CN117083540A

CN117083540A - Adaptive estimation of GNSS satellite bias

Info

Publication number: CN117083540A
Application number: CN202180096636.9A
Authority: CN
Inventors: 戴礼文; 陈义群
Original assignee: Deere and Co
Current assignee: Deere and Co
Priority date: 2021-01-31
Filing date: 2021-12-17
Publication date: 2023-11-17

Abstract

A first pair of wide-lane (WL) homodyne (ZD) bias filters and corresponding supplemental WL bias prediction filters determine a time-varying wide-lane bias of the corresponding received satellite based on adaptive estimates responsive to the tuning dynamic noise (S820). A second pair of Narrow Lane (NL) homodyne (ZD) bias filters and corresponding NL bias filter/code phase bias filters determine time-varying NL bias for the corresponding satellite based on adaptive estimates of adaptive estimates responsive to tuned dynamic noise (S822). The correction signals include WL ambiguity, time-varying WL bias, and NL ambiguity and time-varying NL bias for each received satellite within the corresponding GNSS constellation (S824).

Description

Adaptive estimation of GNSS satellite bias

RELATED APPLICATIONS

The present application is a continuing application of U.S. patent application Ser. No. 17/302,572, filed 5/6/2021, which claims priority and filing date based on U.S. C. ≡119 (e) filed 1/2021, U.S. provisional application Ser. No. 63/143,921, filed 5/6/2021, which is incorporated herein by reference.

Technical Field

The present disclosure relates to methods and systems for providing satellite correction signals using adaptive estimates (adaptive estimation) of GNSS satellite biases.

Background

In some prior art, the service provider provides the correction signal to the end user of the satellite navigation receiver via a wireless signal, such as a satellite wireless signal on the satellite L-band. In a GNSS constellation, each satellite clock may have a satellite clock bias or clock error that may be measured with reference to GNSS system clock time, as well as other alternatives. The correction data includes a respective clock offset or clock solution for each satellite within a field of view or reception range at the mobile receiver of the end user. The satellite clock bias with the corresponding satellite identifier is transmitted to the end user or subscriber of the satellite correction signal.

In some prior art, some GNSS satellites are susceptible to significant code and/or phase bias variations during transient periods due to variations in satellite transmit power, variations in satellite transmit power allocation between or among code components (e.g., pseudorandom noise code components and/or coarse acquisition code components), temperature variations, or switching of redundant transmitters, different antennas, or different hardware signal paths within one or more satellites within any GNSS satellite constellation. Accordingly, there is a need to improve the accuracy and timeliness of code and phase bias in correction signals wirelessly provided to rover stations or mobile GNSS receivers.

Disclosure of Invention

According to one embodiment, a method or system for providing global satellite differential correction signals includes an electronic data processor of a data processing center configured to determine wide-lane fixed ambiguities for corresponding satellites and corresponding time-varying wide-lane offsets. A first pair of wide-lane (WL) homodyne (ZD) bias filters and corresponding supplemental WL bias prediction filters determine time-varying wide-lane bias for each satellite based on adaptive estimates responsive to tuned dynamic noise provided by the supplemental wide-lane bias prediction filters for the respective satellites. The electronic data processor of the data processing center is configured to determine a lane-fixed ambiguity for the corresponding satellite, a satellite slow clock solution, and a time-varying lane-bias. A second pair of lane (NL) homodyne (ZD) bias filters and corresponding NL bias filter/code phase bias filter determine time-varying lane bias (e.g., refraction corrected (NL) code phase bias) for each satellite based on adaptive estimates of adaptive estimates responsive to another tuned dynamic noise within the lane bias/code phase bias filter for the corresponding satellite. The electronic data processor of the data processing center is configured to provide correction signals including a wide lane ambiguity, a time-varying wide lane offset, and a narrow lane ambiguity and a time-varying narrow lane offset and a code offset (e.g., a code phase offset).

According to another aspect of the present disclosure, to properly tune the dynamic noise of both the satellite wide-lane and narrow-lane biases, first and second pairs of filters for each satellite are developed to support adaptive estimation of the satellite wide-lane and narrow-lane biases, respectively. When the satellite refraction corrected code bias is significant, the code phase bias can be integrated into the correction signal (e.g., starFire) in a timely manner through a wireless channel (e.g., satellite L-band communication) or an Internet enabled wireless channel for civilian operation (e.g., real-time or timely control of heavy equipment, agricultural, building, or industrial equipment/vehicle basis) ^TM Correction signal).

Drawings

FIG. 1A is a block diagram of one embodiment of a system for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks, wherein the satellite correction signals are provided via communication satellites.

FIG. 1B is a block diagram of another embodiment of a system for providing satellite correction signals having accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks, wherein the satellite correction signals are provided via a wireless communication system.

Fig. 2A is a graph showing delays associated with the provision of correction signals (and more particularly, correction signals having a set of clock errors for corresponding satellites), wherein measurements are collected with moderate delays.

Fig. 2B is a graph showing delays associated with the provision of correction signals (and more specifically, correction signals having a set of clock errors for corresponding satellites), wherein measurements are collected at lower delays than the measurements of fig. 2A.

FIG. 3 is a block diagram of another embodiment of a system for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks.

Fig. 4 shows an illustrative example of the correction data estimator of fig. 3 in more detail.

Fig. 5 is a diagram illustrating parallel operation of a slow clock process (e.g., a medium delay clock process) and a fast clock process (e.g., a low delay clock process).

FIG. 6 is a flow chart of one exemplary method for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks.

Fig. 7, collectively referred to as fig. 7A and 7B, is a flow chart of another embodiment of a method for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks.

FIG. 8A provides an illustrative graph of calculated time versus GNSS time for providing a slow clock solution.

FIG. 8B provides an illustrative graph of calculated time versus GNSS time for providing a low-delay clock solution.

FIG. 9 illustrates a block diagram of one embodiment of a system for providing satellite correction signals with hot start.

FIG. 10 illustrates a block diagram of one embodiment of a system for providing satellite correction signals with hot start.

FIG. 11 illustrates one embodiment of a method for providing satellite correction signals with hot start.

Fig. 12 shows another embodiment of a method for providing satellite correction signals with a hot start.

Fig. 13 shows yet another embodiment of a method for providing satellite correction signals with a hot start.

Fig. 14 shows yet another embodiment of a method for providing satellite correction signals with a hot start.

FIG. 15A is a block diagram of one embodiment of a WL filtering system.

Figure 15B is a block diagram of one embodiment of an NL filtering system.

FIG. 16A is an illustrative chart comparing the relative signal-to-noise ratio versus time of a satellite transmitting an L1 signal (e.g., an L1P (Y) signal encoded with a pseudo-random noise code (PN) and an L1 CA (coarse acquisition code)), where a correction signal may be applied to account for such transient variations in bias.

Fig. 16B is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting L2 signals (e.g., L2P (Y) and L2C signals encoded with a PN), wherein correction signals may be applied to account for such transient variations in bias.

FIG. 17A is an illustrative chart comparing the relative signal-to-noise ratio versus time for a satellite transmitting an L1 signal (e.g., an L1P (Y) or L2 (P (Y) signal encoded with PN and an L1 CA signal encoded with coarse acquisition code), where a correction signal may be applied to account for such transient variations in bias.

Fig. 17B is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting L2 signals (e.g., P (Y) and L2C signals encoded with PN), where correction signals may be applied to account for such transient variations in bias.

FIG. 18A is an illustrative chart comparing the relative signal-to-noise ratio versus time for a satellite transmitting an L1 signal (e.g., an L1 (P (Y) or L2P (Y) signal encoded with PN and an L1 CA signal encoded with coarse acquisition code), where a correction signal may be applied to account for such transient variations in bias.

Fig. 18B is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting L2 signals (e.g., L2P (Y) and L2C signals encoded with a PN), wherein correction signals may be applied to account for such transient variations in bias.

Fig. 19A is an illustrative chart showing differential code deviation for a relationship of a P1 signal and an L1CA signal, the relationship of the P1 signal and the L1CA signal being associated with a change in power of a carrier or encoded portion of the respective signal.

Fig. 19B is an illustrative chart showing differential code bias for a relationship of a P2 signal relative to an L2C signal, the relationship of the P2 signal to the L2C signal being associated with a change in power of a carrier or encoded portion of the respective signal.

FIG. 20A shows WL bias versus time for a WL filter system operating in a first WL filter mode.

FIG. 20B shows WL bias versus time and WL residual versus time for a WL filtering system operating in a second WL filtering mode.

Fig. 21A shows NL bias versus time for an NL filter system operating in a first NL filter mode.

Fig. 21B shows NL bias versus time and code phase bias versus time for an NL filter system operating in the second NL filter mode.

Fig. 22 is a flow chart of a first illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 23 is a flow chart of a second illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 24 is a flow chart of a third illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 25 is a flow chart of a fourth illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 26 is a flowchart of a fifth illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 27 is a flowchart of a sixth illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 28 is a flow chart of a seventh illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 29 is a flowchart of an eighth illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Fig. 30 is a flowchart of a ninth illustrative embodiment of a method for providing global satellite differential correction signals consistent with or in combination with any of the block diagrams referenced in the above-described figures of the present disclosure.

Like reference numerals in any one of two or more drawings denote like features, methods, steps or elements.

Detailed Description

As used herein, a module or estimator may refer to software, hardware, or both. If the module is implemented as software, it may be stored in the data storage 24 for processing by the data processor 20. Adapted, configured or arranged to mean that the module, estimator or other device is capable of performing the functions or supporting features described in the specification. For example, adapted, configured or arranged to include modules programmed with software instructions stored in the data storage device 24 for processing by the data processor 20 to perform the specific functions set forth herein.

Unless explicitly stated otherwise, about or approximately shall mean a tolerance of plus or minus 25% of any value, quantity, or numerical value.

A position determining receiver or satellite receiver (12, 30), such as a Global Navigation Satellite System (GNSS) receiver, is capable of receiving carrier phase measurements that are affected by ambiguity, such as integer ambiguity (integer ambiguity), in terms of the number of or fractional periods of the received satellite signal. Epoch or measurement time means a particular time instant or time interval of the navigation satellite system during which the mobile receiver measures the carrier phase (e.g., at some corresponding frequency or rate). The receiver (12, 30) determines or resolves ambiguities of the carrier phase measurements to accurately estimate the exact location or coordinates of the receiver. Although the code phase (code) or pseudorange measurements of a GNSS receiver (12, 30) are independent of the integer ambiguity in the received satellite period, the code phase measurements do not provide the centimeter level of position accuracy required for some applications. As used in this document, ambiguity is generally specific to the case of a particular equation that relates to the observations of one or more receivers of carrier phase signals from one or more satellites. Thus, there may be wide-lane (WL) ambiguity, narrow-lane (NL) ambiguity, zero-difference (ZD) ambiguity, single-difference (SD) ambiguity, double-difference (DD) ambiguity, real-time-kinemic (RTK) ambiguity, and refraction-corrected (RC) ambiguity associated with phase measurements from one or more receivers or one or more satellites. Furthermore, some ambiguity will be specific to some modules because different modules or prediction filters (e.g., kalman filters) within these modules adapt to different update rates of the filters and the state of the filters, as well as the communication or state of data between the filters of different modules. Any reference to ambiguity may refer herein to a single ambiguity or to multiple ambiguities.

If the satellite navigation receiver can receive at least two frequencies, such as L1 and L2 frequencies, the difference of the L1 and L2 carrier phase measurements can be combined to form a Wide Lane (WL) measurement (e.g., having a wavelength of about 86.25 centimeters for GPS), and the sum of the L1 and L2 carrier phase measurements can be combined to form a Narrow Lane (NL) measurement (e.g., having a wavelength of about 10.7 centimeters). The wide lane measurements help resolve the wide lane integer ambiguity quickly and efficiently, while the narrow lane measurements help resolve the narrow lane ambiguity with minimal phase noise accurately and precisely. The refraction-corrected ambiguity compensates for the first order atmospheric delay.

Single difference measurements (e.g., of carrier phase or code phase (code)) are typically made with respect to one satellite, reference receiver 30, and rover receiver (12). Alternatively, single difference measurements may be made with respect to one receiver (reference receiver 30 or rover station 12) and a pair of satellites.

In contrast, the double difference measurement is typically formed with respect to two satellites, the reference receiver 30 and the rover receiver (12), or by subtracting two single difference measurements. However, some double difference measurements may be formed using two single difference measurements from the same reference receiver at two different times and associated with a pair of satellites, as will be described later in fig. 6.

Referring to fig. 1A, 1B, 3 and 4, a method or system provides a satellite correction signal having accurate, low-delay satellite clock estimates. The electronic data processing center 18 is arranged to collect raw phase measurements from a plurality of reference receivers 30 located at known respective locations (e.g. three-dimensional coordinates). A measurement preprocessing (MPP) module (36 in fig. 3) or data processor 20 of the data processing center 18 determines the widelane ambiguities and corresponding satellite widelane deviations for each satellite's collected phase measurements to provide assistance or appropriate constraints for efficient or rapid interpretation of the widelane ambiguities. The orbit solution module 38 or the data processor 20 determines the satellite correction data 16 for each satellite in the orbit solution at an orbit correction rate based on the collected raw phase and code measurements and the determined orbital narrow lane ambiguities and the corresponding orbital satellite narrow lane deviations, which can be estimated with the aid of the determined wide lane ambiguities and wide lane deviations. Advantageously, in an embodiment, the widelane ambiguities and corresponding widelane deviations determined by the measurement preprocessing module 36 may be shared and utilized among one or more predictive filters (39, 43, 412) in the other modules (38, 44, 42) of the corrected data estimator 34. The clock solution module 44 or the data processor 20 determines a slow satellite clock correction (e.g., a moderate delay satellite clock correction) at a slow update rate (or a moderate update rate) based on the satellite orbit correction data 50, the raw phase and code measurements collected, the time Zhong Zhai lane ambiguities and the corresponding satellite lane offsets, which may be estimated with the aid of the determined widelane ambiguities and widelane offsets. The low delay clock module 42 or the data processor 20 determines satellite clock correction data 16 or clock delta adjustment with lower delay for the slow satellite clock at a fast update rate based on the collected recent or most recently updated measurements of the raw phase measurements to provide clock correction data 16 with lower delay, which are closer to the current value than a plurality of previous measurements of the collected raw phase measurements for the slow satellite clock correction. In one embodiment, the fast update rate is a fixed rate that is greater than the slow update rate or the track rate (e.g., track update rate). However, in alternative embodiments, the fast update rate may be changed (dynamically) based on: (1) Availability, reliability, or quality (e.g., signal strength, accuracy factor, or other quality metric) of raw carrier phase measurements from a particular satellite or reference station, or (2) actively selecting a subset of satellite measurements or reference stations for estimating correction data from time to time based on the availability, reliability, or quality of raw carrier phase measurements.

The data processing center 18 incorporates satellite orbit correction data 50 and clock correction data 16 with low latency into the correction data 16, which correction data 16 has global availability for GNSS for transmission (e.g., satellite or wireless transmission) to one or more mobile receivers 12, which one or more mobile receivers 12 operate in a precise positioning mode, such as a precise point positioning (precise point positioning, PPP) mode. For example, the data processing center 18 incorporates satellite wide-lane bias, satellite orbit correction data, satellite narrow-lane bias from slow clock solutions, and clock correction data with low delay into correction data encoded on global satellite differential correction signals that are of global significance for GNSS transmission to one or more mobile receivers. The precise positioning mode (e.g., PPP mode) uses precise clock and orbit solutions for the received signals of the satellites and satellite biases to provide precise correction data 16, which precise correction data 16 is globally valid or independent of locally valid differential data, e.g., real-time kinematic (RTK) correction data 16, which real-time kinematic (RTK) correction data 16 is locally valid, accurate (e.g., for applications or off-road vehicles requiring greater than reliable decimeter level accuracy) for a short baseline between the reference station and the mobile station.

In one embodiment, the track rate (e.g., track update rate) is less than (e.g., or less than or equal to) the slow update rate; an orbital homodyne filter 404 is applied to estimate the orbital narrow lane ambiguity and corresponding narrow lane satellite bias by an orbital narrow lane estimator 39 (e.g., a narrow lane filter) at an orbital update rate based on the raw phase measurements collected. In another embodiment, the fast update rate is greater than the slow update rate or the track rate; a clock homodyne filter 408 may be applied to facilitate estimating clock narrow lane ambiguities and corresponding narrow lane satellite biases by the clock narrow lane estimator 43 (e.g., a narrow lane filter) at a slow update rate based on the raw phase measurements collected.

According to fig. 1A, the system is capable of providing correction data 16 encoded with satellite correction signals in real time, the correction data 16 having accurate, low-delay Global Navigation Satellite System (GNSS) satellite clock solutions or accurate, low-delay clock data. The final satellite clock solution includes accurate low-delay clock data representing an integrated solution of two simultaneous or parallel processes for estimating GNSS satellite clock estimates: (1) Slow clock solutions, and (2) low-delay clock solutions or medium-low delay solutions.

The delay may be measured based on a time difference between an earlier measurement time (e.g., epoch) associated with the collection of phase measurements of the satellite signals at one or more reference receivers 30 and a later measurement time (e.g., epoch) of the processed measurements received at the mobile receiver 12 or rover station. For example, fig. 2A and 2B divide the delay time difference value into different time periods for additional analysis, which will be described later. Low delay means a delay lower than the moderate delay of the slow clock solution of the clock solution module 44. The lower delay solution (e.g., of the low delay clock module 42) may be referred to as a fast clock solution (e.g., from a fast clock process), whereas the moderate delay solution (e.g., of the clock solution module 44) may be referred to as a slow clock solution (e.g., from a slow clock process). The low delay clock module 42 or the data processor 20 determines satellite clock correction data 16 or clock delta adjustment for the slow satellite clock having a lower delay based on a recent or most recently updated measurement of the collected raw phase measurement that is closer to the current value than a plurality of previous measurements of the collected raw phase measurement for the slow satellite clock correction to provide clock correction data 16 having a lower delay. Clock correction data 16 having a lower delay refers to total smoothed clock correction data, having a lower delay than the slow clock solution, containing contributions from satellite clock correction data and slow satellite clock correction data having a lower delay.

Low-delay clock data or low-delay clock solutions may refer to either or both of: (1) clock data associated with a low delay process; or (2) a final satellite clock solution generated from an integration solution based on a slow clock solution and a medium-low delay solution. The low-delay clock data improves the accuracy of the satellite clock and reduces the delay of the final solution (or increases timeliness) in real time, which is incorporated into the correction data 16 for allocation to the mobile receiver 12 or rover station.

In one embodiment, the slow clock process may utilize most or all possible measurements (e.g., carrier phase measurements from the reference data network 32 of the reference receiver 30) to estimate the slow clock solution, but with a slow clock delay, or moderate delay (e.g., about 6 seconds to about 10 seconds) associated with: (1) Estimating absolute satellite clocks, tropospheric bias, satellite narrow lane bias and satellite narrow lane ambiguities (e.g., refraction corrected narrow lane ambiguities) by the data processing center 18 and the associated reference data network 32; or (2) collecting raw phase measurements from the reference receiver 30 by the data processing center 18 and the associated reference data network 32; (3) both the above estimation and collection. In one embodiment, a slow clock solution is used to perform ambiguity resolution and evaluate the tropospheric bias and gradient of each reference receiver in order to determine the refraction correction of the clock narrow lane ambiguity and the corresponding narrow lane bias; the resolved ambiguity and/or estimated tropospheric bias from the slow clock process may be shared or used in a low delay clock solution. For example, tropospheric bias can be estimated based on an a priori model and residual tropospheric bias estimates from a slow clock solution. The slow clock process supports collecting and analyzing more measurements than the low delay process to facilitate accurate or absolute accuracy of the clock estimates and finer slow clock models. In one embodiment, the slow clock process has greater data processing capacity or throughput requirements than the low latency clock process, such that the computation of the slow clock solution may take about one second to about two seconds even though the data processing center 18 supports a parallel data processing environment.

Meanwhile, the low-delay clock process uses fewer measurements (e.g., carrier phase measurements from the reference data network 32) than the slow clock process, and the low-delay process has a low delay or low-delay clock delay (e.g., about one to two seconds) to collect satellite clock changes, or to collect and calculate satellite clock changes, at a low delay rate that is greater than the slow clock rate or the moderate delay rate. The data processor 20 or data processing center 18 integrates the low delay clock data with the slow clock data and the track data at a low delay rate to provide a consistent, accurate, and timely set of correction data 16.

Working with an orbit solution and a slow clock solution at a low delay rate, the data processing center 18 may transmit a consistent set of correction data 16 in a timely manner (e.g., with reduced delay or low delay relative to the slow clock process), the set of correction data 16 including satellite orbits, clocks (e.g., absolute clock estimates), wide lane satellite biases, narrow lane satellite biases, and quality information. In particular, the data processing center 18 may transmit the correction data 16 to one or more mobile receivers having the correction wireless device 14 in real-time through satellite signals (e.g., L-band signals) in fig. 1A, through the wireless communication system 135 (in fig. 1B), or through the wireless communication system 57 (in fig. 3) (e.g., through the internet 56). In one embodiment, the correction data is transmitted in real-time without significant delay (between the measurement time or raw phase measurements at the reference receiver 30 and the availability of correction data at the mobile receiver 12) that would tend to reduce the accuracy of a position estimate at the mobile receiver 12 that has a level of accuracy (e.g., channel-to-channel) or better of about 5 centimeters, that has a reliability of about ninety-five percent, and that has a variance of less than one standard deviation.

In FIG. 1A, in one embodiment, the system includes a cluster of satellites (e.g., with satellite transmitters 10) including at least those satellites that are within view or reception range of one or more reference satellite receivers (e.g., reference GNSS receivers). In practice, the reference receivers 30 (e.g., GNSS reference stations) are distributed globally at locations that have good satellite geometry and are visible to a set of satellites.

Each reference receiver 30 has a digital portion of the receiver that includes an electronic data processing system that includes an electronic data processor, a data storage device, a data port, and a data bus that supports communication between the data processor, the data storage device, and the data port. In addition, the receiver includes a measurement module for measuring one or more received satellite signals from the navigation satellite transmitter 10. In one embodiment, the measurement module (e.g., carrier phase measurement module) is associated with baseband or intermediate frequency processing or stored as software instructions in a data storage within the digital portion of the receiver 30.

Each reference receiver 30 has a measurement module that measures an observability, such as carrier phase of one or more received satellite signals from each satellite. The measurement module of the reference receiver 30 may also measure the pseudorange or code phase of a pseudorandom noise code encoded on one or more carrier signals. In addition, a demodulator or decoder (e.g., stored as software instructions in a data storage device) of the reference receiver 30 may decode navigation messages, such as ephemeris data, that are encoded or otherwise combined with pseudorandom noise codes on the received satellite signals. The reference receiver 30 receives and transmits measurements, ephemeris data, other observables, and any information derived from a hub that is transmitted to the data processing center 18 or has similar processing capabilities in real time.

