U.S. patent application number 15/025722 was filed with the patent office on 2017-09-14 for method of reducing inter-channel biases in glonass gnss receivers.
This patent application is currently assigned to TOPCON POSITIONING SYSTEMS, INC.. The applicant listed for this patent is LIMITED LIABILITY COMPANY "TOPCON POSITIONING SYSTEMS". Invention is credited to KONSTANTIN VLADIMIROVICH CHERESHNEV, ANDREY VLADIMIROVICH VEITSEL, VLADIMIR VICTOROVICH VEITSEL.
Application Number | 20170261617 15/025722 |
Document ID | / |
Family ID | 58386731 |
Filed Date | 2017-09-14 |
United States Patent
Application |
20170261617 |
Kind Code |
A1 |
VEITSEL; VLADIMIR VICTOROVICH ;
et al. |
September 14, 2017 |
METHOD OF REDUCING INTER-CHANNEL BIASES IN GLONASS GNSS
RECEIVERS
Abstract
The present invention discloses methods of accuracy improving
for code measurements in GLONASS GNSS receivers. One component of
error budget in code measurements of GLONASS receivers is caused by
a difference in signal delays arising in the receiver analog Front
End and antenna filter on different channel frequencies specific to
GLONASS satellites. Methods to compensate for differences in delays
for different GLONASS channel frequencies have been proposed using
data collected from a GLONASS signals simulator.
Inventors: |
VEITSEL; VLADIMIR VICTOROVICH;
(Moscow, RU) ; VEITSEL; ANDREY VLADIMIROVICH;
(Moscow, RU) ; CHERESHNEV; KONSTANTIN VLADIMIROVICH;
(Moscow, RU) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
LIMITED LIABILITY COMPANY "TOPCON POSITIONING SYSTEMS" |
Moscow |
|
RU |
|
|
Assignee: |
TOPCON POSITIONING SYSTEMS,
INC.
Livermore
CA
|
Family ID: |
58386731 |
Appl. No.: |
15/025722 |
Filed: |
September 23, 2015 |
PCT Filed: |
September 23, 2015 |
PCT NO: |
PCT/RU2015/000598 |
371 Date: |
March 29, 2016 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01S 19/23 20130101 |
International
Class: |
G01S 19/23 20060101
G01S019/23 |
Claims
1. A method of reducing inter-channel bias in a GLONASS receiver,
the method comprising: for each satellite channel, storing
correction data for different temperature in a memory; measuring a
current temperature of an analog Front-End of the GLONASS receiver;
receiving signals from GLONASS satellites and determining the
primary code measurements for each GLONASS satellite; applying the
correction data from the memory to the determined primary code
measurements to compensate for temperature-dependent inter-channel
biases in the primary code measurements using the measured current
temperature; and outputting the corrected primary code
measurements.
2. The method of claim 1, wherein the memory is a local memory of
the GLONASS receiver.
3. The method of claim 2, wherein the correction data is downloaded
into the memory from an external source.
4. The method of claim 1, wherein the memory is located remotely
from the GLONASS receiver and the applying step is performed
remotely from the GLONASS receiver.
5. The method of claim 1, further comprising calculating current
coordinates based on the corrected primary code measurements.
6. The method of claim 1, further comprising obtaining the
correction data for different temperatures of the analog Front-End
placed in a thermal chamber and using a GLONASS simulator, and
taking into account a delay of the signals from GLONASS satellites
in the analog Front-End and an antenna filter.
7. The method of claim 6, wherein a set of temperatures in the
thermal chamber is selected such that the temperature varies in
increments of up to 10.degree. C. over a range of temperature
changes in the analog Front-End of at least 50.degree. C.;
8. The method of claim 7, wherein the value of a compensation for
the inter channel biases for the current reading of the thermal
sensor is calculated by linear interpolation, if the analog
Front-End temperature is within the range of temperature changes,
and by approximation functions if the analog Front-End temperature
is outside the range of temperature changes.
9. The method of claim 8, wherein a sum of a constant and harmonic
functions parameters of which are determined based on measurements
of the correction data, are used as the approximation
functions.
10. The method of claim 1, further comprising outputting the
current temperature and the determined primary code
measurements.
11. The method of claim 1, wherein the determining the primary code
measurements step is based on a GLONASS standard-precision
signal.
12. A GLONASS receiver with reduced inter-channel bias comprising:
an antenna receiving signals from GLONASS satellites; an analog
Front-End receiving and processing the signals from the antenna; a
digital circuit receiving the processed signals from the analog
Front-End; a memory accessible by the digital circuit and storing
correction data, including correction data corresponding to signal
delay in the Front-End and in an antenna filter of the GLONASS
receiver for different temperatures for each satellite channel; and
a temperature sensor measuring a current temperature of the analog
Front-End of the GLONASS receiver; wherein the digital circuit,
using the measured current temperature, applies the correction data
from the memory to the determined primary code measurements to
compensate for temperature-dependent inter-channel biases in the
primary code measurements and outputs corrected primary code
measurements.
13. A method of reducing inter-channel bias in a GLONASS receiver,
the method comprising: generating primary code measurements in the
GLONASS receiver, including using a Delay Lock Loop (DLL) for
tracking GLONASS signals, where a position of a working
discriminator point depends on delays in an analog Front-End of the
GLONASS receiver, delays in antenna filters of the GLONASS
receiver, and on discriminator characteristic slope; for each
satellite channel, applying inter-channel bias correction data from
a memory to the determined primary code measurements to compensate
for the inter-channel biases in the primary code measurements; and
outputting the corrected primary code measurements.
