Note: Descriptions are shown in the official language in which they were submitted.
CA 02497363 2005-02-17
x
r
a
METHOD FOR THE HIGH ACCURACY GEOLOCATION OF OUTDOOR
MOBILE EMITTERS OF CDMA CELLULAR SYSTEMS
FIELD OF THE INVENTION
The present invention relates generally to a high-accuracy method for the
geolocation, without a collaboration of a network, of outdoor mobile emitters
of
CDMA cellular systems, based on an ability to distinguish between line-of
sight and
reflected signals.
BACKGROUND OF THE INVENTION
In all forms of geolocation there currently exist no techniques to determine
if a
first-to-arrive signal reaching an interceptor employed in the geolocation of
a mobile
of a CDMA cellular system is a line-of sight signal or a reflected signal.
Although
some mitigation of the presence of reflected signals is possible by using
techniques
such as spatial filtering or other sophisticated signal processing techniques,
no
technique exists at present to determine if the first-to-arnve signal is a
line-of sight
signal or a reflected signal. This causes a substantial deterioration of any
geolocation
results, as it is impossible to determine if the location is calculated from
valid, line-of
sight signals or from erroneous data originating from reflected signals.
For example, in 'CDMA Infrastructure-Based Location Finding for E911', J.
O'Connor, B. Alexander and E. Schorman, 1999 IEEE 49'h Vehicular Technology
Conference, vol.3., p. 1973-1978, a geolocation method is proposed where the
collaboration of the mobile and of the infrastructure is assumed. In that
technique, no
attempt is made to distinguish if the signal being processed is a line-of
sight signal or
a reflected signal. Similarly, in 'Performance Analysis of ESPRIT, TLS-ESPRIT
and
UNITARY-ESPRIT Algorithms for DOA Estimation in a W-CDMA Mobile System,
K. AlMidfa, G.V. Tsoulos and A. Nix, First International Conference on 3G
Mobile
Communication Technologies, Conference Publication No. 471, 2000, p. 2000-
2003,
various signal processing techniques are evaluated. However, no attempt is
made to
distinguish if the signals being processed are line-of sight signals or
reflected signals.
1
i
4 t
r
CA 02497363 2005-02-17
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a high-accuracy method for
the geolocation, without a collaboration of a network, of outdoor mobile
emitters of a
CDMA cellular system, based on the ability to distinguishing between line-of
sight
and reflected signals. According to one aspect of the invention, it provides a
method
for the outdoor geolocation of a mobile of interest in a CDMA cellular system
comprising steps of: (i) dynamically and wirelessly receiving a signal from a
mobile
whose location is unknown at two or more interceptors, which are located at
different
known geographic locations inside a CDMA coverage area defined by said base
station where a signal from said base station to said mobile is line-of sight;
(ii)
dynamically computing a total time-of flight of said signal from said base
station to
each interceptor via said mobile; (iii) dynamically computing an ellipse of
position of
said mobile for each total time-of flight computation, where each ellipse of
position
has as its foci said base station and said interceptor corresponding to said
total time-
of flight measurement for said interceptor; (iv) dynamically computing
intersection
points) of each possible pair of ellipses of position, if any such
intersection point
exists; (v) dynamically and wirelessly receiving said signal from said mobile
and
measuring an angle-of arrival of said signal received at each of said
interceptors; (vi)
dynamically computing a line of position corresponding to each angle-of arnval
measurements; (vii) dynamically computing an intersection point of each
possible pair
of line of position based on angle-of arrival measurements, if such an
intersection
point exists; (viii) dynamically comparing said intersection points) of each
pair of
ellipses of position, if any such intersection point exists with the
corresponding
intersection point of said lines of position based on angle-of arrival
measurements, if
such an intersection point exists; and (ix) determining either (a) a
geographic area
within which the mobile is located that is defined by the area of intersection
of all
ellipses of position whenever for all possible pair of interceptors, either no
intersection point of the angle-of arrival lines of position corresponding to
a pair of
interceptors coincides with the intersection points) of the ellipses of
position
corresponding to the same pair of interceptors, or no intersection point of
angle-of
arnval lines of position exists, or (b) the actual position of the mobile
whenever the
2
CA 02497363 2005-02-17
4 ~
signal from the mobile to each of any pair of interceptors is line-of sight
which occurs
whenever the intersection point of the angle-of arnval lines of position
corresponding
to the interceptors intersects one of the two intersection points of the
ellipses of
position corresponding to the interceptors, corresponding to the actual
position of the
mobile at that time.