In fig. 1A, a set of reference receivers and communication links is referred to as a reference data network 32. In one embodiment, each reference receiver 30 transmits a set of carrier phase measurements of the received satellite signals (e.g., via a communication link, communication network, wireless channel, communication line, transmission line, or other means), along with associated satellite identifiers and ephemeris data, to an electronic data processing center 18 (e.g., a reference data processing hub).

In one embodiment, data processing center 18 includes an electronic data processor 20, a data storage device 24, and one or more data ports 26 coupled to a data bus 22. The data processor 20, the data storage 24, and the one or more data ports 26 may communicate with each other via a data bus 22. The software instructions and data stored in the data storage 24 may be executed by the data processor 20 to implement any of the blocks, components, or modules (e.g., electronic modules, software modules, or both) described in this disclosure. The data processor 20 may include a microcontroller, a microprocessor, a programmable logic array, an Application Specific Integrated Circuit (ASIC), a digital signal processor, or another device for processing data, manipulating, accessing, retrieving, and storing data. The data storage device 24 may include electronic components, non-volatile electronic memory, optical storage, magnetic storage, or another device for storing digital or analog data on a tangible storage medium (e.g., optical disk, magnetic disk, electronic memory). The data port 26 may include a buffer memory, a transceiver, or both, for interfacing with other network elements, such as a reference receiver 30 or a ground satellite uplink station 28.

In one embodiment, the data processing center 18, the data processor 20, or the correction data estimator 34 receives the phase measurements from the reference receiver 30 and the corresponding satellite identifiers, the reference receiver identifiers (or corresponding coordinates), and processes the phase measurements to estimate a clock bias or corresponding clock solution for each satellite (or more precisely for each satellite signal) for incorporation into the correction data 16. For example, correction data estimator 34 includes software instructions or modules for determining correction data 16 based on phase measurements received from reference data network 32 or reference receiver 30. As shown in fig. 1A, the clock solution, clock bias, or correction data 16 is provided to a ground uplink station 28 or another communication link. For example, the terrestrial uplink communicates or transmits clock solutions, clock bias, or correction data 16 to a communication satellite 35 (e.g., a repeater).

In turn, the communication satellite 35 is adapted to make the correction data 16 available or to transmit the correction data 16 to the correction wireless device 14. The correction wireless device 14 is coupled to a mobile receiver 12 (e.g., a mobile GNSS receiver or a mobile satellite receiver) or a rover station. The mobile receiver 12 also receives satellite signals from one or more satellite transmitters 10 (e.g., GNSS satellites) and measures carrier phases of the satellite signals received from the satellite transmitters 10. In conjunction with the phase measurements of the mobile receiver 12, the mobile receiver 12 may use the accurate clock solutions or clock offsets in the correction data 16 to estimate the accurate position, attitude, or velocity of the mobile receiver 12 or its antenna. For example, the mobile receiver 12 may use a precise location estimator, such as a precise point location (PPP) estimator, that uses a precise clock and orbit solution for the received signals of the satellite transmitter 10.

Herein, the method and real-time Global Navigation Satellite System (GNSS) receiver navigation techniques may enable centimeter-level accuracy positioning by using real-time global differential correction data 16. This correction data 16 is globally available and valid by means of one or more of the following: (1) Satellite communications in fig. 1A (e.g., L-band geostationary communication satellite); (2) The wireless communication system 135 in fig. 1B (e.g., cellular communication); or (3) a wireless communication system 57 (e.g., a cellular wireless system or a WiFi system coupled to the internet 56 for receiving correction data 16 from the server 54). Global differential correction eliminates the need for local reference stations and radio communications that would otherwise be used to establish a short baseline (e.g., less than 20 km to 30 km) between the reference receiver 30 and the rover station for accurate position accuracy, as compared to local reference station correction (e.g., by real-time kinematic (RTK) base stations or some wide area correction that does not strictly adhere to the globally available PPP model).

The system of fig. 1B is similar to the system of fig. 1A except that the ground uplink station 28, communication satellite 35, and correction wireless device 14 of fig. 1A are replaced by wireless communication device 128, wireless communication system 135, and correction wireless device 114 of fig. 1B, respectively. Further, correction wireless device 14 may include a satellite receiver and correction wireless device 114 may include a cellular transceiver, a wireless receiver, or another wireless communication device. Like reference numerals in fig. 1A and 1B denote like elements.

In fig. 1B, the data processing center 18 or the correction data estimator 34 provides the correction data 16 to the wireless communication device 128, either directly or indirectly, through one or more communication networks (e.g., the internet), communication links, data packet networks, or communication channels. In turn, the wireless communication device 128 transmits the data to one or more wireless communication systems 135. If multiple wireless communication systems 135 are used, a communication network, communication link, packet network, switching network, mobile telephone switching office, microwave link, communication line, fiber optic link, or communication channel may interconnect wireless systems 136 to support communication of correction data 16 from wireless communication device 128 to correction wireless device 114. Thus, the mobile receiver 12 obtains correction data from the correction wireless device 114 with an acceptable level of delay.

Fig. 2A is a graph showing delays associated with the provision of correction signals, and more particularly, correction signals having a set of clock errors for corresponding satellites, wherein measurements are collected with moderate delays during a measurement collection time. Fig. 2A is a diagram illustrating delays associated with the provision of correction signals, and more particularly, correction signals having correction data 16, which may include a set of clock errors for corresponding satellites. The delay time is measured along the horizontal axis 60. Correction delay is one of the key factors affecting the overall system performance of providing GNSS correction data services to end users or subscribers. For example, the correction delay may be defined as the time difference between a measurement epoch at the reference receiver 30 or a set of reference receivers 30 from the processed measurements of the reference data network 32 (or a preprocessing completion epoch in the measurement preprocessing module 37 for the set of reference receivers) and the reference epoch of the correction data 16 applied or applied in the one or more mobile receivers 12. In one embodiment, the correction delay of the correction data 16 or the correction delay of the correction signal is a combination of three basic sources: (1) A measurement collection period 62 (e.g., from T0 to T1) for reference receiver 30 measurements (e.g., GNSS data from reference receiver 30, reference receiver 30 may be geographically located worldwide) to reach data processing center 18 (or server); (2) A processing time period or clock estimation processing time period 64 (e.g., from T1 to T2 or T1 to T3) of the data processing center 18; and (3) a correction transmission period 66 (e.g., T3 to T4) transmitted to the end user at the mobile receiver 12.

In an alternative embodiment, the additional delay source is a clock messaging period 65 (e.g., T2 to T3) between the processing period and the correction delivery period.

As shown in fig. 2A, the first delay or measurement collection time 62 is between the measurement time of the reference receiver 30 or set of reference receivers 30 and the reception time at the electronic data processing center 18. The first delay is associated with factors such as the location of the reference receiver 30 and the distance between the data processing centers 18, and propagation delays associated with the transmission of electrical or electromagnetic signals between the reference receiver 30 (at different locations, possibly throughout the world) and the data processing centers 18. In certain embodiments, the first delay may include a preprocessing time for position estimation, ambiguity, tropospheric delay, atmospheric delay, clock bias, receiver bias, or other preprocessing estimates provided by one or more reference receivers 30 or by a measured value preprocessing (MPP) module 36 or by both the reference receivers 30 and the MPP module 36. For example, when the carrier phase measurements and the code phase measurements from the reference data network 32 reach the data processing center 18, the carrier phase measurements and the code phase measurements from the reference data network 32 are collected and preprocessed.

The data processing center 18 may adjust the measurement collection time period 62 over a range to facilitate accuracy or speed. The longer the data processing center 18 waits to collect measurement data (and pre-process data) from the reference network 32 before processing the measurement data, the more measurements the data processing center 18 can collect to support the improved accuracy of orbit, clock, and satellite bias and reliability of network ambiguity resolution. However, because correction data 16 will be stale or outdated when it is ultimately received at mobile station 12, and because other delays or delays in addition to the measurement collection time must be considered in evaluating whether correction data 16 is stale or sufficiently old at mobile station 12, longer data collection times can reduce the accuracy of correction data 16 if measurement collection time period 62 exceeds a maximum threshold or is too long.

The second delay or clock processing period 64 is between the receipt at the data processing center 18 and the processing time at the data processing center 18, which may be affected by: throughput or capability of data processor 20, clock speed of data processor 20, specifications or operations to be performed within each time unit of data processor 20. The processing time of the data processing center 18 or the correction data estimator 34, e.g., the innovative clock track real-time estimator (iCORE), must be minimized as much as possible in order to allow the data processor 20 to output the correction data 16 at a rate of 1 hertz (Hz) or higher and to minimize the final correction delay on the rover side.

The third delay or clock messaging period 65 is associated with the time between the data processor 20 at the data processing center 18 completing data processing and the transmission of data to the ground uplink station 28, the communication satellite 35, or other communication device (e.g., a wireless communication network). For example, the fourth delay or correction transmission period 66 is associated with the transmission of the correction data 16 message from the communication satellite 35 or other communication device. Although the first delay (62) is listed as about 6 seconds in fig. 2A; the second delay (64) is listed as about two seconds; the third delay (65) is listed as about zero to three seconds and the fourth delay (66) is listed as about one second to two seconds; other durations of the delay are possible and may fall within the scope of the claims. Due to limitations in satellite channel bandwidth (e.g., L-band bandwidth) and propagation delay of geosynchronous satellites, some time (e.g., about 4 seconds) is required to transmit the full set of corrections from the data processing center 18 to the mobile receiver 12 or rover station. According to some models of location accuracy, each additional second delay or target delay range of the clock correction above the total target delay (e.g., from the measurement time to the reception time at the rover) may reduce the path-to-path accuracy for rover navigation by up to 5%.

Fig. 2B is a graph showing delays associated with the provision of correction signals, and more particularly, correction signals having a set of clock errors for corresponding satellites, wherein measurements are collected with lower delays than the measurements of fig. 2A. Fig. 2A is similar to fig. 2B except that the first delay 162 and the second delay 164 of fig. 2B are shortened as compared to the first delay 62 and the second delay 64 of fig. 2A, respectively. Like reference numerals in fig. 2A and 2B denote like elements.

As shown in fig. 2B, the first delay 162 or measurement collection time is similar to the first delay 62 of fig. 2A, except that the first delay 162 has a duration of about two seconds or less. The first delay 162 may be reduced relative to the first delay 62 by one or more of the following factors: (1) Collecting measurement data from a set of reference receivers 30 over a short period of time; (2) Ignoring, by the data processing center 18, measurement data from satellites, satellite signals or reference receivers that do not pass quality inspection or statistical analysis, such as standard deviation for ambiguity resolution of widelane ambiguities, narrow-lane ambiguities at one or more reference receivers 30, or refraction-corrected ambiguities; and (3) enhance throughput or data processing capacity of the data processing center 18 by higher clock speeds, parallel data processing, correction of improved effective software instructions within the data estimator 34, and the like.

Similarly, the second delay 164 or clock processing time is similar to the second delay 64 of fig. 2A, except that the second delay 164 has a duration of about ten milliseconds or less. The second delay 164 may be reduced relative to the second delay 64 by one or more of the following factors: (1) Collecting measurement data from a set of reference receivers 30 over a short period of time; (2) Ignoring, by the data processing center 18, measurement data from satellites, satellite signals or reference receivers that do not pass quality inspection or statistical analysis, such as standard deviation for ambiguity resolution of widelane ambiguities, narrow-lane ambiguities at one or more reference receivers 30, or refraction-corrected ambiguities; and (3) enhance throughput or data processing capacity of the data processing center 18 by higher clock speeds, parallel data processing, correction of improved effective software instructions within the data estimator 34, and the like.

FIG. 3 is a block diagram of another embodiment of a system for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks. The system of fig. 3 is similar to the system of fig. 1A or 1B, except that the data storage device 24 stores or supports the correction data estimator 34 and provides two alternative paths for the flow of the correction data 16 to the end user of the mobile receiver 12. Like reference numerals in fig. 1A, 1B, and 3 denote like elements.

The correction data estimator 34, e.g., an innovative clock track real-time estimator (iCORE), includes modules or software instructions executable by the data processor 20 for providing as output correction data 16 based on the input of carrier phase measurements and associated data from one or more reference receivers 30. The associated data may include data such as satellite identifiers, satellite signal identifiers (e.g., frequencies or frequency bands), or ephemeris data associated with received satellite signals, reference station identifiers (e.g., or reference station coordinates), measurement timestamps associated with measured carrier bits, and other assistance data. Furthermore, in other embodiments, the associated data may include pre-processed data, resolved widelane ambiguities, resolved narrow elane ambiguities, or resolved refraction-corrected ambiguities for any reference receiver 30 or reference station having a known or fixed location (e.g., three-dimensional coordinates).

The correction data estimator 34 generates correction data 16, which correction data 16 is capable of ambiguity resolution at one or more mobile receivers 12 or rover stations receiving the correction data 16 via the correction wireless device. The correction data estimator 34 employs an innovative ambiguity resolution algorithm to generate low-delay clock data and satellite bias. For example, the architecture of the correction data estimator 34 or data processing center 18 is well suited to support the processing of measurement data (and associated preprocessing data) for many (e.g., one hundred or more) reference receivers 30 or reference stations, including all necessary computations within fractions of a second, such as measurement preprocessing, orbit and clock determination, ambiguity resolution, and final correction data 16 generation.

In one embodiment, the correction data estimator 34 includes a measurement preprocessing (MPP) module 36, a track solution module, a clock solution module 44, a low delay clock module 42, and a correction distribution module to provide the global differential correction data 16. In one embodiment, the measurement value preprocessing (MPP) module 36 receives reference network data 46 from the reference data network 32 of one or more reference satellite receivers. The reference network data 46 may include one or more of the following: the raw measurements, raw carrier phase measurements from each reference satellite receiver, raw code phase measurements from each reference satellite receiver, reference satellite receiver identifiers, position offsets or position offset vectors of the reference receiver 30 from its known position, phase offsets corresponding to the position offset of the reference receiver 30, atmospheric offset data, satellite offset data, receiver clock offset data, satellite clock offset data, or other data. The raw measurements may refer to raw carrier phase measurements from one or more reference satellite receivers, or raw code phase measurements from one or more reference satellite receivers, or both.

In one embodiment, the correction data estimator 34 may generate the correction data 16 or correction signal in real-time based on the reference network data 46 or collected measurement data from the reference data network 32, wherein the correction data 16 is generated to provide a centimeter-level accuracy estimate comprising: (1) satellite orbit estimates, (2) satellite clock estimates, and (3) satellite phase bias and quality information thereof.

In one embodiment, the measurement preprocessing module 36 accepts the reference network data 46 as input data and applies the wide lane estimator 37 (e.g., ambiguity Resolution Estimator (ARE)) to output the wide lane ambiguities and corresponding wide lane deviations. For example, the preprocessing module 36 or the wide lane estimator 37 (e.g., a wide lane filter or a kalman filter) may output a fixed wide lane ambiguity and associated wide lane bias (48).

The measurement preprocessing module 36 communicates the widelane ambiguity and corresponding widelane bias data to the track solution module 38 and the clock solution module 44. The orbit solution module 38 receives the input of the widelane ambiguity and the corresponding widelane bias data and applies an orbit narrow lane estimator 39 (e.g., a narrow lane ARE) to provide predicted satellite orbit data 50 (e.g., O2C data) output. The predicted satellite orbit data 50 may be used to correct for orbit errors in slow clock and low delay solutions and generate orbit correction signals for incorporation into the correction data 16 for provision to end users.

The orbit solution module 38 communicates the predicted satellite orbit data to the clock solution module 44. The clock solution module 44 receives inputs of predicted satellite orbit data 50, as well as widelane ambiguities and associated widelane deviations (48). The clock solution module 44 applies a clock narrow lane estimator 43 (e.g., a narrow lane (ambiguity resolution estimator), narrow lane filter or kalman filter) and outputs predicted orbit data 50, clock satellite bias data and satellite bias quality data (51). For example, in one embodiment, clock solution module 44 outputs one or more of the following: predicted orbit data, clock satellite bias data, satellite bias quality data, satellite slow clock correction, satellite wide-lane bias correction, and satellite narrow-lane bias correction.

The low delay clock module 42 communicates with the clock solution module 44 to receive predicted orbit data 50, clock satellite bias data, which may include satellite wide-lane bias correction and satellite narrow-lane bias correction, and satellite bias quality data (51). Furthermore, in one embodiment, low-delay clock module 42 receives satellite slow clock corrections. The low-delay clock module 42 outputs low-delay correction data 16 including one or more of the following: low-latency accurate satellite orbit correction data 50 for the respective satellite, low-latency accurate clock data for the respective satellite, wide-lane satellite bias, and narrow-lane satellite bias. In one embodiment, the low-latency data is provided at a data transfer rate greater than similar higher-latency data provided by the clock solution module 44, where the low-latency data is updated periodically, corresponding to the greater data transfer rate, to provide accurate and current correction data 16.

In some configurations, the data processing center 18 may communicate via a communication link with one or more auxiliary data processing hubs (not shown) that are geographically distributed (e.g., on a global basis), where each auxiliary data processing hub is configured with hardware and software similar to the data processing center 18 with the correction data estimator 34, and the data processing center 18 may control the one or more auxiliary data processing hubs.

For example, correction manager 40 may select correction data 16 (e.g., optimal correction data 16 or most reliable correction data 16) provided by data processing center 18 for distribution to end users, alone or in combination with one or more auxiliary data processing hubs. In addition, correction manager 40 may select a geographic range of measurement data or an identity of the satellite (e.g., outliers or unreliable measurements from faulty satellites may be ignored) for distribution to end users of correction data 16 via satellite or wireless communication system 57.

Correction manager 40 is capable of monitoring correction data 16 for error correction and distributing the data to end users or subscribers serviced by data correction data 16. Correction manager 40 may distribute correction data 16 via a satellite communication network, a wireless network (e.g., wiFi,802.11, or cellular network), or both. The broadcast system is capable of transmitting optimal global differential corrections from a plurality of correction generation servers 54 (e.g., data processing center 18 and one or more auxiliary data processing hubs) to the user receiver. For example, the set of global differential corrections may be selected and up-bound to INMARSAT communications satellites via Land-Earth Station (LES), as shown in FIG. 3.

Correction manager 40 is capable of transmitting or distributing correction data 16 to satellite uplink communications devices or distributing satellite data via a satellite communications network. In turn, the satellite uplink communications device utilizes a combination of transceiver, transmitter and receiver to provide signals to the communications satellite 35 to communicate correction data 16 via electromagnetic or satellite signals (e.g., L-band signals) to the mobile receiver 12 or rover station equipped with the correction wireless device 14. In some embodiments, the electromagnetic or satellite signals with correction data 16 may be encrypted or encoded such that only subscribers or licensees may access, decode or decrypt correction data 16, or correction data 16 of some level of accuracy (e.g., SF3 correction data 16).

Correction manager 40 is capable of transmitting or distributing correction data 16 to a server 54 having access to an electronic communications network, such as the internet 56. For example, server 54 may include a computer that accesses the Internet 56 via an Internet service provider to enable correction data 16 to be transmitted in one or more data packets 55 (e.g., internet protocol data packets). The data packets may be processed by a wireless communication network 57 (e.g., a WiFi wireless system, a local wireless network, a wide area wireless network, or a cellular communication system) via the correction wireless device 14, which correction wireless device 14 may include a smart phone, a WiFi-enabled communication transceiver, or another device for receiving the correction data 16 and providing the received correction data 16 to the mobile receiver 12 or the rover station. As in the case of satellite signals having correction data 16, the correction data 16, data packets 55, or both, sent by server 54 may be encrypted or encoded such that only the subscriber or licensee may access, decode or decrypt the correction data 16, or some level of accuracy of the correction data 16.

The end user's mobile receiver 12 is capable of receiving correction data 16, the correction data 16 comprising global differential corrections. The mobile receiver 12 or rover station is able to resolve ambiguities and achieve centimeter level navigation based on the received correction data 16.

Fig. 4 shows an illustrative example of the correction data estimator 34 of fig. 3 in more detail. Like reference numerals in fig. 3 and 4 indicate like elements, modules or features.

The correction data estimator 34 includes a measurement preprocessing module 36, a track solution module 38, a clock solution module 44, and a low delay clock module 42. The measurement preprocessing (MPP) module prepares "cleaned" measurements and provides a wide lane fix ambiguity and wide lane bias product by correcting the data estimator 34 and its other modules. The orbit solution module 38 provides accurate satellite position and velocity estimates to aid in making the appropriate geometric or range estimates between the particular reference receiver 30 and the corresponding satellite. The orbit solution or accurate orbit data is provided for use by the correction data estimator 34 and its other modules. Clock solution module 44 provides satellite slow clock solution estimates and narrow lane offset products at a low rate, slow clock rate, or slow update rate. The low-delay clock module 42 provides the fast satellite clock estimate at a low-delay update rate (e.g., about 1 hertz (Hz) or higher), a fast rate, or a fast update rate, the fast update rate being greater than the slow clock rate. In addition, the low-latency module integrates, manages, and communicates status data and filters the results for sharing among the ambiguity resolution filters to enable the MPP module, the orbit solution module 38, and the slow clock solution module 44 to output or generate a consistent set of correction data 16 or signals in real-time.

Each of the MPP module, the track solution module 38, and the clock solution module 44 includes two parts: (1) A homodyne filter and (2) a network ambiguity resolution module or filter. In one embodiment, each homodyne (ZD) filter (e.g., a kalman filter) performs one or more of the following: (a) processing the ZD measurements; (b) defining or forming state variables of the ZD filter; (c) And (e.g. based on defined state variables and states) performing or processing an update and/or dynamic update of the ZD filter. In one embodiment, the network Ambiguity Resolution Estimation (ARE) module performs or performs ambiguity resolution by one or more prediction filters, such as a wide-lane estimator 37 (e.g., a wide-lane filter), a narrow-lane estimator 39 (e.g., a narrow-lane filter), or another prediction filter (e.g., a kalman filter). The network ambiguity resolution estimation module can resolve the wide-lane ambiguity and the narrow-lane ambiguity. Because the update rates and data states of the different modules may be different, different ZD filters and ARE modules (e.g., wide lane estimator 37 (e.g., wide lane filter), narrow lane estimator 39 (e.g., narrow lane filter), or both) ARE used for the different modules, such as the MPP module, the track solution module 38, and the clock solution module 44.

In one embodiment, the low delay clock module 42 may use only carrier phase measurements to derive clock variations between two different epochs. To improve computational efficiency, a double difference approach between time and satellite is used to reduce the magnitude of estimated states, such as ambiguous states and receiver clock estimates. The tropospheric bias is corrected using an a priori model and residual tropospheric bias estimates from a slow clock solution. The low delay clock module 42 estimates satellite clock variations instead of reference receiver 30 clock variations. The computation of the data processing center 18 is very efficient. For example, if the data processing center 18 is implemented on an existing desktop computer at the time of filing the present disclosure, the data processing center 18 may take several milliseconds for any epoch to complete processing all measurements of many locations or reference receivers 30 (e.g., sixty or more reference receivers 30).