14. The method of claim 13, further comprising adding a constant to
the output signal of a discriminator of the DLL.
15. The method of claim 13, wherein: in a reference strobe sequence
used to obtain the discriminator characteristic slope, the position
of each strobe being fixed to the boundaries of each GLONASS PRN
code chip; and a sign of each strobe corresponds to a sign of a
next PRN code chip; each strobe of the reference sequence includes
a sequence of some positive and some negative elements; and a
number of the positive elements is different from a number of the
negative elements, wherein the sign of each strobe is a sum of the
signs of the elements of the strobe.
16. The method of claim 15, wherein a ratio of the positive
elements and the negative elements in each strobe is 3:1.
17. The method of claim 13, wherein the memory is a local memory of
the GLONASS receiver.
18. The method of claim 13, further comprising calculating current
coordinates based on the corrected primary code measurements.
19. The method of claim 13, further comprising obtaining the
correction data using a GLONASS simulator, and taking into account
a delay of the signals from GLONASS satellites in the analog
Front-End and the antenna filters.
20. A GLONASS receiver with reduced inter-channel bias, comprising:
an antenna receiving signals from GLONASS satellites; an analog
Front-End receiving and processing the signals from the antenna; a
digital circuit receiving the processed signals from the analog
Front-End; a Delay Lock Loop (DLL) for tracking GLONASS signals,
where a position of a working discriminator point depends on delays
in an analog Front-End of the GLONASS receiver, delays in antenna
filters of the GLONASS receiver, and on discriminator
characteristic slope; a memory accessible by the digital circuit
and storing inter-channel bias correction data, including data
corresponding to signal delay in the analog Front-End and in the
antenna filters for each satellite channel; the digital circuit
generating primary code measurements based on an output of the DLL
and applying the correction data to the generated primary code
measurements to compensate for the inter-channel biases in the
primary code measurements; and wherein the digital circuit outputs
the corrected primary code measurements.
21. A method of reducing inter-channel bias in a GLONASS receiver,
the method comprising: generating primary code measurements based
on high-precision GLONASS signals; compensating for inter-channel
biases in the primary code measurements using correction data for
each satellite channel previously stored in a memory of the GLONASS
receiver by adding the corrections to the primary code
measurements, and outputting compensated primary code measurements;
wherein the correction data is obtained using a GLONASS simulator
and takes into account a delay of the high-precision GLONASS
signals in an analog Front-End and antenna filters of the GLONASS
receiver.
22. A GLONASS receiver with reduced inter-channel bias, comprising:
an antenna receiving signals from GLONASS satellites; an analog
Front-End receiving and processing the signals from the antenna; a
digital circuit receiving, processing and generating primary code
measurements based on an output of the analog Front-End; the
digital circuit applying correction data to the generated primary
code measurements to compensate for the inter-channel biases in the
primary code measurements, and outputting the corrected primary
code measurements a memory accessible by the digital circuit and
storing inter-channel bias correction data, including data
corresponding to signal delay in the analog Front-End and in the
antenna filters for each satellite channel.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a US national phase of PCT/RU2015/000598
filed on Sep. 23, 2015.
FIELD OF THE INVENTION
[0002] The present invention relates generally to GNSS receivers,
and more particularly, to enhancing accuracy of code measurements
based on GLONASS signals with a frequency division of channels.
BACKGROUND OF THE INVENTION
[0003] Modern high-precision GNSS receivers are capable of
receiving and processing signals from some different satellite
systems, at least from GPS and GLONASS. A well-known peculiarity of
GLONASS is frequency division multiply access. Each GLONASS
satellite transmits a navigation signal modulated by the same
pseudo-random code but it is emitted on its individual frequency.
These carrier frequencies are usually called channel numbers or
channel frequencies, and they are currently numbered from -07 up to
+07. GLONASS satellites transmit signals into two frequency ranges:
upper (L1) and lower (L2). In the L1, the frequency of 1602 MHz
corresponds to number 00, all the rest carrier frequencies (channel
numbers) are divided into j562.5 kHz, and in the L2 range,
frequency 1246 MHz corresponds to 00, other channel numbers are
divided into j437.5 kHz, where j is the channel number.
[0004] All GLONASS satellites transmit two pseudo-random codes on
each frequency: a standard accuracy code and a high-precision
code.
[0005] It should be noted that in the English language technical
literature, Standard Accuracy code similar to GPS pseudo-random
sequences is designated as CA-code, and High-Precision code is
designated as P-code.
[0006] Both codes are an M-sequence with duration of 1 ms. Clock
rate of the CA-code is 511 kHz, and the clock rate of the P-code is
5.11 MHz. According to these values, the bandwidth, concentrating
practically the whole signal power, for the P-signal, is almost 10
times wider than that of CA-signal. If one considers the whole
constellation of GLONASS satellites, then CA-signals in the L1
range take the bandwidth .apprxeq.9 MHz, and the
P-signals--.apprxeq.18 MHz. In the L2 band, CA-signals use
.apprxeq.8 MHz and P-signals .apprxeq.16 MHz.
[0007] A navigation receiver consists of analog and digital
components. The analog component amplifies signals, produces
heterodyning (down conversion of carrier frequency), and filters
signals, while the digital component separates a signal envelope
(demodulation) and measures signal parameters, including signal
arrival time. This time in navigation receivers means the moment of
beginning/ending of PRN-code, which modulates the carrier signal of
a satellite.
[0008] A common analog component is normally used to receive
signals from all GLONASS satellites, and signal division from
different satellites is implemented in receiver digital component
as a demodulation block.