According to another aspect of the invention, it provides a system for the
outdoor geolocation of a mobile of interest in a CDMA cellular system
comprising of
a plurality of interceptors located at known location inside a CDMA coverage
area of
a base station, wherein each of said interceptors comprising of: (i) a means
for
obtaining a total-time-of flight measurement, which is a total propagation
time of a
signal from said base station to said mobile and from said mobile to said
interceptor;
(ii) a means for obtaining an angle-of arrival measurement of a signal from
said
mobile; (iii) a means for distinguishing whether said signal received from
said mobile
is line-of sight or reflected; and (iv) a means for determining a location of
said
mobile.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention will now be described in more detail with reference to the
accompanying drawings, in which:
Figure 1 illustrates how, according to the method as recited in the present
invention, time-of flight and angle-of arrival information of a line-of sight
signal to a
single interceptor can be used to locate a mobile in a CDMA cellular coverage
area;
Figure 2 illustrates how two interceptors can be used to determine the
possible
positions of a mobile in a CDMA cellular coverage area where the signal
propagation
is line-of sight;
Figure 3 illustrates how the use of two interceptors in a CDMA coverage area
can provide information concerning the boundaries within which a mobile is
located
when a reflection in the signal propagation path between the mobile and one of
the
interceptors is present;
Figure 4 illustrates that if the signal path from a base station defining a
CDMA
coverage area to a mobile is line-of sight and the non-line-of sight signal
path from
3
CA 02497363 2005-02-17
the mobile to an interceptor contains only one point of reflection, the point
of
reflection will lie along the angle-of arnval line of position;
Figure 5 illustrates that the ellipses of position corresponding to the range
of
all possible points of reflection along an angle-of arrival line of position
will be
smaller than the corresponding principal ellipse having its foci at the base
station
defining a CDMA coverage area and an interceptor located in that area;
Figure 6 illustrates how the use of two interceptors in a CDMA coverage area
can provide information concerning the boundaries within which a mobile is
located
when a reflection in the signal propagation path between the mobile and each
of the
interceptors is present;
Figures 7 illustrates that when the intersection point of two angle-of arnval
lines of position does not intersect either of the two intersection points of
two
principal ellipses associated with two interceptors located at different
points, the
signal received at one or more of the two interceptors from a mobile within a
CDMA
coverage area defined by a base station is reflected;
Figures 8 illustrates that when the intersection point of two angle-of arrival
lines of position intersects either of the two intersection points of two
corresponding
principal ellipses associated with two interceptors located at different
points, the
signal received at both interceptors from a mobile within a CDMA coverage area
defined by a base station is line-of sight;
Figure 9 illustrates the use of a four-antenna Watson-Watt array to compute a
coarse angle-of arrival of a signal at one of the antennas;
Figure 10 illustrates the use of a three-antenna Watson-Watt array to compute
a coarse angle-of arnval of a signal at one of the antennas;
Figure 11 illustrates a high gain adaptive threshold determination signal
processing method applied to a weak signal received from a mobile within a
CDMA
system; and
Figure 12 illustrates the relationship between frame rates and power control
groups in an IS-95 CDMA cellular system.
DETAILED DESCRIPTION OF THE INVENTION
4
CA 02497363 2005-02-17
In all forms of geolocation applied to CDMA cellular systems there currently
exist no techniques to determine if the first-to-arrive signal reaching an
interceptor
employed in the geolocation of a mobile is a line-of sight signal or a
reflected signal.
This causes a substantial deterioration of the geolocation results, as it is
impossible to
determine if the location of the mobile is calculated from valid, line-of
sight signals or
from erroneous data originating from reflected signals.
