U.S. patent application number 09/803098 was filed with the patent office on 2001-10-04 for method for improved line of sight signal detection using time/frequency analysis.
This patent application is currently assigned to Lucent Technologies, Inc.. Invention is credited to Tekinay, Sirin.
Application Number | 20010027110 09/803098 |
Document ID | / |
Family ID | 25530805 |
Filed Date | 2001-10-04 |
United States Patent
Application |
20010027110 |
Kind Code |
A1 |
Tekinay, Sirin |
October 4, 2001 |
Method for improved line of sight signal detection using
time/frequency analysis
Abstract
The time-of-arrival of the line-of-sight component of a
received, or incoming, signal is determined by performing a
time/frequency analysis of the incoming signal. The term
time/frequency analysis refers to an analysis of the frequency
components (i.e. the frequency make-up) of the incoming signal at
given instants in time. For example, one form of time/frequency
analysis according to the present invention is to compare the
frequency make-up of the received signal to the frequency make-up
of the transmitted signal. Those points in time in which the
frequency make-up of the received signal matches the frequency
make-up of the transmitted signal are the instants in time at which
a multipath component of the transmitted signal is received. In
such a time/frequency analysis, the first-identified component, in
time, is assumed to be the line-of-sight component of the
transmitted signal.
Inventors: |
Tekinay, Sirin; (Morristown,
NJ) |
Correspondence
Address: |
BIRCH STEWART KOLASCH & BIRCH
PO BOX 747
FALLS CHURCH
VA
22040-0747
US
|
Assignee: |
Lucent Technologies, Inc.
|
Family ID: |
25530805 |
Appl. No.: |
09/803098 |
Filed: |
March 12, 2001 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09803098 |
Mar 12, 2001 |
|
|
|
08984728 |
Dec 4, 1997 |
|
|
|
6272350 |
|
|
|
|
Current U.S.
Class: |
455/506 ;
455/67.16 |
Current CPC
Class: |
H04B 7/01 20130101 |
Class at
Publication: |
455/506 ;
455/67.6 |
International
Class: |
H04B 007/005 |
Claims
What is claimed:
1. A method comprising: receiving an incoming signal comprising at
least one multipath component of a signal transmitted by a wireless
terminal in an RF environment; performing a time/frequency analysis
on the incoming signal, including comparing a set of waveforms
corresponding to frequency components of the incoming signal to a
set of waveforms corresponding to expected frequency components of
the transmitted signal, to identify a time-of-arrival of a
line-of-sight component of the incoming signal; and identifying a
first instant in time at which said set of waveforms corresponding
to frequency components of the incoming signal substantially match
said set of waveforms corresponding to frequency components of the
signal transmitted by the wireless terminal, said first identified
instant in time thereby being the time-of-arrival of the
line-of-sight component of the incoming signal, wherein said set of
waveforms corresponding to frequency components of the incoming
signal substantially matches said set of waveforms corresponding to
frequency components of the signal transmitted by the wireless
terminal when a number of matching corresponding waveforms from the
two sets meets or exceeds a threshold.
2. The method of claim 1 wherein said step of performing a
time/frequency analysis includes the step of comparing a set of
frequency components of the incoming signal to a set of frequency
components of the signal transmitted by the wireless terminal.
3. The method of claim 1 further comprising identifying a first
instant in time at which said set of waveforms corresponding to
frequency components of the incoming signal substantially match
said set of waveforms corresponding to frequency components of the
signal transmitted by the wireless terminal, said first identified
instant in time thereby being the time-of-arrival of the
line-of-sight component of the incoming signal.
4. The method of claim 3 wherein said time/frequency analysis is a
wavelet analysis.
5. The method of claim 4 wherein said incoming signal is a signal
received during a regular communication of said wireless
terminal.
6. The method of claim 1 further comprising adjusting said
identified time-of-arrival of the line-of-sight component of the
incoming signal by an amount based on the value of a parameter that
characterizes the scattering hostility in an RF environment through
which said incoming signal traveled.
7. The method of claim 1 further comprising: generating the set of
waveforms corresponding to frequency components from the incoming
signal.
8. The method of claim 7, wherein the generating step includes
passing the incoming signal through a plurality of linear
filters.
9. The method of claim 7, wherein the waveforms corresponding to
frequency components of the incoming signal and the waveforms
corresponding to expected frequency components of the transmitted
signal are wavelets.
10. The method of claim 3, wherein said set of waveforms
corresponding to frequency components of the incoming signal
substantially matches said set of waveforms corresponding to
frequency components of the signal transmitted by the wireless
terminal when a number of matching corresponding waveforms from the
two sets meets or exceeds a threshold.
11. A method comprising: receiving an incoming signal comprising at
least one multipath component of a signal transmitted by a wireless
terminal through an RF environment; decomposing the incoming signal
into a set of decomposed signals, which when superposed constitute
the incoming signal; comparing the set of decomposed signals to a
set of expected decomposed signals at instants in time; and
identifying a time-of-arrival of a line-of-sight component of the
incoming signal from the comparison, wherein the identifying step
includes determining a first instant in time at which the set of
decomposed signals substantially match the set of expected
decomposed signals, the first instant in time being the
time-of-arrival of the line-of-sight component of the incoming
signal, wherein the set of decomposed signals substantially match
the set of expected decomposed signals when a number of matching
corresponding decomposed signals from the two sets meets or exceeds
a threshold.