[0009] When passing the receiver antenna filters and the analog
front end, the signal is delayed. The value of the delay depends on
the filter's phase-frequency characteristic (PFC). If the PFC were
linear, all the signals irrespective of carrier frequencies would
have the same delay. But in practice, due to non-linearity of the
PFC, signals of different GLONASS satellites have different delays
in the receiver.
[0010] The receiver digital component measures time of arrival of a
signal from each satellite. Such measurements are often called code
measurements, since they relate to the time of delivering the
modulating PRN code. The code measurements are also called
pseudo-ranges, emphasizing that the measurements are coordinated to
the receiver time scale that is different from the transmitter time
scale. As operation of all satellites is synchronized by the
system, one can say that there is an offset between receiver and
GNSS scales.
[0011] The difference in code measurements for various satellites
(when the position of satellites is known) enables to solve the
navigation task and determine receiver location. In solving the
navigation task, a delay in the radio path, common for all the
satellites, affects only the offset between the receiver time scale
and the GNSS time scale. But different delays for different
satellites directly affect positioning accuracy; therefore, they
can be regarded as code errors. Such differences, re-computed into
equivalent ranges and expressed in meters, are called "biases" in
the English technical literature.
[0012] Conventional art and our experience have shown that a
difference in receiving GLONASS signals on various channel
frequencies can achieve 3-6 ns, which is equivalent to biases (code
errors) up to .+-.1 m-.+-.2 m.
[0013] Biases noticeably worsen GLONASS positioning accuracy, and
much attention has been recently paid to methods of reducing these
errors.
[0014] Reference [6] considers technical solutions enabling to
reduce a difference between two different receivers rather than
receiver biases themselves. The paper stated that such an approach
efficiently increases positioning accuracy in solving the
navigation task for differential navigation, i.e., in case of
positioning a receiver relative to another one.
[0015] References [1, 4, 7] describe different methods to determine
biases using different frequencies of the first heterodyne. It is
proposed to adjust frequency of the first heterodyne and compare
measurement results for different GLONASS channel frequencies. It
is noted that the biases obtained in this manner take into account
a difference in signal delays arising in SAW filter on receiver
intermediate frequency. SAW filter parameters strongly depend on
temperature, and therefore one needs to regularly carry out such
temperature measurements.
[0016] References [2, 3, 5] discuss methods of compensating biases
in real-time. To do this, a GLONASS simulator integrated with the
navigation receiver is used. A simulator signal is fed to the input
of the receiver Front-End. To compensate for biases, references [2,
3] suggest that the difference between generation of the simulation
signal and its reception (after it has been delayed in the
Front-End) should be used; and in reference [5] it is proposed to
measure a difference between reception times for different
simulation signals.
[0017] Below there are considered the three methods of obtaining
GLONASS code measurements, which providing a considerable reduction
in errors caused by receiver PFC non-linearity.
SUMMARY OF THE INVENTION
[0018] Three methods of reducing errors in GLONASS code
measurements are proposed. The methods relate to errors caused by a
difference in delays of receiver analog Front End receiving GLONASS
signals on different channel frequencies. Such a difference on
different channel frequencies is caused by non-linearity of radio
path phase-frequency characteristic (PFC). The delay, its frequency
dependence and temperature stability substantially depend both on
filters employed in the receiver analog component and digital
techniques of obtaining code measurements.
[0019] The first of the proposed methods includes measuring and
applying corrections, which compensate for code biases, taking into
account their temperature instability.
[0020] The second method suggests that code measurements be
obtained taking into account a delay change in the receiver analog
block being compensated due to changing an operating point of the
DLL discriminator characteristic.
[0021] The third method proposes the use of high-precision GLONASS
signals as code measurements.
[0022] The proposed methods consider a GLONASS simulator, not being
an integral component of the receiver, to obtain corrections. Such
an approach enables taking into account different delays caused,
for example, in antennas located out of the receiver, eliminate
effects of the simulation signal on signal reception from GLONASS
GNSS satellites and use simpler schematics solution in receiver
designs.
[0023] Additional features and advantages of the invention will be
set forth in the description that follows, and will be apparent
from the description, or may be learned by practice of the
invention. The advantages of the invention will be realized and
attained by the structure particularly pointed out in the written
description and claims hereof as well as the appended drawings.
[0024] It is to be understood that both the foregoing general
description and the following detailed description are exemplary
and explanatory and are intended to provide further explanation of
the invention as claimed.
BRIEF DESCRIPTION OF THE ATTACHED FIGURES
[0025] The accompanying drawings, which are included to provide a
further understanding of the invention and are incorporated in and
constitute a part of this specification, illustrate embodiments of
the invention and, together with the description, serve to explain
the principles of the invention.
[0026] In the drawings:
[0027] FIG. 1A shows a typical functional schematic of navigation
receiver, its analog component and Front-End.
[0028] FIG. 1B shows a functional schematic of navigation receiver,
its analog component and Front-End with a thermometer which is used
to compensate inter-channel biases.
[0029] FIG. 2 shows envelope curves for navigation signals, a
reference code sequence and some variants of reference strobe
sequences.
[0030] FIG. 3A shows a typical functional schematic of receiver
digital component.
[0031] FIG. 3B shows a functional schematic of receiver digital
component using a thermometer to compensate inter-channel
biases.
[0032] FIG. 4 shows discriminator curves of DLL.
[0033] FIGS. 5A, 5B show a dependence of signal delay in the analog
component on GLONASS frequency.
[0034] FIG. 6 shows a schematic of a test-bench to determine biases
of receiver analog component.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0035] Reference will now be made in detail to the preferred
embodiments of the present invention.