According to a method described herein and illustrated in Figure 1, time-of
flight and angle-of arrival information of a line-of sight signal can be used
to locate a
mobile 100. A base station 102 defining a CDMA coverage area 10 dynamically
receives network timing through the synchronization procedure associated with
the
CDMA cellular system (not shown). The base station 102 then distinguishes
itself by
transmitting a short code having a particular offset with respect to the
network timing
of the CDMA cellular system. Upon receiving a base station signal, the mobile
100
will be informed of the base station 102 short code offset. The mobile 100
then
transmits its own short code with zero time-offset with respect to the network
timing.
Of course, this short code will be delayed with respect to the network timing
by the
time it took for the base station 102 signal to propagate to the mobile 100.
An
interceptor 104 capable of receiving a signal from the mobile knows its own
physical
location, as well as that of the base station 102. It will also have access to
the network
timing through the synchronization procedure associated with the CDMA cellular
system (not shown) and accounting for any timing delays resulting from the
separation between the interceptor 104 itself and the base station 102. The
first
approach is preferable since the second assumes line-of sight signal
propagation
between the base station and the interceptor. By knowing the network timing,
and by
receiving the short code from the mobile 100, the interceptor 104 can
determine the
total time-of flight of a signal both from the base station 102 to the mobile
100 and
from the mobile 100 to the interceptor 104. If a single interceptor 104 is
employed, a
mobile 100 will be located on an ellipse of position 106 having the base
station 102
and the interceptor 104 located at the foci, assuming line-of sight signal
propagation
between the mobile 100 and the interceptor 104. Angle-of arrival information
of the
signal at the interceptor 104 can also be determined using Watson-Watt antenna
arrays and monopulse measurements at the interceptor location, as further
described
S
CA 02497363 2005-02-17
below. The location of the mobile 100 can be determined as the intersection of
the
ellipse of position 106 and angle-of arrival line of position 108 passing
through the
mobile 100 and the interceptor 104.
Figure 2 illustrates how two interceptors 200 and 202 can be used to determine
the possible positions of a mobile 204 in a CDMA cellular coverage area 20
where the
signal propagation is line-of sight. In such a case, two ellipses of position
206 and 208
are obtained both having the base station 210 defining the cellular coverage
area 20 at
a common focal point, and interceptor 200 or 202 situated at the remaining
focal point
of each ellipse, respectively. The two ellipses 206 and 208 will always
intersect each
other at two and only two locations 212 and 214. Again assuming line-of sight
signal
propagation, the mobile 204 will be located only at one of these two
intersection
points 212 or 214.
Figure 3 illustrates how the use of two interceptors 300 and 302 located at
different points and capable of receiving a signal from a mobile 306 situated
in a
CDMA coverage area 30 defined by a base station 304 can provide information
concerning the boundaries within which the mobile 306 is located when a
reflection in
the signal propagation path between the mobile 306 and one of the interceptors
300 is
present. It is noteworthy that in such situations, a link between the base
station 304
and mobile 306 is assumed to be line-of sight 308, since the base station 304
is
probably located in a highly visible location. However, the link between the
mobile
306 and an interceptor 300 may be non line-of sight, resulting in a reflection
in the
signal, which follows a "dog leg" path 312 and 314 from the mobile 306 to the
interceptor 300. In such a situation, the corresponding elliptical line of
position 316,
and the angle-of arrival line of position 320 do not yield direct information
on the
location of the mobile 306. However they do provide information on the
boundaries
of the area in which the mobile 306 is located. In the situation where there
is line-of
sight propagation between the mobile 306 and the interceptor 302 (in addition
to line-
of sight propagation from the base station 304 to the mobile 306), the mobile
306 will
be located somewhere on the ellipse of position 318 having the base station
304 and
interceptor 302 at the foci. An ellipse, such as 318, that has the base
station 304 and
an interceptor 302 as its foci, and that encompasses all possible locations of
the
mobile 306 is called a principal ellipse. A principal ellipse, such as 318 or
316 is
6
CA 02497363 2005-02-17
defined through the measurement of the total time-of flight of a signal
between the
base station 304 and the interceptor 300 or 302, respectively via the mobile
306. The
mobile 306 will be located along the portion of the principal ellipse 318
situated
inside the principal ellipse 316. This assumes that the reflection is not
located very far
from an interceptor 300 or 302, which is most likely to be the usual
situation.