12. The method of claim 11 further comprising adjusting said
identified time-of-arrival of the line-of-sight component of the
incoming signal by an amount based on the value of a parameter that
characterizes the scattering hostility in the RF environment
through which said incoming signal traveled.
13. The method of claim 11, wherein the decomposing step includes
passing the incoming signal through a plurality of linear
filters.
14. The method of claim 11, wherein the set of decomposed signals
is a wavelet representation of the incoming signal.
15. The method of claim 11, wherein the set of expected decomposed
signals is a wavelet representation of the signal transmitted by
the wireless terminal.
16. A method comprising: receiving an incoming signal comprising at
least one multipath component of a signal transmitted by a wireless
terminal through an RF environment; decomposing the incoming signal
into a set of decomposed signals, which when superposed constitute
the incoming signal; comparing the set of decomposed signals to a
set of expected decomposed signals at instants in time; and
identifying a time-of-arrival of a line-of-sight component of the
incoming signal from the comparison, wherein the identifying step
includes determining a first instant in time at which the set of
decomposed signals substantially match the set of expected
decomposed signals, the first instant in time being the
time-of-arrival of the line-of-sight component of the incoming
signal, wherein the set of decomposed signals substantially match
the set of expected decomposed signals when a number of matching
corresponding decomposed signals from the two sets meets or exceeds
a threshold.
17. A method for detecting a line of sight signal comprising:
receiving one or more incoming multi-path components of a signal
transmitted by a wireless terminal through a rich-scattering RF
environment; identifying the time of arrival of a first arriving
one of the one or more multi path components using time frequency
analysis; and adjusting the identified time of arrival of the first
arriving multi path components to obtain a time of arrival
associated with a line of sight component.
18. The method of claim 17, further comprising the step of
processing a plurality of adjusted times of arrival associated with
a plurality of first arriving multi path components to determine a
geolocation of the wireless terminal.
19. The method of claim 17 wherein the time frequency analysis
includes wavelet analysis.
Description
RELATED APPLICATIONS
[0001] This invention is related to the invention disclosed in the
applicant's co-pending application Ser. No. 08/984,779, entitled
"Method For Frequency Environment Modeling and Characterization,"
filed on Dec. 4, 1997; Ser. No. 08/985,133, entitled "Detecting The
Geographical Location Of Wireless Units," filed on Dec. 4, 1997;
and Ser. No. 08/984,780, entitled "Method For Improved
Line-Of-Sight Signal Detection Using RF Model Parameters," filed on
Dec. 4, 1997.
FIELD OF THE INVENTION
[0002] The present invention relates to the field of signal
processing, and more particularly to a method for detecting
line-of-sight signals.
BACKGROUND OF THE INVENTION
[0003] In wireless communications systems, wireless terminals and
base stations are designed to transmit and receive radio frequency
(RF) signals that propagate through RF environments. Depending on
the type of wireless communications system and the services offered
by the wireless system, the wireless terminals and base stations
are equipped to perform specific signal-processing functions. For
example, some wireless systems are required to identify the
geographical location (i.e. geolocation) of the wireless terminals
communicating on the system. Such wireless systems have been
referred to as geolocation systems. The term geolocation as used
herein refers to the point in two or three-dimensional space
defined by a set of coordinates, e.g. longitude and latitude,
and/or defined by a vector, i.e. distance and direction, from a
known point in space.
[0004] Some conventional geolocation systems identify the
geolocation of a wireless terminal by determining the
time-of-arrival of the line-of-sight component of a signal
transmitted by the wireless terminal. The line-of-sight component
of the transmitted signal is that component of the signal that
propagated directly from the wireless terminal to the location at
which the signal was received (e.g. base station) without
scattering or reflecting off structures in the RF environment. The
term scattering refers to the phenomenon wherein an RF signal,
traveling in an RF environment, hits and reflects of structures in
the RF environment, thereby causing the RF signal to take random
paths through the RF environment. This so-called multipath
phenomenon can cause the incoming signal to be composed of several
repeated versions of the transmitted signal, each version being a
multipath component of the incoming signal.
[0005] To determine the time-of-arrival of the line-of-sight
component of the transmitted signal, such geolocation systems
receive the transmitted signal and pass the so-called incoming
signal through a matched filter. The matched filter generates a
correlation value based on a comparison of the shape of the
waveform of the incoming signal to the shape of the waveform of the
transmitted signal. The correlation value essentially peaks each
time the matched filter determines that the shape of the waveform
of the incoming signal is similar to or matches the shape of the
waveform of the transmitted signal. Each time the correlation value
reaches a peak value, the geolocation system identifies that time
as the time-of-arrival of a multipath component of the incoming
signal. As a result, since the line-of-sight component of the
transmitted signal travels directly to the location of the
receiving unit, such conventional geolocation systems assume that
the time-of-arrival of the line-of-sight component of the incoming
signal is the time at which the correlation value reaches its first
peak.
[0006] It is the time-of-arrival of the line-of-sight component of
the incoming signal that the above-described conventional
geolocation systems use to identify the geolocation of the wireless
terminal. In particular, the geolocation systems identifies the
time-of-arrival of the line-of-sight component of the incoming
signal received at a plurality of locations, and processes the
various times of arrival to determine the distance of the wireless
terminal from each of, for example, three receiver locations. From
this "distance" information, the geolocation system determines the
geolocation of the wireless terminal itself.