[0036] Two main parts can be separated in navigation receivers:
antenna and the receiver itself. Antennas for high-precision
navigation receivers typically include a receiving antenna element,
a low-noise amplifier and an antenna filter. The antenna filter is
intended for isolating a frequency range for operation with one or
some GNSSs.
[0037] Modern navigation receivers include mainly analog components
and a digital component.
[0038] The analog component in turn includes antenna filters and an
analog Front End. Generally, these include signal amplifiers,
frequency converters, a quartz clock generator, intermediate
frequency synthesizers, and some filtration cascades. Three
filtration cascades are most often used: high-frequency,
intermediate and low-frequency.
[0039] High-frequency filtration is intended for separating signals
of a specific GNSS in the upper and/or lower frequency bands and
has mostly a wide bandwidth. Such a bandwidth can be up to 60 MHz
in the upper band and 140 MHz in the lower band.
[0040] Intermediate filtration should first provide interference
immunity of the receiver from undesired interference, and, second,
secure high accuracy of code measurements. SAW filters with a
bandwidth of 20-30 MHz are often used for intermediate
filtration.
[0041] Low-frequency filtration suppresses parasitic harmonics
which occur during frequency conversion. The bandwidth of these
filters is usually 30-50 MHz.
[0042] A receiver includes a combination of antenna filters and an
analog Front End, determining receiver through common
amplitude-frequency characteristic (AFC) and phase-frequency
characteristic. Common AFC and PFC depend on the predetermined
frequency plan, i.e., frequency-conversion schematics.
[0043] FIG. 1A shows an example of typical receiver analog
components with double frequency conversion. Only those analog
elements that relate to the substance of the invention are shown in
this figure.
[0044] Signals from satellites are received by an antenna element
(100) and after passing through an antenna filter (110) are fed to
a receiver analog Front End (200) that is in a shielded box. The
required frequency plan is guaranteed by a frequency synthesizer
(230) generating frequencies F.sub.LO-1 and F.sub.LO-2 from
frequency F.sub.q of a quartz generator (220). After a first mixer
(230) and intermediate frequency filter (250) the signal is
transferred to the first intermediate frequency f.sub.int-1. After
second mixer (260) and low-frequency filter (270) the signal is
transferred to the second intermediate frequency f.sub.int-2.
Different frequencies f.sub.int-2,j correspond to different GLONASS
satellites. After the signals passed through the analog Front End,
they come into a receiver digital component (300) wherein all
necessary measurements are generated.
[0045] Expressions for the common AFC and PFC for the schematics in
question can be as follows
G.sub.R(f)=G.sub.1(f)+G.sub.2(f)+G.sub.3(f-f.sub.LO
1)+G.sub.4(f-f.sub.LO 1-f.sub.LO 2) (1)
.PHI..sub.R(f)=.PHI..sub.1(f)+.PHI..sub.2(f)+.PHI..sub.3(f-f.sub.LO-1)+.-
PHI..sub.4(f-f.sub.LO-1-f.sub.LO-2) (2)
[0046] where G.sub.R(f) and .PHI..sub.R (f) are common AFC and PFC
for the receiver analog component, correspondingly;
[0047] G.sub.1(f) and .PHI..sub.1(f) are the corresponding AFC and
PFC of the antenna filter;
[0048] G.sub.2(f) and .PHI..sub.2(f) are the corresponding AFC and
PFC of the Front-End high-frequency filter;
[0049] G.sub.3(f) and .PHI..sub.3(f) are the corresponding AFC and
PFC of the Front-End first intermediate frequency filter installed
at the first intermediate frequency;
[0050] G.sub.4(f) and .PHI..sub.4(f) are the corresponding AFC and
PFC of the Front-End low frequency filter installed at the second
intermediate frequency;
[0051] f.sub.LO-1 and f.sub.LO-2 are the frequencies of the first
and second heterodynes correspondingly, providing the predetermined
receiver frequency plan.
[0052] In equations (1) and (2) AFCs of filters should be in dB,
and PFCs in cycles, frequency in Hz. A signal delay of the receiver
analog component (in seconds) can be approximately:
.tau. R ( f ) = d df [ .PHI. R ( f ) ] ( 3 ) ##EQU00001##
[0053] Expression .tau..sup.R(f) in (3) is called "group delay" and
it enables to determine precisely enough a delay only for
narrowband signals, i.e., when PFC non-linearity is negligible. A
more exact calculation of a delay of the modulated signal takes
into account the whole signal spectrum. But expression (3) can be
useful for further description and understanding.
[0054] PFC .PHI..sub.R(f) is typically a non-linear function of
frequency f, and hence signals from each GLONASS satellite take
different delays .tau..sub.j.sup.R(f.sub.j) in the receiver analog
component.
[0055] Curves A) and B) in FIG. 2 illustrate signal conversion in
the analog Front End. Curve A) presents a fragment of the signal
envelope at the antenna input, and curve B) shows the corresponding
fragment at the analog component output for a conventional
navigation receiver. It can be seen from the figure that, first,
the time instant of changing the code sign at the output happens
later than at the input, and second, if the sign of the code
sequence changes fast enough at the input, then at the output the
same process takes some time that is often called the front
duration .tau..sup.front.
[0056] The most important task of the receiver digital component is
to obtain parameters of measurements of the received signal needed
for solving the navigation task and determining a receiver
position.
[0057] Measuring signal parameters in navigation receivers is
described in detail in the technical literature. In particular, in
reference [3] there is described a commonly-used method of building
high-precision digital receivers. A simplified version of the
receiver digital component is shown in FIG. 3A.
