time interval and the other of each
such pair being even with respect to its central coordinate value.
.Iaddend..Iadd. 31. The method of claim 27, further including the
step of ascertaining angular information regarding more than one
angle of at least one of the targets, comprising the steps of:
(a) producing a signal pattern comprising at least two signals each
said signal being a linear pair combination of base signals each
having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of each said pair of base signals is zero;
(3) each base signal of such a pair being in phase quadrature to
the other member of the pair;
(4) each of such pair of base signals being transformable to a
symmetric signal relative to a time reference by adding a constant
phase angle to the phase at each frequency; and
(5) each pair of such base signals occupying at least partially
non-overlapping band of frequency from the other pairs;
(b) transmitting such signal pattern;
(c) introducing phase distortions having a component independent of
frequency which vary in a known and single valued manner according
to desired angular information into the signal propagation path,
differing angular information being introduced in differing
frequency bands;
(d) receiving reflections of the transmitted signal pattern from
the target field; and
(e) processing the received reflections to ascertain the angular
information for at least one of such targets in the target
field.
.Iaddend..Iadd. 32. The method of claim 27, further including the
step of ascertaining angular information regarding more than one
angle of at least one of the targets, comprising the steps of:
(a) producing a signal pattern comprising at least two signals each
said signal being a linear pair combination of base signals each
having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of each said pair of base signals is zero;
(3) each base signal of such a pair being in phse quadrature to the
other member of the pair;
(4) each of such pair of base signals being transformable to a
symmetric signal relative to a time reference by adding a constant
phase angle to the phase at each frequency; and
(5) each pair of such base signals occupying at least a partially
overlapping band of frequency from the other pairs;
(b) transmitting such signal pattern;
(c) phase encoding positional information about plural angles in
each signal;
(d) receiving reflections of the transmitted signal pattern from
the target field; and
(e) processing the received reflections to ascertain the angular
information for at least one of such targets in the target
field.
.Iaddend..Iadd. 33. The method of claim 32, wherein said step of
phase encoding comprises the step of:
introducing phase distortions having a component independent of
frequency which vary in a known and single valued manner according
to desired angular information into the signal propagation path,
differing angular information being introduced into member signals
of the signal pattern
with independent phase encoding. .Iaddend..Iadd. 34. A system for
determining the velocity of a signal reflecting object positioned
in a medium of known signal-propagation characteristics
comprising:
(a) signal transmitter means for transmitting at two distinct
instants of time at least two signals having a predetermined time
interval therebetween, each signal having a continuous amplitude
spectrum between low and high frequency limits, for reflection of
the signals by the object;
(b) signal receiver means for receiving the reflected signals;
(c) processing sequencer means for measuring only the time interval
between the received signals; and
(d) parameter estimator means for determining the relative velocity
of the object based only upon the measured time interval between
the received signals and the predetermined time interval between
the transmitted signals. .Iaddend..Iadd. 35. The system of claim
34, wherein the two distinct instants of time are predetermined,
and wherein:
(a) said processing sequencer means comprises means for measuring
the arrival time of at least one reflected signal and determining
the signal transit time from the measured arrival time and the time
instant of its transmission,
(b) said parameter estimator means comprises means for correcting
the transit time as a function of the determined velocity, and
determining the distnace of said object relative to said
transmitter means and said receiver means as a function of the
corrected transit time and the propagation characteristics of the
medium. .Iaddend. .Iadd. 36. A system for determining the direction
of a signal reflecting object, comprising:
(a) transmitter means for transmitting a signal having a continuous
amplitude spectrum between low and high frequency limits, for
reflection from the object;
(b) receiver means for receiving the reflected signal;
(c) phase lens means for changing the phase of the signal in
dependence upon the direction of its travel while propagating
between said transmitter means and said receiver means, such phase
change characterized other than by a simple time delay;
(d) processing sequencer means for measuring the phase of the
received reflected signal; and
(e) parameter estimator means for determining the direction of the
object based on the measured phase. .Iaddend..Iadd. 37. The system
of claim 36 wherein said phase lens means for changing the phase
comprises means for changing the phase by a constant for each
direction independent of frequency. .Iaddend. .Iadd. 38. The system
of claim 37 wherein said phase lens means for changing the phase of
the signal comprises:
a medium having a dimension which varies in dependence upon the
direction of travel of the signal. .Iaddend..Iadd. 39. The system
of claim 36 wherein said phase lens means for changing the phase of
the signal comprises:
a medium having a dimension which varies in dependence upon the
direction of travel of the signal. .Iaddend. .Iadd. 40. A system
for ascertaining the range of one or more reflecting targets in a
target field referred to a coordinate system, comprising:
(a) transmitter means for producing and transmitting into the
target field a signal pattern having a linear combination of at
least a pair of base signals, such signals having the properties
of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of said pair of base signals is zero;
(3) each base signal of such pair being in phase quadrature to the
other member of the pair;
(b) means for receiving reflections of the transmitted signal
pattern from the target field; and
(c) means for processing the received reflections to ascertain the
range of at least one of such targets in the target field.
.Iaddend..Iadd. 41. A system for ascertaining the range and
velocity of one or more reflecting targets in a target field
referred to a coordinate system of fixed origin, comprising:
(a) transmitter means for producing and transmitting into the
target field a signal pattern having at least two signals during at
least partially different time intervals for each signal of the
signal pattern, each signal being a linear combination of at least
a pair of base signals, said base signals having the properties
of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of said pair of base signals is zero;
(3) each base signal of such pair being in phase quadrature to the
other member of the pair;
(b) means for receiving reflections of the transmitted signal
pattern from the target field; and
(c) means for processing the received reflections to ascertain the
range and velocity of at least one of such targets in the target
field. .Iaddend..Iadd. 42. A system for ascertaining the angular
information of one or more reflecting targets in a target field
referred to a coordinate system, comprising:
(a) transmitter means for producing and transmitting into the
target field a signal pattern being a linear combination of a pair
of base signals, such signals having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of said pair of base signals is zero;
(3) each base signal of such a pair being in phase quadrature to
the other member of the pair; and
(4) each of such pair of base signals being transformable to a
symmetric signal relative to a time reference by adding a constant
phase angle to the phase at each frequency;
(b) phase lens means for introducing a phase distortion, having a
component independent of frequency, which varies in a known and
single valued manner according to desired angular information into
the signal propagation path;
(c) means for receiving reflections of the transmitted signal
pattern from the target field; and
(d) means for processing the received reflections to ascertain the
angular information for at least one of such targets in the target
field.
.Iaddend..Iadd. 43. The system of claim 42, wherein said means for
processing comprises:
means for processing the received reflections to ascertain the
angular information and range for at least one of such targets in
the target field. .Iaddend. .Iadd. 44. The system of claim 42,
wherein:
said transmitter means comprises means for producing and
transmitting into the target a signal pattern comprising at least
two said signals during at least partially different time intervals
for each signal of the signal pattern, and
said means for processing comprises means for processing the
received reflections to ascertain the range and velocity of at
least one of such targets in the target field. .Iaddend..Iadd. 45.