[0007] These geolocation systems, however, are hindered by their
failure to consider the effects of scattering in determining the
times-of-arrival of the line-of-sight components. As described
above, scattering may cause a signal to multipath and thus cause
the incoming signal to be composed of a plurality of multipath
components that arrive at different times. Depending on the amount
of time between the respective multipath components of the incoming
signal, the matched filter of the conventional geolocation system
may not be able to accurately identify the time-of-arrival of the
line-of-sight component. That is, the matched filter may not be
able to distinguish between the time of arrival of the
line-of-sight component and the next-arriving component of the
incoming signal. When this happens, the matched filter may
mistakenly view the line-of-sight component and the next-arriving
component as a single multipath component. As a result, the
correlator value will reach a peak somewhere in between the actual
time-of-arrival of the line-of-sight component and the
time-of-arrival of the next-arriving multipath component. This will
cause the geolocation system to mistakenly assume that the
time-ofarrival of the line-of-sight component was received at a
later, or "time-shifted," time than it actually arrived. Such a
mistake could substantially reduce the accuracy of the processing
performed by the geolocation system in calculating the geolocation
of the wireless terminal.
SUMMARY OF THE INVENTION
[0008] According to the principles of the present invention, the
time-of-arrival of the line-of-sight component of a received signal
is identified with more accuracy than that obtained by merely
passing the received signal through a matched filter and
identifying the time at which a correlation value reaches its first
peak. Instead, the time-of arrival of the line-of-sight component
of the received signal is determined by performing a time/frequency
analysis of the incoming signal to identify the instants in time at
which the frequency components of the incoming signal are similar
to the frequency components of the transmitted signal. The
time/frequency analysis reduces the so-called time-shift of the
identified time-of-arrival of each component of the incoming
signal, and thus increases the accuracy of the identified
time-of-arrival of the line-of-sight component of the incoming
signal.
[0009] The term time/frequency analysis as used herein refers to an
analysis of the frequency components (i.e. the frequency make-up)
of a signal at given instants in time. For example, one form of
time/frequency analysis according to the present invention is to
compare the frequency make-up of the received signal to the
frequency make-up of the transmitted signal. Those points in time
in which the frequency make-up of the received signal matches the
frequency make-up of the transmitted signal are the instants in
time at which a multipath component of the transmitted signal is
received.
[0010] In accordance with a feature of the invention, the
time/frequency analysis is performed using wavelets. A wavelet is a
waveform that is localized in time. That is, a wavelet lasts for
only a few cycles. Analyzing a signal using wavelets (i.e. wavelet
analysis) is similar to analyzing a signal using Fourier analysis.
In particular, wavelet analysis involves using an algorithm to
decompose a signal into a family of wavelets called the wavelet
representation of the signal. It is the wavelet representation that
can be used, in accordance with the principles of the present
invention, to identify the times-of-arrival of the multipath
components of the incoming signal.
[0011] In accordance with another feature of the invention, the
accuracy of a determined time-of-arrival is improved by adjusting
the determined time-of-arrival to reduce the inaccuracies due to
the effects of scattering on the incoming signal. For example, the
determined time-of-arrival can be adjusted according to the value
of a parameter that characterizes the scattering hostility in the
RF environment through which the incoming signal traveled. The
resultant "adjusted" time-of-arrival thereby provides a measure of
the time at which the line-of-sight component would have arrived at
the receiving unit if the RF environment were scatter-free.
[0012] Advantageously, by accurately determining the time of
arrival of the line of sight component of a received signal
according to the present invention at a plurality of locations, the
geolocation of the wireless terminal that transmitted the incoming
signal can be determined with increased accuracy over that
determined by the prior art.
[0013] In particular embodiments, the determination of the
geolocation of the wireless terminal may be advantageously
carried-out using the methods disclosed in my co-pending
application Ser. No. 08/985,133, entitled "Detecting The
Geographical Location Of Wireless Units," filed on even date
herewith.
[0014] In addition, in particular embodiments, an identified
time-of-arrival of the line-of-sight component of the incoming
signal can be adjusted using the methods disclosed in my co-pending
application, Ser. No. 08/984,780, entitled "Method For Improved
Line-Of-Sight Signal Detection Using RF Model Parameters," filed on
even date herewith.
[0015] Also, in particular embodiments, parameters used to adjust
the identified times-of-arrival according to the present invention
can be determined using the methods disclosed in my co-pending
application Ser. No. 08/984,779, entitled "Method For Frequency
Environment Modeling and Characterization," filed on even date
herewith.
BRIEF DESCRIPTION OF THE DRAWING
[0016] FIG. 1 is block diagram of an illustrative embodiment of a
method for determining the geolocation of a wireless terminal.
[0017] FIG. 2 is a top-view of the paths of the line-of-sight and
multipath components of a signal traveling from a wireless terminal
to a plurality of receiver locations in an RF environment.
[0018] FIG. 3 is a graphical view of the various components of an
illustrative incoming signal received at a receiver location shown
in FIG. 2.
[0019] FIG. 4 is a graphical view of the various components of an
illustrative incoming signal received at another receiver location
shown in FIG. 2.
