Telecommunications – Radiotelephone system – Zoned or cellular telephone system
Reexamination Certificate
2001-04-17
2004-02-03
Trost, William (Department: 2682)
Telecommunications
Radiotelephone system
Zoned or cellular telephone system
C455S425000, C455S456100, C455S334000, C342S450000, C375S142000, C375S150000, C375S349000
Reexamination Certificate
active
06687507
ABSTRACT:
FIELD OF INVENTION
The present invention relates to the field of mobile radio telecommunications, and more particularly, to determining the location of mobile stations within the coverage area of a radio telecommunications network using time of arrival (TOA) estimations.
BACKGROUND
The problem of determining the location of a mobile station (MS) is of considerable interest. The primary application that is driving this activity is the positioning of E911 callers in the United States. The United States Federal Communications Commission has imposed a requirement wherein operators, by October 2001, must report the position of emergency callers within their service area. Also, the European Union has proposed a similar law for all 112 callers, which is to take affect by January 2003. In parallel, different vendors of mobile communication equipment have presented solutions to this problem to fulfil these legal requirements.
In GSM, four different position location methods have been standardized to enable operators to offer location-based services. Accordingly, in addition to providing the position of emergency callers, it is likely that mobile positioning will open the door into a new dimension of mobile services and applications that use the subscriber position as input. For example, the position of a subscriber can be used to provide the subscriber with information about restaurants in proximity to the subscriber.
The cellular positioning techniques available today can be divided into network based solutions and terminal based, e.g., mobile station based, solutions. A network-based solution standardized in GSM is the Uplink Time-of-Arrival (TOA) positioning method, which does not require changes to the mobile station. A mobile station based solution standardized in GSM is the Enhanced Observed Time Difference (E-OTD) method.
The core measurements performed by the mobile station to support the E-OTD location method are Time-of-Arrival (TOA) measurements. The mobile station listens to the broadcast control channel (BCCH) carrier of a certain cell and measures the TOA of bursts relative to its own time base. OTD values are formed by subtracting the TOA measurement of a neighbor cell from the TOA measurement of the serving cell. To obtain an accurate position of the mobile station, the TOA's must be estimated with a high accuracy. For example, a TOA error of 1 bit (i.e. 1 sampling point) corresponds to approximately 1100 meters range error in the position estimation.
For TOA estimation, the mobile station can use normal bursts, synchronization bursts, dummy bursts or a combination thereof. It is not necessary to synchronize to the neighbour base station in order to perform the TOA measurements. The TOA measurement strategy is similar to the neighbour cell measurements in GSM, i.e., where the mobile station is required to perform neighbour cell measurements (e.g. signal strength measurements) in order to find possible candidates for a handover. In principle, the TOA measurements and the neighbor cell signal strength measurements can be made in parallel. The mobile station can be provided with assistance data by the network, which allows predicting the TOA value together with an uncertainty. This defines a correlation search window within which the TOA is expected to be. Therefore, the mobile station knows when to measure the TOA for a particular signal and can schedule the TOA measurements for the individual links accordingly. For more information regarding correlation windows, the interested reader should refer to U.S. patent application Ser. No. 09/186,192 “Improvements In Downlink Observed Time Difference Measurements” by A. Kangas et al., which is herein expressly incorporated by reference.
The choice between synchronization bursts or normal bursts depends, e.g., on the requested response time and the mode of the mobile station. Although the synchronization bursts offer the best correlation properties, these bursts occur very infrequently, i.e., only once every 10 TDMA frames, whereas normal bursts are available at most 8 times per frame. To enable a quick measurement response from the mobile station in dedicated mode, e.g. during emergency calls, it may therefore be necessary to measure on normal bursts.
One problem for TOA estimation is that a mobile station must be able to hear a sufficient number of base stations. The signal strength from neighboring base stations may be very low, resulting in a low signal-to-noise ratio, typically −10 dB. Multipath propagation is also a problem. The multipath propagation channel sets the limit on the estimation accuracy. In co-pending U.S. patent application Ser. No. 09/354,175 “Efficient Determination of Time of Arrival of Radio Communication Signals” by E. Larsson et al., which is herein incorporated by reference in its entirety, a simple TOA estimation algorithm with very low complexity is described for estimating TOA at low signal-to-noise ratios. This algorithm is based on the Incoherent Integration (ICI) with Multipath Rejection (MPR) principle presented in International Patent Publication WO-9927738, which is also incorporated herein by reference in its entirety.
In accordance with the ICI principle described in the above-identified International Patent Publication, the received burst i is first correlated with the known training sequence, to obtain the correlation result c
i
(k); as indicated below in equation (1):
c
i
(
k
)={tilde over (b)}
i
(
k
)*
TS
(
k
)
i
(1)
where {tilde over (b)}
i
(k) is the received, de-rotated burst, TS(k) is the known training sequence contained in the burst {tilde over (b)}
i
(k) and * is the correlation operator. This correlation is performed for a number of M received bursts. The absolute squares of the M correlation results c
i
(k) are summed, as shown in equation (2).
&psgr;(
k
)=&Sgr;
i=o
M−1
|c
i
(
k
)|
2
. (2)
The effect of this summation is that the noise in the correlation result is reduced and the maximum (i.e. the TOA) is more likely to be detected. Performing a weighted summation can increase the detection probability, per equation (3):
&psgr;(
k
)=&Sgr;
i=o
M−1
w
i
|c
i
(
k
)|
2
, (3)
where the weights w
i
are based on the estimated SNR. Since the weights w
i
are difficult to estimate, an alternative ICI method based on the maximum likelihood criterion, also described in co-pending U.S. patent application Ser. No. 09/354,175, is presented in equation (4) below:
&psgr;
log
(
k
)=&Sgr;
i=o
M−1
log(
E
s
E
bi
−|c
i
(
k
)|
2
), (4)
where E
S
is the energy of TS(k) and E
bi
is the energy of {tilde over (b)}
i
(k). The sum of logarithms is the logarithm of the product and since the logarithm is a monotonic function, the maximum (or minimum) of log (a·b·c) is the maximum (or minimum) of (a·b·c). Therefore, equation (4) reduces to:
&psgr;
logi
(
k
)=&psgr;
log(i−l)
(
k
)(
E
s
E
bi
−|c
i
(
k
)|
2
), (5)
The minimum value of the cost function, as illustrated above in equation (5), k
min
, is the desired TOA in sampling point units. With the detected k
min
, an estimate of the channel impulse response is performed for each burst and interpolated to give the desired resolution.
FIGS. 1 and 2
respectively illustrate the TOA estimation performance of the ICI algorithm in a static one-peak channel with additive White Gaussian noise (AWGN) and Co-channel interference (CCI). The Figures illustrate the root-mean-square error (RMSE, 90%) in microseconds as function of signal-to-noise ratio E
S
/N
0
(
FIG. 1
) and C/I (
FIG. 2
) for a different number of GSM normal bursts used in the incoherent integration process. The results illustrated in
FIGS. 1 and 2
assume that the transmitted bursts are GSM normal bursts and that the receiver assumes that GSM normal bursts have been transmitted.
As illustrated in
FIG. 1
, the TOA estimation error is characterized by a large scale error region at lo
Fischer Sven
Kangas Ari
Ewart James D
Telefonaktiebolaget LM Ericsson (publ)
LandOfFree
Time of arrival estimation for edge/GSM does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Time of arrival estimation for edge/GSM, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Time of arrival estimation for edge/GSM will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3336408