Measured value preprocessing module

As shown, the measurement preprocessing module 36 also includes a measurement preprocessing homodyne filter 400 and a network wide lane ambiguity resolution estimator (e.g., a wide lane estimator 37 (e.g., a wide lane filter)). Measurement preprocessing homodyne filter 400 supports accurate point positioning (PPP). The optional measurement preprocessing homodyne filter 400 may be used to determine a no-difference or homodyne (ZD) ambiguity state or a floating ambiguity state associated with the raw carrier phase measurement (e.g., L1 raw carrier phase, L2 carrier phase, wide-lane difference of L1/L2 combinations of carrier phases) for one or more reference receivers 30 in the reference data network 32. The homodyne filter is shown in dashed lines in fig. 4, showing that the homodyne filter is optional, and may be included within a wide lane estimator 37 (e.g., a wide lane filter) in alternative embodiments. For example, the homodyne ambiguity state may be determined based on correction data 16 containing satellite bias information from a network or set of reference receivers 30.

After the measurement preprocessing module 36 receives raw measurements from each reference receiver 30 of the reference data network 32, the measurement preprocessing module 36 processes, preprocesses, and "cleans" the measurements at regular or sampling intervals, and resolves Wide Lane (WL) ambiguities associated with the received carrier phase measurements for each satellite in the field of view of each reference receiver 30. The measurement preprocessing module 36 provides support for orbit/clock solution and low delay clocks by providing "clean" carrier measurements, and corresponding fixed widelane ambiguities, and respective satellite WL biases.

In one embodiment, the measurement preprocessing module 36 uses Melbourne-The linear combination is used as a homodyne (ZD) measurement to estimate the following state variables:

1) The ZD float WL ambiguity for each visible satellite and reference position (reference receiver 30) combines receiver wide-lane offset and WL integer ambiguity.

2) One wide lane offset for each satellite.

3) One GLONASS IFB WL coefficient per tracking position.

Using ZD floating WL ambiguity as a constraint or search constraint for a wide-lane estimator 37 (e.g., a wide-lane filter) (e.g., a kalman filter), the measurement preprocessing module 36 or the wide-lane estimator 37 (e.g., a wide-lane filter) resolves the WL ambiguity in double-difference (DD) and single-difference (SD) forms, where the receiver WL bias is cancelled.

In an alternative embodiment of the reference receiver 30 tracking GLONASS satellites, the measurement pre-processing module 36 determines GLONASS IFB WL coefficients for each tracked location, where the sensitivity coefficients of the GLONASS IFB WL coefficients are the number of satellite frequencies. This state variable is only applicable to the GLONASS case and not to other GNSS systems like GPS.

To make the calculation efficient, if no cycle slip is detected, the measurement pre-processing module 36 calculates ZD Melbourne-The measurements were averaged. For example, the reference receiver 30 includes a cycle slip detector for detecting the carrier phase of each received signal from a given satellite or a minimum set of satellites (e.g., five satellites) required to reliably track the three-dimensional position of the reference receiver 30Carrier cycle slip in the bit measurements. For each sampling interval, the measurement pre-processing module 36 or ZD filter (e.g., ZD Kalman filter) processes the smoothed ZD Melbourne-/on a location-by-location basis>And (5) measuring values. At each measurement update interval, the ZD Kalman filter dynamic update and the measurement update are processed to update the state variables.

In an alternative embodiment, the GPS and GLONASS systems have two separate wide lane estimators 37 (e.g., wide lane filters) associated with the measurement preprocessing module 36. Because the bias of the reference receiver 30 is not used for user receiver navigation, the bias of the reference receiver 30 is not the desired global differential product. Therefore, to reduce the filter size and computational complexity, the WL bias of the reference receiver 30 is not explicitly estimated and instead is combined into the ZD floating ambiguity state.

Specifically, the preprocessing module 36 or network wide lane ambiguity filter uses the homodyne (ZD) Melbourne-The linear combination is used as an input measurement to estimate one wide-lane floating ambiguity state for each satellite in view. The wide-lane satellite bias may be broadcast to the mobile receiver 12 in real-time within the correction data 16 or correction signal and this term will be compensated for using equation (2). />

In one embodiment, a wide-lane estimator 37 (e.g., a wide-lane filter) that may be applied to the PPP determination uses the following equations described below. Given code and phase measurements from two frequencies, e.g., L1 and L2 for GPS, G1 and G2 for GLONASS, a Melbourne-Linear combination->

By developing equation (1) above, it can be shown that the geometric distance related terms including distance, receiver and satellite clocks, ionosphere and troposphere errors, and phase end (wind-up) terms are cancelled out. It can be expressed as equation (2):

wherein:

λ _WL is the wide lane wavelength, about 86.4 cm for GPS, and c is the speed of light,

is the full perimeter widelane ambiguity for satellite j,

b _WL is the wide-lane receiver bias (one for each receiver and all visible satellite clusters), which is the combination of the L1 and L2 receiver code bias and the phase bias, as shown in equation (5):

Wherein most of the GLONASS inter-frequency bias in the code measurementsAnd->Typically assumed to be a linear or trigonometric function of the number of GLONASS satellite frequencies; for all satellites in view, it is different from the case of CDMA signals (such as GPS);

wherein, IFB ^j Is the inter-frequency bias of satellite j, e.g., the inter-frequency bias of GLONASS satellites;

wherein,is the wide-lane satellite j bias (one for each satellite); and

wherein,is the wide-lane measurement error of satellite j, including white noise, multipath, and remaining unmodeled error.

Regarding the inter-frequency bias for each satellite, for the GLONASS cluster, the linear model may be approximated as equation (6):

IFB ^j ≈k·n ^j (6)

where k is the IFB coefficient of the receiver code bias. IFB varies from receiver to receiver, as well as from one site (antenna and wiring set up) to another. Modeling in this way, typically k is less than 0.1.

Wide lane satellite j bias(one for each satellite) is a combination of the L1 and L2 satellite code bias and the satellite phase bias, as shown in equation (7); satellite bias varies slowly with time; satellite and receiver wide lane bias will vary over time:

wherein,satellite bias at satellite j, which is the code phase, or at frequency L1 (f ₁ ) Upper encoded pseudo-range signal, whereinSatellite bias or frequency L2 (f) of satellite j, which is the code phase ₂ ) Pseudo-range of>Satellite bias of satellite j, which is the carrier phase on frequency L1, wherein +.>Is the satellite bias of satellite j for the carrier code on frequency L2.

Rail solution module

The track solution module 38 relates to track determination. In the correction data estimator 34, the other main blocks including the slow clock estimation block and the low delay clock block 42 do not estimate the satellite orbit. Other modules rely entirely on predicted tracks within a corresponding valid track period (e.g., a few minutes) from the track solution. Because the GNSS satellite orbit is smooth, the orbit solution module 38 runs at an orbit correction rate or a lower rate, such as 300 seconds per iteration or update of the orbit solution. In the correction data estimator 34, the modules including the measurement value preprocessing module 36, the track module, the clock module, and the low-delay clock module 42 may run in parallel.

The orbit solution uses refraction-corrected codes and carrier-phase measurements from a global reference station network. The following three types of state variables are considered in the orbit solution module 38 and its associated filters (e.g., the orbit homodyne filter 404 and the network NL filter for ambiguity resolution):

1) Satellite-dependent state variables including satellite position, velocity, satellite clock, satellite lane bias, yaw rate, and empirical solar radiation modeling parameters.

2) State variables that depend on the receiver include reference position, receiver clock, residual tropospheric bias and gradient, carrier phase ambiguity.

3) Common state variables include earth orientation parameters such as polar motion and UT 1-UTC.

To provide global differential positioning services, e.g. STARFIRE ^TM Correction data 16 services must estimate and transmit accurate satellite clocks and orbits to end user receivers in real time. STARFIRE correction data 16 service is Dier corporation (Deere) of Morin, ill&Company). In general, these predicted orbits are considered known in clock estimation because the errors of the predicted satellite orbits (e.g., known as O2C data) within a few minutes are quite small and stable and can even be absorbed by the estimated clock. The correction data estimator 34 may use the predicted orbit data to correct for orbit errors in the slow clock solution and the low delay solution and generate the correction data 16 in real time.

In one embodiment, the track solution module 38 may include a track homodyne filter 404 and a network narrow lane Ambiguity Resolution Estimator (ARE). With the benefit of preprocessing the results (e.g., floating ambiguity state) of the homodyne filter 400, the track homodyne filter 404 can be used to determine a no-difference or homodyne (ZD) ambiguity state or floating ambiguity state for one or more reference receivers 30 in the reference data network 32, the no-difference or homodyne (ZD) ambiguity state or floating ambiguity state being associated with an original carrier phase measurement (e.g., an L1 original carrier phase, an L2 carrier phase, a wide-lane difference of an L1/L2 combination of carrier phases, or a narrow-lane difference of an L1/L2 combination of carrier phases).

A network narrow-lane ambiguity estimator (associated with the orbit resolution module 38) may estimate a narrow-lane ambiguity (e.g., a fixed full-cycle NL ambiguity) or a refraction-corrected narrow-lane ambiguity for one or more reference receivers 30 in the reference data network 32 using the benefits of a no-difference or homodyne (ZD) ambiguity state or a floating ambiguity state associated with the raw carrier-phase measurements or the narrow-lane differences of the raw carrier-phase measurements and the resolved WL ambiguities provided by the measurement preprocessing module 36. The resolved WL ambiguity can be used as a constraint in the search process or to assist the NL ambiguity estimator (e.g., NL filter) to converge quickly to a whole-cycle ambiguity solution for the carrier-phase measurements.

Clock solution module

In one embodiment, the clock solution module 44 may include a clock homodyne filter 408 and a network narrow lane Ambiguity Resolution Estimator (ARE). With the benefit of preprocessing the results (e.g., floating ambiguity state) of the homodyne filter 400, the clock homodyne filter can be used to determine a no-difference or homodyne (ZD) ambiguity state or floating ambiguity state for one or more reference receivers 30 in the reference data network 32, the no-difference or homodyne (ZD) ambiguity state or floating ambiguity state being associated with an original carrier phase measurement (e.g., L1 original carrier phase, L2 carrier phase, wide lane difference of L1/L2 combinations of carrier phases, or narrow lane difference of L1/L2 combinations of carrier phases). While the ZD fuzzy floating state and other filter states may be shared to the extent that the shared filter states are timely, the clock ZD filter may operate at a different update rate for states than the other ZD filters in correction data estimator 34.

A network narrow-lane ambiguity estimator (associated with the clock solution module 44) may estimate a narrow-lane ambiguity (e.g., refraction-corrected, fixed whole-cycle NL ambiguities) for one or more reference receivers 30 in the reference data network 32 using the benefits of a no-difference or homodyne (ZD) ambiguities or a floating ambiguities associated with the raw carrier-phase measurements or the narrow-lane differences of the raw carrier-phase measurements and the WL ambiguities provided by the measurement preprocessing module 36 providing constraints.

Slow clock solving module

All or most of the measurements (e.g., carrier phase measurements) from the reference data network 32 are collected, preprocessed, and batch processed as they arrive at the data processing center 18. The longer the data processing center 18 or correction data estimator waits, the more measurements are collected that are available for processing, but the longer the delay and the greater the likelihood that the clock solution will become stale upon reaching the end user's mobile receiver 12. In some embodiments, slow clock solution module 44 or correction data estimator 34 typically estimates thousands of states to determine a clock solution. For example, the data processing center 18 may take several seconds to complete the calculation of the slow clock solution. To reduce correction delay and utilize more measurements, the correction data estimator 34 uses two clock solutions, including a slow clock solution and a low delay clock solution. In the slow clock solution module 44, all measurements are batched as long as they arrive before a fixed delay, e.g. 6-15 seconds.

The slow clock solution module 44 measures and clock homodyne (ZD) filter (e.g., ZD kalman filter) uses similar measures as the orbit solution module 38, except for a few main differences. First, the slow clock solution module 44 runs at a different rate (e.g., every 30 seconds or 60 seconds) or provides updates to the slow clock solution. In contrast, because clock corrections change faster than track corrections, the track solution module 38 runs or provides updates to the track data or track solution every 300 seconds. Second, in the slow clock solution module 44, all or most of the state variables remain the same or similar to the corresponding state variables in the orbit solution module 38, except for the satellite orbit-related states. Instead of estimating the state associated with the satellite orbit, the orbit estimation result of the orbit solution from the orbit solution module 38 is used.

In one embodiment, slow clock solution module 44 may output a complete set of global difference correction or correction data 16 that includes one or more of the following corrections: satellite orbit correction, satellite clock correction, satellite WL bias, satellite NL bias and quality information. The slow clock solution module 44 may pass the correction data 16 and the estimated tropospheric parameters to the low delay clock module 42. As used herein, "bias" that is not considered satellite or receiver bias will refer to satellite bias such as satellite WL and NL bias. The bias of the reference receiver 30 is not of interest for the global differential correction product and is therefore not solved in the reference receiver 30, whereas the bias of the mobile receiver 12 may be processed in the mobile receiver 12.

The slow clock solution module 44 uses the non-differential refraction-corrected codes and carrier-phase observations to estimate satellite and receiver clocks, tropospheric bias, satellite narrow-lane bias. In one embodiment, the track update rate of the track data is very low (e.g., 5 minutes update rate), the slow clock update rate (e.g., 30 seconds), or even longer. Because a large amount of ambiguity must be estimated along with the receiver and satellite clock parameters, the computation of the data processing center 18 is time consuming, especially in the case of ambiguity resolution and satellite bias estimation. The data processing center 18 or the correction data estimator 34 may wait longer to ensure that sufficient measurements from the reference data network 32 (e.g., the global network) are collected and processed as they arrive at the StarFire data processing center 18. The longer the data processing center 18 waits, the more measurements the data processing center 18 collects, which also results in a longer delay in clock correction and timely real-time arrival at the end user's mobile receiver 12. The data processing center 18 and correction data estimator 34 maintain data corrections in real time in sufficient time for use by the rover receiver such that the satellite clock, along with the satellite bias, maintains the full-cycle nature of the ambiguity resolution of the rover receiver. Ambiguity fixing can reduce convergence time and improve accuracy of navigation of the mobile receiver 12 or rover station.

As shown in fig. 4, in one configuration, the low-delay clock module 42 also includes a clock delta filter 412. Although the modules are shown as separate blocks in the block diagram of fig. 4, it is to be understood that the correction data estimator 34 shown in fig. 3 and 4 may be implemented by one or more prediction filters, such as kalman filters, and that these blocks represent one possible explanation of the illustrative software that may be used to facilitate or perform the methods and systems described in this disclosure.

Fig. 5 is a diagram illustrating parallel operation of a slow clock process 500 (e.g., a medium delay clock process) and a fast clock process 502 (e.g., a low delay clock process). The two parallel axes in fig. 5 show that the time will be increased to the right to the same time scale. Clock solution module 44 supports or performs slow clock process 500 (e.g., at slow clock update intervals or moderate update intervals), while low-latency clock module 42 supports or performs fast clock process 502 or low-latency clock process (e.g., at fast clock update intervals or low-latency update intervals). In one embodiment, the clock solution module 44 provides slow clock data (504, 506), satellite wide-lane offset correction data, satellite narrow-lane offset correction data, slow clock data (504, 506) such as satellite slow clock correction data, to the low-delay clock module 42 at regular intervals (e.g., at T0, T30, and T60), such as at slow clock update intervals (e.g., about 30 time units, e.g., about 30 seconds, as shown). In one embodiment, the orbit solution module 38 may provide the satellite orbit correction data 50 alone at an orbit correction rate, while in another embodiment, the clock solution module 44 provides the satellite orbit correction data 50 and the satellite slow clock correction data 16. For example, clock solution module 44 provides slow clock data at regular time intervals, such as at times T0, T30, and T60, as shown in FIG. 4.

At the same time, the low-delay clock module 42 utilizes or uses the slow clock data (504, 506) as base data or input data to calculate the amount of clock delta adjustment (508, 510) for the slow clock correction data 16 at low-delay update intervals or fast clock update intervals. The low-delay clock module 42 outputs the correction data 16 or the clock increment adjustment (508, 510) at a low-delay update interval that is updated at a greater rate or at a shorter update time interval than the slow clock update interval. For example, the low-delay clock module 42 generates the low-delay correction data 16 at a low-delay rate (or a fast clock rate), and the low-delay correction data 16 may be an integer multiple of the slow clock update rate. Further, the low-delay clock module 42 or estimator may allocate a valid period of time to the correction data 16 (e.g., low-delay clock data) or to an amount of clock delta adjustment that is commensurate with (e.g., approximately equal to) the slow clock update interval.

In one embodiment, the low delay clock module 42 or clock delta filter 412 uses (e.g., uses only) the carrier phase measurements to calculate the clock variation (clock delta), as shown in fig. 5. Slowly varying parameters, such as tropospheric and satellite narrow lane bias, are fixed to, synchronized with, or provided by the estimates of the slow clock solution module 44. The slow clock module periodically provides updates from the slow clock module to the low delay module (e.g., from a slow clock process to a low delay clock process) at a slow clock update rate, e.g., once every 30 seconds. The low delay clock module 42 uses the delta carrier phase between T0 and the current epoch Ti to estimate the clock delta. When the slow clock module provides a new reference epoch, such as a T30 epoch, the low delay clock module 42 changes the reference epoch from T0 to T30.

FIG. 6 is one embodiment of a method for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks. The method of fig. 6 begins at block 600.

In block 600, the data processor 20, the correction data estimator 34, or the orbit solution module 38 determines predicted orbit data or updates to predicted orbit data for a respective measurement time (e.g., epoch Ti) based on the reference network data 46 (e.g., batch data or raw measurement data for the time or epoch Ti) received from the one or more reference receivers 30 and previously predicted orbit data (e.g., for time T) from the orbit solution module 38 or stored in the data storage device 24 (e.g., register, electronic memory, or nonvolatile random access memory). The measurement time or epoch (e.g., ti) may be the previous epoch or the next epoch after the first epoch (e.g., T0). Further, the trajectory solution module 38 may provide predicted trajectory data (e.g., predicted trajectory data for a measured time or epoch Ti) or updates to the predicted trajectory data based on the widelane ambiguities and corresponding widelane ambiguities bias data provided by the measurement preprocessing module 36.

In one example of performing block 600, the correction data estimator 34 or the orbit solution module 38 estimates predicted orbit data (e.g., O2C data) at an orbit update rate (e.g., once every 300 seconds) within a few minutes based on the fixed time of the orbit solution in the low delay clock module 42.

In block 602, the data processor 20, the correction data estimator 34, or the clock solution module 44 determines clock input data or updates of the clock input at the measurement time or epoch T0, slow clock solution data, wide lane bias, and narrow lane bias data based on the predicted orbit data (e.g., at the measurement time or epoch Ti or epoch T0) and based on the reference network data 46 (e.g., batch data or raw measurement data for the time or epoch T0) received from the one or more reference receivers 30. As used herein, a measurement time or epoch Ti follows a measurement time or epoch T0.

For example, in block 602, the data processor 20, the correction data estimator 34, or the clock solution module 44 determines clock input data, or updates to the clock input updated at a slow clock rate or at slow clock intervals. Therefore, the transition from measurement time T0 to measurement time Ti does not necessarily trigger an update of the clock input data unless Ti coincides with the next update interval of the slow clock process. For example, for block 604, the pre-processed measurements from the measurement pre-processing module 36 are batched and sent to the low delay clock module 42 after a waiting window of a few seconds (e.g., 1-2 seconds). At the same time, the preprocessed measurement values are sent to the track/clock solution module after a longer period of time (e.g., 6-15 seconds).

After block 602, the method continues in block 604. In block 604, the data processor 20, the correction data estimator 34, or the low delay module selects a reference satellite for each location of the reference network or a pair of reference satellites for each reference receiver 30 of the reference network. For example, in one embodiment, the correction data estimator 34 or low delay module selects the satellite with the highest elevation without cycle slip as the reference satellite for each reference position. For tropospheric bias compensation, any difference in elevation between the reference receiver 30 and the mobile receiver 12 should be considered. The tropospheric bias is corrected using an a priori model and residual tropospheric bias estimates from a slow clock solution.

In block 606, the data processor 20, the correction data estimator 34, or the low delay module determines a double difference between the carrier phase measurements or narrow lane carrier phase measurements at the pair of satellites and the measurement times or epochs T1 and T0. For example, the double difference of the carrier phase measurements at each reference receiver 30 is determined at the measurement times or epochs Ti and T0 and at the pair of satellites. The Double Difference (DD) narrow lane ambiguity is resolved to determine an accurate refraction-corrected carrier-phase measurement for which some deviation is cancelled. For example, in a double differential technique, one or more of the following deviations may be counteracted: receiver code phase bias (e.g., receiver code phase bias and satellite code phase bias), carrier phase bias (e.g., receiver phase bias and satellite phase bias), and clock bias (e.g., receiver clock bias and satellite clock bias), which are common between the satellite and the receiver and can be offset by a double differential operation between the satellite and the receiver. Some ionospheric propagation delay bias is cancelled out in the double difference equation. At different times, after the double differential of the double differential between the same reference receivers 30, the remaining atmospheric errors including ionosphere and troposphere delays may be ignored. However, ionospheric errors between different reference receivers 30 separated by long baselines may be estimated and used by the correction data estimator 34.

In one embodiment, low-delay clock module 42 reduces correction delay to improve clock accuracy with the absolute clock from the slow clock solution. To increase computational efficiency, dual differential measurements between time and satellite are used so that some unnecessary states such as ambiguity and receiver clock are eliminated. The low delay module or clock delta filter 412 only evaluates the state changes of the satellite clock to provide the mobile receiver 12 with processing efficiency and enhanced fast availability/reduced delay of the correction data 16.

In one example, the clock solution module 44 determines predicted orbit data, satellite bias data, and satellite bias quality data (e.g., variance-covariance data) based on the resolved, double-differential refraction-corrected narrow-lane ambiguities.

In block 608, the data processor 20, the correction data estimator 34, or the low delay clock module 42 receives the predicted orbit data, satellite bias data, and satellite bias quality data (e.g., variance-covariance data) based on the resolved, double-differential and refraction-corrected narrow-lane ambiguities and provides an update of the clock delta filter 412. Before the next update of the clock solution module 44 at the slow update rate in block 602, the low-delay clock module 42 only evaluates the satellite clock delta so that the computation can be updated with a low-delay rate greater than the orbit update rate of the orbit solution module 38 and the slow update rate of the clock solution module 44.

In one example of the method of fig. 6, after each iteration of blocks 604, 606, and 608 of each location or reference receiver 30, the method continues to block 604 until all calculations of blocks 604, 606, and 608 have been performed for all locations or reference receivers 30 in the reference data network 32. Further, each iteration of blocks 604, 606, and 608 is consistent with providing low-delay correction data 16 at a low-delay interval or low-delay data rate.