The system of claim 42, wherein:
(a) said transmitter means comprises means for producing and
transmitting into the target field a signal pattern comprising at
least two signals each said signal being a linear combination of a
pair of base signals, each having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of each said pair of base signals is zero;
(3) each base signal of such a pair being in phase quadrature to
the other member of the pair;
(4) each of such pair of base signals being transformable to a
symmetric signal relative to a time reference by adding a constant
phase angle to the phase at each frequency; and
(5) each pair of such base signals occupying at least partially
non-overlapping band of frequency from the other pairs;
(b) said phase lens means for introducing phase distortions
comprises means introducing phase distortions having a component
independent of frequency which varies in a known and single valued
manner according to desired angular information into the signal
propagation path, differing angular information being introduced in
differing frequency bands;
(c) said means for processing comprises means for processing the
received reflections to ascertain the angular information for at
least one of such
targets in the target field. .Iaddend..Iadd. 46. The system of
claim 42, wherein:
(a) said transmitter means comprises means for producing and
transmitting into the target field a signal pattern comprising at
least two signals, each said signal being a linear pair combination
of base signals, each having the properties of:
(1) sharing a common amplitude spectrum which is essentially flat
or smoothly unimodal occupying a contiguous band of
frequencies;
(2) having a finite time interval before and after which each
signal of each said pair of base signals is zero;
(3) each base signal of such a pair being in phase quadrature to
the other member of the pair;
(4) each of such pair of base signals being transformable to a
symmetric signal relative to a time reference by adding a constant
phase angle to the phase at each frequency; and
(5) each pair of such base signals occupying at least partially
overlapping band of frequency from the other pairs;
(b) said phase lens means includes means for phase encoding
positional information about plural angles in each signal; and
(c) said means for processing comprises means for processing the
received reflections to ascertain the angular information for at
least one of such targets in the target field. .Iaddend.
Description
BACKGROUND OF THE INVENTION
Echo location systems are designed to identify some subset of
individual reflector or target parameters. These parameters include
the target position in terms like bearing and elevation angle, the
range, target relative velocity and the impedance contrast which
causes the echo and is also a measure of target quality. Individual
targets distributed with a propagation medium make up a target
field. The propagation medium can be attenuative and preferentially
cause the loss of the higher frequency components of a propagating
signal according to some physical law. Additionally, the
propagation medium can be dispersive causing the different
frequency components of a signal to travel with different
velocities hence, introducing distortions of phase and equivalency
of form into the propagating signal as a function of its
travel.
Such known echo systems emit a signal or signals into the
propagation medium; the identification process involves detecting
the echo train and performing a variety of appropriate analyses.
While this procedure is conceptually straight forward, there are a
number of practical difficulties which act to complicate, degrade
and make ambiguous such identifications.
First, there is a noise background to consider which is almost
always a problem in systems where signals are transmitted and
detected. Noise is defined in this instance as any contribution
which is not a part of the particular identification process and
has as its sources such elements as incoherent scattering by the
propagation medium or even the targets themselves. There are a
variety of techniques for the detection and enhancement of signals
in the presence of noise.
Next, there is the inherent ambiguity between the range and the
relative velocity of a target. A moving target can not only stretch
or shrink a returning echo signature depending on the sense of its
motion, but will also delay or speed up the time of its detection,
hence affecting the range calculation. Once again, there are a
variety of known techniques which can resolve this ambiguity. It is
widely recognized that continuous wave signals, for example, a
persistent sinusoid at a single frequency can provide good
resolution of the target's relative by means of the Doppler
frequency shift. The companion range resolution of such a signal is
necessarily poor since its character is indistinguishable from
cycle to cycle. Very short duration signals are affected only
slightly by target motion and while they provide good resolution in
detection time, they convey little or no information about relative
velocities. The chirp signal described by Klauder, Price,
Darlington and Albersheim in the Bell System Technical Journal,
Vol. 39, pp 745-808, July 1960, represents a compromise having
ambiguity in both velocity and range. Its advantages lie rather in
effectiveness of equipment utilization and the noise suppression of
its companion correlation detection.
Lastly, there is the problem of resolving target angular parameters
such as elevation and/or bearing. Currently, definition of angles
is achieved by the use of arrays of broad-beam source or receiver
elements, or else by means of narrow-beam source or receiver
elements. In both cases, the space in which the target field is
distributed must be scanned or viewed only one small part at a
time. Scanning is accomplished either electronically by steering
array beams or sequencing the operation of large numbers of
elements, or even mechanically by rotating operational narrow-beam
elements to new position.
The energy requirements of a scanned system are usually favorable
since the entire field of potential targets need not be illuminated
at once. On the negative side, however, the individual targets are
then not being continuously monitored.
SUMMARY OF THE INVENTION
This invention employs encoded signals followed by the correlation
of the received echo train with known signal signatures. A
relatively long signal train made up of essentially short signals
is used, thereby encompassing both continuous wave and impulse-like
properties. The correlation step achieves a measure of noise
suppression. Simultaneous high resolution information about range
and relative velocity is achieved essentially by means of
simultaneous solutions involving use of all of the observables
embodied in the signal train after detection by correlation.
Both the sources and receivers operate as broad-beam elements with
simultaneous illumination of all targets. Angular resolution is
achieved by appropriately interposing phase distorting lenses
between the sources and echo receivers. Information about the
angles is encoded into the phase character of the propagating
signal train. Energy requirements are modest despite the
simultaneous illumination of the entire field of targets because
the high repetition rate of the system allows that a rather low
echo signal level be tolerated. Also, since range and velocity
resolution are not directly dependent on the use of high frequency
signal components, lower frequency band signals may be used with
correspondingly less energy loss through attenuation.
Even if the propagating medium does modify the traveling signals by
altering their amplitude and phase spectral properties, the system
described by the invention will be able to function nevertheless.
Some empirical corrections would then have to be made to compensate
for the propagation effects, but these could be easily determined
by calibration studies using targets of known parameters.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows an embodiment of the invention with a single
transmitter, receiver and target;
FIG. 2 shows a processing sequence for the embodiment shown in FIG.
1 with a single transmitter, receiver and target;
FIG. 3 shows another embodiment in which a single phase lens
encodes angular information about targets;
FIG. 4 shows a processing sequence for the embodiment of FIG.
3;
FIG. 5 shows another embodiment utilizing a number of mutually
exclusive signal trains from the transmitter, one phase lens being
interposed for each signal train;
FIG. 6 shows a processing sequence for the embodiment shown in FIG.
5;
FIG. 7 shows another embodiment having different phase encodings of
angular information for each member signal of a single outgoing
signal train;
FIG. 8 shows a processing sequence for the embodiment of FIG.
7;
FIG. 9 shows another embodiment where the propagation phase
distortion and/or the reflecting phase distortion are treated as an
additional lens;
FIG. 10 shows another embodiment with a single transmitter, a
single receiver and a multiplicity of reflecting targets;
FIG. 11 shows the derivation of angles resolution by triangulation
for an embodiment with a single transmitter, two receivers and a
multiplicity of reflecting targets;
FIG. 12 shows the time and Fourier frequency domain properties of a
design base signal pair having respectively odd and even
symmetry;
FIG. 13 shows a typical outgoing signal connection of two
members;
FIG. 14 shows the graphical evaluation of the amplitude spectrum of
a cross-correlation in the Fourier frequency domain between a
detection signal pair and a propagating signal which has been
Doppler distorted;
FIG. 15 shows a computer simulation of the phase encoding of
angular information in a two member signal train;
FIG. 16 shows a simulated development of a correlation amplitude
function for the signal train with phase encoded angular
information;
FIG. 17 shows a sequence of cross-correlations of a sine chirp with
Doppler distortions of its original forms;
FIG. 18 shows the computation of phase and amplitude spectra of
member signals carrying phase-encoded angular information; and
FIG. 19 shows a computer simulation of the invention utilizing
three moving targets in the presence of random noise.