[0058] An analog signal is converted in the digital form with an
analog-digital converter (ADC) (301) and is then fed to digital
tracking loops PLL and DLL.
[0059] Tracking of the input signal phase and frequency is
implemented by PLL consisting of the following main components:
complex multiplier (302), carrier NCO (303), correlators (304) and
(305), discriminator (310) and loop filter (309).
[0060] Omitting PLL structure and operation, one should note that
carrier NCO (303) is used for separate processing of GLONASS
signals. This NCO generates in digital form two quadrature
components (sin and cos) of the harmonic oscillation with nominal
f.sub.p,j, corresponding to one of the GLONASS channel frequencies,
and a difference between the real satellite signal and its nominal
value in frequency and phase is generated by control signals of the
carrier NCO (303) from PLL filter (309). Note that FIG. 3 presents
loops PLL and DLL tracking only one GLONASS satellites, similar
schemes are implemented for other satellites.
[0061] A Delay lock loop (DLL) tracks changes in signal arrival
times. Main elements of DLL are a controlled code sequence
generator (307) and controlled strobe sequence generator (308), two
correlators (305) and (306) generating correlation components I and
dI, respectively, DLL discriminator (311) and DLL loop filter
(312). With signals of the loop filter (312), block (313) generates
code measurements .rho..sub.j.
[0062] To receive signals, the code sequence generator (307)
generates a pseudo-random M-sequence corresponding to the PRN code
transmitted by GLONASS satellites. The position in time of this
sequence is defined by control signals which are fed from the
generator (307) from the DLL loop filter (312). This sequence is
fed to correlator (305), where it is multiplied by the input
in-phase quadrature signal C from the output of complex multiplier
(302) and stored over the pre-determined time T.sub.C (T.sub.C is
often selected equal to 5 ms). The so-obtained number is called
correlation signal I .
[0063] The same code sequence from generator (307) is also fed to
strobe sequence generator (308). The most used and well-known
strobe sequence is a sequence of rectangular pulses, the center of
which coincides with sign changes instants in the reference code
sequence and the polarity (sign) of the pulse match the sign of the
next chip code, i.e., if the code changes its sign from (-) to (+),
the sign of the strobe will be positive, otherwise (from (+) to
(-)--negative. Such strobes and sequences are often called
"simple". A fragment of the simple strobe sequence is shown in
curve C) of FIG. 2, and in curve D) there is the corresponding
fragment of the reference code sequence.
[0064] The strobe sequence generated in block (308) is fed to
correlator (306), at the output of which there is generated
correlation signal dI. Correlator (306) works in the same way as
correlator (305) described above. The output signals I and dI of
correlators (305) and (306), further come to DLL discriminator
(311).
[0065] The most known and used discriminator calculates the ratio
of these two correlation signals, i.e., the generating
discriminator signal according to
z dll = dI I , ( 4 ) ##EQU00002##
[0066] A signal dependence at the discriminator output
z.sup.dll(.tau.) on the time mismatch between the input signal
envelope and reference code sequence .tau. is often used DLL
discriminator characteristic. See FIG. 2. A typical discriminator
characteristic for a "simple" strobe sequence is shown in FIG. 4,
graph (a). The discriminator signal, depending on advancing or
slowing the reference code sequence relative to the input signal
envelope (.tau.>0 or .tau.<0), changes it sign from (+) to
(-), or otherwise. A point wherein z.sup.dll(.tau.)=0 is hereafter
called "working" discriminator point. For the case shown in curves
B), C) and D) of FIG. 2, at z.sup.dll=0, the instants of sign
changes in the reference code sequence and input signal envelope
coincide.
[0067] The signal from discriminator (311) is further fed to DLL
loop filter (312). There are known different variants of building
loop filters references [3, 4], the schematics and parameters of
the filters determining DLL noise and dynamic properties. Signals
at the output of the loop filter are used for controlling the
reference generator and code (307) and strobe (308) sequences. In
accordance with these signals, a time position of the code sequence
and, respectively, strobe sequence changes. A closed tracking
system (DLL) in the steady operation mode keeps the discriminator
signal within a range of the "working" point (z.sup.dll.apprxeq.0)
of the DLL discriminator.
[0068] Control signals generated by loop filter (312) are
simultaneously fed to measurements generator (313)--a block of
generating code measurements. In this block, the current beginning
of the reference code sequence is taken as a current estimate of
signal arrival time (t.sup.Rec) measured according to the receiver
clock scale. A satellite emits the signal at a time instant
(t.sup.Tr) (measured according to the satellite clock scale), and a
difference between the signal arrival time and signal emission time
can be used in calculating the so-called code measurements
.rho.=c(t.sup.Rec-t.sup.Tr) (5)
[0069] where c is the speed of light.
[0070] A range-difference positioning method is used in GNSS
receivers, therefore, only the difference in code measurements
obtained for different satellites affects the positioning accuracy.
Signal delays for different GLONASS satellites can differ due to
non-linearity of PFC in the receiver analog component, and
correspondingly, code measurements are dependent on this delay
resulting in errors in positioning. FIG. 4, graph (b) shows a
change in "working" point positions for the DLL discriminator
characteristic as a function of changing the delay in the receiver
analog component .tau..sup.RF. The position of the reference code
sequence in time also changes, and so do the code measurements.
