Note: Descriptions are shown in the official language in which they were submitted.
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-1-
POSITIONING SYSTEM AND METHOD FOR CELLULAR MOBILE RADIO
CROSS-REFERENCES TO RELATED APPLICATIONS
This Application for Patent claims the benefit of priority from, and hereby
incorporates by reference the entire disclosure of, co-pending U.S.
Provisional
Application for Patent Serial No. 60/067,113, filed December 1, 1997.
This Application is also related by subject matter to commonly-assigned
U.S. Patent Application Serial No. 08/894,466, flied August 18, 1997, which is
1 o incorporated by reference herein in its entirety.
BACKGROUND OF THE INVENTION
The present invention relates generally to the mobile telecommunications
field and, in particular, to a method and system for determining the position
of
mobile stations (MSs) in a cellular Time Division Multiple Access (TDMA)
system.
D~ecrintion of Related Ark
There is a wide range of applications where the global position of an MS in
a cellular system is of great interest. For example, it is important to be
able to
2 o determine the position of MSs involved in emergency calls (police and fire
vehicles) or for fleet management purposes (e.g., taxi companies). As such, a
number of different solutions have been proposed for determining the location
of
MSs. There are terminal-based solutions such as MSs with built-in Global
Positioning System (GPS) receivers, as well as network-based solutions such as
the
one disclosed in Swedish Patent Application No. 9303561-3 to R. Bodin.
In this Swedish Application, a "positioning handover" method is used to
determine the position of an MS. However, notwithstanding the advantages of
this
method, there are still some problems that exist, such as, for example, a risk
of
losing calls, and the relatively long period of time it takes to perform the
3 o positioning procedure. However, such problems are solved by the method
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-2-
disclosed in commonly-assigned U.S. Patent Application Serial No. 08/977,470
filed November 24, 1997, whereby only one positioning handover has to be made.
In both of the above-described cases, access delays are measured that
represent the absolute distance to the MS. The MS's position is then
calculated by _ .
triangulation. In other words, the MS's position is located at or near the
point
where a plurality of plotted circular arcs cross over one another.
An alternative positioning method is to base the position determinations on
Time Difference of Arrival (TDOA) calculations, which are further based on
Time
of Arrival (TOA) measurements. In this case, an MS's position is located at or
near
the point where a plurality of hyperbolic arcs cross over one another. This
approach advantageously provides more accurate positioning information than
non-
TDOA approaches, because the measurements made are dependent only on the
different uplink delays. Another advantage of this approach is that the MS
positioning can be performed without disturbing any ongoing communications.
A TDOA positioning method is used by Trueposition~. This approach uses
separate positioning receivers {Signal Collection System or SCS) to sample the
uplink waveform and transfer all data to a common TDOA Location Processor
(TLP), which correlates the information with known transmitted sequences.
The above-cited, commonly-assigned U.S. Patent Application Serial No.
2 0 08/894,466 (hereinafter, the "'466 Application's discloses a TDOA
positioning
solution which is integrated into the mobile telephone system (e.g., for the
Digital-
Advanced Mobile Phone System or D-AMPS). For this approach, a correlation
with known bit sequences is performed in special purpose receivers that are
integrated into the base station (BS).
2 5 A shortcoming of the prior MS TDOA positioning approaches is that they
describe only generally how a TDOA positioning method can be used with certain
digital TDMA cellular systems. In particular, the prior art does not disclose
how
MS TDOA positioning techniques can be applied to the Global System for Mobile
Communications (GSM). Furthermore, the prior MS TDOA approaches do not
3 0 address certain problems, such as the risk associated with performing
handovers
during the position determination process, or positioning while listening to
channels using logical channel multiplexing and discontinuous transmissions.
CA 02313053 2000-06-O1
WO 99/Z9130 PCT/SE98/02198
-3-
However, as described in detail below, the present invention successfully
resolves
the above-described problems.
SUMMARY OF THE INVENTION
In accordance with the present invention, a method and system are provided
for determining the position of MSs based on TDOA measurements, which can be
applied to digital mobile radiotelephone networks such as, for example, in the
GSM. In a preferred embodiment, the network retrieves the identity of the
serving
cell and serving channel allocated for the MS whose position is to be
determined,
locates a plurality of BSs surrounding the serving cell, allocates a
measurement
1 o channel for each of the surrounding BSs so located, and schedules a
measurement
time for the located surrounding BSs. Each BS then performs a TOA measurement
at the scheduled measurement time, and reports the measurement information to
the
network. The network uses the TOA measurement information to calculate TDOA
information, and thus derives the MS's position.
An important technical advantage of the present invention is that a TDOA
MS positioning method is provided for digital TDMA cellular systems, such as,
for
example, the GSM.
Another important technical advantage of the present invention is that a
method for determining the position of an MS is provided wherein the risk of
2 0 losing connections while performing MS positioning handovers is
significantly
reduced.
Yet another important technical advantage of the present invention is that a
method for determining the position of an MS is provided wherein the risk
associated with determining an MS's position while listening to channels using
2 5 logical channel multiplexing or discontinuous transmissions is
significantly
reduced.
Still another important technical advantage of the present invention is that a
method is provided for determining the position of MSs wherein the positioning
accuracy is significantly higher than prior methods.