DETAILED DESCRIPTION OF THE INVENTION
FIGS. 1 and 2 show an elementary echo location system of the
invention.
A suitable transmitter 1 emits a signal train 1A having at least
two members into a propagation medium 2 in which is embedded a
reflecting target 3. An echo 3A from the target propagates to a
receiver 4 where it is detected and sent on to the processing
sequencer 5. The outputs of the processing sequencer are directly
interpretable as the target parameter estimator 6, which gives the
relative velocities between the source or transmitter 1, receiver
4, and reflecting target 3, and the range to the reflecting target
3 as a sum of the distance to the transmitter 1 and the receiver 4.
It should be understood that transmitter 1 and receiver 4 may be
coincident.
The system shown in FIG. 1 embodies components which are
essentially standard for the particular application. For example,
in the case of a sonar system, transmitter 1 might be a transducer
capable of introducing pressure waves into the water of a form and
sequence prescribed by a control signal. Receiver 4 might be a
hydrophone while a submarine can serve as reflecting target 3. In
this instance the sea water would be the propagation medium 2 and
processing sequencer 5 could be simple electronic network of
alternatively digital logic accomplishing equivalent
operations.
At least two signals are needed in the outgoing train to provide a
calibration standard to separate the contribution of the relative
velocity to the echo arrival time from the effect of target range.
By such means, the echo location system described is capable of
simultaneously making quite simple, highly resolved estimates of
both the target relative velocity and range.
Each member signal of the outgoing train has like polarization
characteristics if these are applicable and a common amplitude
spectrum as defined by the modulus of the exponential Fourier
transform. Further, the common amplitude spectrum F(w) is
essentially flat or smoothly unimodal over the ranges of positive
and negative angular frequencies w.sub.o
.ltoreq..vertline.w.vertline..ltoreq.w.sub.f and zero for practical
purposes outside the band as defined by these constant limiting
frequencies w.sub.o, w.sub.f.
Additionally, each member signal will be a linear combination of a
pair of base signals .function..sub.o (t), .function..sub.l (t)
which have the following properties to the practical approximation
required by the system:
I. .function..sub.o (t), .function..sub.l (t) share the common
amplitude spectrum F(w) as described.
II. There is a finite time interval of duration .alpha., before and
after which both .function..sub.o (t) and .function..sub.l (t) may
be considered to be zero.
III. .function..sub.o (t) and .function..sub.l (t) are in
quadrature. Every Fourier frequency component necessary for the
description of .function..sub.o (t) is displaced in phase as
measured from any origin in time by 90.degree. from its counterpart
in the description of .function..sub.l (t). Property I above
implies that all Fourier frequency components are present in
approximately equal amounts in .function..sub.o (t) and
.function..sub.l (t).
As a mathematical formalism, Property III states that
.function..sub.o (t) and .function..sub.l (t) are a Hilbert
transform pair at least to the accuracy of the system and
##EQU1##
By way of illustration, several pairs of signals which satisfy the
above specified properties are:
I. "Chirp" Signals s(t), c(t) ##EQU2##
II. "Klauder" Signals k.sub.o (t), k.sub.l (t) ##EQU3##
III. "Gabor" Signals g.sub.o (t), g.sub.l (t) ##EQU4##
FIG. 12 shows the time and frequency domain properties of the
Klauder Signals depicted as .function..sub.o (t), .function..sub.l
(t), the base signal pair.
If .function..sub.k (t) is a member signal of the outgoing train
and there are no distortions introduced by the propagation medium
2, then the received echo 3A for a moving reflecting target 3 will
have the mathematical form
T is a time delay associated with the target range at the time of
detection while s is a scale factor which arises from the target
relative velocity. For a reflecting target fleeing detection,
s>1 and a stretched signal signature will result. We have
assumed a positive impedance contrast between the propagation
medium 2 and the reflecting target 3. A "soft" target having a
smaller impedance than the medium for signal travel would cause a
signal reversal (or equivalently a 180.degree. phase shift in the
echo signal 3A.
The stretching or compression of the propagating signal is the
familiar Doppler effect and gives rise to the ambiguity in
resolution between target relative velocity and target range.
Suppose one could recognize some signal characteristic which
occurred at t=t.sub.o. In the echo 3A it would occur at t=st.sub.o
+T. If it were assumed, nevertheless, that it occurred at t=t.sub.o
+T, one would make an error in determining the echo arrival time of
(1-s)t.sub.o which would correspondingly cause an error in the
estimation of target range.
Signals designed as described, however, have a companion processing
sequencer 5 which allows a signal characteristic to be
unambiquously identified despite stretching or compression.
One embodiment of this sequencer is shown in FIG. 2. The use of a
train of signals of known intervals of separation then allows that
s be computed from the changes of such intervals in the echo train
3A. Once s is known, the echo arrival times may be compensated for
the target relative velocity effects so that accurate estimates of
the target range can be obtained.
To illustrate the principle, there is shown in FIG. 13 an outgoing
signal train of two members, these being respectively k.sub.o (t)
and k.sub.l (t), the Klauder Signals with their origins of
definition shifted. In mathematical language, the outgoing train
shown is simply given by
Further generality can be added by defining
where a, b, c, d are constants. Let us also introduce a scaling
such that
The outgoing signal train now will be taken to have the more
general form
The returning echo 3A will then have the following mathematical
form
The processing sequencer 5 is shown in FIG. 2. The received signal
is divided into two parts by signal splitter 5B. The correlation in
parallel of the two parts with .function..sub.o (t) and
.function..sub.l (t), k.sub.o (t) and k.sub.l (t) in this case is
effected by cross-correlators 5C1 and 5C2.
One can evaluate the correlation functions A(t), B(t) and their
squared functions A.sup.2 (t), B.sup.2 (t) very simply by
considering only the first member signal of the returning echo as
given by equation 8. If we relocate the origin of the first member
signal of the echo for convenience, the indicated chain of
operations gives the following
where * denotes a convolution operation (standard signal processing
usage) and -t denotes reversal of the signal in the time variable
t.
Owing to the symmetry properties of k.sub.l (t) and antisymmetry of
k.sub.o (t) it is evident that
and one can write
or
which follows along the processing sequence through 5E and which
has made use of a number of unstated assumptions and extended
properties of k.sub.o (t), k.sub.l (t). These latter points will
now be briefly discussed.
The result of the convolution of two functions both having either a
point of symmetry or antisymmetry is a function which has a point
of symmetry. Similarly, if one function has a point of symmetry
while the other has a point of antisymmetry, the result will have a
point of antisymmetry. If we perform Fourier analyses, using as a
coordinate origin these points of symmetry and antisymmetry, the
phase spectrum of the resulting transform must either be
identically zero, .+-..pi. or else equal to .pi./2 sgn(w).
From the well-known convolution theorem, three results can be
derived. First, if we scale a time variable by a constant s, we
scale the frequency domain variable by 1/s. In mathematical
terms,
are time and frequency domain equivalents. Convolution in the time
domain is equivalent to multiplication in the frequency domain.
Hence,
are time and frequency domain equivalents. In polar form, the
product F(w)G(w) is equivalent to a multiplication of the
respective amplitude spectra and a simple ordered difference of the
phase spectra. Finally, the time and frequency domain equivalents
of reversal in time are as follows:
where + denotes complex conjugation or simply the reversal in sign
of the Phase Spectrum.
With the three results 12, 13 and 14, we return to Equation 11
where we note that they have been used to eliminate cross-product
terms having the coefficient cd. Also we recognize that
which is in fact yet another Klauder Signal of slightly different
frequency content with a point of even symmetry, which we shall
call k.sub.i '(t). We can readily appreciate the make-up of k.sub.l
'(t) by examining FIG. 14 which is a graphical evaluation of the
convolution of Equation 15 for an approaching target performed in
the Fourier frequency domain.