However, in an extreme case, the reflection may be very far from an
interceptor,
making the ellipse 316 computed from the corresponding time-of flight
measurement
so large that ellipse 318 lies completely within ellipse 316. In that case,
intersection
points 324 and 322 will not exist. This extreme scenario is not shown in the
Figures.
Figure 4 illustrates that if the signal path 400 from a base station 402
defining
a CDMA coverage area 40 to a mobile 404 is line-of sight and a reflected
signal path
408 and 410 from the mobile 404 to an interceptor 412 contains only one point
of
reflection 414, that point of reflection 414 will lie along the angle-of
arrival line of
position 416. As before, the total time-of flight of the signal from the base
station 402
to the interceptor 412 via the mobile 404 defines a principal ellipse 418.
Figure 5 illustrates that the ellipses of position 500, 502 and 504 (which is
a
degenerate ellipse) corresponding to the range of all possible points of
reflection
including 506, 508 and 510 along an angle-of arnval line of position 512 will
be
equal to or smaller than the corresponding principal ellipse 500 having its
foci at the
base station 514 defining a CDMA coverage area 50 and at an interceptor 505
located
in that area 50. For each possible point of reflection, such as 506, 508 and
510 along
an angle-of arnval line of position 512, the time-of flight from the base
station 514 to
the assumed point of reflection 506, 508 or 510 via a mobile (not shown) can
be
calculated. For each assumed point of reflection, such as 506, 508 and 510, an
elliptical line of position 500, 502 and 504 can be established having the
base station
514 and the assumed point of reflection 506, 508 or 510 as the foci. As the
assumed
point of reflection moves away from the interceptor 506 (along the angle-of
arnval
line of position 512) to the edge of the principal ellipse 510, the
corresponding
ellipses of position (from 500 to 504) all remain within the principal ellipse
500 and
become more elliptical until the ellipsis of position degenerates into a
straight line 504
connecting the foci 514 and 510 when the point of reflection 510 intersects
the
principal ellipse 500. A point of reflection cannot exist beyond the
intersection 510 of
7
CA 02497363 2005-02-17
the angle-of arrival line of position 512 and the principal ellipse 500.
Figure 6 illustrates how the use of two interceptors 600 and 602 situated at
different points and capable of receiving a signal from a mobile located in a
CDMA
coverage area 60 defined by a base station 604 can provide information
concerning
the boundaries within which a mobile 606 is located when a reflection in the
signal
propagation path between the mobile 606 and each of the two interceptors 600
and
602 is present. Once again, the link from the base station 604 to mobile 606
is
assumed to be line-of sight 608, since the base station 604 is probably
located in a
highly visible location. However, the link between the mobile 606 and each of
interceptors 600 and 602 may be non line-of sight, resulting in a reflection
in the
signal, which follows a "dog leg" path 612 and 614 from the mobile 606 to the
interceptor 600, and a "dog leg" path 616 and 618 from the mobile 606 to the
interceptor 602. The two principal ellipses generated from the measured total
time-of
flight between the base station 604 and the intercept sites 600 and 602 via
the mobile
606 will intersect each other at only two points 624 and 626, and the mobile
606 will
lie somewhere in the intersection area 628 of the two principal ellipses 620
and 622.