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
BRIEF DESCRIPTION OF THE DRAWINGS
A more complete understanding of the method and apparatus of the present
invention may be had by reference to the following detailed description when
taken
in conjunction with the accompanying drawings wherein:
FIGURE 1 is a diagram of a mobile radiotelephone system overview that
can be used to illustrate a number of methods for determining the position of
MSs,
in accordance with preferred embodiments of the present invention;
FIGURE 2 is a flow diagram that can be used to implement an exemplary
embodiment of the present invention for an ongoing connection in a synchronous
network similar to that shown in FIGURE 1;
FIGURE 3 is a diagram that helps to better illustrate the exemplary TOA
measurement method described with respect to FIGURE 2;
FIGUREs 4A and 4B are related diagrams that illustrate how a TN can be
calculated for allocation in a detector's BS;
FIGURE 5 is an exemplary frame timing diagram that illustrates how to
provide a guaranteed measurement window, in accordance with the second
embodiment of the present invention;
FIGURE 6 is a flow diagram that illustrates a method for determining the
position of an MS using controlled transmission of access bursts in a
synchronous
2 0 network, in accordance with the third exemplary embodiment of the present
invention;
FIGURE 7 is a diagram that illustrates how an existing handover command
method can be implemented for the third exemplary embodiment of the present
invention;
2 5 FIGURE 8 illustrates a situation when an access burst transmitted by a MS
is received by a serving BS and detectors; and
FIGURE 9 is a diagram that illustrates how TOA measurements can be
made, in accordance with the fourth exemplary embodiment of the present
invention.
3 0 DETAILED DESCRIPTION OF THE DRAWINGS
The preferred embodiment of the present invention and its advantages are best
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-5-
understood by referring to FIGURES 1-9 of the drawings, like numerals being
used for
like and corresponding parts of the various drawings.
Essentially, in accordance with the present invention, a method and system are
provided for determining the position of MSs based on TDOA measurements, which
can be applied to digital mobile radiotelephone networks such as, for example,
in the
GSM. In a preferred embodiment, the network retrieves the identity of the
serving cell
and serving channel allocated for the MS whose position is to be determined,
locates
a plurality of BSS surrounding the serving cell, allocates a measurement
channel for
each of the surrounding BSS so located, and schedules a measurement time for
the
located surrounding BSS. Each BS then performs a TOA measurement at the
scheduled measurement time, and reports the measurement information to the
network.
The network uses the TOA measurement information to calculate TDOA
information,
and thus derives the MS's position.
Specifically, FIGURE 1 is a diagram of a mobile radiotelephone system
overview that can be used to illustrate a number of methods for determining
the
position of MSS, in accordance with preferred embodiments of the present
invention.
At this point, it is useful to discuss some basic assumptions that may be made
with
regard to the system shown in FIGURE 1. For example, all TOA and/or TDOA
measurements can be made at receivers located at the different BSS in the
network.
2 0 As a synchronous network, all of these receivers are highly accurately
time-
synchronized to a single global time reference. However, the actual
synchronization
accuracy required should depend on the specific network operating scenarios
and
applications involved. An exemplary method that can be used to attain highly
accurate
global time synchronization is to locate a GPS receiver (highly accurate time
base) as
2 5 a time reference at each BS's site.
As a basis for obtaining accurate calculations of the MS' positions, the
geographic location of each BS should be known. More precisely, the exact
locations
of the phase centers of the antennae that feed the receivers' detectors should
be known,
as well as the exact signal delays from the antennae to the respective
receiver
3 o detectors. Notably, if these signal delays are known, then their effects
can be
adequately compensated for.
As such, an exemplary method that can be used for determining the position
CA 02313053 2000-06-O1
WO 99/29130 PGT/SE98/OZ198
-6-
of an MS in the system shown in FIGURE 1 is described in the '466 Application.
This
exemplary method can be summarized as follows. First, a positioning request
for an
MS is issued (by a user or network operator) at the user interface (L1I) of a
mobile
positioning service node 10. The request is then forwarded from the service
node to
a mobile services switching center (MSC) 12 in the cellular network. In a GSM
network, the MSC can be combined with a base station controller (MSCBSC). At
least three BSs (e.g., BS 1, BS2, BS3) with antennae at different locations
detect a
signal (waveform) transmitted from the MS 14 whose position is to be
determined.
Each such BS then measures the TOA of the detected signal at the BS's
respective site.
The waveform used by the MS 14 may be ordered by the network (e.g., when a new
connection is to be established, or a handover has been ordered), or it can be
in the
form of signals that are normally transmitted from the MS (e.g., speech
frames).
Notably, in order to be able to accurately measure the TOA of a transmitted
signal, the
existence of the signal should be known in advance by the BS involved.
Alternatively,
the incoming signal waveform can be sampled and conveyed to a central location
for
further processing. This is similar to an approach used by TruePosition~.
For the next step, information related to the TOA measurements is conveyed
from the BSs to the mobile positioning service node 10 (via the MSCBSC 12). A
triangulation procedure to determine the MS's position is performed in the
mobile
2 0 positioning service node 10 based on the TOA information received from at
least three
BSs, and the known coordinate information for those BSs. The MS's position is
then
conveyed to the user or operator via the UI of the service node 10.
As such, the present invention provides a number of alternative methods
(embodiments) for deterniining the position of an MS, which improve
significantly on
2 5 the above-described method. FIGURE 2 is a flow diagram that can be used to
implement an exemplary embodiment of the present invention in a network
similar to
that shown in FIGURE 1. Essentially, the method illustrated by FIGURE 2 can be
used to determine an MS's position by listening to an ongoing connection in a
synchronous network. The network can thus be assumed to be synchronized for
this
3 0 embodiment (i.e., the timing of the air interface is common to all cells
in the network).