Using analogous reasoning we find also that
which is the antisymmetric counterpart of k.sub.l '(t) and shall
herein be called k.sub.o '(t). We shall now state as a general
principle that any signal base pair having Properties, I, II and
III, thereabove, and further exhibiting respectively antisymmetry
and a symmetry characteristic will have a sum of squares which is
sharply peaked about the respective point of symmetry and
coincident point of antisymmetry. This property holds for all
Klauder Signals 2B including.
and all Gabor Signals 2C. Before presenting some analysis which
suggests the validity of this general principle, we should note
that equation 11 may be rewritten as
k.sub.o '(t) and k.sub.l '(t) as functions with coincident points
of antisymmetry and symmetry must have Fourier series
representations of the form ##EQU5## The sum of the squares of
equation 18 is then simply ##EQU6## Since the common amplitude
spectrum F(w) of k.sub.o '(t) and k.sub.l '(t) is smooth and
unimodal, we can take all the h.sub.i which are nonzero to be
approximately equal. This allows us to recognize that equation 20
has a maximum at t=0 where all the cosine harmonics are in phase.
We also recognize that equation 20 is symmetrical about t=0.
An analysis of .function..sub.j (s(t-.alpha./2-.tau.)-T), the
second member signal of the echo train described by equation 8
through the processing sequence of FIG. 2 would parallel the
considerations of equations 9 through 20. Hence the net effect of
the processing sequency of FIG. 2 on the echo 3A described
mathematically for the particular example by equation 8 is to
identify two well-defined correlation peaks of mathematical form of
equation 18 which occur at times.
in the correlated variable time scale.
The interval between the peaks is simply s.tau. and since .tau. is
known, s can be determined thus providing the estimate of the
target relative velocity. Once s is known, the target range may be
computed from T and a knowledge of the velocity V of the signal 1A
in the propagation medium 2 (FIG. 1).
FIGS. 15 and 16 further illustrate the technique through the
results of a digital computer simulation. In FIG. 15 we observe
seven returning two-member Klauder echo signal trains of the
mathematical form of equation 8. The various trains correspond to
differing choices of the constants c, d, a, b as they are defined
by equations 5 and 6. For the particular illustration
note that the signal echo train identified by "Return Azimuth"
0.degree. is in fact of the form illustrated by FIG. 13. Results of
the processing sequencer of FIG. 2 are shown in FIG. 16. The
interval between the peaks is simply s.tau. in all cases and it can
easily be determined as described.
Application of the technique using signals like the chirps of
equation 2A are similar, but require somewhat more subtlety in
their explanation. It is well-known that the cross- and
auto-correlation signatures of the chirp signals 2A in the absence
of any stretching or compression are in fact to a good
approximation the Klauder Signals 2B. While this is well-known, we
illustrate this point again by computer simulation. The bottom
curve of FIG. 17 is in fact the auto-correlation of a sinusoidal
chirp or sweep where
The cross-correlations of the undeformed sweep with stretched
versions of itself are also shown.
It is believed that the processing sequence of FIG. 2 when applied
to the deformed echo chirps will nevertheless produce a result as
in FIG. 17 with easily identifiable major peaks. As before, the two
peaks would have similar form and be separated by an interval
s.tau.. There now will be, however, a correction to be made to the
peak arrival times which depends on the computed value of s prior
to any estimation of the target range. The nature of the correction
and its variation with s may be determined by calibration studies
using correlations as illustrated in FIG. 17.
As a final note, we emphasize that additional member signals in the
outgoing train will provide redundant information about the target
relative velocity and range which will enable the echo location
technique to function even in noisy environments.
FIG. 3 shows another embodiment of the invention. The major
difference between the embodiments of FIGS. 1 and 3 in terms of
physical elements is the interposition of a phase lens 7 somewhere
in the signal path from the transmitter 1 to the reflecting target
3 to the receiver 4. We term the lens a "phase lens" since its
primary purpose is to introduce into the propagating signal train a
phase distortion which varies in a known and single valued manner
with certain desired angular information. The outgoing signal 1A
will now be designed in such a manner that after the received echo
3A passes through the processing sequencer 5, outlined in some
detail in FIG. 4, recoverable target parameters will include range
and relative velocity as in the embodiment of FIG. 1, as well as
the angular information which may, for example, relate to the
target azimuth or bearing.
Again the transmitter 1 emits a signal train 1A with at least two
members. The signal may encounter the phase lens 7 at this time or
else intercept it prior to detection by the receiver 4. A
reflecting target 3 as before is imbedded in the propagation
medium. The principal difference in design of members of the signal
train for the application in FIG. 3 and the one in FIG. 1 is that
we now require a fourth property in that .function..sub.o (t),
.function..sub.l (t) must have properties of antisymmetry and
symmetry respectively or be "odd" or "even" respectively about the
points of antisymmetry or symmetry.
Formally stated, the fourth property is
IV. .function..sub.o (t) and .function..sub.l (t) must be odd (or
antisymmetric) and even (symmetric) respectively about the central
coordinate value in their interval of definition of duration
.alpha..
Let us for the moment assume that the phase lens 7 introduces a
constant phase shift .theta. independent of w which varies only as
described with the desired angular information. We shall describe
the operation of the processing sequencer 5 of FIG. 3 for this
circumstance and then indicate those modifications which would be
required for more complex phase shifts. Recall the signal train 7
and its manner of definition from Klauder signals 2B and the
constants c, d, a, b according to equations 5 and 6. We shall use
such a signal train or one designed analogously from Gabor signals
2C with the following modification
Note that owing to property IV, the chirp signals 2A are
disqualified from direct use in this technique.
Let us consider .function..sub.k (t) and .function..sub.j (t)
(equation 5) individually for the moment and their Fourier
frequency domain equivalents. In the Fourier frequency domain:
##EQU7## In equation 25, F(w) is the common amplitude spectrum
(Property I), .+-. is a shorthand notation for -sgn(w), and
equations 24 have been applied. The properties of .function..sub.k
(t) and .function..sub.j (t) in the complex Fourier frequency
domain are much like those of .function..sub.o (t),
.function..sub.l (t) except that a constant phase angle .theta. has
been introduced into each frequency component.
In fact, a phase lens 7 able to introduce phase distortions
independent of w would convert the wave train 4 into the form of
the wave train 7. If now this phase distortion were diagnostic of
the desired angular information, then recovery of it would be
tantamount to recovery of the angular information. Assuming such a
phase lens 7, we now change our notation for the propagating signal
train 7 to
to give expression to the variation of the train with the desired
angular information. The received echo (3A and 4) can
correspondingly be expressed as
--refer to equation 8. It is important to recognize that .theta. is
unaffected by the Doppler effects owing to the special properties
of our signals.
For an understanding of the essentials of processing of FIG. 4, it
will be more convenient to look at the mathematical analyses
entirely in terms of their Fourier frequency domain equivalents. To
facilitate comprehension, we refer to the results of equations 12,
13, 14, 25 and FIG. 14.