Figure 7 illustrates that when the intersection point 700 of two angle-of
arrival
lines of position 702 and 704 does not coincide with either of the two
intersection
points 706 or 708 of two corresponding principal ellipses 710 and 712
determined
from the total time-of flight of signals received at one or more of two
interceptors 714
and 716, respectively, via a mobile 718 from a base station 720 defining a
CDMA
coverage area 70, the signal from the mobile 718 to one or both of the
interceptors
714 and 716 is reflected, assuming that the signal from the base station 720
to the
mobile 718 is line of sight 722. It is also possible for the angle-of arnval
lines of
position 702 and 704 not to intersect each other (not shown). When this
occurs, it is
also an indication that there is a reflection between the mobile 718 and one
or both of
the interceptors 714 and 716. Since the intersection point 700 of the lines of
positions
based on angle-of arrival measurements does not coincide with the intersection
points
of two corresponding principal ellipses 710 and 720, the mobile of interest is
deemed
to be located inside an intersection area 730 of two corresponding principal
ellipses
710 and 720. It is also possible that no intersection of a corresponding pair
of lines of
position of angle-of arrival measurements may be found. In such case, as well,
a
8
CA 02497363 2005-02-17
mobile of interest is deemed to be located within an area of intersection
areas of a
corresponding pair of principal ellipses of position.
Although the discussions of Figures 3 to 7 only illustrated single reflections
in
the path between a mobile of interest and an interceptor, the principles
described with
reference to those Figures also apply if there are multiple reflections in
such a path.
Figure 8 illustrates that when the intersection point 800 of two angle-of
arnval
lines of position 802 and 804 intersects either of the two intersection points
800 or
806 of two corresponding principal ellipses 808 and 810 determined from the
total
time-of flight of signals received at one or more of two interceptors 812 and
814,
respectively, via a mobile 816 from a base station 818 defining a CDMA
coverage
area 80, the signal from the mobile 816 to both of the interceptors 812 and
814 is line-
of sight. This situation not only confirms that line-of sight propagation has
taken
place, it also will identify which of the intersection points 800 or 806 of
the principal
ellipses is the actual location of the mobile.
The techniques described above can be used to determine if signals received
from a mobile are line-of sight as in Figure 8 or reflected signals as in
Figure 7. If the
signals are line-of sight, the actual location of the mobile can also be
determined. The
technique can be applied to mobiles that are moving. In such cases, the angle-
of
arrival lines of position and the principal ellipses will be dynamically
changing. As
the mobile moves it may enter a location for which both of the paths from the
mobile
to an interceptor becomes line-of sight. At this instant, the intersection
point of the
angle-of arnval lines of position will cross one of the current intersection
points of the
principal ellipses, and establish the location of the mobile.
Although the preceding discussion has only described the use of two
interceptors, in practice more interceptors will usually be used, and the
methods
described above will be applied to the each possible pair of interceptors, in
turn. The
greater the number of interceptors used, the greater will be the probability
that the
signal from the mobile of interest to each of at least one possible pair of
interceptors
will be line-of sight thereby yielding the actual location of the mobile. Even
if that is
not the case, by comparing the intersection areas of each possible pair of
ellipses of
positions generated for the interceptors employed to locate the mobile of
interest at a
point in time, it is possible to define the possible area within which the
mobile is
9
CA 02497363 2005-02-17
located at that point in time as the intersection area of all of the possible
pair of
ellipses of position generated for the number of interceptors employed. This
defined
area will typically decrease as the number of interceptors is increased.
In sum, this mobile geolocation method depends on the ability to monitor total
time-of flight and angle-of arrival information of a desired signal. It is
possible to
continuously monitor the time-of flight of a mobile's forward link signal
through
acquisition of the mobile's reverse link channel, as described below. It is
also
possible to instantaneously determine the angle-of arrival at interceptor
sites using
well-known direction-finding techniques, also described below.
In order to obtain the total time-of flight as well as the angle-of arrival
information required to apply the techniques discussed above, the reverse link
access
channel or traffic channel must be acquired. To achieve this, each interceptor
must
first acquire the base station pilot channel consisting of one or more short
codes with
a network timing-offset associated with the base station or its particular
sector.
This general technique applicable to any CDMA cellular system can be
illustrated in the following description of a preferred embodiment for
computing total
time-of flight in an IS-95 CDMA cellular system. In such a case, the short
codes
employed would be the I and Q short codes.