In block 610, the data processor 20, the correction manager 40, or the low delay clock applies a RAIM (receiver autonomous integrity monitoring) algorithm to the clock delta filter 412. The RAIM algorithm includes software that uses overdetermined solutions or redundant calculations to check the consistency of satellite measurements, such as carrier phase measurements and code phase measurements, of one or more satellites of each reference receiver 30 in the network. The RAIM algorithm requires at least five satellites in reception range to detect significant carrier phase error measurements or significant errors in the clock correction of any satellite in the cluster. The correction manager 40 or the data processor 20 may delete, pause or flag (as suspicious or unreliable) the low-latency clock correction data 16 for one or more satellites so that the mobile receiver 12 or rover station may ignore the low-latency clock correction data 16 that has been marked as suspicious or unreliable or provide a lower weight to the low-latency clock correction data 16 that has been marked as suspicious or unreliable.

In one example of performing block 610, the received satellite signal, low delay clock module 42, or clock delta filter 412 uses a priori satellite clock rates from broadcast ephemeris to estimate satellite clock delta as an error checking mechanism, such as a mechanism supporting the RAIM algorithm. Within the low-delay clock module 42, an additional prediction filter (e.g., a Kalman filter or a least squares estimator) may be used to estimate the clock increment of the RAIM algorithm. Further, estimated satellite clock deltas derived from broadcast ephemeris may be compared to estimated satellite clock deltas associated with a prediction filter or a least squares estimator. The number of estimated state variables or unknowns is equal to the number of active satellites. The RAIM algorithm is used to ensure that any measurements with cycle slip are detected and removed.

In block 612, the data processor 20, the correction data estimator 34, or the low delay clock module 42 accumulates the clock delta data and calculates the clock data corresponding to the measured time or epoch Ti for incorporation into the correction data 16 or the low delay correction data 16. For example, the low-delay correction data 16 includes accurate orbit correction data 50, accurate low-delay clock data, accurate low-delay clock quality data, and wide-lane satellite bias data and narrow-lane satellite bias data on a satellite-by-satellite basis, which may be applied to particular satellites in view or reliable reception range of the mobile receiver 12. In one configuration, the correction data 16 may be globally valid in the GNSS system for each respective measurement time or epoch and for each satellite to which it applies.

Fig. 7 discloses a flowchart of another embodiment of a method for providing satellite correction signals with accurate, low-latency Global Navigation Satellite System (GNSS) satellite clocks. The method of fig. 7 starts at step S800.

In step S800, the plurality of reference receivers 30 or the measurement modules of the plurality of reference receivers 30 are located at known respective locations (e.g., geographically distributed locations, such as locations worldwide, to receive satellite signals from one or more GNSS systems), and the reference receivers 30 measure raw phase measurements, code phase measurements, or both for the respective locations. The measurement module takes raw phase measurements at measurement times called epochs. For example, the reference receiver 30 may collect raw phase measurements and code phase measurements (e.g., pseudorange measurements) at one or more measurement times or epochs that are indicative of Global Navigation Satellite System (GNSS) system time. The code phase measurements are measurements of codes (e.g., pseudorandom noise codes) encoded on one or more received satellite signals or carriers of the received satellite signals. The reference receiver 30 sends or transmits the collected raw phase measurements to the data processing center 18 to estimate correction data 16, such as correction data 16 for Precision Point Positioning (PPP).

In step S802, the data processing center 18 collects raw phase measurements, code phase measurements, or both, and corresponding reference receiver 30 identifiers or location identifiers from a plurality of reference receivers 30. The data processing center 18 may use the raw phase measurements and the code phase measurements for estimating the correction data 16. In addition to raw phase measurements, code phase measurements, the reference receiver 30 may provide pre-processing data or other reference network data 46 including any of the following: widelane ambiguities, narrow elane ambiguities, estimated position errors from reference stations of known positions derived from received satellite signals, tropospheric bias, satellite clock bias, bias of satellite transmitters 10, ephemeris data and navigation data.

In step S804, the measurement preprocessing module 36 or the correction data estimator 34 determines the widelane ambiguities and satellite widelane deviations of the collected phase measurements for each satellite. For example, the measurement preprocessing module 36 or the correction data estimator 34 determines a fixed full-cycle widelane ambiguity and satellite widelane bias for each satellite's collected phase measurements to assist (e.g., provide constraints for rapid or efficient convergence) in estimating narrow-lane ambiguities within one or more modules of the correction data estimator 34.

In one example of performing step S804, the measurement preprocessing module 36 includes a prediction filter (e.g., a wide-lane estimator 37 (e.g., a wide-lane filter) or a kalman filter) for estimating the wide-lane ambiguity or position thereof of the received satellite signals for each reference receiver 30 based on the collected raw phase measurements and code phase measurements alone or in combination with assistance data as constraints. Furthermore, the determined widelane ambiguity facilitates efficient and fast convergence of one or more narrow lane estimators 39 (e.g., narrow lane filters) (e.g., kalman filters) for determining the track solution and the slow clock solution, which are described in other steps of the method of fig. 7.

In step S806, based on the collected raw phase measurements and code measurements, the orbital solution module 38, the narrow-lane ambiguity resolution estimator, or the correction data estimator 34 determines (or applies a previously determined) an orbital narrow-lane ambiguity and an orbital satellite narrow-lane bias that are consistent with the determined wide-lane ambiguity and corresponding satellite wide-lane bias for each satellite in the orbital solution. In one example, a fixed narrow-lane ambiguity may not need to be updated without cycle slip and significant tropospheric delay changes over a period of time, but the set of carrier-phase measurements and code-phase measurements are updated at a track correction rate (e.g., approximately once every 300 seconds) for the track solution, which may be different from the slow update rate (e.g., approximately once every 30 seconds to 60 seconds) of the set of carrier-phase measurements and code-phase measurements for the slow clock solution. In the presence of carrier phase cycle slip or any carrier signal loss lock or low signal quality of the received signal of any satellite, the track narrow lane ambiguity may be updated with a track correction rate by means of the input state or clock narrow lane ambiguity, which may be different from the slow update rate of the slow clock solution.

In step S808, the orbit solution module 38 or the correction data estimator 34 determines satellite orbit corrections (e.g., at an orbit correction rate) that are consistent with the determined orbital narrow lane ambiguities and corresponding orbital satellite narrow lane deviations based on the raw phase measurements and code measurements collected.

In step S810, based on the collected raw phase measurements and code measurements, the clock solution module 44, the narrow lane ambiguity resolution estimator or the correction data estimator 34 determines (or applies a previously determined) a clock narrow lane ambiguity and a corresponding clock satellite narrow lane bias that is consistent with the determined wide lane ambiguity and corresponding satellite wide lane bias for each satellite in the slow clock solution. In one example, a fixed narrow lane ambiguity may not need to be updated without cycle slip and significant tropospheric delay changes over a period of time, but the set of carrier phase measurements and code phase measurements are updated at a slow update rate (e.g., about once every 30 seconds to 60 seconds) for a slow clock solution. In the event of a cycle slip of the carrier phase or any loss of lock of the carrier signal or low signal quality of the received signal of any satellite, the clock narrow lane ambiguity may be updated by virtue of the input state with a slow update rate, which may be different from the orbit correction rate of the orbit solution.

In step S812, based on the collected raw phase measurements and code measurements (e.g., the previously collected raw data and the code measurements updated at the slow update rate) and derived data from the raw phase and code measurements, the clock solution module 44 or the correction data estimator 34 determines a slow satellite clock correction (e.g., and tropospheric delay bias and gradient) for each reference receiver 30, wherein the derived data includes one or more of: the determined satellite orbit correction data 50, the determined wide-lane integer ambiguity and corresponding satellite wide-lane bias, and the determined clock narrow-lane integer ambiguity and corresponding satellite narrow-lane bias data.

In step S814, the low-delay clock module 42 or the correction data estimator 34 determines satellite clock correction data 16 or clock delta adjustments for the slow satellite clock having a lower delay based on the collected recent or most recently updated measurements of the raw phase measurements (e.g., updated at a fast update rate or a low-delay update rate that is greater than the slow update rate) to provide clock correction data 16 having a lower delay, the collected recent or most recently updated measurements of the raw phase measurements being closer to the current value than a plurality of previous measurements of the collected raw phase measurements for the slow satellite clock correction. In some embodiments, the low-delay clock module 42 or the clock solution module 44 may apply a clock delta adjustment to the slow clock solution.

Step S814 may be performed according to various techniques, which may be applied individually or cumulatively.

Under the first technique, the clock solution module 44, the low-delay clock module 42, or the correction data estimator 34 estimates the relative clock error at a first rate (e.g., a relatively high rate (e.g., about 1Hz or greater)) for lower-delay correction data (e.g., selecting the satellite with the highest elevation without cycle slip as the reference satellite for each position) using a double difference technique (e.g., between time Ti and T0 and between the reference satellites for each reference position of the local reference station).

Under a second technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 (e.g., the moderate delay correction data 16) includes a slow clock solution; the clock solution module 44 or correction data estimator 34 integrates the fast clock solution and the slow clock solution over an integration period (e.g., about 30 seconds) to provide an absolute satellite clock estimate (e.g., GPS reference system time).

Under a third technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or the correction data estimator 34 provides a longer delay for the slow clock solution, including a longer GNSS raw data collection time (e.g., about 6 seconds to about 10 seconds) for more GNSS raw data and a data processing time of a few seconds or seconds for the complex slow clock solution.

Under a fourth technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or the correction data estimator 34 provides low delays for the satellite clock correction data 16 (e.g., fast clock solution) with low delays, including short GNSS raw data collection times (e.g., about 1 second to about 2 seconds) and data processing times (e.g., a few milliseconds) for very efficient estimation of the satellite clock correction data 16 (e.g., fast clock solution) with low delays.

Under a fifth technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or the correction data estimator 34 uses tropospheric estimates from a slow one of the fast clock solutions. (e.g., estimating tropospheric bias based on an a priori model and residual tropospheric bias estimates from a slow clock solution).

Under a sixth technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or correction data estimator 34 uses orbit correction data 50 (e.g., common orbit correction data 50) from the orbit solution in the slow clock solution and the low delay satellite clock correction data 16 (e.g., fast clock solution).

Under a seventh technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or correction data estimator 34 updates the slow clock solution using clock increments from the satellite clock correction data 16 (e.g., the fast clock solution) with lower delays. For example, correction data estimator 34 updates the slow clock solution at an update interval or a slow update rate, such as about 30 seconds to about 60 seconds (e.g., to predict satellite clock dynamics in the slow clock solution or accumulated clock increments or changes from the fast clock solution.) with any clock incremental adjustments to the slow clock solution, or both are updated by the orbit solution or orbit correction data 50 at an orbit correction rate, such as about once every 300 seconds, to predict satellite clock dynamics associated with the orbit solution.

Under the eighth technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or the correction data estimator 34 uses a lower delay correction that includes a change in satellite clock correction based primarily on raw carrier phase measurements (e.g., GNSS carrier phase data only).

Under the ninth technique, the lower delay correction data includes a fast clock solution and the higher delay correction data 16 includes a slow clock solution; the clock solution module 44 or correction data estimator 34 estimates the moderate delay data to include absolute satellite clocks, tropospheric bias, satellite narrow lane bias, and narrow lane ambiguity solutions.

Under the tenth technique, the lower delay correction data includes a fast clock solution and the moderate delay correction data 16 includes a slow clock solution; the clock solution module 44 or correction data estimator 34 estimates correction data 16, the correction data 16 including satellite orbit correction data 50, satellite clock correction data 16 with lower delay, satellite wide lane offset data, satellite narrow lane offset data, and satellite quality indicators with lower delay clock and narrow lane offset. Correction data 16 may represent integrated and absolute satellite clock data available within a satellite correction signal having a lower delay.

In step S816, the correction manager 40 or correction data estimator 34 combines the satellite orbit correction data 50 and the low delay clock correction data 16 into the correction data 16, which correction data 16 is encoded on the global satellite differential correction signal with global availability for GNSS transmission to one or more mobile receivers 12. For example, the correction data estimator 34 or data processor incorporates satellite wide-lane offset, satellite orbit correction data, satellite narrow-lane offset from slow clock solution, and low-delay clock correction data into correction data encoded on global satellite differential correction signals that are of global significance for GNSS transmission to one or more mobile receivers.

In step S818, the correction manager 40, the wireless communication system 57, or the correction data estimator 34 sends the satellite correction data 16 (e.g., satellite orbit, clock, satellite wide and narrow lane offset and quality signals for each satellite) with lower delay to one or more mobile receivers 12 via a correction data 16 message (e.g., via satellite L-band signals or cellular network feeding/correcting data 16 with the internet 56).

As used herein, the delay is based on the time difference between the earlier epoch associated with the collection of measurements for processing (and observation at the reference station) and the later epoch when the processed measurements are applied in the mobile rover receiver.

FIG. 8A provides an illustrative chart of typical time intervals or delays versus GNSS time for providing a slow clock solution. As shown in fig. 8A, the vertical axis 700 provides an indication of slow clock delay time or slow clock update interval. At the same time, the horizontal axis 702 provides GNSS time, such as GPS time for the respective satellite. In one embodiment, an illustrative chart is provided for a cluster of one or more reference stations (e.g., about 60 reference receivers 30) and at least five satellites in view or reliable range. In the illustrative example of fig. 8A, the slow clock solution takes approximately 2 seconds to 3 seconds to complete the one epoch measurement process. After the measurement process for each epoch is completed, the satellite clock changes (e.g., clock delta data) are integrated with the absolute clock from the slow clock solution to get an accurate absolute clock for any epoch. Finally, these orbital, clock correction and satellite Wide Lane (WL), narrow Lane (NL) offset products with quality information are transmitted to the user in real time over a satellite channel (e.g., an L-band channel) or wireless communication network (e.g., over the internet 56).

FIG. 8B provides an illustrative graph of time lag or delay that provides a low-delay clock solution versus GNSS time. As shown in fig. 8B, the vertical axis 704 provides an indication of a low delay clock delay time or a fast clock update interval or a low delay update interval. At the same time, the horizontal axis 706 provides GNSS time, such as GPS time for the respective satellite. In one embodiment, an illustrative chart is provided for a cluster of one or more reference stations and at least five satellites in view or reliable range. If the data processing center 18 includes a desktop computer or server 54, the data processing center 18 may take several milliseconds to complete processing all measurements of the low-latency solution in FIG. 8B, which is 300 times faster than the slow-clock solution of FIG. 8A. The low delay calculation is very efficient.

The illustrative example of FIG. 8B shows a typical correction period from the real-time receiver for correction data 16 of approximately 4 seconds, which includes the time that the network data arrives at data processing center 18 (e.g., computer or server 54), the network processing time of data processing center 18; and a correction transmission time for transmitting the correction data 16 or correction data 16 message from the data processing center 18 to the end user's mobile receiver 12. Whether or not any reverse language exists, the results or latencies of the correction data 16 shown in fig. 8A and 8B are shown for illustrative purposes only and even when the methods, systems, or information set forth in the present disclosure are employed, real world latencies may be different than those latencies shown based on many possible factors.

According to one embodiment shown in fig. 9, a system 911 for providing satellite correction signals includes a satellite receiver for receiving a series of raw satellite signal measurements. The system 911 of fig. 9 is similar to the system 11 of fig. 1A, wherein like reference numerals indicate like elements.

In FIG. 9, the data processing center 118 is similar to the data processing center 18 of FIG. 1A, except that the data processing center 118 of FIG. 9 includes a data collector 19, a data source selector 15, and offline data 23 or stored data. The data collector 19, data source selector 15, and data storage 24 may be coupled to a data bus 22 for communication with each other, the electronic data processor 20, and one or more data ports 26.

In other embodiments, for example, a virtual data communication path, a physical data communication path, or both, are possible between the data collector 19, the data source selector 15, the data port 26, and the data storage device 24. For example, a virtual data communication path may represent a logical communication path via software, a link, or a call. The physical data communication paths may include transmission lines, cables, data buses, striplines, microstrip lines, circuit board traces, or other physical communication paths for the transfer of signals or data between modules or components of the data processing center 118.

In one embodiment, the data collector 19 may communicate with a reference data network 32 or one or more reference receivers 30 via the data port 26. In turn, the data collector 19 may store or facilitate the storage of the recorded offline data, historical measurement data, and associated ephemeris data in the data store 24 for later reference during a startup mode or a warm startup mode, thereby reducing startup time for convergence. The recorded offline data 23 may include one or more of the following: historical measurement data from one or more reference receivers 30, raw measurement data from one or more reference receivers 30, and associated ephemeris data from the same measurement epoch or sampling time interval as the measurement data.

In one embodiment, the data storage device 24 stores offline data 23 recorded or raw satellite signal measurements received over a series of time windows prior to a current time (e.g., a current GNSS measurement time). The time window may be selected according to various techniques that may be used cumulatively or alternately. Under a first technique, the series of time windows is based on available data and a target time for convergence of the warm boot. Under a second technique, the series of time windows is cumulative over a duration of at least 24 hours. Under a third technique, the series of time windows have a cumulative or complete range from about 24 hours to about 48 hours in duration.

Each of the satellite signal measurements (e.g., carrier phase signal measurements associated with each received satellite signal at the respective reference receiver 30) is associated with a respective stored measurement time tag. During the start-up mode or warm-up mode, the estimator 34 is adapted to estimate satellite correction data or satellite correction signals based on satellite orbit data, satellite clock data and satellite bias correction data derived from stored raw satellite signal measurements or recorded offline data 23. In some configurations, the correction data or satellite correction signals are valid on a global basis for accurate point positioning, while in other configurations, the satellite correction signals may be valid for a defined geographic area less than worldwide.

In one embodiment, the data source selector 15 may communicate with one or more reference receivers 30 via a data port 26. Further, the data source selector 15 may communicate with the data storage 24 via the data bus 22 or via another virtual data path or physical data path. The data source selector 15 may seamlessly switch or change the measurement data source from the stored received raw satellite signal measurements (e.g., recorded offline data 23) to live, real-time raw satellite signal measurements if or when the corresponding measurement time tag of the last processed one of the stored received satellite signal measurements approaches or reaches the current time, wherein the difference between the corresponding measurement time tag of the corresponding last processed, stored received raw satellite signal measurement and the current time is less than the threshold time range.

As shown in fig. 9, the correction data estimator or estimator 34 is stored in the data storage device 24. For example, estimator 34 may include software instructions for interpretation or processing by electronic data processor 20 to estimate correction data such as satellite orbit data, satellite clock data, satellite bias data, or other bias data. In one configuration, the estimator 34 is adapted to determine when satellite orbit data, satellite clock data, and satellite bias data (e.g., wide-narrow lane bias data) have converged to reliable satellite correction data with ambiguity resolution.

Once the correction data is estimated or determined by the correction data estimator 34, the satellite uplink station 28 or satellite uplink transmitter may transmit reliable correction data with ambiguity resolution to the end user over a satellite communication channel via the communication satellite 35 and the wireless correction device 14 (e.g., satellite downlink receiver).

In an alternative embodiment, once the correction data is estimated or determined by the data estimator 34, a wireless communication system, such as a cellular system, a Code Division Multiple Access (CDMA) system, or a time division multiple access system (TDMA), may transmit reliable correction data with ambiguity resolution to the end user over a communication channel via the wireless communication system.

In another alternative embodiment, once the correction data is estimated or determined by the data estimator 34, the server may distribute or make available the correction data to authorized subscribers over the Internet via a virtual private network, an encrypted communication channel, or another communication channel.

In one embodiment, the data processing center 118 updates or refreshes the correction data at a rate of at least 1 hertz or greater. In one embodiment, as shown in FIG. 10, data processing system 218 includes a clock solution module 44 for providing slow clock estimation and a low delay clock module 42 for providing low delay clock estimation, wherein the data processing system integrates two parallel clock estimation processes including slow clock estimation and low delay clock estimation.

The data processing system (118 or 218) may operate in a startup mode or a warm startup mode or a normal mode. During the startup mode, the data storage device 24 or the data processor 20 feeds approximately six hours of stored data measurements into the slow clock module 44 (e.g., slow clock estimation module) to achieve steady state clock estimation. However, during the start-up mode, historically stored data measurements are not required to feed the low delay clock module 42. During the normal mode of operation, the estimator 34 is adapted to estimate the differential correction signal using satellite orbit data, satellite clock bias data and satellite bias data based on raw measurements received live in real time.

First, the data collector 19 may facilitate the storage of historical measurement data and ephemeris data for about 24 hours to about 48 hours. Second, in one example, the estimator 34 or data processor 20 inputs a first portion (e.g., about 24 hours to about 48 hours) of the stored historical measurement data and associated stored ephemeris data (alone or together, the recorded offline data 23) to the MPP module 36 and the orbit solution module 38. Third, the estimator 34 or the data processor 20 inputs a second portion (e.g., about the last two hours) of the stored historical measurement data and associated stored ephemeris data (alone or together, the recorded offline data 23) into the low delay clock module 42, wherein the second portion is less data than the first portion, or wherein the second portion is collected over a shorter period of time than the longer period of time of the first portion. Fourth, the estimator 34 or the data processor 20 only inputs the real-time measurement data and the associated ephemeris data to the low delay clock module 42. Fifth, the estimator 34 seamlessly switches the input source of the measurements from the recorded offline data 23 to real-time data collected from one or more reference receivers 30 or the global GNSS reference data network 32 when the last GNSS measurement time tag fed from the recorded offline measurement data or the stored measurement data reaches or approaches the real-time GNSS measurement time tag.

In one embodiment, an estimator 34, such as an innovative clock orbit real-time estimator 34 (iCORE), estimates Global Navigation Satellite System (GNSS) orbit data and clock data, wherein the estimator 34 typically takes up to 24 to 48 hours of measurement to reach a pull-in or stable and reliable state of the position estimate of the reference receiver 30 and associated correction data (e.g., differential correction data or accurate point positioning data). In one configuration, when the software of the estimator 34 is started, the last 24 to 48 hours of offline measurements are stored in an hourly file in the data collector 19 (e.g., the server of the data collector 19) and are first fed into the estimator 34 for data processing.

When the last GNSS measurement time tag read from the offline data reaches or approaches the real-time GNSS measurement time tag with minimal lag, the data selector switches the measurement input source from the offline data or historical measurement data in the data storage device 24 to real-time data collected from the global GNSS reference station network receiver. After the estimator 34 or its quality module determines convergence of the orbit and clock solutions, the estimator 34 or data processing center (118 or 218) may provide or transmit a consistent set of correction signals (e.g., starFire) in real-time in a timely manner via wireless communication ^TM Correction signals), including satellite orbit, clock, wide-narrow satellite bias, and quality information, such as L-band satellite channels, cellular channels, code Division Multiple Access (CDMA) communication channels, time Division Multiple Access (TDMA) channels, the internet. The warm start may significantly reduce the correction signal start-up time from about 24-48 hours to about 1.5-3 hours to support improved operating efficiency and productivity.