In the Fourier frequency domain, the received signal 5A is
simply
The signal splitting operation 5B simply requires division of
equation 28 by 2. The cross-correlations 5C1 and 5C2 simply require
that the halved components of equation 28 be multiplied
respectively by F.sub.o.sup.+ (w) and F.sub.1.sup.+ (w). Evaluation
of the amplitude portion of the convolutions proceeds analogously
to the illustration of FIG. 14 and we are again led to define the
signals .function.f.sub.o '(t) and .function..sub.l '(t)
analogously to k.sub.o '(t) and k.sub.l '(t) as described in the
discussions surrounding equations 15 and 16. Recall that
.function..sub.o '(t), .function..sub.l '(t) are much like
.function..sub.o (t),.function..sub.l (t), but span a slightly
different frequency range.
In detail, and in the frequency domain, the two convolution
products are simply
Now making use of the signal pair .function..sub.o '(t),
.function..sub.l '(t) we can write explicitly the time domain
equivalents of the expressions of equation 29 or
In deriving the results of equation 30, we have defined (and
employed) the signal pair f.sub.k '(t,.theta.), f.sub.j
'(t,.theta.) in a manner analogous to the definitions implied by
equations 26 and 27. Also, the notation T.sup.-1 {} denotes the
operation of an inverse Fourier transform.
At this point, following the sequence of FIG. 4, we must take A(t)
and B(t), the two expressions of Equation 30, square them (5D1 and
5D2) and add them (5E). Note that owing to a sign difference no
"cross-terms" of the form .function..sub.k '(t,.theta.)
.function..sub.j '(t,.theta.) appear. We are left simply with the
quantity
Before interpreting the significance of equation 31, it is
desirable to comment on the nature of the expression which has the
form
Recalling all of the arguments of similar nature surrounding the
understanding of equation 11, and in specific the results developed
through equations 18, 19 and 20, we recognize that equation 31 will
have well defined maxima at times.
We can now determine the target relative velocity and range
precisely as in the application of FIG. 1. Before we consider some
alternate methods of recovery of the target angular information 5G1
and 5G2, it is instructive to refer back to FIG. 15 where we now
recognize a family of two member signal trains with encoded values
of .theta. ranging from 90.degree. to -90.degree. in 30.degree.
increments. Note that the computer simulation of the sequence of
FIG. 4 through 5F (Equation 33) shows that the detected peak
interval is for practical purposes independent of the encoded phase
information.
We shall now consider some methods of recovery of .theta. or
equivalently the target angular information by the alternate 5G1 of
FIG. 4. On the stretched time scale of equation 31 consider the
arctangent of the ratio of ordinate values of the return signal
(equation 27) at the two particular time values corresponding to
the correlation maxima. In detail,
where ##EQU8## Phase Angle=arctan {-sin.theta./cos.theta.}=arctan
{-tan.theta.}=-.theta.34
To derive the result of equation 34 we have used the definitions of
equation 5 and property IV or .function..sub.o (O)=0 and
.function..sub.l (O)=1. Analogous reasoning allows additional
computations of .theta. using the correlation components A(t), B(t)
certain of which are illustrated in FIG. 4 as 5G1.
The alternate 5G2 of FIG. 4 for determining .theta. or the target
angular information is also readily accomplished. Using the peak
positions determined by 5F (equation 33) as coordinate origins,
Fourier transforms of the correlated member signals of the trains
A(t), B(t), or the returning echo train itself may be computed and
the phase spectra computed. Over the band of frequencies
encompassed by f(w) (property I), the phase sepctrum will be
constant and equal to .+-..theta. or .+-..theta. .+-..pi./2. We
illustrate this point again by computer simulation.
FIG. 18 illustrates a nine member returning signal train in which
.theta. is successively encoded at 10.degree. increments from
-90.degree. to 0.degree.. Note that sign convention used here is
the opposite of the one used in FIG. 15. An amplitude and phase
spectrum was computed for each member signal using as the
coordinate origin the peak indicated by processing according to
FIG. 4 up through element 5F. The train is denoted by Sequence B
and it is readily observed that the amplitude spectra are identical
as they should be. For the meaningful frequency band defined by the
amplitude spectrum we see that the phase spectrum is in fact
constant and equal in value to .theta..
We now address the question of the introduction by the phase lens 7
of FIG. 3, phase variations which vary with the target angular
information but which are not independent of the frequency. It
should be noted, however, that for restricted frequency bands the
approximation
may be expected to reasonably represent phase effects. The linear
variation of phase with frequency inherent in equation 35 and
governed by .theta..sub.l necessarily implies that some timing
adjustment depending on .theta..sub.l will be needed which will
change the target range computation. Target relative velocity will
be unchanged as all member signals undergo the same phase
distortion for any given value of the target angular information
hence suffering also the same delays.
To illustrate the previous discussion we note that substituting
.theta.(w) for .theta. in equation 25 would give in the time domain
in place of equation 26
The received echo (3A and 4) would be correspondingly modified. If
the desired angular information is functionally related to
.theta.(w) then it is similarly related to .theta..sub.o,
.theta..sub.l through equation 35.
Referring to FIG. 4 in this circumstance, it is evident that the
alternate computation for .theta., 5G1 will yield a single
characteristic .theta..sub.o value which can uniquely be related to
the desired target angular information for reasonably behaved phase
lenses 7 as described by empirical procedures which make use of
reflecting targets of known parameters. The same calibration
procedures can be used to relate .theta..sub.l to the measured
.theta..sub.o so that the echo timing correction can be made. Of
course, alternate 5G2 will similarly allow a direct estimation of
.theta..sub.o, and .theta..sub.l may then be related to
.theta..sub.o by empirical calibrations. .theta..sub.l is again
directly interpretable as the timing correction needed for
determining the target range, and .theta..sub.o and .theta..sub.l
are now both available to define the approximation to .theta.(w)
thus allowing as before determination of the target angular
information.
Yet another embodiment of the invention similar to that illustrated
in FIG. 3 is indicated in FIG. 5. Here a reflecting target 3 is
identified in terms of its relative velocity, range and angular
information in more than one angle as perhaps the bearing and
elevation angle. The technique uses a number of the embodiments of
the type of FIG. 3 in concert, each realization of the former
technique occupying different but non-overlapping bands of
frequency or polarization (if applicable). The signal bands are
selected to retain their exclusion in polarization and/or frequency
despite all Doppler effects and effects introduced by the
propagation medium 2 or reflecting target 3. Angular information is
encoded in each signal band or polarization direction by an
appropriate phase lens as in FIG. 3 except that the lenses may
affect only certain of the signal bands or polarizations, and
different angular information may be encoded in the different
bands.
FIG. 5 illustrates the technique in terms of a number of different
frequency bands, but it should be understood that mutual exclusion
may still be achieved within a single band if differing directions
of polarization are employed. In this case, separation of the
signal bands is accomplished by filtering according to polarization
first.
The transmitter 1 of FIG. 5 emits the coincident signal trains as
described which may have any relative alignment in time including
simultaneity. Prior to encountering the receiver 4 a series of
phase lenses 7A, 7B, 7C, etc. are interposed or some other
mechanism employed which encodes desired angular information as a
phase distortion in the manner previously described. The lenses 7A,
7B, 7C, etc. or equivalent mechanisms act independently upon the
differing signal bands so that differing angular information may be
encoded in each of them. Encoding of the same information in
differing bands would, of course, provide redundancy and improved
recovery of such information noisy environments.
The propagation medium 2 and reflecting target 3 function
analogously to their respective roles in FIG. 3. The processing
sequence 5 functions much as in FIG. 3 but in parallel for each
signal band. FIG. 6 diagramatically illustrates the parallel
processing for the configuration outlined by FIG. 5. The target
parameter estimator 6 might encompass certain added features to
allow for the effective use of redundant information about target
range and relative velocity.