However, the cells may be offset in Frame Number (FN) counting. After
receiving the
positioning request, the serving cell and actual serving channel in use by
that MS (e.g.,
CA 02313053 2000-06-O1
WO 99/29130 PGT/SE98/02I98
_'j_
14) are noted by the network (e.g., service node 10). The network then
allocates
detectors in a number of the BSS surrounding the serving cell. For this
embodiment,
from one to several Normal Bursts can be transmitted from the MS and used for
the
TOA measurements.
Referring to FIGURES l and 2, at step 100, the positioning request for a MS
(e.g.,14) is made. At step 102, if the network (MSCBSC 12) determines that the
MS
14 is in the idle mode, then at step 104, the network initiates a call setup
procedure for
that MS. Otherwise, at step 106, the network continues the positioning
procedure.
Using identity information about the MS (e.g., IMSI in the GSM), the network
retrieves the MS's occasional cell location (e.g., stored in the MSCBSC in the
GSM).
The network also retrieves a description of the channel allocated to the MS
14.
Although it is possible that the MS can leave the channel before the end of
the
measurement period; an option can be exercised to inhibit a channel change or
release
during this period. In other words, handovers from the channel should be
avoided
where possible, and releases, for example, could be delayed. In any event, if
the MS
has to leave the channel for some reason, then the positioning procedure
should be
intemipted, and the cause of the interruption should be determined before
proceeding
further.
In the GSM, the Channel Description contains the . following parameters:
2 0 Channel Type and TDMA Offset information indicate the type of channel
involved
(e.g., TCH/F, SDCCH/8) and subchannel number; TN is the Timeslot Number in the
air interface (0-7); H is the Hopping Channel, where "0" represents a Single
RF
Channel, and "1" represents an RF Hopping Channel; H~ represents an Absolute
Radio Frequency Channel Number (ARFCN); H=1 represents a Mobile Allocation
2 5 Index Offset (MAIO) and Hopping Sequence Number (HSN), which are
parameters
that describe the hopping sequence {in this case, the Mobile Allocation (MA)
with all
frequencies in the hopping set should also be retrieved); and Training
Sequence Code
(TSC) (0-7), which identifies any of 8 possible fixed Training Sequences (26
bits)
used in all Normal Bursts transmitted over the air interface. The Training
Sequence
3 0 is then used in the detector for correlating with the received information
(i.e., for
determining the position of the burst in time).
In general, any information that describes the incoming bursts can be
CA 02313053 2000-06-O1
WO 99/Z9130 PCT/SE98/OI198
_g_
forwarded to the detector. A longer Training Sequence can improve the
receiver's
sensitivity. In other words, in general, the MS is expected to transmit a
code, x, at
time T. The code, x, can be loaded into the detectors prior to time T. The
detectors
can then capture the hansmitted code subsequent to time T, and measure the
delay
between time T and the instant when the code is captured.
As mentioned earlier for this method, the radio network should be
synchronized. In other words, the air interface timing for all cells is
synchronized.
However, the synchronized cells are preferably offset in time (FN counting),
because
in the GSM, the MSs measure signal strength and decode information about
1 o neighboring cells. Consequently, the measurements would take much longer
to
perform if all of the cells had the same FN time. Each cell is thus offset in
time by use
of the parameter, FN Offset, which is retrieved from the serving cell and
forwarded to
the respective detector.
At step 108, the network {e.g., accessing a lookup table) determines what BSs
(detectors) surround the serving cell. Preferably, these surrounding BSs are
located
at sites other than that of the serving BS. It is advantageous if all BSs
likely to be able
to detect the MS's transmitted bursts within their respective measurement
windows
can be identified. One approach is to use the list of neighbor cells produced
by
conventional methods as candidates for handover. A neighbor cell list can be
2 0 produced as a result of MS measurements. As such, the neighbor cells on
the list at
the same sites as the serving cells are canceled, and out of the remaining
cells, at least
one cell (e.g., the "best" one) for each BS site is selected. However, a
shortcoming
of this approach is that the neighbor cell list might exclude some BSs that
could be
used for performing TOA measurements. For example, the sensitivity of those
BS's
2 5 detectors could be better than the sensitivity of the MS's itself.
Another approach to identify such BSs is for the network to lookup in a
database, which stores pertinent information about the positions of all BSs in
the
network, what surrounding BSs are within a probable "listening distance" from
the
involved MS. The positions may be known to the network, or the MSC/BSC (12)
3 0 could request this information from the mobile positioning service node (
10) after the
serving cell has become determined. If directional antenna information
(antenna
directions of sector sites) is also stored by the network, then this
information can be
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-9-
used for selecting the "best" detector cell (possibly in conjunction with
information
provided by the neighbor cell lists).
Note that the serving BS will normally also participate in making TOA
measurements on the serving channel. If for some reason this is not possible
(e.g., due
to limited processing capabilities), another detector channel can be allocated
in the
same or another cell at the serving BS's site.
At step 110, the network allocates a TOA measurement channel (detector
channel) in each of the surrounding BSs found. If there are no channels
available in
one or more of the surrounding cells, then as an option, active MSs can be
handed over
l0 to other cells to allow measurements to be made in the surrounding cells.
This is
especially important if less than three detectors are available, because at
least three
detectors (at different sites) are needed for triangulation. Note that two
consecutive
timeslots can be allocated for long distances between the MS and certain
detectors
(e.g., to compensate for longer propagation times).