[0020] FIG. 5 is a graphical view of a wavelet representation of a
sawtooth signal helpful in explaining how, in preferred
embodiments, the above-described time shift is determined.
[0021] FIG. 6 is a block diagram of a set of linear filters that
can be used to decompose an incoming signal into a set of frequency
components in order to identify the time-of-arrival of a component
of the incoming signal.
[0022] FIG. 7 is a block diagram of an illustrative embodiment of a
method for using wavelet analysis as a form of time/frequency
analysis, in accordance with the principles of the present
invention, to identify the line-of-sight component of an incoming
signal.
[0023] FIG. 8 is block diagram of an illustrative embodiment of a
method for determining whether a wavelet representation of an
incoming signal matches a wavelet representation of a transmitted
signal.
[0024] FIG. 9 is a block diagram of an illustrative embodiment of
an apparatus for adjusting the identified time-of-arrival of an
incoming signal.
[0025] FIG. 10 is a block diagram of an illustrative embodiment of
a method for determining a basis parameter used to adjust an
identified time-of-arrival of an incoming signal.
[0026] FIG. 11 is a block diagram of an illustrative embodiment of
a method for forming an RF model of an RF environment.
[0027] FIG. 12 is a block diagram of an illustrative embodiment of
a method for determining the geolocation of a wireless terminal
based on adjusted times-of-arrival
DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS OF THE
INVENTION
[0028] Referring now to FIG. 1 there is shown a method 10 for
determining the geolocation of a wireless terminal. As shown,
method 10 begins with step 11 wherein a signal transmitted by the
wireless terminal is received at a plurality of receiver locations.
The signal received at each receiver location is referred to herein
as the incoming signal. A time/frequency analysis, in accordance
with the principles of the present invention, is performed on each
incoming signal, at step 12, to identify the time-of-arrival of the
first-arriving multipath component of the respective incoming
signal at each receiver location. The identified time-of-arrival of
the first-arriving multipath component at each location is
adjusted, at step 13, by an amount based on the value of at least
one parameter of an RF model that characterizes the scattering
hostility of the RF environment in which the respective incoming
signal traveled. The resultant "adjusted" time-of-arrival of the
first-arriving component of each incoming signal more accurately
reflects the time the line-of-sight component of the incoming
signal would have reached the respective receiver locations if the
RF environment were scatter-free. The various "adjusted"
times-of-arrival are then processed, at step 14, to determine the
geolocation of the wireless terminal.
[0029] In order to more fully explain the operation of method 10,
the following topics will be addressed: (1) the incoming signal;
(2) the line-of-sight and multipath components of the incoming
signal; (3) the "time-shift" of the line-of-sight component of the
incoming signal due to scattering, (4) time/frequency analysis of
the incoming signal in accordance with the principles of the
present invention, (5) identifying the time-of-arrival of the
line-of-sight component of the incoming signal in accordance with
the principles of the present invention; (6) using an RF model to
reduce a "time-shift" of the identified time-of-arrival of the
line-of-sight component of the incoming signal, due to scattering;
(7) determining parameters that form the RF model used to reduce
the time shift; and (8) using the time-of-arrival of the
line-of-sight component of the incoming signal received at a
plurality of locations to determine the geolocation of the wireless
terminal.
[0030] The Incoming Signal
[0031] The incoming signal recited in method 10 can be any signal
transmitted by the wireless terminal including those signals
transmitted during regular communications. The term regular
communications as used herein refers to the communications the
wireless terminal regularly performs when in operation, for
example, communications involving a call between the user of a cell
phone and another party.
[0032] Line-of-sight and Multipath Components
[0033] As described above, the incoming signal may be composed of
any number of multipath components of the transmitted signal,
depending on the scattering hostility in the RF environment in
which a signal travels. An illustration of the physical process by
which various multipath components arrive at different receiver
locations in an RF environment 20 is shown in FIG. 2. As shown, a
wireless terminal 21, buildings 24-27 and receiver locations 22 and
23 are all located at different positions in RF environment 20.
Depending on the position of wireless terminal 21 with respect to
buildings 24-27 and receiver locations 22 and 23, a signal
transmitted from wireless terminal 21 travels a plurality of paths
28-32 therefrom. Specifically, a signal transmitted from wireless
terminal 21 travels to receiver location 22 along line-of-sight
multipath, or line-of sight path, 30 and multipath 29, and to
receiver location 23 along multipaths 28 and 32.
[0034] Time-shift Due to Scattering
[0035] The incoming signal received at receiver location 22 has
both a line-of-sight multipath component (i.e. the component that
traveled along line-of-sight path 30) and a multipath component
(i.e. the component that traveled along multipath 29). Since the
multipath component travels a longer distance than the
line-of-sight component, the time-of-arrival of the multipath
component is later than the time-of-arrival of the line-of-sight
component.