StarFire is a real-time Global navigation satellite augmentation System (GNSS) that achieves centimeter-level accuracy positioning by using real-time global differential corrections. Such corrections are available worldwide through Internet Protocol (IP) or L-band geostationary communication satellites. Global differential correction eliminates the need for a local reference station and radio communications as compared to local reference station correction.

The hot start correction system with correction estimator 34 is well suited to minimize or reduce downtime from software updates and maintenance. The performance of the correction system may be based on an evaluation of the start-up time. The start-up time may be defined as the period of time required from the start of the calculation of the correction estimate and the server software to the time when the correction signal is ready to be broadcast or wirelessly transmitted in real time to the satellite mobile receiver 12 (of an end user or subscriber) equipped with the wireless correction device 14 (e.g., satellite receiver or cellular transceiver). The start-up time is mainly affected by the long GNSS orbit and clock convergence time, especially by the GNSS orbit estimation.

There is often a conflict between the fast start-up time and the reliable steady state performance of the correction estimation. A stable and reliable GNSS orbit and clock estimation requires more GNSS measurements than is required for achieving a fast start time. For example, in practice, an accumulated data window of historical measurement data of at least 24 hours is required in order to obtain reliable steady state performance. However, the more GNSS measurements that are used, the longer it will take for the estimator 34 to process the measurements to obtain reliable clock data and orbit data estimates.

According to fig. 10, in one configuration, the architecture of the correction system 1011 is constructed as follows: the GNSS reference receivers 30 are distributed worldwide with good satellite geometry and visibility. The reference receiver 20 is associated with a satellite uplink station 28 or other communication system for transmitting measurements and ephemeris data (individually and collectively, offline data 23) to the data collector 19 or data processing center 118 in real time. The data collector 19 may store the measurement data and ephemeris data (individually and collectively, offline data 23) in a data store 24.

In one embodiment, the data storage device 24 may include a magnetic disk, an optical disk, an electronic memory, a non-volatile random access memory, or any other data storage device 24. The collected data is stored in the data storage 24 for access during a warm start mode for offline processing or reprocessing. In one embodiment, the data collector 19, alone or in conjunction with the data source selector 15, may record data in the data storage device 24 as well as send real-time data to the estimator 34 of the data processing center 118 in real-time.

In one embodiment, when the GNSS measurements reach the data processing center (118 or 218), the one or more data processing centers (118 or 218) collect, store and pre-process the GNSS measurements from the reference receiver 30. The longer the data processor 20 or estimator 34 waits before processing the measurements, the more measurements are collected. By collecting more measurements, the data processor 20 or estimator 34 may improve the accuracy of orbit, clock, and satellite bias and the reliability of network ambiguity resolution. If the data processing center (118 or 218) is distributed around the world with the reference receiver 30, the total processing time for achieving a converged solution may be long; particularly for multi-GNSS position solution estimation with ambiguity resolution. Multiple GNSS location solution estimates are required for applications such as global positioning system GPS (united states); GALILEO (europe), quazi-Zenith satellite system QZSS (japan); beidou satellite navigation system BDS (China); the global navigation satellite system GLONASS (russia) and other different GNSS satellite systems.

Because the correction data is provided via satellite communications, such as an L-band satellite channel with an associated bandwidth limitation, there may be a delay in providing the correction data to the end user's mobile receiver 12. The estimator 34 is configured to reduce the processing time of the estimator 34 to minimize delays in the correction data provided to the mobile receiver 12 to allow the processor 20 to operate at 1Hz with at least 24 hours of data for startup. The estimator 34 minimizes the processing time for generating a correction signal that enables ambiguity resolution at the user receiver. The estimator 34 uses an innovative ambiguity resolution algorithm to generate the clock and satellite biases. Furthermore, estimator 34 facilitates processing data from up to one hundred reference receivers 30 from one or more of the following data sources: (a) Live or real-time GNSS measurements (e.g., live measurement data streams) from the active GNSS reference receiver 30; (b) Stored historical or offline recorded GNSS measurements (e.g., recorded survey data streams) from the GNSS reference receiver 30 (e.g., at the last or previous 24 hours to 48 hours); (c) Any mix or combination of real-time GNSS measurements and recorded GNSS measurements.

According to one example, as shown in step S914 of fig. 14, the data source selector 15 is adapted to use a mix or combination of measurement data sources comprising stored raw satellite signal measurements and raw satellite signal measurements in live real time, if or when the respective measurement time tag of the last processed satellite signal measurement of the stored received satellite signal measurements has not been close to or reached the current time (e.g. the current GNSS measurement time), wherein the difference between the respective measurement time tag of the corresponding last processed stored raw satellite signal measurement and the current time is larger than the threshold time range. In one embodiment, estimator 34 may provide all necessary calculations such as measurement preprocessing, orbit and clock determination, ambiguity resolution, and final correction generation in fractions of a second.

In one embodiment, estimator 34 includes a clock solution module 44 and a low-delay clock module 42 such that estimator 34 provides an integrated clock solution of two parallel clock estimation processes (e.g., the slow clock of clock solution module 44 and the delay clock of low-delay clock module 42). The clock solution module 44 and the low delay clock module 42 collectively facilitate the accuracy of absolute clock estimation and the reduction of clock delay in position estimation using correction data. In one embodiment, the clock solution module 44 determines a slow clock solution that uses all possible GNSS measurements (from the reference receiver 30) such that the GNSS measurements include longer delay measurements (e.g., up to 6 seconds) from the reference receiver 30 or associated network to estimate absolute satellite clocks, tropospheric bias, satellite narrow lane bias and ambiguity and to do ambiguity resolution. In some examples, the clock solution module 44 requires a large amount of computation to determine the slow clock estimate, which may take several seconds to process.

Instead, the low-delay clock module 42 uses only the last available measurement with a shorter delay (e.g., receiver measurement with a delay limited to about 1 second to about 2 seconds) to calculate satellite clock changes at a high rate. The low-delay clock module 42, the data processor 20, or the estimator 34 integrates the low-delay clock estimate with the slow clock estimate and the low-rate orbit to communicate a consistent set of correction signals, including satellite orbit, clock, wide-narrow satellite bias, and quality information, over an L-band wireless channel, wireless communication system, electronic communication network, or the internet in real-time in a timely manner.

In one embodiment, the data processing center (118 or 218) includes hub navigation algorithm software, such as clock track real-time estimator 34, which may be referred to as an innovative clock track real-time estimator (iCORE). The estimator 34 includes a measurement preprocessing (MPP) module 36, a track solution module 38, a clock solution module 44, and a low delay clock module 42, which are key components of global differential correction. The data processing center (118 or 218) or estimator 34 generates correction signals to provide centimeter level accuracy estimates, including: (1) satellite orbit estimates; (2) satellite clock estimates; and (3) satellite phase bias and associated quality information.

When the estimation software of the estimator 34 is started, the historically collected measurement data (if available) and associated ephemeris data, e.g., collected data or measurement data of about 24 hours to about 48 hours, are fed into the MPP module 36 and the orbit solution module 38 for enhancing or accelerating the convergence of the GNSS orbit estimation. In addition, the last six hours of historical collected data is also fed to estimator 34 or clock solution module 44 which estimates the slow clock estimate. Typically, the estimator 34 requires about 24 to 48 hours of historical measurement data to achieve steady state, reliable orbit estimation, and about 6 hours of historical measurement data for slow clock estimation. Because the low-delay clock does not require time to pull in steady state, the low-delay clock module 42 does not need historical measurement data or collected data to estimate the low-delay clock solution.

In one embodiment, the estimator 34 seamlessly switches the measurement input sources of all modules from the offline data 23 or stored data in the data storage 24 to real-time data collected from the global GNSS reference station network receiver when the last GNSS measurement time fed from the offline data reaches the current or real-time GNSS measurement time. The correction system uses a communication system or broadcast system to communicate global differential corrections from one or more data processing centers (118 or 218) distributed throughout the world to the mobile satellite receiver 12. For example, the set of global differential correction data may be selected and transmitted to a communication satellite 35 (e.g., an inarsat communication satellite) via a satellite uplink transmitter or a land-to-ground station (LES). In one embodiment, the user satellite mobile receiver 12 is capable of receiving the set of global differential corrections, resolving ambiguities and achieving centimeter-level accuracy of position estimation for navigation.

Fig. 10 shows a data correction system 1011 associated with the data processing center 218 and the estimator 134. The estimator 134 includes the following modules: a real-time data collector 19, a measurement preprocessor module 36, a track solution module 38, a clock solution module 44, and a low delay clock module 42. In some configurations, the real-time data collector 19 also includes a data source selector 15 (in fig. 9), but in other configurations, the data source selector 15 may be a separate module.

As shown, the data collector 19 (e.g., a real-time data collector) communicates with one or more reference receivers 30 and with data storage 24 in a data processing center 218. The data collector 19 or the data storage device 24 may provide measurement data, ephemeris data, or both (e.g., offline data 23) to the MPP module 36. For example, the data collector 19 may provide measurement data, ephemeris data, or both to the MPP module 36 via the data source selector 15. In turn, the MPP module 36 may provide the measurement data and the pre-processing data derived from the measurement data to one or more of: a track solution module 38, a clock solution module 44, and a low delay clock module 42. For example, the MPP module 36 may communicate with the track solution module 38, the clock solution module 44, and the low-delay clock module 42 via the data bus 22, a virtual communication path, a physical communication path, or some combination of the preceding.

In one embodiment, the real-time data collector 19 may collect measurement data from the satellite reference receiver 30 in real-time and may facilitate storing the collected measurement data in the data storage device 24. Further, in some embodiments, the real-time data collector 19 may include a data selector module 15, which data selector module 15 distributes the collected data or measurement data to other modules in the data processing system 1011. In other embodiments, the data collector 19 may distribute the collected data or measurement data to other modules or data processing servers.

In one configuration, a measurement preprocessing (MPP) 36 module prepares "cleaned" measurements and provides a wide lane fixed ambiguity and wide lane offset product based on one or more data sources, such as real-time data measurements, offline or stored data measurements (e.g., last 24-48 hours of stored data measurements), or mixed offline data (e.g., last 24-28 hours of stored data measurements), or stored data and real-time data measurements (e.g., real-time streaming data measurements).

As shown, the orbit solution module 38 is adapted to provide accurate satellite position and satellite velocity estimates based on one or more data sources, such as real-time data measurements, offline or stored data measurements (e.g., last 24-48 hours of stored data measurements) or hybrid offline data (e.g., last 24-28 hours of stored data measurements) or stored data and real-time data measurements (e.g., real-time streaming data measurements).

In one embodiment, the clock solution module 44 is adapted to provide satellite slow clock solution estimates and lane offset products at a low rate based on one or more data sources, such as real-time data measurements, offline or stored data measurements (e.g., last 6 hours of stored data measurements) or hybrid offline data (e.g., last 6 hours of stored data measurements), or stored data and real-time data measurements (e.g., real-time streaming data measurements).

The low delay clock module 42 or estimator 134 is adapted to provide a fast satellite clock estimate at 1Hz and integrate the output MPP solution, orbit solution, and slow clock solution to generate a consistent set of correction data signals in real time. In one configuration, the data source for the low delay clock module 42 is accepted only from the real-time measurement stream.

In one embodiment, the low delay clock module 42 uses only carrier phase measurements to derive clock variations between two different epochs. To increase computational efficiency, a double difference approach between time and satellite is used to reduce the magnitude of estimated states such as ambiguity and receiver clock. The model (e.g., a priori model) and the residual tropospheric bias estimates from the slow clock solution are used to correct the tropospheric bias. The low delay clock module 42 evaluates the satellite clock variations rather than the absolute value of the clock data. Thus, low latency clock computation is very efficient. For example, the low-delay clock module 42 may determine or estimate satellite clock variations or low-delay data for all relevant measurements of the reference receiver 30 for any epoch in a few milliseconds or less.

In one embodiment, each of the MPP module 36, the track solution module 38, and the clock solution module 44 includes two portions or components: (1) A homodyne (ZD) kalman filter or data processor 20; (2) A network ARE (e.g., ambiguity Resolution Estimator (ARE)) module. Each homodyne kalman filter processes the received measurement data to form ZD measurements, thereby defining a state variable and (e.g. dynamically) updating the kalman filter measurements. The network ARE module processes the measurement data to perform ambiguity resolution. In one embodiment, the measurement preprocessing module accepts measurement data from the reference data network 32 as input data and applies the wide lane estimator 34 to output the wide lane ambiguity and corresponding wide lane bias.

Measured value preprocessing module

In one embodiment, the measurement preprocessing module 36 preprocesses and purifies raw measurement data from one or more satellite reference receivers 30 or reference station networks 32 based on real-time or live measurement data, offline recorded or stored measurement data, or both at regular intervals (e.g., once per second). Further, the MPP module 36 may resolve Wide Lane (WL) ambiguity. In one configuration, the MPP module 36 provides filtered or "cleaned" measurements, fixed widelane ambiguity, and satellite WL bias to the low delay clock module 42, clock solution module 44, and orbit solution module 38.

In some configurations, the MPP module 36 uses Melbourne-Linear combination->The following state variables are estimated as homodyne (ZD) measurements:

ZD floating WL ambiguity for each visible satellite and location, which combines receiver wide-lane bias and WL integer ambiguity. Resolving ambiguity in the form of Double Difference (DD) and Single Difference (SD), where receiver WL bias is cancelled;

one wide lane offset per satellite; and

one GLONASS inter-frequency bias (IFB) WL coefficient per tracking position. The sensitivity coefficient is the number of satellite frequencies. This state variable is only applicable to the GLONASS case.

To be computationally efficient, if there is no cycle slip, ZD Melbourne-The measurements were averaged. Cycle slip refers to, for example, the loss of lock or pull-in of the carrier phase of any satellite signal. For each time interval, the filtered or smoothed ZD Melbourne @ is processed in the ZD Kalman filter position by position for each satellite reference receiver>And (5) measuring values. At each measurement update interval, in the MPP module 36, ZD Kalman filter dynamic updates and measurement updates are processed to update the state variables.

In one embodiment, each GNSS system satellite may be modeled with a separate and distinct wide-lane filter. For example, the GPS, GLONASS, GALIEO and BEIDOU systems with the Quazi-Zenith satellite system (QZSS) each have separate wide-lane filters. QZSS is a satellite based augmentation system for supplementing asian regions with GPS. Note that the reference receiver bias is not an ideal global differential product because it is not used for user receiver navigation. Thus, in some embodiments, to reduce the size and computational complexity of the filter, the reference receiver WL bias is not explicitly estimated; instead, WL bias is combined into the ZD float ambiguity state.

Rail solution module

The orbit solution module 38 is the main module of the estimator 34, which estimates the satellite orbits of one or more constellations or sets of GNSS satellites. The clock solution module 44 and the low-delay clock model 42 rely on the predicted orbit provided from the orbit solution module 38 over time intervals (e.g., every few minutes). The orbit solution module 38 provides an orbit solution at a lower rate (e.g., once every 300 seconds) because the GNSS satellite orbits are substantially smooth.

The data source of the orbit solution module 38 may be from real-time measurements, offline data 23 (e.g., recorded measurements or stored measurements, such as offline historical measurements for about 24 hours to about 48 hours), or any mix of real-time measurements and offline data 23 (e.g., stored measurements, for example, about 24 to about 48 hours). In one embodiment, the MPP module 36 requires at least 24 hours of offline data 23 to bring the track solution to steady state. In one configuration, in estimator 34, MPP module 36, track solution module 38, slow clock solution module 44, and low delay clock module 42 may run in parallel according to a parallel data processing process.

In one embodiment, the orbit solution module 38 uses the refraction-corrected code and carrier-phase measurements from one or more reference receivers 30 or the global reference station network 32. The orbit solution module 38 may be configured as a filter that processes three types of state variables, such as one or more of the following:

(1) Satellite-dependent state variables including satellite position, velocity, satellite clock, satellite lane bias, yaw rate, and empirical solar radiation modeling parameters;

(2) State variables dependent on the receiver including reference position, receiver clock, residual tropospheric bias and gradient, carrier phase ambiguity; and

(3) Common state variables include earth orientation parameters such as polar motion and universal time-universal coordinated time (UT 1-UTC).

To provide correction data for the global differential positioning service, the orbit solution module 38 evaluates the precise orbit and the uplink station 28 or other wireless communication device transmits the precise orbit to the end user mobile receiver 12 in real time. In general, the predicted satellite orbits (e.g., referred to as o2c data) within a certain maximum time interval (e.g., a few minutes) are considered known in clock estimation because the errors of these predicted orbits are quite small and stable and can even be absorbed by the estimated clock. The predicted satellite orbit (e.g., orbit o2c data) may be used to generate an orbit correction for the estimated precise orbit. In addition, the predicted satellite orbit may be used to correct orbit errors in the slow clock solution module 44 and the low delay clock module 42, such as by providing the slow clock solution module 44, the low delay clock module 42, or both with orbit correction or with estimated precise orbit after correction.

Clock solution module

In one embodiment, the measurements from the reference receiver 30 are collected, preprocessed, and batched for processing by the correction system 1011 as they arrive at the one or more data processing centers 218. The longer the data processing center 218 waits, the more measurements are collected, which delays the output and distribution of correction data. The clock solution module 44 evaluates a number of states to determine or evaluate a clock solution, which may lead to delays in the computation process. To reduce correction delay and utilize more measurements, two clock solutions are considered, including a slow clock solution and a low delay clock solution. In slow clock solutions, all measurements are batched as long as they arrive before a fixed delay in a time range such as about 6 seconds to about 15 seconds.

The clock solution module 44 determines or evaluates the slow clock solution based on a data source, such as real-time measurements of the reference receiver 30, offline recorded measurements or stored measurements (e.g., 2-3 hours of offline storage), or a combination or mix of real-time measurement data and offline recorded measurements or streams. For slow clock solutions, the clock solution module 44 typically takes at least 6 hours to reach steady state. In one example, slow clock solution module 44 uses similar measurements and ZD Kalman filters as orbit solution module 38, except for several main differences: first, the clock solution module 44 runs at a different rate, such as about 30 seconds or 60 seconds, that is different from the rate of about 300 seconds of the track solution module 38, because the clock correction changes faster than the track correction. Second, in the slow clock solution module 44, all state variables remain the same as in the orbit solution module 38 except for the satellite orbit-related state which is not estimated but uses the orbit estimate from the orbit solution. Third, the slow clock solution module 44 or estimator 34 outputs a complete set of global differential corrections including satellite orbit corrections, satellite clock corrections, satellite WL bias, NL bias and quality information. The previously corrected and estimated tropospheric parameters are sent to a low delay clock module 42. Note that because receiver bias is not of interest to the global differential correction product assigned to the mobile receiver 12, all biases are referred to as satellite WL and NL biases.

In the clock solution module 44, the satellite and receiver clocks, tropospheric bias, satellite narrow lane bias are estimated using the non-differential refraction-correction codes and carrier-phase observations. The update rate is typically low, such as a low update rate (e.g., a track update rate of 5 minutes, or a slow clock update rate of 30 seconds, or even longer). Because a large amount of ambiguity must be estimated along with the receiver and satellite clock parameters, the clock solution module 44 uses ambiguity resolution and satellite bias estimates to determine and update the slow clock solution at a low update rate. Estimator 34 or slow clock solution module 44 may wait a longer time before participating in the batch computation or calculation of the slow clock solution to ensure that sufficient measurements from reference receiver 30 or the reference network are collected and processed when reaching one or more data processing centers 218. The longer the clock solution module 44 waits, the more measurements are collected by the data collector 19 and the longer the delay for generating clock corrections. In one configuration, the clock solution module 44 is configured to determine a real-time satellite clock and satellite bias to preserve the full-cycle nature of the ambiguity of the mobile receiver 12. For example, ambiguity fixing can reduce convergence time and improve navigation accuracy.

Low delay clock module

The low delay clock module 42 determines changes or adjustments to the slow satellite clock to reduce correction delays. To improve clock accuracy with consistent satellite bias products and also to reduce correction delays, the clock solution module 44 and the low delay clock module 42 operate two parallel clock processes in cooperation with each other. The low delay clock module 42 uses only carrier phase measurements to calculate clock variations (clock increments). Slowly varying parameters such as tropospheric and satellite narrow lane bias are fixed as estimates from a slow clock solution module 44, which slow clock solution module 44 updates periodically (e.g., periodically) from slow clock process to low delay clock process at update time intervals, such as every 30 seconds update interval.

The low delay clock module 42 may estimate the clock delta using the delta carrier phase between T0 (at zero seconds) and the current epoch Ti. When the slow clock provides a new reference epoch, e.g., a T30 epoch (30 seconds), the low delay clock changes the reference epoch from T0 to T30.

In one embodiment, the low-delay clock fixes the predicted track (O2C data) in the low-delay clock solution a few minutes after the track solution is reached, e.g., 300 seconds. The low delay clock only evaluates the satellite clock delta so that the computation (of the satellite clock delta) can be updated at a very high rate. After a few seconds waiting window (e.g., 1 to 2 seconds), the preprocessed measurement values from the MPP module 36 are batched and sent to the low delay clock module 42. Meanwhile, the track solution module 38 and the clock solution module 44 provide a track solution and a clock solution, respectively, after a longer period of time (e.g., 6 seconds to 15 seconds). The low-delay clock solution is designed to reduce correction delay and improve clock accuracy by the absolute clock from the slow clock solution. To improve computational efficiency, double difference measurements between time and satellite are used in order to remove some unnecessary states, such as ambiguity and receiver clock. In one configuration, satellites that are highest in view (e.g., within a reliable reception range and with sufficient azimuth) and that have no cycle slip (e.g., in phase of one or more received carrier phase signals of the reference satellite) are selected as reference satellites for each respective reference receiver 30 or respective location.

The tropospheric bias is corrected using an a priori model and residual tropospheric bias estimates from a slow clock solution. The only state of satellite clock variation is estimated. The a priori satellite clock rate from the broadcast ephemeris is loosely used to estimate satellite clock delta. Of course, RAIM (receiver autonomous integrity monitoring) algorithms are used to ensure that any measurements with cycle slip are detected and removed.

The data processor 20, correction manager 40, or low delay clock applies the RAIM algorithm to the clock delta filter 412. The RAIM algorithm includes software that uses an over-determined solution or redundancy calculation to check the consistency of satellite measurements, such as carrier phase measurements and code phase measurements for one or more satellites of each reference receiver 30 in the network.

In one configuration, a Kalman filter or least squares method may be used to estimate the clock increment. The number of estimated state variables or unknowns is equal to the number of active satellites. The data processing center 218 may typically process the low delay clock measurements in a matter of milliseconds or less, which is faster (e.g., hundreds of times faster) than the clock solution module 44, which provides a slow clock solution.