[0071] To measure delay in the receiver analog component, a GLONASS
simulator can be used. The so-called "double differences" allow
eliminating a clock offset between satellite and receiver scales,
i.e., generating a mathematical combination
DD.sub.j=[(.rho..sub.j.sup.Rec-.rho..sub.j=0.sup.Rec)-(.rho..sub.j.sup.S-
IM-.rho..sub.j=0.sup.SIM)] (6)
[0072] where .rho..sub.j.sup.Rec=ct.sub.j.sup.Rec; t.sub.j.sup.Rec
is the time of signal arrival measured according to the receiver
clock scale;
[0073] .rho..sub.j.sup.SIM=ct.sub.j.sup.SIM; t.sub.j.sup.SIM is the
time of signal arrival measured according to the simulator clock
scale, generated by the simulator for the given receiver
location;
[0074] j is the satellite number or the channel number;
[0075] j=0 means that the measurement has been obtained at a
randomly-selected channel frequency, for example, at the zero
channel number.
[0076] If analog delays are different for different GLONASS channel
numbers, the value DD.sub.j is other than zero. The combination (6)
is often called GLONASS GNSS biases.
[0077] An external (relative to the receiver) signal simulator is
proposed to be used to measure analog biases. Such a simulator can
be employed, first, for bias measurements in receivers of different
marks and units, and second, for determining biases caused by a
non-linear PFC of the antenna filter in addition to Front-End
biases. To solve the second task, an additional filter (110) used
in the utilized antenna can be installed in the measurement path
between the simulator output and Front-End input.
[0078] The previously-measured bias DD.sub.j can be further used as
a correction data to compensate for the difference in the analog
component delays, thereby leading the measurements to the combined
value corresponding to a delay on a chosen channel frequency
.rho..sub.j.sup.corr=.rho..sub.j-DD.sub.j (7)
[0079] To apply the above-described approach, DD.sub.j should be
unchangeable in time. A change in DD.sub.j biases first of all
relates to possible temperature variation during following receiver
service.
[0080] A maximal temperature operation range for navigation
receivers is -40.degree. C. . . . +75.degree. C., sometimes this
range is narrower: -20.degree. C. . . . +50.degree. C. Such
considerable temperature variations are explained by possible use
in different climate zones, and seasonal, weather, day and night
changes in ambient/environmental temperatures. Also, receiver
internal heat emission due to different operation modes and design
heat production features can affect temperature changes.
[0081] Below there are considered different variants of possible
solutions to providing temperature stability of corrected code
measurements .rho..sub.j.sup.corr.
[0082] One embodiment suggests a method of obtaining such
corrections that consider current temperature of the receiver. In
this case, expression (7) is as follows
.rho..sub.j.sup.corr=.rho..sub.j(T)-DD.sub.j.sup.T(T) (8)
[0083] where T is the temperature of the receiver or its
components;
[0084] DD.sub.j.sup.T(T) is the correction data for temperature
T;
[0085] .rho..sub.j(T) is the primary code measurement at
temperature T.
[0086] A study of the corrections-temperature dependence
DD.sub.j.sup.T(T) for different receivers has shown that a
variation of 25.degree. C.-35.degree. C. in temperature results in
a change in the correction data no more than 10-15 cm for some
GLONASS channel numbers, but for other channel numbers the change
in the correction data reaches 0.8-1.2 m.
[0087] Analyzing delay behavior in different filters of the
receiver analog Front End has enabled to state that some SAW filter
types (250), which are installed on the first intermediate
frequency, mainly contribute to the temperature dependence of
biases for CA signal (FIG. 1). An oscillation type of
delay-temperature variation is specific for this filter group. If
the temperature varies by 50.degree. C.-70.degree. C., delay values
are periodically repeated. Biases have the same periodic behavior
in this case.
[0088] Research has shown that the bias-temperature dependence is
in a good correlation with a function:
DD.sub.j.sup.T(T)=DD.sub.j.sup.0+DD.sub.j.sup.1(T-T.sub.0)+DD.sub.j.sup.-
maxcos(.OMEGA..sub.jT+.PSI..sub.j) (9)
[0089] where T is the current Front-End temperature;
[0090] T.sub.0 is the nominal working Front-End temperature;
[0091] DD.sup.0, DD.sup.1, DD.sup.nax, .OMEGA., .PSI. are the
function parameters which are specific for each channel number j.
It should be noted that for some channel numbers biases are small
(do not exceed .+-.0.1 m), and they are practically independent of
temperature.
[0092] Bias periodicity and the possibility of describing biases
behavior depending on temperature allow measuring biases within a
temperature range of at least 50.degree. C. in increments of
maximum 10.degree. C., and then predicting a bias value at
different temperatures. Well-known interpolation methods help in
such prediction within the temperature range, and extrapolation
methods with a predetermined approximation function, for example,
(9) are helpful out of this range.
[0093] To implement this method, temperature shall be measured at
the location of intermediate frequency (IF) SAW filters (250) along
with bias measurements. The analog Front End (200) is normally
shielded, and the temperature of the analog Front End is higher
than the ambient temperature. The implementation of this method is
made by supplementing typical block-diagrams of the analog and
digital components shown in FIG. 1A and FIG. 3A by elements
presented in FIG. 1B and FIG. 3B. A temperature sensor (280) is
installed in the shielded analog Front End (200) to measure IF SAW
filter temperature. Readings of this sensor in (T.sup.0C) are fed
to the receiver digital component (300). In the receiver digital
component (300) the readings are converted into digital form with
the help of ADC (314), and then processed in the measurements
generator (313).