We can readily appreciate that polarization filtering (if
applicable) or the cross-correlation operations 5CA, 5CB, 5CC, etc.
of FIG. 6 do in fact achive a separation of the signal bands. For
example, each such operation develops correlation components
analogous to A(t), B(t) of FIG. 4 (also Equation 30). Equation 29
shows the counterpart Fourier transforms A(w), B(w) in polar form.
If the amplitude spectrum of the received return signal does not
overlap that of the base function pair of the cross-correlation
5CA, 5CB, 5CC, etc., no output whatsoever will result. Hence the
coincident signal bands do not interfere and may be treated quite
independently.
FIG. 7 illustrates another embodiment of the invention which is
like the embodiment of FIG. 5 in that the reflecting target 3 is
again identified in terms of its relative velocity, range, and
angular information in more than one angle as perhaps the bearing
and elevation angle. Unlike the embodiment of FIG. 5 which utilizes
simultaneously many signal bands, the angular information in this
approach is phase encoded differently in the various member signals
of the outgoing signal train 1A. The transmitter in this instance
functions as in FIG. 3 utilizing only a single signal band defined
as having uniform polarization (if applicable) with impulsive
member signals of four properties as described having a frequency
content limited to w.sub.o
.ltoreq..vertline.w.vertline..ltoreq.w.sub.f.
One possible mechanism for accomplishing the diverse phase
encodings is shown in FIG. 7 where differing phase lenses 7 are
interposed in turn between the transmitter 1 and the reflecting
target 3 such that each member signal of the outgoing signal train
1A encounters a different lens. If information is desired about two
angles, at least two such lenses are required. Additional angles
would require additional lenses and at least an equivalent number
of member signals in the outgoing train.
Let each phase encoding be characterized by a single phase value
.theta..sub.i recognizing that as in preceding discussions the
characterization can be generalized to accommodate more realistic
encodings of general linear form. We shall now assume that the
desired angular information .psi..sub.j is related to .theta..sub.i
by the known or empirically determined relation
where i denotes the member signal of the train and j denotes the
particular angular information in a plane j containing the
transmitter 1 and referred to a line in this plane.
For a more concrete discussion, we shall consider the specific case
in which there are only two outgoing member signals, =1,2 and one
value of .psi..sub.j or .psi..sub.l. g.sub.1 and g.sub.2 shall be
defined as implied by the equations
We may now similarly interpose other phase lenses 8 between echo 3A
from reflecting target 3 and receiver 4. One such lens is depicted
in FIG. 7. As before, the single value of phase .phi..sub.i will
characterize the encoding, where i again refers to the member
signals, but now of the echo train 3A. The .phi..sub.i are related
to the desired angular information referred to a specified line in
a plane k through the receiver 4 by the known or empirically
determined relation
Again, for illustrative purposes and with no loss of generality, we
shall assume a single lens 8, a single desired angle .gamma..sub.1
such that for the two returning member signals, the phase encoding
by the lens 8 would be
where h.sub.1, h.sub.2 have implicit definition in equations 39. It
is important to add that for each desired angle beyond the first,
either g.sub.i (.psi..sub.j) or h.sub.i (.gamma..sub.k) for member
signal i must have a single value form more general than the simple
addition of a constant value.
The processing sequence 5 described in greater detail in FIG. 8 is
analogous to the sequence of FIG. 4 through element 5F of that FIG.
where the timing of peaks in the correlation component square sum
leads to estimates of the target relative velocity and range. FIG.
8 illustrates alternate 5G1 of FIG. 4 being employed to estimate
the phase characteristic for each member signal. Referring to
equations 36 and 37 these are given generally by
which are a set of i simultaneous equations for j+k unknowns. The
unknown .psi..sub.j and .gamma..sub.k constitute the desired
angular information and may be determined when i.gtoreq.j+k.
In the simplified case cited for illustration in equations 37 and
39, the counterparts to equations 40 are
whose solution can be directly obtained as a simple sum and
difference of observed phases or ##EQU9## Note that the successive
phase encodings imposed on the propagating signal by the cascaded
phase lenses 7 and 8 are simply additive.
Where the phase distortion imparted by any lens requires a more
general description such as a linear approximation as described by
equation 35, the procedure is similar except that after determining
the desired angular information from member signal i, a timing
correction is needed leading once again to a correction to the
target range and in this embodiment also to the target relative
velocity. Note that in other embodiments of the invention where all
member signals passed through common phase lenses, no correction
was needed for the target relative velocity. The timing correction
appertaining to each lens may be determined empirically as a
function of phase as before using targets of known parameters.
These same calibration techniques would define the relations
(equations 37 and 38) unless they were determined theoretically or
on some other basis.
It should also be recognized that alternate 5G2 of FIG. 4 for each
member signal would again provide values of .theta..sub.i
+.phi..sub.i which could be similarly employed to derive the
desired angular information and to correct the target relative
velocity and range should such corrections be needed.
Another embodiment of the invention which follows from the previous
discussions is one which incorporates the elements exemplified by
both FIGS. 5 and 7 and their complementary processing sequences
illustrated by FIGS. 6 and 8. Here again, a reflecting target 3 (of
FIGS. 5 or 7) would be identified in terms of its relative
velocity, range and angular information in more than one angle as
perhaps the bearing and elevation angle. However, the phase
encoding of the desired angular information would not be the same
for all member signals of the outgoing train as in FIG. 7, and,
more than a single band would also be utilized, such bands being
mutually exclusive even after Doppler effects, by virtue of
polarization and/or frequency differences, with each band having
encoded differing angular information.
One practical objective in combining the approaches of FIGS. 5 and
7 as described is to achieve even greater redundancy of all the
reflecting target parameters yet without increasing the period of
time which is needed to make such identifications. Details of
accomplishing the processing for such identification follows
precisely from both FIGS. 6 and 8.
FIG. 9 depicts an embodiment of the invention much as described by
FIG. 7, where again a reflecting target 3 is identified in terms of
its parameters as target range, relative velocity and angular
information in many angles, but now an accommodation is made also
for a phase distortion 2A which is a function of the target range
introduced by the propagation medium 2 and/or the reflecting target
3 as a function of the angle of signal incidence. For illustrative
purposes in this discussion the phase distortion will be assumed
attributable in its entirety to the propagation medium 2, with no
loss of generality of the technique implied by such assumption.
For practical materials or propagation media 2, the phase
distortion 2A will be smoothly varying and represented well by an
analytic expression like equation 35. Such a phase distortion was
specifically discussed and treated in all of the previous
embodiments of the invention where angular information was to be
determined. As a consequence of a distortion of this form a target
range correction will result, however, no relative velocity error
would be introduced since all of the member signals see the same
phase distortion 2A. The constants which approximate the phase
distortion in the form of equation 35 will also be estimated.
In FIG. 9, the role of the propagation medium 2 with regard to
phase is analogous to the interposition of one additional phase
lens 7 or 8. Unlike the phase lenses 7 or 8, the phase effects to
be introduced are not designed as a part of the invention and vary
with the target range or better the travel distance of the signal
rather than any angular information. Let us designate the phase
distortions 2A for a given value of target range R as
Since phase effects are additive, and if the phase distortion
introduced by the phase lenses 7 and 8 have form analogous to
equation 43, then for the i.sup.th member signal of the echo train
3A after being received 4 and processed according to the processing
sequence 5 of FIG. 8, the detected phase would be
where g.sub.i and h.sub.i are defined as in equations 36 and 38 and
where .lambda.(R) is a function of the target range.