With time synchronization of the pilot channel, the interceptor can also
easily
obtain the forward link sync channel. In the IS-95 CDMA cellular system, this
consists of time-offset I and Q short codes with a Walsh 31 code overlay (at
the same
chip rate) carrying convolutionally encoded and interleaved data at a base
rate of 1.2
kbps. The information on the sync channel consists of the network timing, and
position of the long code at the start of 4th 80 milliseconds super frame
following the
super frame in which the information is being transmitted. The network timing
determines the short code offset used by the base station.
With the long code position known, the interceptor can receive the forward
link paging channels. These channels consist of time-offset I and Q short
codes, with
known Walsh code overlays and (decimated) long code scrambling. The channels
carry channel assignment data and other system overhead information. This
channel
assignment data can be used to build the mobile's mask for generating its
unique
offset of the long code.
CA 02497363 2005-02-17
With the mobile's mask and network timing, the interceptor can receive the
access channel and the traffic channel from the mobile in the reverse link
band.
The reverse link traffic channel from the mobile provides the interceptor with
a continuous stream of concatenated Walsh codes modulated with zero time-
offset
(but symbol offset) I and Q short code and long code spreading using the
mobile's
unique mask. With knowledge of the mobile's long code mask, the time-of arnval
of
the first signal to arrive at the interceptor can be determined.
From knowledge of the network timing obtained from GPS and of the offset
used by the base station in the transmission of the I or Q short code, the
time of
transmission at the base station can be determined. By taking the difference
of this
time and the time-of arrival of the start of the I or Q short code, the total
time-of
flight from the base station to the mobile to the interceptor can be
determined.
The angle-of arnval measurements are performed in two stages; a fast
(instantaneous) coarse measurement followed by an accurate, monopulse
measurement, whose accuracy can be improved even further through the
application
of digital signal processing techniques, all as more specifically described
below.
Coarse angle-of arrival measurements can be made with the use of one of
three well known techniques: antenna main beam or null pointing direction,
Doppler
measurement from a revolving antenna or from a ring of commutating antennas,
and
phase measurement between separate receiving antennas. Of these, the first two
approaches require a large physical antenna, or ring around which an antenna
is
revolved or around which many antennas are commutated. The phase measurement
technique provides the angle-of arrival measurement of comparable accuracy
with a
much smaller physical size. Further it does not require rotating mechanisms.
In
addition, measurements performed using this technique are instantaneous.
In a preferred embodiment, the coarse angle-of arnval measurement is
accomplished using a well-known technique based on measuring the relative
phase of
a received signal between two or more separate antennas, called the Watson-
Watt
array. For azimuth determination (in the presence of an accompanying elevation
component), either an array of 3 or 4 antennas can be used.
Figure 9 shows the use of a four-antenna array 900. In such an array, the
distance L between antennas 1 and 3 is the same as the distance between
antennas 2
11
CA 02497363 2005-02-17
and 4. The angle-of arnval 8 of a signal is the angle at which a signal
arrives relative
to a straight imaginary line 902 passing through antennas 1 and the centre of
the
antenna array 904. The angle-of arrival 8 can be determined indirectly using
the phase
differences (not shown) measured between each of two pair of antennas using
the
following equations:
8 = atan ((p24 / ~13)
or
8 = atan (cp23 / y2) - 45
where cp;~ is the phase difference of a signal between antennas i and j
It should be noted that antenna pairs 1-4 and 4-3 give the same equation as
antenna pairs 2-3 and 1-2. This provides a third redundant measurement.
It is interesting to observe that since their spacing is 0.707L, the
sensitivity of
pairs 1-2, 2-3, 3-4, and 4-1 is reduced by a factor of 0.707 compared to that
of pairs 1-
3 and 2-4, and the standard deviation of their errors is increased by 1.414.
However
this is compensated for by the fact that they form redundant pairs.
In order to minimize interceptor's receiver complexity, an angle-of arrival
determination based on the 4-antenna array could use just the phase difference
measurements between antennas 2 and 4, and between antennas 1 and 3.