[0036] Referring now to FIG. 3 there is shown a graphical view of
the line-of sight component S.sub.30 and the multipath component
S.sub.32 of an incoming signal received at receiver location 22,
where the incoming signal is shown in FIG. 3 as signal 36. As
shown, the line-of sight component S.sub.30 of signal 36 has a
time-of-arrival 34 and the multipath component S.sub.32 of signal
36 has a time-of-arrival 35. As stated above, if time-of-arrival 34
and time-of-arrival 35 are very close in time, conventional
geolocation devices that use matched filters to determine the
time-of-arrival of the line-of-sight component may inaccurately
determine that the line-of-sight component S.sub.30 of incoming
signal 36 arrived at time 31, which is .DELTA.t later than it
actually arrived. When this happens, the identified time-of-arrival
(i.e. time 31) of the line-of-sight component S.sub.30 is said to
be "time-shifted" by the time shift .DELTA.t due to scattering
and/or multipathing in the RF environment
[0037] The time shift .DELTA.t due to such close-arriving
multipaths can be reduced by performing a time/frequency analysis
on the incoming signal, as described below.
[0038] Referring now back to FIG. 2, the incoming signal received
at receiver location 23 has two multipath components, i.e., the
components that traveled along multipaths 28 and 32. However, it
has no line-of-sight component because line-of-sight path 31 is
blocked by building 27. Since such multipath components travel a
farther distance in RF environment 20 than would the line-of-sight
component if it had not been blocked by building 27, the
time-of-arrival of each multipath component is therefore, by
definition, later than the time at which the line-of-sight
component would have arrived at receiver location 23.
[0039] Referring now to FIG. 4 there is shown a graphical view of
an incoming signal 33 received at receiver location 23. As shown,
incoming signal 33 includes a multipath component S.sub.38 that
traveled along path 28, a multipath component S.sub.39 that
traveled along path 32, and an expected line-of-sight component
S.sub.3, that never reaches location 23 because it is blocked by
building 27. The multipath components S.sub.38 and S.sub.39 are
shown to have times-of-arrival 38 and 39, respectively, and line-of
sight component S.sub.37 is shown to have an expected
time-of-arrival 37 Since a conventional geolocation system, as
described above, would assume that the line-of-sight component of
incoming signal 33 is the first-arriving component, such a
geolocation system would incorrectly assume that the
time-of-arrival of the line-of-sight component of incoming signal
33 is time 38, as opposed to time 37. When this happens, the
geolocation system would incorrectly assume that the
time-of-arrival of the line-of-sight component S.sub.37 was a
time-shift .DELTA.t, i.e. time difference between time 37 and time
38, later than it should have arrived. As described above, such a
time-shift .DELTA.t would reduce the accuracy of the geolocation
system in determining the geolocation of wireless terminal 21.
[0040] The time shift .DELTA.t, due to such a blocked line-of-sight
path, can be reduced by adjusting the identified time-of-arrival of
the first-arriving component of the incoming signal by an amount
based on a parameter that characterizes the scattering hostility of
the RF environment in which the incoming signal traveled.
[0041] Time/frequency Analysis
[0042] The term time/frequency analysis as used herein refers to an
analysis of the frequency components of a signal (e.g. the
magnitude of the waveform at each component frequency of the
signal) at given instants in time. For example, one form of
time/frequency analysis, in accordance with the principles of the
present invention, is to compare the frequency components of the
incoming signal to the frequency components of the transmitted
signal at given instants in time. (Although the exact nature of the
transmitted signal is not typically known, its frequency components
can indeed be known to a great extent since they are a function of
the carrier frequency and the modulation used) In such a
time/frequency analysis, those instants in time in which the
frequency components of the incoming signal match the frequency
components of the transmitted signal are the instants in time at
which a multipath component is received.
[0043] One illustrative method for performing a time/frequency
analysis on the incoming signal, in accordance with the principles
of the present invention, is to perform a so-called wavelet
analysis on the incoming signal. Wavelet analysis involves the act
of breaking down a signal into a set of simpler elements, called
wavelets. The wavelets are basically localized waveforms that last
for only a few cycles. Thus, according to wavelet analysis, a
wavelet representation of a signal is the set of wavelets that can
be superposed to form the waveform of the signal.
[0044] Wavelet analysis can be explained by analogy to Fourier
analysis. A Fourier transform represents a signal as a
superposition of sinusoids with different frequencies, and the
Fourier coefficients represent the contribution of the sinusoid at
these frequencies Similarly, a wavelet transform represents a
signal as a sum of wavelets with different widths, called
dilations, and amplitudes, called scalings, and the wavelet
coefficients provide a measure of the contributions of each wavelet
at these dilations and scalings.
[0045] For example, referring now to FIG. 5 there is shown a
sawtooth signal 52 that is represented by a family of wavelets, or
wavelet representation, 51. As shown, wavelet representation 51
includes 16 different wavelets, each having a different dilation
and scaling. For example, wavelet 53 has a dilation 54 and a
scaling 55, and wavelet 56 has a dilation 57 and a scaling 58.
Wavelet representation 51 is referred to as the wavelet transform
of sawtooth signal 52, and sawtooth signal 52 is referred to as the
inverse transform of the wavelet representation 51.
[0046] Both the wavelet transform and the inverse transform are
arrived at according to known algorithms. For example, one
illustrative algorithm used to compute both the wavelet transform
and the inverse transform of a waveform is the fast pyramid
algorithm described by A. Bruce, D. Donoho and H. Y Goo, in
"Wavelet Analysis," IEEE Spectrum, October 1996, and incorporated
herein by reference. The fast pyramid algorithm provides a "forward
algorithm" that serves to compute the wavelet transform, and a
"backward algorithm" that serves to compute the inverse transform.