As another possibility, for example, if there are no idle channels available
for
measurements in the surrounding BSs, then the network can temporarily suspend
communications on (the uplink o~ a channel (e.g., a channel serving a low
priority
MS). The connection will not be lost, but there will be a slight interruption
in the user
speech and/or data. Alternatively, a special temporary detector channel can be
2 0 allocated in parallel with the existing channel.
The complete Channel Description (retrieved at step 106) is conveyed to the
detectors' BSs together with the FN Offset from the serving cell. The uplink
(reception) part of the described channel is then activated in the detector's
BS on the
indicated TN and ARFCN alternative hopping sequence and, possibly, on two
2 5 consecutive timeslots (the FN Offset can be used to adjust the detector's
timing). The
TSC is loaded into the det~tor's correlator, which is now prepared to listen
for bursts
with the correct Training Sequence. Reception is started at the activated
channel, and
if hopping is activated (e.g., H=1 ), the ARFCN is given by the hopping
parameters and
the actual FN. Note that if a temporary listening channel has been allocated,
then that
3 0 channel is not activated until the beginning of the Perform TOA
measurement step
( 114) below. As such, communications on the original channel continues up to
this
moment.
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE9$/02198
-10-
Notably, the problems described above are typically due to the use of a
conventional h~ansceiver for positioning MSs (i.e, equipment that is normally
used for
maintaining traffic betwcen the MS and the network). Theoretically, a similar
problem can exist if there are concurrent positioning tasks attempting to
allocate to the .
same receiver. An alternative approach is to use a separate positioning
receiver (not
normally used for traffic). In that case, there is no longer a conflict
between traffic and
positioning applications that would have been attempting to allocate the same
communication resources.
More importantly, a separate positioning receiver can be designed with better
1 o performance (e.g., higher sensitivity) than a receiver used for ordinary
traffic. This
approach typically implies higher costs, which may be problematic if imposed
for all
transceivers of a BS. The above-described '466 Application describes such a
solution
using a separate positioning receiver (ModR7~. However, this special
positioning
receiver is integrated into the BS. Nevertheless, the '466 Application does
not
describe such a solution for the GSM or similar digital TDMA cellular systems.
At step 112, the network schedules the timing for the TOA measurements. As
such, the network reads the actual Global Time. The Global Time should be
known
before the TOA measurements are performed. The Global Time can be made
available at all BSs for this embodiment, because it is assumed that they are
all time-
2 0 synchronized (e.g., using the GPS-derived time reference). However, the
Global Time
can also be made available at a central location in the network (e.g., at a
BSC's site).
In any event, it is advantageous if the Global Time is represented by the
timing used
for the air interface (i.e., global FN for a cell with an FN Offset=0), which
enables
synchronization of the measurements with timeslots and logical channel
multiplexing.
2 5 From an actual global FN value, a subsequent "measurement FN" is
calculated
with a margin that considers the time it takes to convey a measurement order
to each
detector involved. Next, a TOA Measurement Command is conveyed to the
allocated
detectors, which includes a measurement start time and measurement period.
Alternatively, instead of conveying a separate measurement command, this order
can
3 0 be carried within the same message as the message that activates the
detector channel
(e.g., in accordance with step 110 above).
FIGURE 3 is a diagram that helps to better illustrate the exemplary TOA
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/OZ198
-11-
measurement method described above with respect to FIGURE 2. Referring to
FIGURES 2 and 3, at step 114, after the detectors involved receive the TOA
measurement order, they await the TOA measurement start time (FN). At the
measurement start time, which coincides with a timeslot border, sampling of
the
incoming data begins with a rate which is at least equal to the symbol rate
(e.g., 270
kbits/s in the GSM). However, over-sampling is preferred for higher accuracy.
The
sampling process continues for the complete measurement window, as defined by
the
number of allocated consecutive timeslots. At the end of the sampling period,
the
recorded measurement data is matched (correlated) with the training sequence.
The
1 o results of this correlation provide the measured TOA of the received
signals.
As mentioned earlier, if using a temporary channel, the communications on the
original channel are suspended. The (temporary) detector channel is activated
immediately prior to the start of the measurement process. Note that as an
option, the
communications on the downlink of the original channel can be continued.
The TOA of a received burst can be measured as the time difference between
the central part of the training sequence and the sampling time T
(corresponding to tl,
t2 and t3 in FIGURE 3). In addition to the TOA measurement, the signal
strength and
quality of the burst can also be measured. An example of such a quality
measure is
the relative strength of the correlation peak. These additional measurement
values can
2 0 be helpful in calculating the MS's position, including the positioning
accuracy.
For the serving cell, the burst is located within the timeslot corresponding
to
the allocated channel. In the GSM, the burst's location is regulated by the
Timing
Advance (i.e., if the burst slides towards an end of the timeslot, the MS is
commanded
to advance its transmission of the burst, and vice versa for an earlier
burst).
2 5 Consequently, only one timeslot is needed in the serving cell to obtain an
accurate
TOA measurement.
Normally, the distances to the separate detectors are different. Consequently,
the burst will slide into a subsequent timeslot for detectors that are farther
away from
the MS than the serving BS, and into the preceding timeslot for the closer
detectors.
3 0 The first situation is more likely, because the serving BS is normally
closest to the
MS. However, exceptions can occur, such as for example, if the closest BS is
overshadowed by a building.