[0094] To measure biases, the receiver is placed into a thermal
chamber/oven enabling to fix needed temperature in its volume. A
simulator sequentially or simultaneously generating signals on all
operating GLONASS channel frequencies is connected to the input of
the receiver analog component. Bias estimates are then generated
for each channel frequency based on code measurements and according
to equation (6). At the same time, the readings from the thermal
sensor installed in the shielded analog Front End are stored in
memory. Then, the temperature of the chamber is changed, and
measurements are made again. The temperature in the chamber is
changed at 10.degree. C. increments within an adjusting range of at
least 50.degree. C. The obtained data set is stored in the receiver
memory for further use during receiver service. In addition, based
on the obtained data, the coefficients of the approximation
function (9) are calculated and also stored in memory. To compute
these coefficients, different mathematical methods can be used, for
example, Ordinary Least Squares method (OLS). A schematic of a test
bench to do measurements is shown in FIG. 6.
[0095] When a receiver receives GLONASS code measurements, current
temperature is determined for the analog Front End, and a
correction is generated depending on the measured temperature (for
each channel frequency). Different interpolation and extrapolation
methods can be chosen by the experienced engineer to compute these
corrections. The corrections are further used according to (8) to
diminish receiver code errors.
[0096] Another method to reduce code errors in navigation receivers
is based on a considerable difference spectral characteristics of
CA and P GLONASS signals and the specific characteristics of AFC
and PFC SAW filter.
[0097] As has been previously said, the GLONASS satellites transmit
both standard accuracy signals (CA) with the clock rate of 511 kHz
and high-precision signals (P) with the clock rate of 5.11 MHz.
This P-signal power is distributed in a wider bandwidth and takes
about 5.11 MHz.
[0098] Researching AFC and PFC of SAW filters used in navigation
receivers have shown that group delay (3) has a strong oscillation
pattern. An example of changing the group delay (expressed in
meters) depending on signal frequency is shown in FIG. 5A by a
dotted line. Delays on frequencies corresponding to GLONASS channel
numbers in the L1 band are marked with diamond symbols in this
graph. The graph illustrates well, and researches prove that for
narrow-band signals, depending on frequency, a filter delay varies
according to the almost harmonic law with the amplitude of 0.5-0.7
m and period 630-700 kHz.
[0099] By applying P-signals with bandwidth .about.5.0 MHz to
obtain code measurements, one can considerably average delay
variations, and the frequency-delay dependence in the filter
reaches a comparatively monotonic pattern. A graph of such a
dependence is presented in FIG. 5A with a solid line. Triangle
symbol show delays corresponding different GLONASS channel
numbers.
[0100] A study of temperature stability of corrections measured
with the help of P-signals has shown that a temperature change in
the receiver in the range up to 50.degree. C. results in a delay
change no greater than .+-.0.15 m.
[0101] To implement this method using the receiver diagram shown in
FIG. 1B and FIG. 3B, the code sequence generator (307) has to
generate a reference sequence corresponding to high-precision
GLONASS signals. These signals should be generated by a GLONASS
simulator. Computation of corrections in this case can be made
without a thermal chamber at the standard operation temperature of
the receiver, and the corrections may be applied without a
temperature sensor even if the temperature intensely varies.
[0102] One more method of reducing code errors in case of GLONASS
signals is based on the fact that when a signal passes through a
filter its delay is proportional to the duration of the (wave)
front, i.e., .tau..sup.RF.about..tau..sup.front. As was mentioned,
the signal is converted in the analog Front End, and the sign of
the input signal envelope after the filter changes gradually,
curves A) and B) in FIG. 2 illustrates this fact. It is also known
that the slope of the DLL discriminator characteristic
d ( z dll ( .tau. ) ) d .tau. ##EQU00003##
for simple strobe sequences, shown in curve C) in FIG. 2, is
inversely proportional to the front duration .tau..sup.front.
Keeping the above in mind, one can write
d ( z dll ( .tau. ) ) d .tau. ~ 1 .tau. front ~ 1 .tau. RF ( 10 )
##EQU00004##
[0103] FIG. 4, graph (b), in particular, shows that when delay
.tau..sup.RF increases, not only the working point of the
discriminator characteristic shifts but also its slope
decreases.
[0104] Expressions (10) enable to propose two ways of stabilizing
the working discriminator point when the delay in the filter
changes.
[0105] One way is to use a DLL discriminator as follows
z dll = dI I + h ( 11 ) ##EQU00005##
[0106] Where correlation signal dI is generated with a simple
strobe sequence shown in curve C) of FIG. 2, and the parameter h is
selected according to the condition below depending on the filters
used
h .apprxeq. .tau. strobe 2 .tau. chip ( 12 ) ##EQU00006##
[0107] A comparison of FIG. 4, graph (c) and FIG. 4, graph (a)
shows that the discriminator working point can be shifted in
.delta..tau. by varying parameter h. FIG. 4, graph (d) clearly
illustrates the fact that shift .delta..tau. depends on both the
parameter h and the slope of the discriminator characteristic
d ( z dll ( .tau. ) ) d .tau. , ##EQU00007##
and the sign of shifting is inverse to the sign of the delay change
.tau..sup.RF. FIG. 4, graph (d) also shows that parameter h allows
a remarkable reduction in position change of the working
discriminator point when filter delay .tau..sup.RF changes
considerably. Therefore, both temperature changes in the receiver
analog component delay can be reduced and delay differences in PFC
non-linearity-related delays of the receiver analog component on
different channel numbers can be compensated.