The forward algorithm uses a series of linear filters to decompose
a signal into a set of filtered components. It is the waveforms of
these filtered components that form the wavelet representation of
the signal. The reverse algorithm uses a set of linear filters to
combine the wavelets comprising the wavelet representation to form
the signal.
[0047] Referring now to FIG. 6 there is shown a block diagram of an
illustrative set of linear filters 600 used to decompose a signal
601 into a set of filtered components 615622. As shown, the set of
linear filters 600 has a first-line filter 602 which is connected
to second-line filters 603 and 604. Second line filter 603 is
connected to third-line filters 605 and 606, and second-line filter
604 is connected to third-line filters 607 and 608.
[0048] In operation, signal 601 is input into first-line filter 602
which decomposes signal 601 into a high-frequency filtered
component 609 and a low-frequency filtered component 610.
Second-line filter 503 decomposes high-frequency component 609 into
high frequency filtered components 611 and 612, and second-line
filter 604 decomposes low frequency filtered component 610 into
low-frequency filtered components 613 and 614. Third-line filter
605 decomposes high frequency filtered component 611 into high
frequency filtered components 615 and 616. Third-line filter 606
decomposes high-frequency filtered component 612 into high
frequency filtered components 617 and 618 Third-line filter 607
decomposes low-frequency filtered component 613 into low-frequency
filtered components 619 and 620. Third-line filter 608 decomposes
low-frequency filtered component 614 into low-frequency filtered
components 621 and 622.
[0049] The waveforms of the set of filtered components 615-622 that
result by inputting signal 601 into the set of linear filters 600
is shown. The waveforms of frequency components 615-622 are a set
of waveforms that can be superposed to form signal 601 As a result,
it is the waveforms of frequency components 615-622 that form a
wavelet representation of signal 601. Alternatively, a different
set of linear filters may be chosen to decompose signal 601 into a
different set of frequency components that provide a different set
of waveforms, or wavelets. Thus, it can be understood that
different sets of linear filters can decompose signal 601 into
different wavelet representations.
[0050] Identifying the Time-of-arrival
[0051] The above-described time/frequency analysis can be used, in
accordance with the principles of the present invention, to
identify the time-of-arrival of the line-of-sight component of an
incoming signal. For example, wavelet analysis can be used as a
time/frequency analysis to identify the time-of-arrival of the
line-of sight component of an incoming signal. Referring now to
FIG. 7 there is shown one illustrative embodiment of a method 70
for using wavelet analysis for identifying the time-of-arrival of
the line-of-sight component of an incoming signal in accordance
with the principles of the present invention.
[0052] As shown, method 70 begins at step 71 wherein the incoming
signal is passed through a set a set of linear filters to obtain a
wavelet representation of the incoming signal. The wavelet
representation of the incoming signal identified at each instant in
time is then compared, at step 72, to a wavelet representation of a
transmitted signal. As stated above, although the exact nature of
the transmitted signal is not known, the waveform of its frequency
components can indeed be known to a great extent since they are a
function of the carrier frequency and the type of modulation used.
Thus, since the wavelet representation of a signal is dependent on
the waveform of its frequency components, the wavelet
representation of the transmitted signal can be estimated with
great accuracy.
[0053] The instants in time wherein the wavelet representation of
the incoming signal is substantially similar to, or matches, the
wavelet representation of the transmitted signal (called the
expected wavelet representation) are identified, at step 73, as the
times-of-arrival of the multipath components of the incoming
signal. Since the line-of-sight component of the incoming signal
travels the shortest distance, the first-arriving, or earliest,
time-of-arrival is identified, at step 74, as the time-of-arrival
of the line-of-sight component of the incoming signal. As stated
above, determining the time-of-arrival of the line-of-sight
component using such a time/frequency analysis, reduces the
inaccuracies associated with using a matched filter. Thus, method
70 increases the accuracy of the identified time-of-arrival of the
line-of-sight component of the incoming signal over the prior art
methods that utilize such a so-called matched filter method.
[0054] An illustrative embodiment of a method 80 for determining
whether the wavelet representation of the incoming signal matches
the expected wavelet representation of the transmitted signal is
shown in FIG. 8. As, shown, method 80 begins at step 81 wherein the
dilations and scalings of the wavelets which represent the incoming
signal are compared to the dilations and scalings of the
"corresponding wavelets" which represent the transmitted signal.
The term "corresponding wavelets" as used herein refers to the
wavelet of the incoming signal and the wavelet of the transmitted
signal that represent the waveform of the frequency band of its
respective signal. At step 82, the corresponding wavelets that have
dilations and scalings that are within some tolerance of each other
are identified, and labeled as matching wavelets. The number of
matching wavelets is then compared to a threshold number, at step
83. If the number of matching wavelets is equal to or greater than
the threshold number then, at step 84, the wavelet representation
of the incoming signal is said to match the wavelet representation
of the transmitted signal. If, however, the number of matching
wavelets is less than the threshold number then, at step 85, the
wavelet representation of the incoming signal is said to not match
the wavelet representation of the transmitted signal. As stated
above, each instant at which a match is identified can therefor be
identified as a time-of-arrival of a multipath component of the
incoming signal
[0055] Using an RF Model to Adjust an Identified
Time-of-arrival
[0056] As described above and shown in FIGS. 2 and 4, the
identified time-of-arrival of the line-of-sight component of the
incoming signal can be "time-shifted" when the line-of-sight
component is blocked (i.e. prevented from arriving at the receiver
location).