CA 02313053 2000-06-O1
WO 99!29130 PCT/SE98/OZ198
-12-
The Training Sequence is locate in the center of the burst, and the remaining
part of the burst is not needed for TOA measurements. Consequently, a large
part of
the burst can slide out of the timeslot without creating a problem. Also, it
is possible
to allocate a subsequent timeslot for the measurement when the distance to the
serving .
cell is much shorter than the distance to the other detectors.
At step 116, the measured TOAs are reported to the network, the TDOAs are
calculated, and the MS's position is then determined. The measurement period
requested can cover one or more bursts. If the measurement period covers a
number
of bursts, a new TOA measurement (e.g., including signal strength and quality
measurements) can be made in every TDMA frame during the period, and each
measurement is logged separately along with an identification tag (e.g., FN).
Subsequent to the measurement period, the full set of measurements is reported
to the
MSCBSC (12), which collects the measurement information from all of the
detectors
involved, and forwards the collected information to the positioning service
node (10).
The positioning service node calculates a set of TDOA values for each set of
TOA
measurements taken (i.e., measurements with the same identification tag). As
such,
data indicating false bursts may be discarded. The calculated TDOAs are then
averaged, and the MS's position is calculated based on the known positions of
the
detectors' BSs and the average TDOA values. If a temporary channel is being
used
2 0 for positioning, communications can be resumed in the original channel at
the end of
the measurement period.
The present method can also be used for MS positioning while using logical
channel multiplexing and discontinuous transmissions. For example, in the GSM,
an
air interface timeslot can be used for communications with a single MS's full
rate
2 5 traff c channel (TCH/F). In a different configuration, two half rate
channels (TCH/H)
share the same basic physical channel (sequence of timeslots). Timeslots
configured
with a Stand-alone Dedicated Control Channel (SDCCH) contain a plurality (4 or
8)
of SDCCH subchannels. For example there are only 4 such subchannels when the
SDCCH is multiplexed with the common Broadcast Control Channel (BCCH) or
3 0 Common Control Channel (CCCH). If applicable, the subchannel can be
identified
in the Channel Description message. The problem that arises is how to ensure
that the
measurement is made only on the correct subchannel. Note that the Traffic
Channels
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/OZ198
-13-
(TCHs) and the SDCCHs are also associated with the Slow Associated Control
Channel (SACCH), respectively, which can also be used for TOA measurements.
Additionally, the transmission in a GSM network on a specific timeslot can be
discontinuous. For speech (on a TCH), a Discontinuous Transmission (DTX) mode
can be used, which means that if a speaker is quiet, the corresponding
transmitter is
turned off most of the time. Also, data traffic (also on the TCH) is naturally
intermittent, which creates a similar problem.
A solution to these problems is to synchronize the TOA measurements with the
known multiplexing scheme used for the network's air interface. In the GSM, a
specific SDCCH subchannel is allocated to a specific part of the multiplexing
scheme;
namely, four consecutive TDMA frames of the so-called 51-frame multiframe.
Consequently, measurements are allowed only at the FNs representing this
subchannel
(including the associated individual SACCH).
Although the TCH transmission can be intermittent, there is associated
information which is always sent at a lower rate. In the DTX mode, Silence
Descriptor information (SID frames) is sent every 104-frame period at a
specific
"location". Also, the SACCH channel is transmitted periodically and can be
used, for
example, for reporting MS measurements. Consequently, it is always possible to
find
information with which to perform measurements in a GSM or similar system.
2 o In conciusion, for this embodiment, when a measurement time and period are
scheduled, the following information should be considered. There should be
allowed
FNs to measure on (e.g., belonging to the correct subchannel), and for
discontinuous
transmissions, TDMA frames that are always used for transmission (e.g., the
SACCH
block) should be identified. Note that for discontinuous transmissions, some
TDMA
2 5 frame positions may or may not contain bursts. If bursts are received at
these
positions, the TOA measurements can be performed there anyway, in order to
help
improve the overall measurement statistics.
A second embodiment of the present invention is now described. Essentially,
for this exemplary embodiment, the method illustrated by FIGURE 2 is modified
and
3 0 used to determine an MS's position by listening to an ongoing connection,
but in an
asynchronous network. The network can thus be assumed to be not synchronized
for
this embodiment (i.e., the timing of the air interface can be different for
all cells in the
CA 02313053 2000-06-O1
WO 99/Z9130 PCT/SE98/02198
-14-
network). However, each of these cells should be accurately time-synchronized
(e.g.,
using a GPS receiver at each BS's site).
In this case, subsequent to receipt of a Positioning Request {from a user or
network operator), the positioning service node (10) makes note of the serving
cell and
actual serving channel involved. The network (MSCBSC for the GSM) then
allocates
a plurality of detectors in the surrounding BSs for use in performing the
positioning
procedure. Since (as assumed) the detectors' cells are asynchronous with
respect to
the serving cell, at least two detectors should be allocated. The TOA sampling
process, which is synchronized with the air interface timing in the serving
cell, is
l0 performed simultaneously in all detectors involved. From one to several
Normal
Bursts from the MS of interest can be used for the TOA measurements.
Specifically, for this second embodiment of the present invention (referring
again to the flow chart shown in FIGURE 2), upon receipt of the Positioning
Request
at the MSCBSC (12), the following steps can be performed. At step 102, if the
MS
(14) whose position is to be determined is in the idle mode, at step 104, a
call is set up
between the network and the MS by the MSCBSC. At step 106, the MS's occasional
cell location and description of the allocated channel is retrieved from local
storage in
the MSCBSC. As such, the same step followed for the first embodiment described
above is again followed here, but with the following exception. Since the
network is
2 o not synchronized for this embodiment, it is not particularly useful to
retrieve the FN
O~set values. However, the sync position, T1, can be read (i.e., the actual
global time
at the start of a TDMA frame). Again, the global time is assumed to be
available at
each BS's site (e.g., using a GPS receiver for a time reference). As such, the
sync
position, T1, is thus sampled at the start of an arbitrary TDMA frame. In
order to be
2 5 able to attain synchronization with logical channel multiplexing involved,
the actual
FN at the sync position (T 1 ) can also be retrieved and used.
