Note: Descriptions are shown in the official language in which they were submitted.
3 ~ Y~
APD-84-2
58-231F
ENHANCED GLOBAL POSITIONING
SYSTEM DELTA RA~G~ PROCESSING
By ~1 i son Brown
Field of the Invention
The present inventicn relates to an apparatus and
method for improving the accuracy obtained from a global
positioning system which comprises a plurality of
satellites each broadcasting two or more navigational
signals.
Background of the Invention
The NAVSTAR Global Positioning System (GPS) is a
satellite-based radio-navigation system intended to
provide highly accurate three-dimensional position and
precise time on a continuous global basis. When the
system becomes fully operational in late 1988, it will
consist of l~ satellites in six orbital planes, and three
active spares.
Each satellite will continuously transmit
navigation signals at two carrier frequencies, Ll =
1575.42 MHz and L2 = 1227.6 MHz consisting of the P-code
ranging signal (a 10.23 MBPS pseudonoise code), the
C/A-code ranging signal (a 1.023 MBPS pseudonoise code),
and 50 BPS data providing satellite ephemeris and clock
bias information. Unbalanced Quadri-Phase Shift Keying
(UQPSK) modulation is utilized with the clata bits added
to the ranging codes, the C/~-cocle signal lagging the
P-code signal by 90; and the C/A-code signal power
nominally exceeding the P-code signal power by 3dB.
Navigation using GPS is accomplished by passive
triangulation. The GPS user equipment measures the
Pseudo-Range to four satellites, computes the position of
the four satellites using the received ephemeris data;
and processes the Pseudo-Range measurements and satellite
positions to estimate three-dimensional user position and
--1-- ~r,
q~
3 ~'
APD-84-2
58-231F
precise time.
GPS receiver signal processing can be partitioned
into three parts: RF signal processing, estimation of
In-phase (I) and Quadrature-phase (Q) signals, and
subsequent processing of these I and Q signals to
implement code and carrier tracking, data demodulation,
SNR estimation, sequential detection, and lock detection
functions. Traditionally, all three parts are
implemented using analog signal processing techniques.
10A paper briefly reviewing analog RF signal
processing implementation which describes three digital
signal processing (DSP) techniques for implementing the
I/Q generation function, and describes DSP algorithms for
implementing the subsequent processing functions may be
15found in the NTC '83 lEEE_1983 National Teles~stems
Conference papers, entitled "Digital Signal Processing
Techniques For GPS Receivers," by Mark A. Sturza,
November 1983.
A Delta-Range (DR) measurement is derived from the
difference in carrier phase over a fixed time interval.
Position can be estimated using Delta-Ranges through the
change in the satellite position over the observation
interval. In the prior art, it would generally take
approximately 24 hours of processing Delta-Range
measurements to obtain a reasonable accuracy for the
location of the receiver utilizing a global positioning
system.
Accordingly, it is an object of the present
invention to reduce the amount of time required to obtain
an accurate location of the receiver utilizing a global
positioning sytem.
~5~37
2196-472
Summary of the Invention
The present invention replaces the measurement of
the Delta-Range over a plurality of predetermined time intervals
with an improved configuration. Under tne improved configuration,
the Delta-Range measurement, generally measured over a one-second
period, is accumulated for the first second. Therea~ter, tne
Delta-Range is accumulated for the first and second second;
followed by accumulation of the Delta-Range for the first, second
and third second; and so on. The heart of the invention is the
realization that such accumulation of all Delta-Range measurements
for each new time interval and all old time intervals eliminates
many of the errors created when measuring a Delta-Range for an
individual time period. By adding information from previous
Delta-Range measurements to the newest Delta-Range measurement,
it is possible to determine the position of the receiver utilizing
a global positioning system in a much faster time time frame.
For example, an accurate position can be determined utilizing the
global positioning system and the present invention in less than
eight hours. This is an improvement of better than 3:1.