Figure 10 shows the use of a three-antenna array 1000. In such an array, the
distance L between each pair of antennas 5-6, 6-7 and 7-5 is the same. The
angle-of
arrival B is the angle at which a signal arnves relative to a straight
imaginary line
1010 passing through antenna 6 and the center of the antenna array 1020. The
angle-
of arrival 8 of a signal relative to an antenna 6 can be determined indirectly
using the
phase differences (not shown) measured between the various possible pair of
antennas
using the following equation:
8 = atan { 1.732 cp5~ / (cp76 - cp6s)}
12
CA 02497363 2005-02-17
More accurate angle-of arnval measurements can be made using well-known
monopulse techniques when the mobile is line-of sight and its coarse location
is
known. In theory, monopulse can be either amplitude or phase comparison in
nature.
In practice, amplitude comparison monopulse provides better performance than
phase
comparison, being less sensitive to mechanical tolerances. Accuracy of 0.01
degree is
achievable, particularly when digital signal processing techniques designed to
increase signal to noise ratio are applied, as described below, to the signal
received
from the mobile to be located when the monopulse technique is applied to the
signal.
Another problem that must be overcome by the present invention, is ensuring
that the signal received at the interceptor, which can be quite weak, can
actually be
distinguished from any associated noise so that the time-of flight and angle-
of arnval
data yielded by the techniques described above will be reliable. The following
discussion describes the manifestation of the problem in the context of an IS-
95
CDMA cellular system. However, the problem may arise in any CDMA cellular
system and the technique employed to overcome the problem described below can
be
applied in general to any CDMA system.
In an IS-95 CDMA system operating at full rate, the reverse link of the system
has a signal processing gain of 21 dB. After demodulation, the signal to noise
ratio of
the data demodulated by a base station is expected to be of the order of 6 to
7 dB.
This means that the signal to noise ratio of the signal reaching the base
station is of
the order of -15 dB. It is to be noted that the base station controls the
power emitted
by the mobiles in such a way that the base station receives the same power
from all
the mobiles. This is to minimize the mutual interference of the mobiles and to
achieve the maximum capacity of the cell. Consequently, the power transmitted
by a
mobile located close to the base station is likely to be much smaller than the
power
transmitted by a mobile located far from the base station.
A receiver, such as an interceptor, trying to intercept the signal from a
mobile
located close to the base station is likely to have very little power to work
on. In such
a case high-gain signal processing will be required to handle the situation.
Receiving
a signal from a mobile located far from the base station should be less
problematic as
the mobile is likely to emit more power. Consequently, in order to be able to
operate
on a good selection of mobile positions, an interceptor should be able to
provide a
13
CA 02497363 2005-02-17
large gain as it is likely to have to process signals with a much smaller
signal to noise
ratio that the -15 dB expected at the base station.
Figure 11 illustrates an adaptive high-gain signal processing method applied
to
a signal 1100 received from a mobile (not shown) at the interceptor. The first
step
1102 is the despreading (i.e., stripping of the long code and the short codes)
of the
signal 1100 using a stored reference signal (not shown) having the long code
offset
mask used by the mobile of interest and the time-of arrival of the first-to-
arnve signal
of the mobile of interest. The despreading 1102 produces an output signal 1104
consisting of concatenated Walsh codes with Walsh chips that have a duration
of four
spreading chips. The next operation 1106 consists of integrating the four
spreading
chips included in each Walsh chip. The knowledge of the network timing
previously
acquired during the synchronization permits to determine the location of the
boundaries of the Walsh chips. After this operation, the signal 1108 consists
of Walsh
chips that are either positive or negative. Up to this point, the signal
processing is
identical to what the receiver of a base station normally performs. The
following steps
are novel and essential to the proper functioning of the direction finding
operation.
The next step 1110 is the squaring of the Walsh chips. The resulting output
signal 1112 will have a signal to noise ratio that is twice the signal to
noise ratio of the
input signal 1108 processed in this manner. Thus, for example, a signal to
noise ratio
of -20dB before squaring would be -40 dB after squaring. The next step 1114 is
the
integration over the transmitted power control groups of one frame. This
integration
produces the gain that is required to overcome the very negative signal to
noise ratio
and produce a high gain detectable signal 1116.