[0108] The other way of obtaining a similar discriminator
characteristic is to use strobe sequences of special types. The
main feature of these sequences, unlike those described earlier is
strobe position, which is related not only to the transition
moments (i.e., a change of the code sign) but also to
non-transitions. In other words, the position of each strobe in the
sequence should be related to the PRN chip boundaries. The number
of transitions and non-transitions in the GLONASS PRN code differs
by the value of unity/one, correlation signal dI also changes by
some value .delta., and value
h = .delta. I .apprxeq. .tau. strobe 2 .tau. chip ##EQU00008##
is added to the signal at the DLL discriminator output. Note that,
similar to the previous case, the position of the working
discriminator point does not practically change.
[0109] An additional possibility of adjusting the compensation
degree can be achieved by applying the special sequence of
sign-variable strobes. Each strobe in this sequence is presented as
a sequence of different sign elements. An example of such a
sequence is given in FIG. 2, curve (E). By selecting the ratio of
positive and negative elements in the sign-alternating strobe one
can adjust the degree of compensation for delay changes in the
receiver analog component, but a certain condition is to be met in
this case: the area under curve following the strobe shape is to be
different from zero, i.e., the number of positive and negative
elements in the strobe is unequal.
[0110] An example of the two methods is shown in FIG. 5B. A
dependence of estimates for analog Front End delays is symboled
with squares on GLONASS channel numbers in the L1 band. In the
graph one can see that delays on different channel numbers are
changed over than .+-.0.1 m, but their average value noticeably
differs from the average obtained other previously-considered
methods.
[0111] As already stated, in navigation receivers a
range-difference positioning technique is used, the average
estimate bias does not therefore affect positioning errors, and
does not regard as measuring errors.
[0112] Researches have shown that the above-mentioned methods allow
obtaining a difference between code measurements on different
GLONASS channel number frequencies no more than .+-.0.1 m . . .
.+-.0.15 m even if the temperature varies in a wide range.
[0113] When using this method the code sequence generator (307) can
generate both the standard-precision signal and high-precision
signal. The strobe sequence generator (308) is to generate a strobe
sequence according to one of the mentioned methods. Similar to the
previous example, the receiver can be built in accordance with FIG.
1B and FIG. 3B, and correction measurements are performed without a
thermal chamber at standard operation temperature of the receiver.
The corrections can be used without a temperature sensor even if
the temperature considerably varies.
[0114] A block-diagram of the test bench enabling an implementation
of the methods is shown in FIG. 6.
[0115] Simulation satellite signals s are fed from a GLONASS
simulator (400) to the input of an antenna filter (110), and then
to the input of the receiver analog Front End (200). Code
measurements .rho. are fed from the output of the digital component
(300) and stored in a computer (500). In addition, some digital
information about the simulated signal arrival time
.rho..sup.SIM=ct.sup.SIM is transmitted from the simulator (400) to
computer (500). DD biases for all GLONASS channel number
frequencies are calculated in the computer (500) (see above) and
written in the receiver memory to use further for primary code
measurement corrections.
[0116] For the first method, Front-End temperature data in
T.sup.0C, at which code measurements .rho. have been made, is
additionally transmitted from the digital component (300) to the
computer (500). As said, in this case, all equipment including the
antenna filter (110), analog Front End (200), and the receiver
digital component (300) are placed into the thermal chamber (600),
to obtain a dependence of DD.sup.T(T) biases in the computer (500).
The described-above method allows specialists to understand the
operation procedure and peculiarities of this test bench.
[0117] A developer of navigation receivers can select one or other
proposed methods, their combinations and parameters based on his
own considerations and experience depending on the filters used in
the receiver analog Front End and technological features of digital
component design.
[0118] Having thus described a preferred embodiment, it should be
apparent to those skilled in the art that certain advantages of the
described method and apparatus have been achieved. It should also
be appreciated that various modifications, adaptations, and
alternative embodiments thereof may be made within the scope and
spirit of the present invention. The invention is further defined
by the following claims.
REFERENCES
[0119] 1. U.S. Pat. No. 6,608,998 B1, Neumann et al., Method for
reducing inter-frequency bias effects in a receiver;
[0120] 2. EP2204664 A2, Yudanov et al., Inter-channel bias
calibration for navigation satellite system;
[0121] 3. U.S. Pat. No. 6,266,007 B1, Gary R. Lennen, Code group
delay calibration using error free real time calibration
signal;
[0122] 4. EP 1031845 A2, Miroslaw Balodis, Receiver calibration
technique for glonass, Leica Geosystems Inc.
[0123] 5. US 20070008216 A1, Ganguly et al., GPS receiver with
calibrator;
[0124] 6. Algorithms to Calibrate and Compensate for GLONASS Biases
in GNSS RTK Receivers working with 3.sup.rd party Networks, Aleksey
Boriskin, Gleb Zyryanov, Magellan, Russia, ION GNSS 21.sup.st.
International Technical Meeting of the Satellite Division, 16-19,
September 2008, Savannah, Ga.
[0125] 7. GLONASS Receiver Inter-frequency Biases--Calibration
Methods and Feasibility, J. B. Neumann, M. Bates, R. S. Harvey
Novatel Inc. ION GPS '99, 14-17 September 1999, Nashville,
Tenn.
[0126] 8. A. D. Boriskin, A. V. Veitsel, V. A. Veitsel, M. I.
Zhodzishsky, D. S. Milyutin, High precision positioning equipment
for global positioning navigation satellite systems: receivers-end
users of navigation information, M. I. Zhodzishsky, Ed., Moscow,
MAI-Print, 2010
[0127] 9. GLONASS. Design concepts and operation, A. I. Perov, V.
N. Kharisov, Eds., 3.sup.rd edition, Moscow, Radiotechnika,
2005.
* * * * *