Thus, in accordance with a broad aspect oE the
invention, there is provided a method for enhancing the information
received from a global positioning system which includes a
plurality of satellites each transmitting at least two carrier
frequency signals, comprising the steps of:
a) identifying a plurality of selected satellites;
b) setting information storage units for accumulated
Delta-Range information for each selected satellite -to zero;
~S~S63~
2196-472
c) tracking the selected satellites;
d) receiving said carrier frequency signals from said
selected satellites;
e) measuring the phase of said carrier frequency signals
from said selected satellites;
f) storing changes in phase of said carrier frequency
signals over a predetermined time interval in said storage units
to create a Delta-Range measurement for each selected satellite;
g) accumulating the Delta-Range measurement by adding the
above-mentioned accumulated Delta-Range information for each sat-
ellite to the Delta-Range measurement from step f for each
satellite;
h) processing the accumulated Delta-Range information
from each selected satellite through a filter which receives
said input information and produces said information into
output information, said output information being position
information; and
i) repeating steps f, g and h until said position informa-
tion is refined to the accuracy desired.
In accordance with another broad aspect of the
invention there is provided apparatus for enhancing the infor-
mation received from a global positioning system which includes
a plurality oE satellites each transmitting at least two carrier
fxequency signals, comprising:
a) means for selecting a plurality of satellites whose
transmitted signals are to be received;
b) means for storing;
~L~54~637
2196-472
c) means for setting each means for storing to zero;
d) means for receiving said carrier frequency signals from
said selected satellites;
e) means ~Gr measuring the pha~e of said carrier
frequency signals received from said selected satellites;
f) means for measuring changes in phase of said carrier
frequency signals over a predetermined time interval;
g) means ~or placing said measured changes in phase in
said means for storing over said time interval to create a
Delta-Range measurement for each selected satellite;
h) means for accumulating said Delta-Range measurement
by adding the above mentioned accumulated Delta-Range information
for each satellite to said Delta-Range measurement;
i) filter means for processing said accumulated Delta-Range
information from each selected satellite for producing an
output representing position information.
In accordance with another broad aspect of the invention
there is provided an apparatus for improving the accuracy of
measurement informati.on obtained from a plurality of satellites
of a global positioning system, said apparatus comprising:
a satellite selectable global positioning system receiver,
said receiver being capable of receiving A-t least two carrier
frequencies from each selected satellite and tracking the differ-
ence between their phase measurements, to produce Delta-
Range signals ~-hich are derived from said phase measurements;
sai.d global positioning system receiver being operatively
associated with a resettable data storage and accumulation means,
-3b-
Pr 5 ~r~37
2196-472
said receiver providing said data`storage and accumulation means
with said Delta-Range signals;
said resettable data storage and accumulation means storing
and accumulating said Delta-Range signals over a ~irst time
interval;
said resettable storage means processing said Delta-Range
signals to develop a first set of Accumulated Delta-Range signals,
said first set of signals being a function of the accumulation of
Delta-Range signals at said storage means during said first time
interval;
said storage means operatively connected to a filter means,
said storage means transmitting said Accumulated Delta-Range
singals to said filter and simultaneously feeding back said
slgnals to the storage means, so that said Accumulated Delta-
Range signals may be stored in said storage means during a sec-
ond time interval and processed together with a second set of
Delta-Range signals, for deriving a second set of Accumulated
Delta-Range signals at the end of said second time interval,
at which time said second set of Accumulated Delta-Range signals
are transmitted to said filter means;
said filter means further processing said :Eirst and said
second sets of Accumulated Delta-Range signals from each selected
satellite to develop an output signal representative of enhanced
position information.
In accordance with another broad aspect of the inven-
tion, there is provided a method for improving the accuracy
~L~S~ç~37
2196-472
of position measurement information obtained from a global
positioning system having a plurality of satellites, eac~ of
said satellites transmitting at least two carrier frequency
signals, comprising the steps of:
a) receiving the transmitted satellite carrier frequency
signals;
b) processing said satellite signal through a receiver
means capable of identifying and tracking each of the
satellites according to their transmitted carrier frequency sig-
nals;
c) deriving an output signal from said receiver means
which is characteristic of Delta-Range information for each of
said satellites;
d) processing said output signal through a data
accumulation and storage means, deriving an Accumulated Delta-
Range information signal, by storing said signal for a pre-
determined time interval, creating an Accumulated Delta-Range
information signal corresponding to each satellite;
e) repeating said step (d) to further refine the signal
so processed, whereby an Accumulated Delta-Range signal is
derived which corresponds to an accurate position :Eo:r each
satellite to the degree desired.
Description of the Drawings
A better understanding of the benefits and
advantages of the present invention will be available after a
careful review of the following specification and drawing,
wherein:
-3d-
~5~1~37
2196-472
Figure 1 is a block diagram showing the enhanced
glo~al positioning system Delta-Range positioning of the
present invention.