At step 108, the network locates the surrounding BSs (similar to step 108
employed in the first embodiment). At step 110, the network allocates the TOA
measurement channels to be used (as in the first embodiment), but with the
following
3 0 exception. For this embodiment, the FN Offset information is not conveyed
to the
detectors' BSs. Instead, the sync positions of each detector's BS are
retrieved similar
to the method used for the serving cell (step 106 above). For example, if T2
is the
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-15-
sync position for a specific detector's BS, and T&""~ is the periodic time for
the TDMA
frame, then the number of TDMA frames between sync positions T2 and T1 can be
expressed as:
No. o TDMA _ (T2 -Tl ) -
f frames - ,
1
T~~~ ( )
If the integer part of the equation (1) is subtracted out, the remainder
represents the
phase difference between the two cells. Using this phase information and the
known
TN value from the Channel Description for the serving cell, the timeslot
numbers) to
allocate in a detector can be calculated. As such, if the timeslots of the
detector and
serving cell are nearly in-phase, then only one timeslot has to be allocated
for the
measurement. Otherwise, two timeslots have to be allocated to obtain the same
1 o performance as obtained for the synchronized network embodiment.
More specifically, FIGURES 4A and 4B are related diagrams that illustrate
how the TN can be calculated (step 110) for allocation in a detector's BS. For
example, referring to FIGURE 4A, if the timeslots of the detector BS2 and
serving
BS 1 have a phase difference as shown (and the phase difference is defined as
$,2, e.g.,
in cosecs), then the TN can be calculated by:
812 =MOD((T2 -Tl );T~~~).
FIGURE 4B (e.g., in which several TDMA frames are expanded to show
timeslots) also illustrates how to calculate the TN if the timeslots of the
detector BSZ
and serving BS 1 have a phase difference. For example, assume that a
measurement
is made on an MS which is allocated to Timeslot Number 5 in the serving BS 1
(shown
2 0 as TNl). That timeslot lags TNl *T~o, behind the TDMA frame border, where
TS,o~ is
the duration of a timeslot (e.g., about 0.6 cosec in the GSM). Then two
consecutive
timeslots can be allocated (starting with TN2) in the detector BS2, which has
to start
before or at the same time as TNl. This calculation is based on the time
difference
between TNl and the start of a TDMA frame in the detector BS2, or t~;f In
accordance
2 5 with FIGURE 4B, the difference can be expressed as:
CA 02313053 2000-06-O1
- WO 99/29130 PCT/SE98/02198
-16-
td~=TNl *Tjlor-slr (3)
As such, the number of timeslots between the start of TN1 and the start of a
TDMA frame in the detector BS2 is the integer part of the expression,
td;~'T$,or, or
IN'TEGER(t~;~T',,o~. However, a problem arises when TNl *Ta,o~ is smaller than
8,2
when t~;f becomes a negative value. In that case, TN2 also becomes a negative
value,
which corresponds to a timeslot in the preceding TDMA frame. Consequently, the
number "8" is added to the negative TN2 to provide the correct value for TN2.
A
general expression for TN2 is thus given by:
TN2 =MOD(INTEGER( tdrj )~8).
(4)
Tslor
However, Equation (4) can be better expressed as:
TN2 =INTEGER( t~ ), if tdr~0
T riot
= 8 +INTEGER( tdlj ), if t <0. (5)
dlj
T slot
At step 112, the schedule time for TOA measurements is calculated. However,
the global time has already been retrieved (step 106) as the sync position,
T1. As
such, the TOA measurement time can be calculated by:
Twsoi =Tl +n *T jrowts'~TN*Tilor~
where T=,ot is the duration of the timeslot. The FN can then be recalculated
as:
FN=FN+n,
where FN is in the serving cell at Tm~, and the value of "n" is set high
enough to
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE9$/02198
-17-
provide enough margin for the time that elapses from the sampling of Tl
through the
allocation and activation of detectors and ordering of the measurements to be
made.
The sampling time is aligned with the start of the serving channel's timeslot
(TN).
The Measurement Commands are conveyed to the detectors using a similar method
_
as described for step 112 for the first embodiment. The calculated FN is also
conveyed to the detectors.
At step 114, the method is the same as described above for step 114 for the
first
embodiment, but with the following exception. For example, referring to the
exemplary frame timing diagram illustrated in FIGURE 5, it can be seen that
although
1 o Tm~s is aligned with the serving channel's timeslot, T",~ can fall
anywhere within a
detector's (BS2 or BS3) timeslot. However, at the bottom portion of FIGURE 5,
a
number of possible window locations are shown with two consecutive timeslots.
The
present method's use of two consecutive timeslots for TOA measurements in a
detector (e.g., in BS2 or BS3) guarantees a measurement window, which
corresponds
to the unshaded zones shown in FIGURE 5.
At step 116, the same method is followed as step 116 in the first embodiment
(reporting TOA measurements, and calculating TDOAs and the MS's position), but
with the following exception. For this embodiment, the tagging of TOA
measurements should be based on the FN associated with the serving cell.