In this case g.sub.i, h.sub.i and .lambda.(R) are all determined
theoretically or else by empirical studies using a reflecting
target 3 of known parameters. The equations 44 may be solved
simultaneously or by least squares for a sufficient number of
member signals (i.ltoreq.j+k+1) to give .psi..sub.j, .gamma..sub.k
and R which in turn can give values for .theta..sub.i, .phi..sub.i,
.lambda..sub.o as distinct from their sum. For phase
characterization of the general form specified it should be
recognized that a timing correction would be needed for each member
signal which would modify the determinations both of target
relative velocity and target range.
The timing correction as described in certin of the previous
embodiments can be computed theoretically or else can be determined
by some empirical method using a known reflecting target 3 so that
for values of .theta..sub.i, .phi..sub.i, .lambda..sub.o, or
.psi..sub.j, .gamma..sub.k, R an appropriate sum of corrections may
be applied. Of course, once .lambda..sub.o and the timing
correction for the propagation medium 2 are known, the phase
distortion 2A of the medium is known according to equation 43 as is
all other desired information. Note that the timing correction
associated with .lambda..sub.o or R gives rise to no change in the
computation for target relative velocity since, as was mentioned,
all member signals are subjected to the same timing adjustment.
FIG. 9 represents the most general embodiment of the invention up
to this point in a single signal band as defined by frequency range
and polarization direction, if the latter is applicable. FIG. 5, on
the other hand, depicts an embodiment in which the phase encoding
of target parameters and propagation material properties (if
applicable) would utilize a number of signal bands. It follows then
that a more comprehensive embodiment may yet be envisaged in which
a number of signal ands are employed, the technique in each single
signal band being represented as in FIG. 9.
An embodiment as described would provide the greatest redundancy
yet for the identification parameters of the reflecting target and
would also allow the caracterization of the phase distortion of the
propagation medium and/or the reflecting target in a number of
frequency bands and polarization directions.
Another embodiment of the invention complementing all other
variatioons described up to this point would include the use of
amplitude spectral information determined either collectively from
the received return signal or else computed from individual member
signals as one additional target identification parameter
describing its "quality". Referring to FIG. 4 in alternate 5G2 we
first noted the use of Fourier analysis in the processing sequence
for the received return signal 5A.
Throughout preceding discussions, the only modification admitted
for the amplitude spectrum of the base signal pairs had been a
Doppler induced effect. We shall now want to give cognizance to
variations in the amplitude spectrum caused by frequency dependent
alternative mechanisms of the propagation medium and also the
frequency dependent reflective properties of the target.
A preferential loss of higher frequency commponents as a function
of the length of propagation path is a common characteristic of
most propagation media. Particular media can exhibit "window"
effects where for certain frequencies or polarizations (if
applicable) anomalous attenuation or lack of attenuation will
occur. For all of the permissible base signals having either three
or four fundamental properties, these effects should not alter in
most circumstances the basic smooth and unimodal character of the
amplitude spectrum and so the mathematical approximations presented
would retain their validity. No effects would be induced on phase
as it is measured here, since the base signals share a common
amplitude spectrum.
Empirical studies have been cited in a number of previous
discussions as a means for establishing standards or functional
relationships necessary of the determination of certain of the
target identification parameters. Clearly, by empirical studies it
is possible to determine the absolute amplitudes and changes in
form of the amplitude spectrum of member signals caused by a
propagation medium so that these effects might be removed from
consideration. Other differences of magnitude and form must then be
diagnostic in some sense of the target, describing its quality.
Peculiarities of the target figure or shape, as well as perhaps a
transitional character in reflective properties can cause
modifications to the amplitude spectrum which might be unique to
certain targets hence facilitating positive identifications, or
else simply assisting in their categorization. We must also note in
the context of target quality determination that phase plays some
role, since it is a 180.degree. phase shift or reversal of
arithmetic sign of the echo which allows distinction between "hard"
and "soft" targets where the magnitudes of the reflective contrasts
between the target and the propagation medium are equal. The
terminology "hard" is being applied to targets of materials in
which the signal propagation velocity exceeds its velocity in the
propagation medium.
Target quality information is contained in the member signals of
the signal train as well as in the entire train itself. Hence as in
the case of most of the other parameters describing the target, a
measure of redundancy is again present.
FIG. 1 illustrates the most elementary embodiment of the invention
described, but can be used to help explain the most encompassing
embodiment of the invention yet to be described. All variations of
the invention discussed so far have included a single transmitter
1, a single receiver 4, and single reflecting target 3. In fact,
any number of any of these elements can be used, yet allowing all
other essentials of the invention so that the plurality of
reflecting targets can each be identified individually according to
their parameters with whatever degree of redundancy the particular
embodiment allows. FIG. 10 exemplifies such an embodiment based on
the variation shown in FIG. 1.
In FIG. 10, three distinct reflecting targets 3, two distinct
receivers 4A, 4B and a single transmitter 1 are shown. The
propagating signal 1A is taken as in FIG. 1 and the propagation
medium 2 is also as in FIG. 1. Each reflecting target returns an
echo to each of the receivers 4A, 4B. The processing sequences 5
are essentially as depicted in greater detail in FIG. 2 except that
each of the reflecting targets 3 now corresponds to a sequence of
peaks in the step 5E of FIG. 2 which may have any arbitrary
relation in time, one sequence to another, hence necessitating also
the inclusion of some logic to separate the member peaks in each
sequence so that target parameter estimates 6 can be made for each
reflecting target. Since there are two receivers, the target
parameter estimator 6 may now include angular information, even
though the embodiment of FIG. 1 upon which we based this
illustrative case made no provision for the inclusion of such
information.
FIG. 11 suggests a familiar analytic basis by which the use of
information from the two receivers 4A, 4B may be handled to yield
angular information. The transmitter T and receivers R1, R2 of FIG.
11 occupy known relative positions which are either fixed or
changing in a known manner in time. The processing sequence 5 as
described by FIG. 2 is capable of estimating for each reflecting
target only a range and relative velocity. For any estimated range,
the permissive target locus is an ellipse with the transmitter and
receiver which detected the particular echo at its focii. The
intersection of the elliptical locii will define the target
position and thus provide the angular information about the target
expressed in the coordinate network of the transmitter and two
receivers. Ambiguity of position can be eliminated as indicated
again in FIG. 11 by designing the configuration so that certain
positions are disallowed, as for example those to the left of the
dashed line AA'.
It is important to mention that the logic by which the sequences
corresponding to the differing targets are isolated can include
clues about the consistency of relative velocity of the targets and
even amplitude spectral information which in fact was not mentioned
in the embodiment of FIG. 1, but was described in a later, more
sophisticated technique.
To further illustrate the technique in which a plurality of
elements is permissible, we address now the embodiment of FIG. 3,
where angular information in one angle is one of the target
identification parameters, and appeal also to a computer
simulation. For illustrative purposes we adopt the configuration
shown in FIG. 3 but declare only a plurality of reflecting targets,
specifically three. The signal 1A from transmitter 1 is taken to be
made of a Klauder base signal pair as described by equation 4 and
depicted in FIG. 13.
Amplitudes of the three reflecting targets are taken to be in the
ratio 4, 3, 2 while their initial ranges, angular information and
relative velocities tabulate as:
______________________________________ Range Relative (Arbitrary
Angular Velocity Amplitude Units) Information (Ratio) (ratio)
______________________________________ Target 1 150 0.degree. 1 4
Target 2 300 -45.degree. 1.05.sup.a 3 Target 3 450 90.degree. .95#
2 ______________________________________ .sup.a approaching
#fleeing
Note that we are using the same convention for encoding the angular
information as indicated in FIG. 15. We are assuming constant phase
distortions imparted by the phase lens 7 of FIG. 3. Also, the
target relative velocities are expressed as 1 and their ratios with
the speed at which echo location signal 1A travels.