Description of the Preferred Embodiment
. _
A GPS receiver makes a measurement of Pseudo-Range
(PR), the difference between the time of reception and time
of transmission of the GPS signal, from tracking the
-3e-
APD-84-2
58-23lF
Pseudo Random Noise (PRN) codes modulated on the GPS
carrier. The Delta-Range ~DR) measurement is derived
from the difference in carrier phase over a fixed time
interval, generally one second, measured using a Phase
LocXed ~oop ~PL~). The Pseudo-Ranges and Delta-~anges
are related to user position error, Ru, velocity error,
Vu, and user clock and frequency offsets, Bu and ~u
through equations l.l - 1.5.
~ ~c~ 1.1
~ k k (~ k ) Bk 1.2
DRi ~ PRk _ p~ l 1.3
P~i PRi ~ PR~ ku + Bk 1.4
DRi ~ DRl - DRi C (~ U + ~ + ~ku 1.5
11 L~S from user to ith satellite at time tk.
R Position of user.
--4--
~.S~3 ~
APD-84-2
58-231F
Velocity of user.
~ ~ Position of ith satellite.
PR~ Pseudo Range to ith satellite at time tk.
DRk Delta-Range to ith satellite at time tk.
~u Users clock offset from GPS time.
bB~ Chan~e in user's c~oc~ offset over one sec.
LOS is an abbreviation for Line Of Sight Vector
which incorporates the elevation angle and azimuth angle
of ~ satelliteO
For static point positioning the user's velocity is
known to be zero and so equation 1A5 simplifies to
D-Rk ~ (~ 1 )- R + ~Bk 1.6
The LOS vector to a GPS satellite does not change
significantly over a one second interval, so to improve
the geometry of the navigation solution it is advan-
tageous to accumuiate Delta-Ranges even longer time
intervals to form a new measurement variable, Accumulated
Delta-Ranges (ADR).
Since the Delta-Ranges are formed from the
L~37
APD-84-2
58-231F
difference between two phase measurements from the
carrier tracking phase-locked loops, it is important to
note that Accumulating Delta-Ranges to form a new
measurement do not increase the measurement error.
DRk , ~ (nk + ~2~ ~ n 2~ ) 1.7
ADR ~ ~ DR ~(n 2~ 2
~=1
+ nk-l + ~k 1 nk-2 ~ 2
. . , . ~ nl + ~2 ~ n c )
ADRk y ~(n~ + 2~ ~ n ~ 2~ ) 1.8
Note, that the term n represents an integer number
of cycles whereas the term 0k represents the number of
radians in each part cycle of the phase shift. The term
nk l, represents the number of integer cycles one second
before, whereas the term 0k l represents the number of
radians within the partial cycle from the sample the
second before. These Delta-Range differences are added
together wherein the final Accumulated Delta-Range
includes all of the information from the initial start
through to the final measurement at n time period.
$~
APD-84-2
58-231F
Defining the phase measurement errors to be:
E[02] = u~ E~0k0j] = O j~k
E[0~ = o
Then
i
15DRk O ~ Var ~DR~ - 2a~ 7 DRkDRk 1 _ at 1. 9
ADR ~0 Var ~ADR~ - a~ 1.10
i
Since adjacent Delta-Range measurements share a
common phase measurement, the phase errors associated
with these common measurements cancel out in the
Accumulated Delta-Range measurement. The Delta-Ranges
have zero mean error but are correlated between adjacent
time intervals. Alternatively, the Accumulated Delta-
Ranges share a common bias error, the initial carrier
phase error, but apart from this bias all samples are
uncorrelated.
The measurement equation for the Accumulated Delta
Range measurement becomes
3 7
APD-84-2
58-231F
A~k ADR~ ADRk , (~ 1) ~ k ~0 1.11
As time increases, the satellite geometry for
estimating position error improves. Because of the
independence between each measurement, it is possible to
process an Accumulated Delta-Range every second from each
satellite tracked. This allows position to be determined
to far greater accuracy than by processing Pseudo-Range
Measurements as in the prior art.
As shown in equation 1.10 the Accumulated
Delta-Range measurements, generated every second, are
independent of each other and each have the same
measurement noise as the one second Delta Range
Measurements. Four Accumulated Delta Ranges from four
satellites being tracked are generated every second and
can be processed optimally using a Kalman filter to
estimate user position and satellite errors.