In some configurations, the slow clock solution module 44 may take two to three seconds to complete the one epoch measurement process. After the first epoch, the low delay clock module 42p provides a satellite clock variation that integrates with the absolute clock from the slow clock solution to derive the absolute clock for any epoch. For the low-delay clock module 42, only the real-time data source of the reference receiver 30 is accepted for real-time processing, since there is no pull-in time for the low-delay clock solution. Finally, the orbit data, clock correction data, and satellite WL/NL offset product with quality information are transmitted to the user at the mobile receiver 12 in real time over the satellite L-band, wireless communication system, or Internet.

FIG. 11 illustrates a flow chart of one embodiment of a method for providing satellite correction signals. The method of fig. 11 begins at step S900.

In step S900, one or more reference receivers receive a series of raw satellite signal measurements. Raw satellite signal measurements are transmitted or passed from one or more reference receivers 30 to a data processing center (118, 218) (e.g., a data processing hub). For example, the reference receiver 30 may be in communication with one or more data ports 26 of the data processing system.

In step S902, the data storage device 24 stores raw satellite signal measurements received within a series of time windows (or time intervals) preceding the current time (current GNSS measurement time). Each satellite signal measurement is associated with a respective corresponding stored measurement time tag. Step S902 may be performed by using time intervals or time windows of various durations. In one example, the data storage device 24 stores the received raw satellite signal measurements according to a time window, where the time window is accumulated over a duration of at least 24 hours. In another example, the time window is accumulated over a duration ranging from about 24 hours to about 48 hours.

In step S904, during the warm start mode, the data processor 20, the data processing center 118, or the estimator 34 estimates or determines a satellite correction signal based on the satellite orbit data, the satellite clock data, and satellite bias correction data derived from the stored raw satellite signal measurements. During the warm start mode or start mode, the data processor 20 or estimator 34 feeds approximately six hours of stored data measurements into the slow clock estimation module to achieve steady state orbit estimation. During the start-up mode or the warm start mode, no historically stored data measurements are required to feed the low delay clock.

In step S906, the data processing center 118, the data processor 20 or the data source selector 15 seamlessly switches the measured value data source from the stored received raw satellite measurement values (e.g., switch) to a live real-time raw satellite signal measurement value if or when the corresponding measurement time tag of the last processed satellite signal measurement value of the stored received satellite signal measurement values approaches or reaches the current time, wherein the difference between the corresponding measurement time tag of the corresponding last processed, stored received raw satellite measurement value and the current time is less than the threshold time range. For example, after a warm start mode, during a normal mode of operation, the data processing center (118 or 218), the data processor 20, or the data source selector 15 seamlessly switches the measurement data source from the stored received raw satellite measurements (e.g., a handoff) to a live real-time raw satellite signal measurement if or when the corresponding measurement time tag of the last processed one of the stored received satellite signal measurements approaches or reaches the current time.

FIG. 12 illustrates a flow chart of one embodiment of a method for providing satellite correction signals. The method of fig. 12 is similar to the method of fig. 11 except that the method of fig. 12 further includes steps S908 and S910.

In step S908, the data processing center (118 or 218) or the data processor 20 determines when satellite orbit data, satellite clock data, and satellite bias data (e.g., wide-narrow lane bias data) have converged to reliable satellite correction data with ambiguity resolution.

In step S910, the data processing center (118 or 218) or the data processor 20 provides reliable correction data with ambiguity resolution to the end user via the satellite communication channel. For example, the data processing center (118 or 218) or the data processor 20 wirelessly provides and updates correction data to the mobile satellite receiver at a rate of at least 1 hertz or greater. The satellite uplink transmits correction data and updates to the communication satellite that repeats or resends the uplink transmission for receipt by the terrestrial wireless correction device 14, which terrestrial wireless correction device 14 is associated with or co-located with the mobile receiver 12.

Figure 13 shows a flow chart of one embodiment of a method for providing satellite correction signals. The method of fig. 13 is similar to the method of fig. 11 except that the method of fig. 11 further comprises step S912.

In step S912, during the normal mode of operation, the data processor 20 or the data processing center 118 estimates a differential correction signal using the satellite orbit data, the satellite clock bias data, and the satellite bias data based on the raw measurements received live in real time. The clock data and satellite clock bias data are the results of integrating two parallel clock estimation processes including a slow clock solution and a low delay clock solution, where the low delay clock solution is estimated as an increment or change relative to the slow clock solution.

The correction system and method support reliable, accurate and fast hot starts because both offline recorded data and real-time data are fed into the estimator or associated server software to accelerate GNSS correction convergence. The hot start is associated with the use of recorded offline measurement data and associated ephemeris data to reduce the start-up time required to provide correction data and distribute the correction signal to the mobile receiver, or subscriber or end user of the mobile receiver. In some examples, for a local correction system having a single location data processing center, a hot start may reduce the start-up time for providing reliable steady state accurate correction data from about 24 hours to about 48 hours to about 1.5 hours to about 3 hours. For a correction global correction system with a multi-location data processing center, a warm start may reduce the start-up time for providing reliable steady state accurate correction data from about 14 days to about two to three days. After convergence of the orbit and clock solutions, a set of consistent correction signals, including satellite orbit, clock, wide-narrow satellite bias and quality information, may be transmitted or communicated wirelessly in real time in a timely manner over the satellite L-band, wireless communication or the internet.

Figure 14 shows a flow chart of one embodiment of a method for providing satellite correction signals. The method of fig. 14 is similar to the method of fig. 11 except that the method of fig. 14 replaces step S906 with step S914.

According to one example, as shown in step S914 of fig. 14, the data source selector 15 is adapted to use a mix or combination of measurement data sources comprising stored raw satellite signal measurements and raw satellite signal measurements in live real time, if or when the respective measurement time tag of the last processed satellite signal measurement of the stored received satellite signal measurements has not been close to or reached the current time (e.g. the current GNSS measurement time), wherein the difference between the respective measurement time tag of the corresponding last processed stored raw satellite signal measurement and the current time is larger than the threshold time range. In one embodiment, estimator 34 may provide all necessary calculations such as measurement preprocessing, orbit and clock determination, ambiguity resolution, and final correction generation in fractions of a second. Advantageously, step S914 may use parallel processing techniques within each data processing center 218 to process concurrently stored signal measurements and current live signal measurements, where the measurements include satellite carrier phase measurements, to facilitate hot start of correction data and/or potentially reduce convergence time for ambiguity resolution associated with the correction data. For example, parallel processing may require the data processing center 218 to use multiple servers at a time or simultaneously to execute one or more modules (34 or 134) of an estimator or multiple electronic data processors 20 (e.g., multiple microprocessor cores) within the data processing center 218 to execute software instructions associated with one or more modules of the estimator (24 or 134).

WL deviation filtering system

In fig. 15A, WL (wide lane) bias filtering system 505 includes a main WL (wide lane) ZD (homodyne) filter (400, 501) and an additional or supplemental WL bias prediction filter (504) for each satellite (e.g., in view of being subject to any hardware data processing or memory limitations within a reliable signal reception range).

In one embodiment, the master WL ZD filter 501 may include the optional mppzd filter 400 of fig. 4, as indicated by the dashed line in fig. 15A. The optional MPP ZD filter 400 of fig. 15A may be configured to be linked to a corresponding optional wide lane estimator 37 (e.g., WL deviation filter) shown in dashed lines in fig. 4 or in communication with the optional wide lane estimator 37, wherein a column or set of WL filters 504 may supplement the optional wide lane estimator 37. For example, a column or set of WL filters 504 may be available for each satellite in view or within reliable signal reception range.

Cumulatively or separately, WL bias filtering system 505 may replace, merge or link to WL ZD filter 400 or communicate with WL ZD filter 400 alone or with its corresponding optional WL estimator 37 (e.g., wide lane bias estimator) with any of the figures in this disclosure. In practice, each GNSS reference receiver has a column of supplemental WL bias prediction filters (504) and a column of corresponding WL ZD filters (400, 501). Each WL ZD filter (400, 501) may include a prediction filter or a first WL kalman filter. Each supplemental WL deviation prediction filter (504) may include a second WL kalman filter. In the WL filter system, a main WL ZD filter (400, 501) is coupled to a supplemental WL deviation prediction filter (504) for communication of status, parameters or data. For example, the primary ZD filter or the first WL kalman filter provides WL residuals 502 (e.g., post-fit WL residuals) for a respective satellite to a supplemental WL bias prediction filter (504) for the respective satellite. In turn, the WL bias prediction filter 504 provides dynamic noise 503 for one or more WL bias.

The WL bias of a satellite may be mapped to the corresponding resolved DD integer ambiguity for that satellite or satellite cluster. The satellite wide lane offset is not constant, but often varies slowly over time, e.g., over a period of time or over a series of epochs, from 0.05 cycles to 0.2 cycles (at WL frequency or bandwidth) from the average satellite wide lane offset. However, in some cases, satellite WL bias may change rapidly and/or deviate significantly from average satellite WL bias due to external conditions that are not controllable by the receiver (e.g., GNSS mobile receiver and GNSS reference receiver); thus, there are two WL bias filtering modes in conjunction with WL bias filtering system 505: (1) A first WL filter mode (e.g., a conventional WL bias filter mode) for slowly varying WL bias of any corresponding satellite; and (2) a second WL filtering mode (e.g., a time-varying WL bias filtering mode) for rapid and/or greater changes in WL bias that may occur from time to time or over a transient period for any respective satellite. For any epoch or series of consecutive epochs, two mutually exclusive filtering modes may be invoked separately and independently for any satellite within the satellite constellation, or may be invoked separately and independently for a group of satellites within the satellite constellation. In one embodiment, a first WL filter mode (e.g., a conventional WL bias filter mode) is associated with a concomitant first set of kalman filter constraints, parameters, and states of the WL bias filter system 505; a second WL filtering mode (e.g., a time-varying WL deviation filtering mode) is associated with a concomitant second set of kalman filter constraints, parameters, and filter states of the WL deviation filtering system 505, which WL deviation filtering system 505 may include a pair of main WL ZD filters (400, 501) (e.g., a first WL kalman filter) and a supplemental WL deviation prediction filter (504) (e.g., a second WL kalman filter) for each respective satellite.

In a first WL filter mode, in one configuration, a master WL ZD filter (400, 501) assumes or receives verification that satellite bias varies slowly over time, and/or a supplemental WL bias prediction filter (504) may verify that verification's assumption or variation. Thus, in a first filtered WL mode (e.g., a conventional WL bias mode), a main WL ZD bias filter or WL ZD filter (400, 501) applies dynamic noise of a first WL tuning level or dynamic noise of a first WL tuning range based on an evaluation of WL residuals of the same satellite by the supplemental WL prediction filter. In a first example, a supplemental WL bias prediction filter (504) provides or outputsA first WL tuning level of dynamic noise, which may include constant dynamic noise for wide lane bias (e.g., 1e ^-8 Watts or 10,000 picowatts). In a second example, the supplemental WL bias prediction filter (504) provides or outputs dynamic noise for the first WL tuning range, which may include constant dynamic noise for wide lane bias (e.g., about 5,000 picowatts to 15,000 picowatts). Dynamic noise may be provided in watts or dBm (decibel milliwatts). In one embodiment, dynamic noise may be modeled as white noise with a gaussian distribution.

However, in a second WL bias filtering mode (e.g., a time-varying WL bias mode), the additional or supplemental WL bias prediction filter (504) may include a second kalman filter (e.g., a one-dimensional filter) or a residual WL bias filter to monitor satellite residual WL bias to trigger/support application of a second WL tuned dynamic noise (e.g., to replace the first WL tuned dynamic noise), wherein the second WL tuned dynamic noise is greater than the first WL tuned dynamic noise. For example, the second WL tuned dynamic noise may include dynamic noise (e.g., about 5 e) in a WL ZD bias filter (e.g., a kalman filter) ^-6 Watts or 5,000,000 picowatts). In conjunction with or separately from the selection of the second WL tuned dynamic noise, the additional or supplemental WL bias prediction filter (504) may include a second kalman filter (e.g., a one-dimensional filter) or a residual WL bias filter to monitor satellite residual WL bias to trigger/support the application of dynamic noise for a second WL tuning range, where the second WL tuning range of dynamic noise is greater than the dynamic noise for the first WL tuning range.

The WL bias filtering system 505, the first WL kalman filter, or the electronic data processor manages the transition between the first WL bias filtering mode and the second WL bias filtering mode (e.g., from a conventional WL bias filtering mode to a time-varying WL bias filtering mode) based on an evaluation of satellite WL residuals for a particular satellite by a supplemental WL bias prediction filter (504) (e.g., a residual WL bias filter) and generating corresponding dynamic noise (e.g., first WL tuned dynamic noise or second WL tuned dynamic noise). In one embodiment, the WL bias filtering system 505 applies the following equation to tune or select WL dynamic noise and associated WL bias (filtering) modes.

The fitted WL residuals for a given satellite from all receivers (e.g., in equations A8 or B2Is fed as a measurement into WL bias filter system 505 or supplemental WL bias prediction filter (504):

wherein,is a satellite wide lane residual (e.g., a WL residual associated with the resolved WL ambiguity, WL bias, or both for a given satellite) or a wide lane bias residual for the corresponding satellite; />Is the variance of the wide-lane bias residual error of the corresponding satellite; t and Δtt, t+1 are the measurement time and the sampling interval, respectively; and +.>Is dynamic noise. Supplemental WL bias prediction filter (504) (e.g., residual WL bias filter) applies the following test statistic or derived metric (T or T) _WL ) Dynamic noise for determining wide lane offset>Whether it is set correctly (e.g. in the WL ZD filter (400, 501) or the first WL kalman filter):

for example, if the WL bias prediction filter or electronic data processor determines the metric or test statistic T (e.g., T _WL ) Greater than a specified threshold of dynamic noise (e.g., a first WL tuned dynamic noise or a first WL tuned range of dynamic noise), then (a) residual wide lane biasIs very remarkable; and (b) it indicates a first tuned dynamic noise + >Is set too small or too low; and (c) requires that the dynamic noise increases in concert with the transition from the first WL bias filtering mode (e.g., conventional WL bias filtering mode) to the second WL bias filtering mode (e.g., time-varying WL bias filtering mode) (e.g., from the first tuned dynamic noise to the second tuned dynamic noise, or from the first tuned dynamic noise range to the second tuned dynamic noise range).

In the second WL bias filtering mode of the WL bias filtering system 505, the expansion scale may be set to T ² (e.g., or bounded by a limit of 100 times the original dimension) to allow WL bias to fluctuate over one or more epochs with greater bias or limit relative to its average WL bias. For example, the corrected dynamic noise may allow the WL bias to fluctuate over one or more epochs with a greater bias of 0.2 to 0.5 cycles (at WL frequency or wavelength) from the average WL bias of the particular satellite. Further, an increase in dynamic noise (e.g., in the dynamic noise of the first tuned dynamic noise) may be bounded by a range (e.g., the second tuning range of the dynamic noise) such that the next dynamic noise may be equal to the previous dynamic noise multiplied by a suitable multiplier (e.g., 100) that is characteristic of the carrier signal and the operating environment of the code modulation, such as modulating any C/a and/or PN code of each carrier signal.

However, if the supplemental WL bias prediction filter (504) determines the derived metric orTest statistics T (e.g., T _WL ) Less than a specified threshold of dynamic noise or a first tuned dynamic noise (e.g., about 1e ^-8 Watts or 10,000 picowatts)), then the residual wide lane biasIs not significant and it confirms the previous dynamic noise +.>Or first tuned dynamic noise (e.g. 1e ^-8 Tile) or dynamic noise is set correctly; thus, the WL bias filtering system 505, e.g., supplementing the WL bias prediction filter (504) (e.g., the second kalman filter), the WL ZD filter (400, 501), or both, may remain in the first WL bias filtering mode (e.g., the conventional WL bias filtering mode).

The WL bias filtering system 505 is configured to determine whether to operate in a first wide lane filtering mode in which a next wide lane bias correction for a given satellite has a corresponding first limited bias (e.g., a first WL threshold) related to a previous mean of WL biases from a series of previous consecutive epochs or a second wide lane filtering mode in which a next wide lane bias correction for the given satellite has a corresponding second limited bias (e.g., a second WL threshold) related to a previous mean that is greater than the first limited bias (e.g., the first WL threshold) (beyond the limits of the first limited bias) for each corresponding satellite.

For each satellite within the reception range, a supplemental WL bias prediction filter 504 (or an electronic data processor of the data processing center) is configured to determine a derived metric (T) based on the WL residuals and their variances _WL ). Further, the supplemental WL bias prediction filter 504 is configured to provide the second WL tuned dynamic noise (as a next setting for a transition or next epoch) to the WL ZD filter (400, 501) to replace the WL ZD filter if the first WL tuned dynamic noise is greater than the derived metricA first WL tuned dynamic noise (e.g., as a current setting of a current epoch and transitioning from a first WL filter mode to a second WL filter mode) of the filter (400, 501), wherein the second WL tuned dynamic noise is greater than the first WL tuned dynamic noise, and wherein a WL residual associated with the WL bias and a corresponding resolved WL ambiguity are correlated with a predetermined satellite.

For each satellite within the reception range, a supplemental WL bias prediction filter 504 (or an electronic data processor of the data processing center) is configured to determine a derived metric (T) based on the WL residuals and their variances _WL ). The supplemental WL bias prediction filter 504 is configured to provide the first WL tuned dynamic noise (as a next set maintenance for the next epoch to remain in the first WL filtering mode) to the WL ZD filter (400, 501) to maintain the first WL tuned dynamic noise of the WL ZD filter (400, 501) if the first WL tuned dynamic noise is equal to or less than the derived metric, and wherein the WL residual and corresponding resolved WL ambiguity associated with the WL bias are correlated with a predetermined satellite.

The above-described test is used to make a transition between a first WL bias filter mode and a second WL bias filter mode (e.g., from a conventional WL bias mode to a time-varying WL bias mode) with respect to a WL bias filter system 505, a WL ZD filter (400, 501) (e.g., a first kalman filter) and a supplemental WL bias prediction filter (504) (e.g., a residual WL bias filter or a second kalman filter). For example, the above-described testing and techniques may be applied to a variety of different operating conditions of satellites and associated GNSS receivers on a satellite-by-satellite basis, including one or more of the following: (1) a change in transmit signal power of the satellite carrier signal; (2) A change in the received signal-to-noise ratio of the one or more satellite carrier signals; (3) interference or attenuation of satellite carrier signals; (4) A change or allocation of signal power between or among C/a code modulation, P1 (Y code on L1) code modulation, or P2 (Y code on L2) code modulation; (5) the signal path of the transmitted signal varies with the satellite; (6) switching to a redundant satellite transmitter on the same satellite; (7) a change in active antenna or antenna array element; or (8) switch to or activate a different satellite signal path having a different thermal noise energy or perceived satellite clock difference associated with the satellite signal path.

In alternative embodiments, the transition between WL filtering modes may be further managed by not only requiring the above-described tests, but also requiring the addition of additional factors that may be applied to any satellite, either alone or in addition, including any of the following: an observed substantial change per unit time in the signal-to-noise ratio of the carrier signal, and/or an observed substantial change in the differential code bias between the different code components, wherein the additional factors may be measured in the measurement module or carrier phase measurement module of one or more reference receivers.

NL deviation filtering system

In fig. 15B, NL (narrow lane) bias filter system 507 comprises a pair or set of main NL (narrow lane) ZD (homodyne) filters (404, 408, 506) and NL bias filter/code phase bias filter 512 for each corresponding satellite. For example, the master NL ZD filter 506 is configured to communicate with a column or set of code bias filters 512 for each satellite in view or reliable signal reception range. In one embodiment, the master NL ZD filter 506 may include a track ZD filter 404, or a clock ZD filter 408, or both the track ZD filter 404 and the clock ZD filter 408. The orbital ZD filter 404 may be configured to be linked to or in communication with a corresponding optional narrow lane estimator 39 (e.g., NL bias filter), as shown by the dashed lines in fig. 4, wherein a column or set of code bias filters 512 may supplement the optional narrow lane estimator 39 for each satellite within view or reliable signal reception. Similarly, clock ZD filter 408 may be configured to be linked to, in communication with, a corresponding optional narrow lane estimator 43 (e.g., NL bias filter), as shown by the dashed lines in fig. 4, wherein a column or set of code bias filters 512 may supplement optional narrow lane estimators 39 for each satellite in view or within reliable signal reception.

The clock solution module 44 and its clock ZD filter 408 use the same measurements as the track solution module 38 and its track ZD filter 404 except that the clock solution module 44 can operate at a higher rate than the track solution module because the clock correction tends to change faster than the track correction. The track solution module 38 may operate at a track refresh rate, a track update rate, or a track data processing clock rate that is less than the slow clock refresh rate, the slow clock update rate, or the slow clock processing clock rate and less than the low delay clock refresh rate, the low delay clock update rate, and the low delay processing clock rate. Further, when operating at a slow clock refresh rate, a slow clock update rate, or a slow clock processing clock rate, the clock solution module 44 may reference, access, rely on, or utilize the track estimation results (of the track solution module 38), such as: (a) Modeling and estimation of filter state variables, sensitivity coefficients of the track ZD filter (404); and (b) refining the epoch or prior epoch prior entry bias estimate, such as satellite entry bias, receiver entry bias, satellite entry phase bias, receiver entry phase bias, satellite entry code bias, and/or receiver entry code bias.

Cumulatively or separately, the NL bias filter system 507 may replace, merge or link to (NL) ZD filters (404 or 408) or communicate with (NL) ZD filters (404 or 408), respectively, alone or with its corresponding optional NL (bias) estimator (39 or 43) in any of the figures in this disclosure. In practice, each GNSS NL bias filter system comprises an NL bias prediction filter, a code phase bias filter, or both. The NL bias/code phase bias filter 512 may be configured as a second NL kalman filter. In an NL filtering system, a master NL ZD filter (404, 408, 506) is coupled to an NL bias filter/code phase bias filter 512 for communication of status, parameters, or data. For example, the primary ZD filter or the first Kalman filter provides the RC NL code residual 508 of the respective satellite to the NL bias filter/code phase filter 512 of the respective satellite. In turn, the NL bias filter/code phase filter 512 provides dynamic noise 510 for one or more of the following biases: NL bias, refraction Corrected (RC) NL (code) bias and/or code phase bias.

The NL bias of a satellite may be mapped to a corresponding resolved DD integer ambiguity for that satellite or satellite cluster. Satellite lane deviations are not constant, but typically vary slowly over time, e.g., within 0 to about 0.2 cycles (at NL frequency or bandwidth) from the average satellite lane deviation over a period of time or over a series of epochs. However, in some cases, due to external conditions that the receiver (e.g., GNSS mobile receiver and GNSS reference receiver) is not capable of controlling, satellite NL bias may change rapidly and/or deviate significantly from average satellite NL bias, e.g., within a period of time or a series of epochs from average NL bias within 0 to about 2 cycles (at NL frequency or bandwidth); thus, there are two NL bias filtering modes: (1) A first NL filtering mode (e.g., a conventional NL bias filtering mode) for slowly varying the NL bias of any corresponding satellite; and (2) a second NL bias filtering mode (e.g., a time-varying NL bias filtering mode) for rapid and/or greater changes in NL bias that may occur from time to time or over transient periods for any corresponding satellite. For any epoch or a series of consecutive epochs, two mutually exclusive NL filtering modes may be invoked separately and independently for any satellite within the satellite constellation, or may be invoked separately and independently for a group of satellites within the satellite constellation. In one embodiment, a first NL filtering mode (e.g., a conventional NL bias filtering mode) is associated with a concomitant first set of kalman filter constraints, parameters, and states; a second NL filtering mode (e.g., a time-varying WL bias filtering mode) is associated with the accompanying second set of kalman filter constraints, parameters, and filter states.