2 o Alternatively, the global time of each sample can be used instead.
There are also problems with logical channel multiplexing and discontinuous
transmissions for this second embodiment. The resolution of these problems is
similar
to the resolution described above for the first embodiment, but with the
following
exception. For this embodiment, the FN used in a detector has no fixed
relationship
2 5 with the FN of the serving cell. Instead, the channel multiplexing used on
the
timeslot(s) allocated for detection should be based on the FN delivered in the
TOA
Measurement Command. It is known that at T",~, this FN is equal to the FN in
the
serving cell. Then, if the measurement period is longer than one burst, the FN
should
be incremented every TDMA fi~ame, starting with the delivered value of FN.
3 o In accordance with a third exemplary embodiment of the present invention,
TDOA positioning of an MS can be performed using controlled transmissions of
access bursts in a synchronous network. Essentially, as in the first
embodiment, the
CA 02313053 2000-06-O1
WO 99/29130 PGT/SE98/OZ198
-18-
network is assumed to be synchronized (i.e., the air interface timing is
common for all
cells in the network). However, the cells can be offset in the FN counting.
Notably,
this method is similar to a small extent to the method described in the above-
cited U.S.
Patent Application Serial No. 08/977,470 filed November 24, 1997 (hereinafter,
the
"'470 Application"). A significant difference is that the method used in this
third
embodiment measures TOAs instead of access delays. An access delay is the time
between the start of a timeslot and the time the access burst is received. As
such, the
access delays can be used to calculate absolute distances, instead of the
relative
distances calculated by using the prior TDOA methods. However, in a
synchronous
network, the common timeslot border is the same as the global measurement
time, so
the access delay may be interpreted as a form of TOA measurement. As such, a
TDOA method can also be used for calculating an MS's position, which would
improve the measurement accuracy because the downlink's and MS's parts of the
access delay are insignificant when measuring the different arrival times.
For this exemplary embodiment, after receipt of the Positioning Request for
an MS ( 14), the serving cell and actual serving channel are noted. Next,
detectors to
be involved are allocated in a number of surrounding BSs. The MS is then
commanded (e.g., by a Handover command) to begin transmitting access bursts.
Sampling of the TOA of the transmitted signals is performed simultaneously in
all of
2 o the allocated detectors. From one to several of the transmitted access
bursts can be
used for the TOA measurements. Alternatively, a variation of this method is to
configure a common detector channel in all cells. and order the MS (whose
position
is to be determined) to transmit access bursts on this common channel.
There are a number of additional advantages of this method compared with the
2 5 methods described above for the first and second embodiments. For example,
the
Training Sequence for this method is longer for an access burst than for a
Normal
Burst (e.g., in the GSM, 41 bits as opposed to 26 bits), which implies much
higher
sensitivity for this method. Furthermore, no burst will arrive before the
beginning of
the timeslot, because the transmitted access bursts are not Time Advance
regulated.
3 0 Moreover, the access burst transmissions are controlled. Consequently,
there are no
problems with intermittent transmissions. Access bursts arriving from MS's
other
than the MS involved can be discriminated by use of a control parameter (e.g.,
CA 02313053 2000-06-O1
WO 99/19130 PCT/SE98/02198
-19-
Handover Reference). Consequently, there is no need to synchronize with
logical
channel multiplexing. On the other hand, for this embodiment, the position of
an MS
can be determined, but the ongoing communications between the MS and the
serving
cell/channel may be disturbed in the process.
Specifically, FIGURE 6 is a flow diagram that illustrates a method for
determining the position of an MS using controlled transmission of access
bursts in
a synchronous network, in accordance with the third exemplary embodiment of
the
present invention. For steps 202 and 204, these steps are similar to steps 102
and 104
in the first embodiment shown in FIGURE 2. At this point, for the next step,
there are
two options available. One option is to make TOA measurements on a common
detector channel which is kept active in all cells. For this option, steps 206-
210 can
be bypassed (dotted line). The second option is to allocate detectors in the
surrounding cells, as performed in the first and second embodiments.
In that event, at step 206 (similar to step 106 in the first embodiment) the
serving cell information is retrieved. If the actual channel used for
communication
with the MS involved is to be used for the transmission of access bursts for
measuring
TOAs, then the Channel Description information is also retrieved at this step.
Otherwise, if this communication channel is not to be used for TOA
measurements,
then the Channel Description information is not retrieved. However, the TSC is
not
2 0 needed in this event. Step 208 is similar to step 108 employed for the
first
embodiment.
At step 210, the method is similar to step 110 performed for the first
embodiment, but with the following exceptions. Since the TOA measurements in
this
third exemplary embodiment use a different type of burst than used for the
first and
2 5 second embodiments, the detectors' correlators have to be loaded with this
information. As such, in the GSM, the access bursts have 41 predefined "synch
sequence bits" that can be used for this purpose. At each BS involved, the
reception
and sampling of the incoming data is started immediately after activation of
the
measurement channel on the allocated TN in all TDMA frames (regardless of the
FN).
3 0 The Normal Bursts h~ansmitted from the served MS or other MSs on the same
TN can
be heard but discarded by the correlator(s) involved. The access bursts are
detected
(step 214 below).