As is well-known, a Kalman filter is a device which
receives inputs from two or more variables and processes
those variables to produce an output which is a function
of the input variables. For example, if pressure and
temperature were known inputs, it would be possible to
feed this information into a Kalman filter to produce an
output representing altitude. A similar process is used
3~
APD-84-2
58-231F
here. Wherein the Accumulated Delta-Range Measurements
from the four satellites are applied to a Kalman filter
which produces an output representing the position of the
receiving station.
The most significant error sources are the
0 satellite ephemeris errors and the satellite clock error.
These errors can be modeled by including four additional
states per satellite in addition to the four user
position and clock frequency offset states, as shown in
Table 1. The observation equation now becomes
ADRk y tlk ~ li) ~ (Ru ~ R~) + Bk _ Bu _~k + a~ 1.12
The clock errors are all lumped together in one clock
offset term per satellite which estimates the change in
the relative user and satellite clock error over the
accumulation time interval.
~B~ Y (~k _ B0) _ (~k _ B0) i 1
APD-84-2
58-231F
TABLE 1 - KAL~N FILTER STATES
_ __ _ ._
State Description nitial State
No. . ariance Noise
- ---- _ _
1 Xu ~ Position Error lOOCm
2 Yu - Position Error lOOOm
3 Zu - Position Error lOOOm
4 ~Bu - Clock Frequency lOOOm 0.67mm
~a - - - - -- -
5,9,13,17 aB Change in Clock Offset 0 2mm
over accumulation inverva
6,10,14,1a Alon~ Trac~ Satellite 0.8
7,11,15,19 Radial Satellite Error 6.3
8,12,16,20 Cross Track Satellite 3.0
15 ___ . Error . _
This term can also be used to include noise terms
to model the range errors introduced by atmospheric
effects. The relative clock offset, B, is propagated
using the estimate of the users clock frequency offset
every second.
State Propagation ~Bi 1 ~ ~Bi + ~Bu 1.1
--10--
#~
APD-84-2
58-231F
The measurement equation can be written in vector
format, as shown for ~he second sa~e71ite tracked.
2 ~lN , oT ~ T , QT ' T]- 1 15
Q1N is the delta LOS vector from the satellite to the
u~er, ( 1 - 1), expressed in local level coordinates.
~ 1~ is the same ~ector but expressed in orbital frame
coordinates for the second GPS satellite selected, that
is with XO in the Along Track Direction, YO in the Radial
direction and ZO in the Cross Track direction.
Processing the Accumulated Delta-Ranges using this
Kalman filter gives position to significantly greate~
accuracy ~han that normally achievable by GPS, even for
P-code users. After tracking for two hours at lower
latitudes the Root Sum Square (RSS) position error drops
to around 1.0 m RSS and continues to improve gradually,
reaching about 0.3 m RSS after about seven hours. At
higher latitudes the position accuracy degrades slightly
due to the different satellite geometry, but approxi-
mately 1 m RSS is achievable in around five hours.
Referring now to the drawing, Fig. 1 shows a
receiver 10 which receives transmitted navigational
signals froM the satellites that comprise the global
positioning system via an antenna 12. The transmitted
signals are applied to a satellite select circuit 14
which selects the most desirable satellites to be tracRed
by the global positioning system receiver 16. Once each
satellite has been selected, the storage units 18 for
each selected satellite are set to zero or cleared of any
stored information contained therein. The global
63'7
APD-84-2
58-231F
positi~ning receiver then tracks the particular
satellites selected and receives the carrier frequency
signals from each. After measuring the phase of each of
the received carrier signals, the change of phase in the
carrier signals is stored within the information storage
units 18. In the preferred embodiment, this information
is stored over a one second time interval. The Delta-
Range measurements are accumulated within the storage
units 18 so that the first Delta-Range dstermined by the
first one second interval is accumulated with the
Delta-Range information from the second time interval and
so on. The outputs of the storage units 18 are applied
to a Kalman filter 20 and back to the input of storage
units 18 to create Accumulated Delta-Range information.
At each time interval, the step of reapplying the output
of the storage unit 18 to its input is repeated so there
is an accumulation of the Delta-Range information. These
steps are repeated as often as necessary until the
position information at the output 22 of the Kalman
filter represents the accuracy desired.
It will be understood that the present invention
may be practiced through the utilization of software or
hardware; hardware being represented by the block diagram
of Fig. 1.
3~