As illustrated in Figure 12, in an IS-95 cellular system a frame 1200 contains
16 power control groups 1210 and when the system is operating at rate 1220 l,
1/2,
1/4 or 1/8, either 16, 8, 4 or 2 power control groups 1210 are transmitted,
respectively.
The power control groups that are not transmitted are simply gated off at the
transmitter in order to reduce the overall noise level of the system. The time
of
transmission of the power control groups is determined by the long spreading
code
and can be determined once the timing of the system is acquired and once the
long
code offset of the mobile of interest is known. The substantial gain produced
by the
integration over one frame, even when the system operates at 1/8 rate and
transmits
14
CA 02497363 2005-02-17
only 2 power control groups, should produce a significant extension of the
range of
operation of an interceptor over the range of operation of the base station.
Simulation
results suggest that integration over the two power control groups from a
frame
transmitted at 1/8 rate could provide useful results even with a signal
undergoing
Ricean fading with a low power specular component.
It is well known that when an IS-95 CDMA system is operating at rate 1, 1/2,
1/4 or 1/8, either 16, 8, 4 or 2 power control groups are transmitted,
resulting in the
integration of 24576, 12288, 6144 and 3072 spreading chips, respectively. The
resulting gain is 43.9 dB, 40.9 dB, 37.9 dB and 34.9 dB respectively.
In order to maximize the capacity of the reverse link, an IS-95 CDMA mobile
adjusts its transmission rate for each frame according to the quantity of
information to
be transmitted. Therefore the transmission of a mobile is comprised of a few
selected
power control groups whose time of transmission depends on the long code
offset
used by the mobile and on the quantity of data that it is transmitting. This
makes the
measurement of the noise level a difficult matter, since sometimes the signal
of the
mobile of interest is present and sometimes it is not present. The noise
level, once
measured, is used to establish an adaptive detection threshold for the desired
signal.
With reference, once again, to Figure 11, the method employed herein to
measure the noise level consists of the same signal processing operations
1106, 1110,
and 1114 previously used to produce the high gain detectable signal 1118, but
in this
case the despreading 1120 occurnng before these other steps, 1106, 1110, and
1114,
is performed using a stored reference signal (not shown) that uses an
incorrect offset
of the long code, i.e. whose offset is not the offset of the long code used by
the mobile
of interest, although the timing of the first-to-arnve signal of the mobile of
interest is
still used. This procedure ensures that the noise has been integrated only
over the
power control groups transmitted by the mobile. The output signal 1122 from
the
cumulative procedure of steps 1120, 1106, 1110 and 1114 serves as a minimum
signal
threshold. That minimum signal threshold signal 1122 is then processed by a
threshold setting process 1124 that takes into account desired probabilities
of
detection and false alarms in order to determine the actual threshold 1126
that is then
employed in an adaptive signal threshold detection process 1118 applied to the
high
gain detectable signal 1116 derived using a stored reference signal (not
shown) that
CA 02497363 2005-02-17
uses the same offset of the long code used by the mobile of interest in order
to remove
the noise present in that signal and extract the final output signal with
sufficient gain
1128 corresponding to the signal transmitted by the mobile.
A peculiarity of IS-95 is that the mobile does not inform the base station of
the
rate at which each frame is transmitted. Consequently, the base station has to
process
the signal for the four possible rates and then selects the rate producing the
best
results. The interceptor should do the same and perform the processing for the
four
possible rates, both for the production of the correlation peak and for the
integration
of the noise for the setting of the adaptive threshold. The rate giving the
best results
for the production of the correlation peak should also be used for the setting
of the
adaptive threshold.
Although some of the embodiments and variations described herein were
applied to the IS-95 CDMA cellular system, the invention described herein can
be
applied to any CDMA system.
It is to be understood that the embodiments and variations shown and
described herein are merely illustrations of the principles of this invention
and that
various modifications may be implemented by those skilled in the art without
departing from the spirit and scope of the invention.
16