In a conventional NL bias filtering mode, a first master NL ZD filter (404, 408, 506) assumes or receives verification that satellite bias from the NL bias filter/code phase filter of the same particular satellite varies slowly over time. In the first NL filtering mode, the NL bias filter/code phase bias filter 512 provides or outputs a first NL tuning level of dynamic noise to the master NL bias filter or NL ZD filter (404, 408, 506) based on the NL bias filter/code phase bias filter's evaluation of satellite RC code residuals. In the first viewFor example, the NL bias filter/code phase bias filter 512 provides or outputs a first NL tuning level of dynamic noise, which may include a small and constant dynamic noise for narrow lane deviations (e.g., 1e ^-7 Watts or 100,000 picowatts). In a second example, the NL bias filter/code phase bias filter 512 provides or outputs a first NL tuning range for dynamic noise for NL bias. Dynamic noise may be provided in watts or dBm (decibel milliwatts). In one embodiment, dynamic noise may be modeled as white noise with a gaussian distribution.

However, in the second NL bias filter mode, the NL bias filter/code phase bias filter 512 may comprise a second NL kalman filter or residual NL code phase bias filter 512 to monitor satellite RC code phase residuals to trigger/support application of a second NL tuned dynamic noise (e.g., to replace the first tuned dynamic noise), wherein the second NL tuned dynamic noise is greater than the first NL tuned dynamic noise. The second NL tuning dynamic noise (e.g., about 5e ^-5 Tile) may include greater dynamic noise than the first NL tuned dynamic noise applicable in the NL ZD bias filter (e.g., the first NL kalman filter) or in the NL bias filter system 507. The NL bias filter/code phase filter may comprise a second kalman NL filter (e.g., a one-dimensional filter) or a code phase bias residual filter, either together or separately from the selection of the second NL tuning dynamic noise, to monitor the satellite residual RC code phase to trigger/support the application of the dynamic noise for a second NL tuning range, which is larger than the first NL tuning range of the dynamic noise.

The NL bias filter system 507, a first NL kalman filter, or an electronic data processor manages the transitions between the first NL bias filter mode and the second NL bias filter mode (e.g., from a conventional NL bias filter mode to a time-varying NL bias filter mode) based on an evaluation of satellite refraction-corrected (RC) code phase residuals of the particular satellite by the supplemental NL bias filter/code phase filter, and generating corresponding dynamic noise (e.g., first NL tuned dynamic noise or second NL tuned dynamic noise). In one embodiment, the NL bias filtering system 507 applies the following equations to tune or select the NL dynamic noise and associated NL bias (filtering) mode:

Fitted NL residuals for a given satellite from all receivers (e.g., in a15, a16The residual in C34) is fed as a measured value into the NL bias filter system 507 or the NL bias filter/code phase filter.

Wherein,is a satellite RC phase code residual (e.g., satellite NL bias residual for any leading indicator of the corresponding satellite or satellite NL bias +.>)；/>Is the variance of satellite NL bias residuals; t and Deltat _t，t+1 The measurement time and the sampling interval, respectively; and wherein->Is dynamic noise. The NL bias filter/code phase filter applies the following derived metrics or test statistics (T or T _NL ) Dynamic noise for determining narrow lane departure +.>Whether or not to be correctly set (e.g. in the main NLZD filter (404, 408, 506) or in the first NLZD filters (404, 408, 506).

For example, if the NL bias filter/code phase filter determines the derived metric or test statistic T (e.g., T _NL ) Greater than a specified threshold of dynamic noise (e.g., a first NL tuned dynamic noise level or a first NL tuned dynamic range), then: (a) RC code phase residuals are significant; (b) Which is indicative of dynamic noiseIs set too small or too low; and (c) the dynamic noise is required to increase in correspondence with the transition from the first NL filter mode to the second NL filter mode.

Alternatively, the NL bias filter/code phase filter is based on bias from the narrow laneAny substantial variation or limitation of excess in the associated RC code phase residual to determine that the derived metric or test statistic T in equation 57 is greater than the specified threshold for dynamic noise such that the lane bias +.>May be subject to greater deviations (predicted in the future) from their average narrow lane deviation.

Within the second NL filtering mode (e.g., a time-varying NL bias filtering mode), the dilation scale may be set to T ² (e.g., or bounded by a limit of 100 times the original dimension) to allow NL deviations to fluctuate over one or more epochs with greater deviations or limits relative to their average NL deviations. For example, the modified dynamic noise may allow the NL bias to fluctuate over one or more epochs with a greater bias of 0 to about 2 cycles (at NL frequency or wavelength) of the average WL bias from a particular satellite. Furthermore, the increase in dynamic noise (e.g. of the first NL tuned dynamic noise) may be in a certain rangeAs a boundary, such that the next dynamic noise (e.g., the second NL tuned dynamic noise) may be equal to the previous dynamic noise multiplied by a suitable multiplier (e.g., 100) that is characteristic of the operating environment of the carrier signal and the signal-to-noise ratio of the code modulation, such as any C/a and/or PN code modulating each carrier signal.

However, if the derived metrics or test statistics T (e.g., T _NL ) Less than a specified threshold of dynamic noise (e.g., 1e ^-7 Tile or 10,000 picowatts), then deviate from NLThe associated RC code phase residual is not significantly changed, which indicates or confirms the previous dynamic noise of NL bias>(e.g. first NL tuning dynamic noise, e.g. 1 e) ^-7 Tile) is properly set; thus, the NL bias filter (e.g., kalman filter) or the NL ZD filter (404, 408, 506) can remain in a first NL bias filter mode (e.g., a conventional NL bias filter mode).

The data processing center or NL bias filter system 507 has a column of complementary NL bias filters/code phase bias and a column of corresponding NL ZD filters (404, 408, 506). In one embodiment, the master NL ZD filter (404, 408, 506) comprises a prediction filter, such as a first NL kalman filter; NL bias filter/code phase bias filter 512 comprises NL.

The NL bias filtering system 507 is configured to determine whether to operate in a first narrow lane filtering mode in which a next narrow lane bias correction for a given satellite has a corresponding first limited bias (e.g., a first NL threshold) related to a previous mean of NL biases from a series of previous consecutive epochs or a second narrow lane filtering mode in which a next narrow lane bias correction for the given satellite has a corresponding second limited bias (e.g., a second NL threshold) related to a previous mean that is greater than the first limited bias (e.g., the first NL threshold) (beyond the limits of the first limited bias) for each corresponding satellite.

For each satellite in the reception range, a supplemental NL bias filter/code phase filter (or an electronic data processor of a data processing center) is configured to determine a derived metric (T) based on the refraction-corrected code residuals (e.g., RC code phase bias residuals) and their variances _NL ). The NL bias filter/code phase filter is configured to provide a second NL tuning dynamic noise (as a next setting for a transition or a transition of a next epoch from a first NL filtering mode to a second NL filtering) to the NL ZD filter to replace the first NL tuning dynamic noise of the NL ZD filter (e.g., as a current setting for a current epoch), if the first NL tuning dynamic noise is greater than the derived metric, wherein the second NL tuning dynamic noise is greater than the first NL tuning dynamic noise, and wherein the RC code residuals associated with the NL bias and corresponding resolved NL ambiguities are related to predetermined satellites.

For each satellite in the reception range, determining that the supplemental NL bias filter/code phase filter (or electronic data processor of the data processing center) is configured to derive a metric (T) based on the refraction-corrected code residuals (e.g., RC code phase bias residuals) and their variances _NL ). The NL bias filter/code phase filter is configured to provide the first NL tuning dynamic noise (as the next setting for the next epoch) to the NL ZD filter if the first NL tuning dynamic noise is less than or equal to the derived metric to maintain the first NL tuning dynamic noise (as the current setting for the current epoch) of the NL ZD filter, wherein the RC code residuals associated with the NL bias and corresponding resolved NL ambiguities are related to a predetermined satellite. Generation/enhancement of RC NL code phase residual and correction signal

The NL filtering system comprises a pair of main NL ZD filters (404, 408, 506) and NL bias filters/code phase bias filters 512 for each corresponding satellite. In one embodiment, the master NL ZD filter (404, 408, 506) comprises a prediction filter, such as a first NL kalman filter; the NL bias filter/code phase bias filter 512 comprises a prediction filter, such as a second NL kalman filter. In an NL filtering system, a master NL ZD filter (404, 408, 506) is coupled to an NL bias filter/code phase bias filter 512 for communication of status or data. For example, the primary NL ZD filter (404, 408, 506) or the first kalman filter provides the RC NL code residuals for the respective satellites to the NL bias filter/code phase filter for the respective satellites. In turn, the NL bias filter/code phase filter provides dynamic noise for the NL bias (e.g., first NL tuned dynamic noise, second NL tuned dynamic noise, first NL tuned dynamic noise range, or second NL tuned dynamic noise range). Furthermore, the NL bias filter/code phase filter may determine the content of the correction signal based on an evaluation of the Refraction Corrected (RC) code phase residual by the NL bias/filter code phase filter.

In one embodiment, one or more GNSS constellations, such as GPS, GLONASS, GALILEO, BEIDOU and QZSS, may be integrated in one prediction filter (e.g., a kalman filter) related to updating and managing the narrow-lane floating ambiguity variance-covariance matrix. The floating ambiguity variance-covariance matrix update should include a receiver bias update for each GNSS constellation, as set forth below:

wherein:

is the current epoch, current measurement interval, or current updated residual satellite RC NL bias for the prediction filter;

is the previous epoch, previous measurement interval or previous updated residual satellite RC NL bias for the prediction filter;

is dynamic noise for NL receiver bias; and

Δtn-l, n is the previous epoch, previous measurement interval, or previously updated measurement time or sampling interval for the prediction filter.

The satellite NL bias variance is also updated as follows:

dynamic noise for satellite NL biasA reliable and accurate estimation of satellite NL bias is supported. In addition, dynamic noise has a significant impact on GNSS receiver navigation performance. If the dynamic noise is estimated (e.g., or set) too large, rapid changes in NL satellite bias (including code bias and phase bias) may be reflected in the estimation. However, NL bias accuracy may deteriorate; therefore, it may be susceptible to deteriorated navigation performance of the GNSS receiver. If the dynamic noise is estimated (e.g. or set) too small, then the estimated NL bias may be an undue bias, which will not reflect a rapid change or attenuation of the satellite transmit power, e.g. caused by the GPS elastic power of the carrier. Furthermore, after initialization and analysis of the NL ambiguity, the RC NL code may deviate from the true steady state value in the GNSS receiver. To optimally tune the dynamic noise of satellite NL bias, a series of one-dimensional filters (including one filter for each satellite) may be configured to support adaptive estimation of satellite narrow lane bias.

The residual Refraction Corrected (RC) code phase bias for any corresponding satellite exhibits a high degree of time correlation over a short time span. The RC code phase bias of each respective satellite may be expressed as a function of time for a short period of time up to a few minutes. The same or similar assumptions support a systematic biased gaussian markov process as per the following equations (55-56).

Wherein,and->Is the residual satellite RC code phase bias and variance; t and Deltat _t，t+1 The measurement time and the sampling interval, respectively; />Is the dynamic noise of the RC code phase bias. The NL bias filter/code phase bias filter 512 determines the following derived metrics or test statistics (T or T) _NL ) To determine dynamic noise of the lane departure +.>Whether to indicate augmentation of correction signals with code phase deviation data for any respective satellite or to transmit or broadcast such augmented signals to a GNSS mobile receiver or rover station with correction service subscription:

in one embodiment, the fitted RC NL code residuals (e.g., in A15, A16) from any given satellite of a set or cluster of reference receivers (e.g., all reference receivers)Residual in component(s) or C34) of the blockThe measurements are fed to an NL filter system (e.g., NL bias/code phase bias filter 512) to estimate the residual satellite RC code phase bias for a given satellite. In equation 55, the RC code phase bias should be close or near zero under normal conditions. For correction signal enhancement or broadcasting, the default dynamic noise of code phase deviation or RC code phase deviation +. >May be set to a dynamic noise value such as (e.g., about 5e ^-5 Watts or 50,000,000 picowatts), which is typically greater than the dynamic noise of the lane departure +.>(e.g., about 1 e) ^-7 Watts or 100,000 picowatts). Dynamic noise may be provided in watts or dBm (decibel milliwatts). In one embodiment, dynamic noise may be modeled as white noise with a gaussian distribution.

For phase code bias, derived metrics or test statistics may be used to determine data processing and computer resource allocation associated with any of the following phase code bias correction parameters: (1) A phase code bias correction parameter for determining a phase code bias of one or more respective satellites; (2) A phase code bias correction parameter for determining a phase code bias or differential phase code bias of a corresponding coarse acquisition code (C/a), PN code, P (Y) code, P (1) or P (2) for modulating one or more carrier signals of one or more respective satellites; (3) A phase code bias correction parameter for incorporating phase code bias data into the correction signal; and (4) a phase code bias correction parameter for broadcasting phase code bias data. Separately or cumulatively, the above-mentioned phase code deviation correction parameters are not transmitted with the correction signal in the primary mode, but are transmitted to the correction signal in the auxiliary mode.

If the NL bias filter/code phase bias filter 512 determines the derived metric or test statistic T (e.g., T _NL ) Less than a threshold of dynamic noise (e.g., about 5e ^-5 Tile), then residual error for the corresponding satelliteRC NL code phase deviationIs not significant and it confirms the previous dynamic noise +>(e.g., about 5 e) ^-5 Tile) is correctly set; thus, the NL bias filter system 507 (e.g., kalman filter) or the NL ZD filter (404, 408, 506) may remain in a master mode in which the above-described phase code bias correction parameters are not sent to rover stations or mobile GNSS stations subscribed to the correction service. Furthermore, in the main mode, the residual satellite RC NL code bias is not incorporated into the correction signal because the residual RC NL code bias is not significant or substantial. For example, uncertainty in the code phase offset or code phase offset correction parameters may reduce data processing requirements, energy consumption, and required wireless bandwidth (e.g., satellite bandwidth) for transmitting correction signals from the ground station to the communication satellite that is repeating or re-transmitting the signals to a rover station or mobile GNSS receiver that subscribes to the satellite correction service.

However, if the NL bias filter/code phase bias filter 512 determines that the derived metric or test statistic T (e.g., T _NL ) Greater than a specified threshold, residual RC code phase biasIs remarkable; therefore, this indicates dynamic noise +.>(e.g., about 5 e) ^-5 Watts) is set too small and needs to be increased in correspondence with the auxiliary mode. In the second mode, the expansion scale may be set to T ² (e.g., or bounded by a 100-fold limit on the original scale) to allow code phase deviation (e.g., or RC code phase deviation) to deviate or limit (e.g., typically in meters) over one or more epochs by a greater amount relative to its average code phase deviation (e.g., or RC code phase deviation)Unit measurement) fluctuation. For example, the increase in dynamic noise may be bounded by a range such that the next dynamic noise may be equal to the previous dynamic noise multiplied by a suitable multiplier (e.g., 100) that is characteristic of the operating environment of the carrier signal and the code modulation, such as any C/a and/or PN code that modulates each carrier signal. In the assist mode, when a metric or test statistic T (e.g., T _NL ) Notably, the residual satellite RC code phase bias from the clock solution in equation 55 will be broadcast along with the orbit, clock, and WL/NL bias and used by the rover station as a known parameter to assist in real-time positioning and navigation. Equation 57 is used to adaptively increase the dynamic noise of satellite narrow lane bias states in orbit and clock solutions. It also determines whether to broadcast satellite code phase bias.

The first WL filter includes a main WL ZD filter (400, 501) for the corresponding satellite and the second WL filter includes a supplemental WL bias prediction filter (504) (e.g., a residual WL bias filter) for the corresponding satellite. In one embodiment, a supplemental WL bias prediction filter (504) (e.g., a residual WL bias filter) determines an operational dynamic noise level or range of the WL ZD filter (400, 501), or whether the WL ZD filter (400, 501) is operating in a first filtering mode (e.g., conventional WL bias filtering) or a second filtering mode (e.g., time-varying WL bias filtering mode). In one embodiment, the WL ZD filter (400, 501) and the supplemental WL deviation prediction filter (504) each comprise a kalman filter. In general, a column of first WL ZD filters (400, 501) and supplemental WL bias prediction filters (504) are configured such that a pair of WL ZD filters (400, 501) and supplemental WL bias prediction filters (504) (e.g., residual WL bias filters) are available for each respective satellite.

The first NL filter comprises NL ZD filters (404, 408, 506) for the corresponding satellites, and the second NL filter comprises NL bias filters/code phase filters (e.g., refraction corrected code residual NL bias filters) for the corresponding satellites. The NL bias filter/code phase filter determines the dynamic noise level, constraints, or parameters of the NL ZD filter (404, 408, 506), or whether the NL ZD filter (404, 408, 506) is operating in a first NL filtering mode (e.g., a conventional NL bias filtering mode) or a second NL filtering mode (e.g., a time-varying NL bias filtering mode). In one embodiment, both the NLZD filter (404, 408, 506) and the NL bias/code phase filter comprise Kalman filters. Typically, a column of first NL filters and second NL filters is configured such that a pair of NL ZD filters (404, 408, 506) and NL bias filters/code phase filters are available for each respective satellite.

The NL bias filter/code phase filter (e.g., RC code residual filter) monitors or evaluates the code phase residual provided by the master NL ZD filter (404, 408, 506) to the NL bias filter/code phase filter (e.g., RC code residual NL filter). Furthermore, the RC code residual filter may provide code phase bias information to be combined or available for incorporation into correction signals or correction data messages for broadcast or wireless transmission to subscribers of mobile or rover GNSS receivers.

Wide-lane or wide-lane carrier phase refers to the combination or difference of carrier phase measurements of the L1 and L2 signals, where the wide-lane wavelength is subtracted from the L1 and L2 carriers. An alternative wide-lane carrier phase may refer to a combination of the L2 carrier phase signal and the L5 carrier phase signal, rather than the conventional wide-lane carrier phases of the L1 signal and the L2 signal.

Narrow band or lane refers to the combination or sum of carrier phase measurements of the L1 and L2 signals, where the lane wavelength is derived from the addition of the L1 and L2 carriers. An alternative narrow lane carrier phase may refer to a combination of the L2 carrier phase signal and the L5 carrier phase signal instead of the conventional narrow lane carrier phases of the L1 signal and the L2 signal.

The differential code bias includes any of the following: (1) Differences between code phase pairs of P1 carrier signals and P2 carrier signals of respective satellites or a group of satellites of one or more GNSS systems of a common epoch or sampling interval (e.g., GPS P1-GPS P2; GLONASS P1-GLONASS P2); (2) Differences between pairs of P1 carrier and L1C1 carrier signals (e.g., GPS P1-GPS L1C1 or P1-Cl) for respective satellites or a set of satellites of one or more GNSS systems of a common epoch or sampling interval; (3) The difference between the code phase pairs of the P2 carrier signal and the L2C carrier signal of the corresponding satellite or set of satellites of one or more GNSS systems of a common epoch or sampling interval (e.g., GPS P2-GPS L2C1 or P2-C1). P1 represents a pseudo-random noise code or P (Y) code on the L1 carrier; p2 represents a pseudo-random noise code P (Y) code on the L2 carrier; cl represents the coarse acquisition code (C/a) code of the L1 carrier. Within a reference receiver (GNSS reference receiver), a baseband/IF processing module, a measurement module, or a code phase measurement module may support or facilitate determination of phase code bias or differential code bias.

The received signal-to-noise ratio of any of the L1, L2, and L5 carrier signals of the respective particular satellite may be used to detect whether the particular satellite employs elastic power to redistribute energy or transmission power among or among different signal components to increase interference immunity or to protect against interference. For example, the L1 (P (Y)) and L2 (P (Y)) signal components (or pseudo-random noise (PN) codes on the L1 carrier and the L2 carrier) may have increased signal power, which results in a corresponding decrease in L1 (CA) or L1 carrier coarse acquisition code power. Within a reference receiver (e.g., a GNSS reference receiver), a baseband/IF processing module, a measurement module, or a carrier phase measurement module may support or facilitate estimation of signal-to-noise ratio of one or more carrier signals.

In practice, the P1-C1 or P1-P2 differential code bias can be used to determine clock estimates (and orbit estimates) consistent with ionospheric-free linear combinations of code observations provided by interrelated receivers.

In a data processing center, the electronic data processor of the data processing center may use steps, offsets or discontinuities in the observed differential code bias alone or with steps, offsets, discontinuities or other rapid changes in the signal-to-noise ratio of certain carrier satellite signals (e.g., L2 (Y) or L5P (Y)) reported by one or more reference receivers in order to detect potential discontinuities or desired changes in WL bias correction or NL bias correction. For example, if, for any given satellite, the change in signal-to-noise ratio of the carrier signal exceeds a threshold change in signal-to-noise ratio, or if the change in differential code bias exceeds a threshold change in differential code bias, or if both thresholds are exceeded, the WL filter may adjust WL filter bias provided in the correction signal for such satellite (e.g., according to the time-varying WL bias constraint of the kalman filter) in a time-varying WL bias mode, and the NL filter may adjust NL filter bias provided in the correction signal for such satellite (e.g., according to the time-varying NL bias constraint of the kalman filter) in a time-varying NL bias mode, as reported by one or more reference receivers. In contrast, in a conventional WL bias mode, the WL filter may only allow for a small deviation, range, or variation of the average or mean WL bias (e.g., according to the slowly varying WL bias constraint of the kalman filter). Similarly, in a conventional NL bias mode, the NL filter may only allow for a small deviation, range, or variation of the average or mean WL bias (e.g., according to the slowly varying NL bias constraint of the kalman filter).

The position estimator or ambiguity resolution engine estimates the wide-lane carrier-phase ambiguities by determining a double-difference wide-lane combination of the L1 and L2 carrier-phase measurements of the reference station and the mobile receiver. The double difference wide lane carrier-phase combination is typically associated with a double difference residual. The double-difference wide-lane carrier-phase and double-difference narrow-lane code-phase combinations are differenced to determine a wide-lane ambiguity residual. The wide lane ambiguity residuals may be averaged over time to provide an interpretation or estimation of the wide lane whole-cycle ambiguity. For example, the wide lane whole-cycle ambiguity may be resolved by statistical analysis of the average (e.g., mean) of the RMS errors in the double-difference residual, where the reference noise covariance of the code and carrier phases is provided (e.g., from a look-up table or data store). After resolving the ambiguity, the dual differential carrier phase measurements can be used to estimate the relative range or vector measurements between the reference station and the mobile receiver. In one embodiment, a prediction filter (e.g., a Kalman filter) or relative position estimator may apply a least squares estimation technique or similar search algorithm to the relative range or vector measurements to determine the relative position of the mobile receiver with respect to a reference station or other reference coordinates, where such estimation may be further adjusted for refraction correction and atmospheric propagation.