[0057] The present inventor has found that the amount of time-shift
due to such a blocked line-of-sight path directly depends on the
scattering hostility of the RF environment in which the incoming
signal traveled. In particular, the present inventor has found that
the amount of time-shift due to such a blocked line-of-sight path
can be reduced by adjusting the identified time-of arrival of the
first-arriving component of the incoming signal by an amount based
on a parameter that characterizes the scattering hostility of the
RF environment in which the incoming signal traveled.
[0058] Referring now to FIG. 9 there is shown one illustrative
embodiment of a device 90 for adjusting the identified
time-of-arrival of the line-of-sight component of an incoming
signal. As shown, device 90 has an RF model 91 connected to a
processor 92. RF model 91 has a set of parameters, each parameter
characterizing the scattering hostility of a given region of an RF
environment. Processor 92 has inputs 93 and 94, and output 95.
[0059] In operation, processor 92 obtains, at input 93, the
time-of-arrival identified for the first-arriving component of an
incoming signal that traveled through a given region of an RF
environment. Processor 92 obtains, at input 94, a parameter that
characterizes the scattering hostility in the given region of the
RF environment, from RF model 91. The given region through which
the incoming signal traveled is determined by identifying the
direction from which the incoming signal traveled and/or the
strength of the incoming signal. The methods by which such signal
direction and signal strength are determined are well known in the
art. Based on the obtained parameter, processor 92 determines the
amount of time-shift of the obtained time-of-arrival that is due to
scattering and/or multipath in the RF environment. Then, based on
the time-shift information, processor 92 computes and outputs from
output 95 an adjusted time-of-arrival that more accurately reflects
the time at which the line-of-sight component would have arrived in
a scatter-free environment.
[0060] As stated above, each parameter is a measure of the amount
of multipathing, and the amount of "time-shift" that a given signal
would incur in a given region of the RF environment. As a result,
each parameter of RF model 91 indicates the relative amount of
"time-shift" that would occur if a signal were to travel in the
respective region. It can therefore be understood that a parameter
of RF model 91 that characterizes a given region as having a
greater scattering hostility than another region, necessarily
indicates that the amount of "time-shift" that would occur if the
incoming signal were to travel in that given region is greater than
the amount of time shift that would occur in the other region.
[0061] Processor 92 can determine the amount of such a time shift
that would occur in a given region by comparing the value of a
parameter that characterizes the scattering hostility of the given
region to the value of a basis parameter. The term basis parameter
as used herein refers to a parameter that characterizes the
scattering hostility in a given region wherein the amount of
time-shift that results when a signal travels in that given region
is known. Thus, a basis parameter is a parameter that has a known
associated time-shift.
[0062] Referring now to FIG. 10, there is shown an illustrative
embodiment of a method 100 for determining a basis parameter. As
shown, method 100 begins at step 101 wherein a signal is
transmitted in a region characterized by the basis parameter. The
time at which the signal is received at a known distance from the
wireless terminal is determined at step 102. The difference between
the actual time-of-flight (the time it took for the first-arriving
component of the signal to actually travel the given distance) and
the expected time-of-flight (i.e. the time the signal would have
traveled the given distance along the line-of-sight path) is
determined at step 103. The calculated difference is thereby the
known time-shift associated with the basis parameter.
[0063] Referring now back to FIG. 9, processor 92 determines the
amount of time shift of the time-of-arrival obtained at input 93 by
comparing the parameter obtained at input 94 to a so-called basis
parameter, and adjusting the value of the known time-shift
associated with the basis parameter by an amount based on the
difference between the value of the basis parameter and the value
of the given parameter. For example, the known time-shift
associated with the basis parameter can be adjusted in direct
proportion to the difference between the value of the basis
parameter and the value of the given parameter. The resultant
adjusted time shift is the time shift of the identified time-of
arrival of the first arriving component of the incoming signal, due
to such a line-of-sight path, as described above. Processor 92 can
then adjust the identified time of arrival, obtained at input 93,
by the time-shift, and output the adjusted time-of-arrival through
output 95.
[0064] Determining the Parameters
[0065] The term parameter as used herein refers to any parameter or
so-called dimension that is capable of defining or describing a
chaotic process or system in terms of a measure of some aspect of
that system. One type of parameter or dimension that can be used to
define such a chaotic process is a fractal dimension. Fractal
dimensions are described by A. P Pentland in "Fractal-Based
Description of Natural Scenes," IEEE Transactions on Pattern
Analysis and Machine Intelligence, Vol. PAMI-6, No. 6, November
1984, and incorporated herein by reference.
[0066] As noted above, a fractal dimension is a parameter that
defines a chaotic system by characterizing the system in terms of a
measure of so-called self-similarity. For example, a fractal
dimension has been used to define the shape of a mountainous
landscape by characterizing the amount of self-similarity that
exists in the shape of the landscape. The amount of self-similarity
in the shape of the landscape is the number of times a particular
shape is repeated in the shape of the landscape itself. The
particular shape is the largest shape found in the actual landscape
that can be used to define or re-create each piece of the actual
landscape. As a result, the fractal dimension determined for any
given landscape provides a measure of the size of the particular
repeated shape with respect to the size of the landscape itself,
and thus a characterization of the landscape itself.