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-20-
Step 212 is substantially different than step 112 in the first embodiment. For
example, the '470 Application describes two methods for commanding an MS to
transmit access bursts. One such method is to issue a Handover Command, and
the
second method is to issue a special, non-standard Positioning Command. FIGURE
7
is a diagram that illustrates how the Handover Command method in the '470
Application can be implemented for this embodiment. Referring to FIGURE 7, the
Handover Command contains, for example, a Cell Description which is equal to
the
allocated TOA measurement channel, and a Handover Reference. An asynchronous
handover procedure is indicated, which commands the MS to transmit access
bursts
continuously until a time-out period has elapsed. After the time-out period,
the MS
reverts to the original channel used. A special case (preferred) is to perform
a
handover to the MS's own cell and channel, which minimizes the risk of losing
the
MS's connection.
In using this handover method, the new channel is synchronized with the
original channel. Consequently, a synchronous handover procedure can be used
instead. In this case, the MS can transmit access bursts in only four
consecutive
TDMA flames, and then it switches to transmitting Normal Bursts. With this
method,
it is possible to command a handover to an original channel in the serving
cell, which
implies a shorter interruption of the ongoing connection than for an
asynchronous
2 0 handover procedure. Note, however, that four access bursts may not be
sufficient to
attain the accuracy required.
At step 214, the same method followed in step 114 of the first embodiment can
be used, but with the following exceptions. The measurement process was
already
begun in step 210. At step 214, the access bursts transmitted at step 212 are
detected.
2 5 FIGURE 8 illustrates the situation when the access burst transmitted by a
MS (as a
result of step 210) is received by the serving BS and the detectors. The TOA
values
(tl, t2, t3) are measured with respect to the timeslot border. Using this
method, the
access burst cannot arrive prior to the beginning of the timeslot (since the
network is
synchronized). In the serving BS, the TOA value, tl, conds to the access
delay,
3 0 which can be used to command the MS to advance its timing when Normal
Bursts are
being transmitted.
Step 216 is similar to step 116 in the first embodiment, but with the
following
CA 02313053 2000-06-O1
WO 99/29130 PCT/SE98/02198
-21-
exception. The measurement period is not related to any fixed point in time.
Instead,
the detectors can be directed to measure the TOA of a number of consecutive
access
bursts from the MS involved, or during a certain period of time, start
measuring with
the first detected and accepted access burst. The acceptance of an access
burst can be
based on the Handover Reference or a similar reference. Advantageously, there
are
no problems related to logical channel multiplexing or discontinuous
transmissions
when the present method of controlled transmission of access bursts is used.
In accordance with a fourth exemplary embodiment of the present invention,
TDOA positioning of an MS can be performed using controlled transmissions of
access bursts in an asynchronous network. Essentially, this method is a
variant of the
method described above for the third embodiment, except this method is for
asynchronous networks. Referring to FIGURE 6 (which can also be used to
illustrate
the fourth embodiment), upon receiving a Positioning Request at the MSCBSC,
the
following steps can be performed.
At steps 202 and 204, similar to the steps described above for the third
embodiment, if the MS involved is in the idle mode, than a connection is setup
with
the MS and the network. Step 206 is similar to step 206 in the third
embodiment, but
with the following exception. If the network is not synchronized, then it is
not useful
to retrieve an FN Offset. However, the "sync position" Tl can be read (i.e.,
the actual
2 0 global time at the start of a TDMA flame). The global time is assumed to
be available
at each site (e.g., using GPS receivers). The sync position, T1, is thus
sampled at the
start of an arbitrary TDMA flame. In order to be able to synchronize the
network with
logical channel multiplexing, the actual FN at the sync position can also be
retrieved.
Step 208 is similar to step 208 for the third embodiment (the surrounding BSs
2 5 are located). Step 210 is similar to step 210 for the third embodiment,
but with the
following exception. The reception and sampling of incoming data are not
started
immediately at activation. Instead, the method used for step 110 in the second
embodiment is used (i.e., the start time of the TOA measurements is
scheduled). For
example, step 110 in the second embodiment describes how the sync position of
each
3 0 detector's BS in the asynchronous network is calculated. The same
description can
be used at step 210 herein for the fourth embodiment.
At step 212, the MS involved is commanded to transmit access bursts on the
CA 02313053 2000-06-O1
WO 99!29130 PCT/SE98l02198
-22-
measurement channel. The method used for step 112 in the second embodiment
(calculating the measurement time, Trod is used for step 212 in the fourth
embodiment. When calculating the global sampling time, an additional margin
should
be calculated for the time it takes to command the transmission of access
bursts. Next, .
the step 212 for the third embodiment is performed.
At step 214, the method described above for step 114 in the second
embodiment is used to perform the TOA measurements. FIGURE 9 is a diagram that
illustrates how the TOA measurements can be made, in accordance with the
fourth
exemplary embodiment of the present invention. The description provided above
for
step 114 in the second embodiment can also be used herein with respect to
FIGURE
9. As a result of using two-slot measurement windows in the detectors' BSs for
this
step, a guaranteed measurement window can be provided (unshaded portion) for
each
detector. Finally, for step 216, the measured TOAs are reported, and the TDOAs
and
the MS's position are calculated. This step is similar to step 116 described
above for
the second embodiment. As such, in accordance with fourth embodiment, there
are
no problems associated with logical channel multiplexing and discontinuous
transmissions when the present method with controlled transmission of access
bursts
is used.
Although a preferred embodiment of the method and apparatus of the present
2 0 invention has been illustrated in the accompanying Drawings and described
in the
foregoing Detailed Description, it will be understood that the invention is
not limited
to the embodiment disclosed, but is capable of numerous rearrangements,
modifications and substitutions without departing from the spirit of the
invention as
set forth and defined by the following claims.