Since target 2 is approaching the detection system centroid while
target 3 is fleeing, the expected target ranges for the assigned
target relative velocities are 280 and 475 units respectively.
Also, if all other parameters can be correctly identified, then a
reconstruction of the received return signal 4 may be accomplished
for comparison, with the observed one. Such a sequence is shown by
the computer simulations of FIG. 19 where the noise-free and noisy
case are examined.
In FIG. 19, curve A is the observed received return signal 4. The
reconstruction from detected parameters is denoted curve D. Curve B
shows the sum of the squares of the correlation components, curves
C1 and C2. Referring to FIG. 4 which describes the particular
processing sequence 5, we recognize curve C1 as 5D1, curve C2 as
5D2 and curve B as 5E. The family of curves with primes represent
results from the circumstance where background noise is present.
For this case note that there is little difficulty in
distinguishing among the reflecting targets.
Hence these illustrations make clear certain of the advantages of
the embodiments which function in environments with a plurality of
reflecting targets and utilize where beneficial, pluralities of
transmitters and receivers.
A more general variation of the invention, encompassing all other
variations described and which can be developed starting with any
of these techniques, employs a base signal pair in the processing
sequence which can differ from the base signal pair of the outgoing
signal design. The two distinct base signal pairs shall be
designated as the design signal pair and the processing signal pair
respectively. For any given design signal pair, the admissible
processing signal pairs must represent only rotations of the design
signal pair phase spectra by a constant angle, and all differences
in amplitude spectrum must be constrained such that the product of
the design signal pair common amplitude spectrum and the processing
signal pair common amplitude spectrum in itself has a form
appropriate to a base signal pair.
All embodiments described heretofore employed a common base signal
pair for the design of the outgoing signal train and the processing
sequence. In fact, this restriction need not exist and we may with
appropriate planning use an outgoing signal train developed with
Klauder signals, yet select a processing base signal pair
constructed with Gabor signals to extract desired target parameters
(refer to equations 2B and 2C which specifically define Klauder and
Gabor signals.)
Should the particular application of the invention not make use of
phase encoding of information as the variation described in FIGS. 1
and 2, then both the design signal pair and processing signal pair
would each have to satisfy only the first three fundamental
properties which were outlined in that discussion. Where angular
resolution via phase encoding is called upon, the fourth
fundamental property introduced in the discussion of the
application depicted in FIGS. 3 and 4 is needed. A modified form of
this property may now be taken for both the design signal pair and
the processing signal pair. Calling such a signal pair
.function..sub.k (t), .function..sub.j (t) the revised statement of
Property IV reads:
IV. .function..sub.k (t) and .function..sub.j (t) must be
transformable to respective odd and even form about the central
coordinate value in their interval of definition of duration a, by
a constant shift of phase at all frequencies where the coordinate
origin of definition of such phase is again taken at the same
central coordinate value.
In reviewing the mathematical discussions of the simpler techniques
it becomes clear that the processing sequences as described can
progress up to determinations of phase using differing design and
processing signal pairs having three or four fundamental properties
as the application requires, and having permissible departures in
amplitude spectra. Specifically, in the embodiment shown in FIG. 1,
any appropriate processing signal pair allows completion of the
entire processing sequence shown in FIG. 2 in its entirety with no
modification. Alternatively, in the embodiment shown in FIG. 3
where phase encoding is employed to achieve angular resolution, the
processing sequence of FIG. 4 would be unmodified through element
5F. Thereafter, an adjustment in phase related to the constant
shift of the processing base pair relative to the design base pair
described by the revised Property IV would be required.
For purposes of having a more concrete illustration, consider an
outgoing signal train developed using as the design signal pair
.function..sub.k (t), .function..sub.j (t) defined by equation 5
with the constraint of equation 6 and subject to equation 24. In
the Fourier frequency domain following equation 25, we have the
counterparts
If we envisage a phase encoding mechanism as in the technique
described by FIGS. 3 and 4, the received return signal 5A having
undergone a constant phase shift now termed .beta. for the
particular angular resolution, would have a form similar to
equation 27 but specifically
In the Fourier frequency domain the received return signal 5A (also
equation 46) is now
(Compare with equation 28).
Let us select a processing signal pair g.sub.k (t), q.sub.j (t)
defined analogously to .function..sub.k (t), .function..sub.j (t)
having as Fourier frequency domain counterparts
A(t), B(t) or 5D1, 5D2 of the processing sequence of FIG. 4 would
then have as frequency domain equivalents
(Compare with equation 29).
From equation 49 onward this illustrative analysis may proceed in
parallel with the development based on FIGS. 3 and 4 with two
provisions. First, the product G(w)F(w/s) must define an amplitude
spectrum which is essentially smooth and unimodal so that a signal
pair analogous to .function..sub.k '(t,.theta.), .function..sub.j
'(t,.theta.) of equation 30 may be defined. Second, the phase
encoded angular information relating to .beta. can be determined
only after compensating for the phase rotation of the design signal
pair by .theta. and the processing signal pair rotation by .gamma..
Note that the relative rotation between these two pairs is again
constant and equal to .theta.-.gamma.. In particular, alternates
5G1 and 5G2 of FIG. 4 would both yield phase determinations in this
case of -(.theta.+.beta.+.gamma.) which would give a value for
.beta. when corrected as necessary for the known phase rotations
.theta. and .gamma.. (See for example equation 34).
The nature of the amplitude spectral differences permitted between
the design signal pair and the processing signal pair is now
greatly clarified. G(w) and F(w/s) must for all realistic Doppler
variations governed by s overlap sufficiently in frequency w so
that the product G(w)F(w/s) has a band width sufficiently broad to
correspond to a signal of finite duration and impulsive character
when transformed to the time domain with a constant zero phase
spectrum. (This requirement is analogous in some measure to
fundamental property II). Also, the product G(w) F(w/s) must have a
character essentially as demanded in fundamental Property I.
It should be apparent to anyone with some background in signal
processing that the permissible differences in the amplitude
spectra of the design signal pair and the processing pair can often
be used to great advantage. Techniques utilizing some
"conditioning" of amplitude spectra in association with a
correlation or convolution process are widely used and even
standard in the treatment of signals for detection and other
applications (see for example Phillip E. Panter, Modulation, Noise
and Spectral Analysis, McGraw Hill, 759 P, 1965). By analogy,
similar enhancement techniques can be designed to function in the
context of this invention.
As a most elementary example of such a method, one may consider a
practical environment which is attenuative in nature and
preferentially removes the high frequency content of F(w/s) as the
propagation distance to the target increases. For some
approximation to the expected target range, G(w) might conversely
give appropriate emphasis to the high frequencies so that the
product F(w/s)G(w) is again almost flat. Such a method would
improve both the resolution obtainable in the relative velocity
determination and range calculation.
Hence this embodiment of the invention endows great flexibility in
all the alternative variations making possible a number of
advantages which can arise from a judicious manipulation of the
amplitude and phase spectral character of the design signal pair
and the processing signal pair. Introduction of a time variation in
the definition of the processing signal pair might enhance
detectability through amplitude spectral "whitening" while also
compensating for the changing constant phase characteristic
introduced by a propagation medium. The scope and significance of
such possibilities can be appreciable.
* * * * *