[0067] Just like a fractal dimension can be used as parameter to
characterize the a chaotic system such as a mountainous landscape,
so can a fractal dimension be used as a parameter to characterize a
chaotic system such as an RF environment. In particular, a fractal
dimension can be used to characterize the scattering hostility of
an RF environment by providing a measure of the number of times a
similar shape (i.e. the various multipath components of the
incoming signal) is repeated in the waveform of the incoming signal
that traveled in the RF environment. By providing a measure of the
number of multipath components of the incoming signal, the fractal
dimension actually provides a measure or characterization of the
scattering hostility of the RF environment. Thus, a set of such
fractal dimensions forms an RF model of the RF environment.
[0068] Referring now to FIG. 11 there is shown an illustrative
method 110 for forming such an RF model of an RF environment. As
shown, method 110 begins at step 111 wherein an RF signal having a
given waveform is transmitted at a known time from a given region
of an RF environment. The RF signal, after traveling a given
distance in the given region of the RF environment is received,
step 112. The received signal, or so-called incoming signal, is
analyzed, step 113, to determine a parameter that characterizes the
scattering hostility of that region of the RF environment. Then, at
decision step 114, method 110 checks whether a parameter has been
determined for each region of the RF environment. If such a
parameter has been determined for each region, method 1100 ends,
otherwise steps 111-113 are repeated until such time that each
region has a parameter determined therefor.
[0069] It should be understood that no region of the RF environment
is limited to being characterized by one such parameter. Rather
each region can be characterized, for example, by a plurality of
parameters, or a single parameter that is an average of a plurality
of parameters. In addition an RF environment is not limited to a
specific number of regions. The RF environment, for example, can be
a single region or one hundred regions.
[0070] Moreover, each parameter of a given RF model is not limited
to characterizing a region of any particular size. For example, a
given RF model can be compose of a set of parameters wherein each
parameter characterizes the scattering hostility in a region having
the same size and shape as every other region. Or, for example,
each parameter can characterize a region having an arbitrary size
and shape. Or, for example, each parameter characterizes a region
having size and shape based on some criterion such as the physical
profile (i.e. rural, urban, suburban, etc.) of the region.
[0071] In addition, a set of such parameters that characterize any
given region can be determined as a function of time. That is, each
determined parameter may be a time-varying function of the
scattering hostility of a given region of the RF environment.
[0072] Advantageously, a set of such parameters can be used as an
RF model of the RF environment to aid in the design of a wireless
communication system. For example, since each parameter defines the
amount of multipath that a given signal would incur if the signal
were to travel in a region of the RF environment, each parameter
can be used to predict the amount a signal would multipath if the
signal were to propagate in that respective region. Based on the
prediction, a system designer could estimate the amount the
waveform of the RF signal would change shape as a result of
traveling in the given region, and thus could determine whether a
given receiver would be capable of detecting and/or recognizing the
transmitted signal after traveling in the RF environment. Such
information, as described above, may be critical in testing the
design of a wireless system before incurring the cost of building
the system itself.
[0073] In addition, a set of such parameters can advantageously be
used to adjust the identified time of arrival of the line-of-sight
component of the incoming signals received at a plurality of
locations in an RF environment, and thus provide a geolocation
system with more-accurate time-of-arrival information for
determining the geolocation of a wireless terminal operating in the
RF environment.
[0074] Determining the Geolocation of a Wireless Terminal
[0075] Referring now to FIG. 12, there is shown a method 120 for
determining the geolocation of a wireless terminal. As shown,
method 120 begins at step 121 wherein the adjusted time-of-arrival
of the line-of-sight component is identified for the incoming
signal received at a plurality of receiver locations. Then, step
122, the various times-of-arrival are processed to determine the
distance of the wireless terminal from at least three receiver
locations. From this distance information, the geolocation of the
wireless terminal is identified, step 123.
[0076] The processing performed to determine the geolocation of a
wireless terminal, based on the time-of-arrival of the
line-of-sight component of the incoming signal received at the at
least three receiver locations, is well-known in the art. For
example, one illustrative method for using such times-of-arrival
information to determining the geolocation of a vehicle is
disclosed by J. Brooks Chadwick and J. L. Bricker in "A Vehicle
Location Solution Approach," IEEE Position Location and Navigation
Symposium, 1990, and incorporated herein by reference.
[0077] It should be noted that using an adjusted time-of-arrival of
the line-of-sight component of the various incoming signals, as
computed above, increases the accuracy of the just-described
processing for determining the geolocation of the wireless terminal
This is due to the direct dependence of the accuracy of the
geolocation calculation on the accuracy of time-of-arrival of the
line-of-sight component of the incoming signal. Advantageously,
determining the geolocation of a wireless unit as described above,
does not require the consumption of additional bandwidth, or the
increased in cost associated with adding hardware to the wireless
terminal, as in some of the prior art solutions.
[0078] While the invention has been particularly shown and
described with reference to various embodiments, it will be
recognized by those skilled in the art that modifications and
changes may be made to the present invention without departing from
the spirit and scope thereof. As a result, the invention in its
broader aspects is not limited to specific details shown and
described herein. Various modifications may be made without
departing from the spirit or scope of the general inventive concept
as defined by the appended claims.
* * * * *