The column of WL filters includes a pair of WL homodyne filters and a supplemental WL bias prediction filter (504) for each satellite. For the first column, each WL ZD filter (400, 501) is associated with a different corresponding satellite. Each WL ZD filter (400, 501) is coupled to a corresponding supplemental WL deviation prediction filter (504). The WL ZD filter (400, 501) provides or outputs WL bias associated with ambiguity resolution of WL ambiguities for the respective satellite. The WL ZD filter (400, 501) provides or outputs satellite WL bias residuals to a corresponding supplemental WL bias prediction filter (504) for the respective satellite. In turn, the supplemental WL bias prediction filter (504) provides the satellite with dynamic noise for estimation of the WL bias of the corresponding satellite to the WL ZD filter (400, 501).

In one embodiment, the supplemental WL bias prediction filter (504) includes a prediction filter such as a kalman filter. Similarly, the NL bias filter/code phase bias filter may comprise a prediction filter such as a kalman filter. Kalman filters typically use a summation of signal, delay, and feedback to process measurement samples (e.g., WL residuals and/or RC code residuals) and allow for noise and uncertainty. Here, for WL estimation, for each given satellite, a kalman filter receives measured samples of the satellite WL bias and satellite WL residual and outputs estimated dynamic noise (e.g., with a known satellite type and a known satellite identifier) for the satellite WL bias for the corresponding satellite. The WL bias filter or kalman filter is configured to: WL bias in the correction signal is generated or adjusted for the corresponding satellite to be in a time-varying WL mode that is affected by time-varying WL constraints, parameters and corresponding states or in a conventional WL mode that is affected by conventional WL constraints, parameters and corresponding states for slowly varying WL bias.

In alternative embodiments, the kalman filter may enable or activate the first WL filter mode (e.g., time-varying WL filter mode for a period of time or a series of epochs) based on one or more of the following factors: (1) Evaluating a derived metric relative to a dynamic noise threshold, the derived metric being based on WL residuals (e.g., WL bias residuals and/or WL ambiguity residuals) and variances thereof for each satellite; (2) Detection of discontinuities in signal-to-noise ratio or rate of change per unit time of one or more carriers of a GNSS system; (3) Detection of a discontinuity or rate of change per unit time of a code deviation or differential code deviation of one or more coded signals (e.g., a PN signal of a modulated carrier or a coarse acquisition code of a modulated carrier).

Dynamic noise in the determined or estimated WL bias for each respective satellite is based on one or more of: (a) Measurement noise in the estimated WL bias observed by the WL bias filter or the kalman filter; (b) A change, range or limitation of WL bias over GNSS time over one or more epochs or an extended period of time to its mean (average); (c) The change, range or limitation of WL (mean) over GNSS time over one or more epochs or an extended period of time, wherein distortion of WL bias measurements observed during the lifting or setting (setting) of satellites from the field of view of the reference receiver is eliminated, adjusted or compensated for to avoid such distortion that would otherwise be caused by atmospheric propagation delay.

An epoch is a measurement time interval of a GNSS system, such as a time interval between samples of phase measurements of a carrier phase signal or code phase.

Each WL filter system (e.g., WL ZD filter (400, 501)) may provide or estimate dynamic noise for a corresponding WL bias for a respective satellite. The estimated dynamic noise of the WL bias filter may replace default dynamic noise or reference dynamic noise stored in the WL bias filter, WL ZD filter (400, 501) (e.g., which includes a residual WL bias filter) or the data storage of the receiver. The default dynamic noise may be based on empirical measurements, historical measurements to or last measurement of the WL bias filter stored prior to shutting down the receiver. In one embodiment, the default dynamic noise is selected to align with typical dynamic noise (such as the first WL tuned dynamic noise) associated with operating in a conventional WL bias filter mode.

In an alternative embodiment, for operation in the conventional WL bias filtering mode, the default dynamic noise may be based on a noise temperature or average operating noise temperature of a satellite system including a pair of specific satellites at each carrier frequency and corresponding receivers to estimate the first WL tuned dynamic noise, where the noise temperature or average operating noise temperature depends on noise contributions from background noise, atmospheric noise, and receiver front-end noise.

In one configuration, the WL filter or kalman filter provides default dynamic noise or estimated dynamic noise to the WL ZD filter (400, 501). In transient conditions, where there is a fluctuation in the signal-to-noise ratio of the carrier, encoded signal, or code phase offset that would otherwise disrupt or distort the WL bias correction provided for the correction signal, the WL filtering system or WL ZD filter (400, 501) may reduce, increase, or adjust the weights (and would otherwise be incorporated into the correction signal) of the WL bias estimated for one or more corresponding epochs based on the applicable dynamic noise of the conventional WL bias filtering mode, the estimated dynamic noise of the time-varying bias filtering mode, or both, to determine or limit the average or mean of the WL bias for the respective satellite over time (e.g., over a sequence of consecutive epochs).

Under a first technique, in a first filtering mode (e.g., a conventional WL bias filtering mode), a WL filter or WL ZD filter (400, 501) may determine a respective limit, range, maximum limit, maximum range, or maximum variance associated with an average or mean of WL bias of a corresponding satellite associated with a default dynamic noise or a lower level of dynamic noise. For example, the associated WL residuals in the WL ambiguity may form a parameter or index for adjusting the maximum range or maximum limit in the dynamic noise such that WL deviations above or below the maximum range or maximum limit are rejected for use in the estimation of the mean or mean of the WL deviations in the WL ZD filter (400, 501).

For example, in a first WL filtering mode, the WL bias filtering system 505 may apply a kalman constraint in an associated limit, lower limit, upper limit, maximum range, or maximum variance of the average or mean of WL bias to reject WL bias determined or estimated that exceeds the limit, range, or variance for the purpose of determining the average or mean of WL bias for the respective satellite over each epoch or corresponding time period. Thus, dynamic noise may be applied to enhance the accuracy of WL bias of the respective satellite over each epoch or corresponding time period.

In the first WL filtering mode, in the WL bias filtering system 505 or WL ZD filter (400, 501), the WL bias for the corresponding satellite can be modeled as a constant WL bias that slowly varies over time for some epochs or in a conventional WL bias mode, wherein no discontinuity in WL bias occurs. Similarly, in the WL bias filter and the residual WL bias filter, the WL bias for a particular satellite may be modeled as a constant WL bias that varies slowly over time for some epochs (e.g., WL bias varies within 0.1 to 0.2 cycles at WL frequency) if the particular satellite does not support (e.g., dynamic) transmit power adjustment of the L1 or L2C carrier signal to compensate for interference. However, even if the particular satellite does not support transmit power adjustment, the particular satellite may have WL bias discontinuities or variations in WL bias per unit time exceeding a threshold rate of change of WL bias per unit time due to switching of thermal variations within different signal paths, satellite transmitters, or hardware portions of any satellite.

More generally, in a second WL filtering mode (e.g., time-varying WL mode), the WL bias filter and supplemental WL bias prediction filter (504) (e.g., residual WL bias filter), the WL bias is modeled as a time-varying WL bias (e.g., varying by up to about 0.3 to 0.5 cycles, inclusive of the present value, at WL frequency or wavelength), at least for a transient interval of one or more epochs, such as when the transmit power of the satellite is adjusted due to interference, to thereby switch to a redundant transmitter of the satellite (e.g., having different hardware or thermal characteristics), or to change a signal path associated with a transmit antenna or antenna array of the satellite. The WL deviation prediction filter is supplemented during a second WL filter mode or a transition from the first WL filter mode to the second WL filter mode (504).

Fig. 16A is an illustrative graph comparing the relative signal-to-noise ratio versus time on a common time axis 902 for a satellite transmitting an L1 (P (Y) signal 904 encoded with a pseudo-random noise code (PN) and an L1 CA signal 903 encoded with a coarse acquisition code, where a correction signal may be applied to account for such instantaneous changes in bias.

Fig. 16B is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting a PN encoded L2 (P (Y) signal 906 and an L2C signal 905 on a common time axis 902, where a correction signal may be applied to account for such transient variations in bias.

Fig. 17A is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting either an L1 (P (Y) signal 904 or an L2 (P (Y) signal 906 encoded with a PN and an L1 CA signal 903 encoded with a coarse acquisition code on a common time axis 902, where a correction signal may be applied to account for such transient variations in bias.

Fig. 17B is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting a PN encoded L2 (P (Y) signal 906 and an L2C signal 905 on a common time axis 902, where a correction signal may be applied to account for such transient variations in bias.

Fig. 18A is an illustrative graph comparing the relative signal-to-noise ratio versus time for a satellite transmitting either an L1 (P (Y) signal 904 or an L2 (P (Y) signal 906 encoded with a PN and an L1 CA signal 903 encoded with a coarse acquisition code on a common time axis 902, where a correction signal may be applied to account for such transient variations in bias.

Fig. 18B is an illustrative graph comparing the relative signal-to-noise ratio of satellites transmitting PN coded L2 (P (Y) 906 and L2C signals 905 over a common time axis 902, where correction signals may be applied to account for such transient variations in bias.

Fig. 19A is an illustrative chart showing differential code bias 910 for a P1 signal versus L1CA signal associated with a change in power of a carrier or encoded portion of the corresponding signal over time 912. The vertical axis shows differential code bias 910 (e.g., in meters) and the horizontal axis indicates time 912 in military or 24 hour time format. One illustrative example of an estimated differential code deviation 911 is shown in solid lines, while the raw differential code deviation measurement 914 is shown as a dot or dashed line region.

Fig. 19B is an illustrative chart showing differential code bias 916 for a P2 signal versus L2C signal associated with a change in power of a carrier or encoded portion of a corresponding signal over time 912. The vertical axis shows differential code bias 910 (e.g., in meters) and the horizontal axis indicates time 912 in military or 24 hour time format. One illustrative example of an estimated differential code deviation 911 is shown in solid lines, while the raw differential code deviation measurement 916 is shown as a dot or dashed line region.

Fig. 20A shows WL bias estimate 920 versus time for a WL filter system operating in a first WL filter mode. The vertical axis shows WL deviation 917 (e.g., in cycles at WL frequency or fractional cycles). Meanwhile, the horizontal axis shows time 919 in days.

Fig. 20B shows WL bias 923 (e.g., WL bias estimate) versus time 919 and WL residual 921 versus time 919 for a WL filter system operating in the second WL filter mode. Similar to fig. 13A, on the vertical axis, fig. 13B shows WL bias 917 (e.g., in cycles at WL frequency or fractional cycles). Note that the spike 922 in the WL residual 921 (e.g., WL residual spike or WL transient peak) is consistent with a relatively rapid deviation (and corresponding near vertical slope, as shown, at least for the time scale of fig. 13B) in the WL deviation 923 (e.g., WL deviation estimate).

Fig. 21A shows NL bias 918 versus time 919 for an NL filter system (e.g., 507) operating in a first NL filter mode. The vertical axis shows NL deviation 918 (e.g., in cycles at NL frequency or fractional cycles). Meanwhile, the horizontal axis shows time 919 in days. For example, estimated NL bias 930 (e.g., in terms of periods or fractional periods at NL frequencies or their corresponding NL wavelengths) over a two-day period is shown.

Fig. 21B shows NL bias 932 versus time 934 and code phase bias 932 versus time 934 for an NL filter system (e.g., 507) operating in a second NL filter mode. Similar to fig. 14A, on the vertical axis, fig. 14B shows NL bias 932 (e.g., in cycles or fractional cycles at NL frequency) and code phase bias (e.g., in meters). Note that the code phase deviation 937 (which may be expressed in meters, for example) may provide a leading indicator (e.g., an indicator that leads in time) or leading correlation/covariance for a consistent or later epoch that follows earlier changes in the code phase deviation 937, accompanied by later changes in the estimated NL deviation 935. As shown, for example, estimated NL bias (e.g., in terms of periods or fractional periods at NL frequencies or their corresponding NL wavelengths) over a period of two days is shown.

In alternative embodiments that are separate from or additive to any other embodiment (e.g., of the NL bias filtering system 507), the WL bias filtering system 505, any of the WL bias filtering system 505's constituent kalman filters, can estimate WL filter bias to fall within limits or constraints based on the following kalman filter states, factors, and constraints: (1) One or more prior satellite WL deviations during the sampling period, e.g., a lowest satellite WL deviation value or a highest WL deviation value within a range of satellite WL deviations during the sampling period; (2) A series of nearest WL bias values preceding the current WL bias value; (3) Average, mode, median, probability density function, and/or mean of a priori satellite WL bias during a sampling period of one or more epochs; (4) The difference, limit, upper limit, lower limit, maximum deviation or maximum range between the estimated satellite WL bias and the last a priori satellite WL bias or the average or mean of a priori satellite WL bias during the sampling period; (5) Maximum rate of change of satellite WL bias or satellite final WL bias from average or mean; and/or (6) a revised limit that replaces the limit (previous limit) and is commensurate with any increased dynamic noise threshold. Thus, in one embodiment, the WL bias filtering system applies the above-described limitations or constraints in the estimation of WL bias values (e.g., in combination with transient conditions associated with power variations or code phase bias variations in the carrier signal and the encoded components).

In yet another alternative embodiment, separate or additive from any other embodiment, the WL bias filtering system 505, any of the filters in the WL bias filtering system 505 that make up the kalman filter, may be constrained to a compensated WL bias or smoothed WL bias estimate, comprising a series of interconnected and continuous polynomial segments (e.g., a maximum slope constraint or a maximum curve constraint consistent with a history or (artificial intelligence) training of transient conditions associated with previous power changes or code phase bias changes of the carrier signal and code components).

In yet another alternative embodiment, separate from or additive to any other embodiment, after detecting a discontinuity in the residual WL bias (e.g., a temporally aligned discontinuity in the carrier phase measurements at the one or more reference receivers 130 providing the reference phase measurements to the data processing center 118), the WL bias filtering system 505, any of the WL bias filtering system 505 constituent kalman filters may adjust or estimate the compensated WL bias (e.g., for transient conditions associated with power changes or code phase bias changes in the carrier signal and code components) by reducing, increasing, constraining, limiting, or fixing the maximum rate of change per unit time of WL bias observed over some period (e.g., one or more epochs).

In alternative embodiments, separate or additive from any other embodiment, the NL bias filter system 507, any of the filters in the NL bias filter system 507's constituent kalman filters, can estimate the NL filter bias to fall within limits or constraints based on the following kalman filter states, factors, and constraints: (1) One or more a priori satellite NL offsets during the sampling period, e.g., a lowest NL offset value or a highest NL offset value within a range of NL offsets during the sampling period; (2) A series of recent satellite NL bias values preceding the current satellite NL bias value; (3) Average, mode, median, probability density function and/or mean of a priori satellite NL bias during a sampling period of one or more epochs; (4) The difference, limit, upper limit, lower limit, maximum deviation or maximum range between the estimated satellite NL bias and the last a priori satellite NL bias or the average or mean of the a priori satellite NL bias during the sampling period; (5) Maximum rate of change of satellite NL bias from the average or mean or the last satellite NL bias; and/or (6) a revised limit that replaces the limit (previous limit) and is commensurate with any increased dynamic noise threshold. Thus, in one embodiment, the NL bias filter system 507 applies the above-described limitations or constraints in the estimation of the NL bias value (e.g., in combination with transient conditions associated with power variations or code phase bias variations in the carrier signal and the code components).

In yet another alternative embodiment, separate or additive from any other embodiment, the NL bias filtering system 507, any of the filters making up the kalman filter of the NL bias filtering system 507, may be constrained to a compensated NL bias or smoothed NL bias estimate, comprising a series of interconnected and continuous polynomial segments (e.g., a maximum slope constraint or a maximum curve constraint consistent with a history or (artificial intelligence) training of transient conditions associated with previous power changes or code phase bias changes of the carrier signal and the code component).

In yet another alternative embodiment, separate from or additive to any other embodiment, after detecting a discontinuity in the residual NL bias (e.g., a temporally aligned discontinuity in the carrier phase measurements at the one or more reference receivers 130 providing the reference phase measurements to the data processing center 118), the NL bias filtering system 507, any of the NL bias filtering systems 507 comprising kalman filters, may adjust or estimate the compensated NL bias (e.g., for transient conditions associated with power changes or code phase bias changes in the carrier signal and code components) by reducing, increasing, constraining, limiting, or fixing the maximum rate of change per unit time of the NL bias observed over some period (e.g., or a number of epochs).

In one embodiment, the electronic data processor 20 or the clock solution module 44 (e.g., the orbit/clock satellite bias quality module) of the data processing center 118 is adapted to determine one or more quality metrics indicative of the quality of the GNSS satellite orbit solution, satellite clock, satellite wide-lane bias, and narrow-lane bias according to the following equations:

wherein:

N _sat is the number of satellites used;

N _ref is the number of reference positions used;

Q _i the (narrow lane departure quality value) may be one of the following:

=0 (use_sat_bias), if the satellite BIAS is not ready for use;

=l (poor_sat_bias), if the satellite BIAS is POOR and is removed in partial fixes;

=2 (good_sat_bias), if satellite BIAS is appropriate for SD fixing;

=3 (clear_sat_bias), if the satellite BIAS is very good and can be used as a reference satellite.

Further, in one configuration, the electronic data processor 20 of the data processing center 118, or the clock solution module 44 (e.g., an orbit/clock satellite bias quality module), or the correction manager 40, is configured to incorporate one or more of the quality metrics into the correction signals for the respective satellites.

Fig. 22 shows a flowchart of one embodiment (e.g., a first embodiment) of a method for providing a global satellite differential correction signal. The method starts at step S820 and includes improvements to providing global satellite correction signals.

In step S820, one or more of the electronic data processor (e.g., 20), the measurement preprocessing module 36, or filters thereof (400, 37, 501) determine a wide-lane fixed ambiguity for each satellite and a corresponding time-varying wide-lane bias based on adaptive estimates responsive to the tuned dynamic noise within the supplemental wide-lane bias prediction filter 504 for that satellite. For example, one or more of the electronic data processor (e.g., 20), the measurement preprocessing module 36, or filters thereof (400, 37, 501), such as WL bias filtering system 505, determine the time-to-widen lane bias according to the following equation:

/>

wherein,is a satellite wide-lane bias (one for each satellite for all receivers); />Is the L1 satellite code bias; />Is the L2 satellite code bias; />Is the L1 satellite phase bias; />Is the L2 satellite phase offset; />Is the L1 frequency; />Is the L2 frequency and wherein both satellite wide-lane bias and receiver wide-lane bias followThe time is not constant. Furthermore, the time-varying wide-lane bias may include additional wide-lane inter-frequency bias for the corresponding satellites and receivers, particularly for the GLONASS GNSS constellation.

In certain configurations of step S820, one or more of the electronic data processor (e.g., 20), the measurement preprocessing module 36, or the filters thereof (400, 37, 501) sets or establishes satellite WL bias applicable to each respective satellite of the first column WL ZD filters The WL bias is determined as follows:

inputting the fitted wide-lane residual errors of the corresponding satellites into a second column of complementary WL deviation prediction filters comprising WL Kalman filters;

for each satellite, determining, among a plurality of filters in the supplemental WL bias prediction filter, a variance of the wide-lane bias residual input at the corresponding measurement time (t)And

based on the sum of T (T _WL ) Defined tuned dynamic noise consistent with dynamic noise constraints (determined separately and independently for WL kalman filters of respective satellites)To estimate a second variance or a next variance associated with a sampling time (t+1) after the measurement time>

In step S822, an electronic data processor (e.g., 20); the orbit solution module 38 or a filter (404, 39, 506) thereof; or clock solution module 44 or its filter (408, 43, 506) determines a narrow lane for each satellite based on an adaptive estimate of the adaptive estimate responsive to tuned dynamic noise within the narrow lane bias/code phase bias filter 512 for the corresponding satelliteFixed ambiguity, satellite slow clock solution, and time-varying narrow lane bias. For example, an electronic data processor (e.g., 20); the orbit solution module 38 or a filter (404, 39, 506) thereof; or the clock solution module 44 or a filter (408, 43, 506) thereof, such as the NL bias filtering system 50, determines the time-varying narrow lane code bias according to the following equation

Wherein,is the narrow lane L1 code bias for receiver r and satellite s; />Is the narrow lane L2 code bias for receiver r and satellite s; />Is the L1 frequency; />Is the L2 frequency; />(e.g.)>) Representing a generic item for any given frequency; b (B) _r，NL Is the narrow lane code bias of the receiver r (one for each receiver and constellation for all satellites in view); />Is the narrow lane code bias of satellite s; and->Is an error in the refraction-corrected narrow lane code bias, such as an unmodeled error.

In certain configurations of step S822, an electronic data processor (e.g., 20); the orbit solution module 38 or a filter (404, 39, 506) thereof; or the clock solution module 44 or its filter (408, 43, 506) by setting or establishing satellite NL bias applicable to each respective satellite of the first column of NL ZD filtersOr satellite RC code phase deviation->To determine NL bias as follows: />

Inputting refraction corrected code (NL) residuals for the corresponding satellites to a second column NL bias filter/code phase filter comprising an NL kalman filter;

for each satellite, determining, among a plurality of NL bias/code phase filters, the variance of the narrow lane bias residual input at the corresponding measurement time (t) And

based on the sum of T (T _NL ) Defined tuned NL dynamic noise consistent with dynamic noise constraints (determined separately and independently for NL Kalman filters of respective satellites)Or->To estimate a second variance or a next variance associated with a sampling time (t+1) after the measurement time>

In step S824, the low latency clock module 52, the correction manager 40 or the electronic data processor 20 of the data processing center 118 provides correction signals including the widelane ambiguity, the time-varying widelane bias, and the narrow elane ambiguity and the time-varying narrow elane bias.

Step S824 may be performed according to various programs that may be applied individually or cumulatively. Under the first procedure, the low-latency clock module 52, the correction manager 40, or the electronic data processor 20 of the data processing center (18, 118) determine one or more quality metrics indicative of the quality of the GNSS satellite orbit solution, satellite clock, satellite wide-lane bias, and narrow-lane bias according to the following equations:

wherein:

N _sat is the number of satellites used;

N _ref is the number of reference positions used;

Q _i the (narrow lane departure quality value) may be one of the following:

=0 (use_sat_bias), if the satellite BIAS is not ready for use;

=2 (good_sat_bias), if satellite BIAS is appropriate for SD fixing;

In a second procedure, the data processor 20, low delay clock module 52 or correction manager 40 incorporates the quality indicator into the correction signal for one or more epochs or sampling intervals.

In a third procedure, the data processor 20, low delay clock module 52 or correction manager 40 incorporates the correction signal alone or with the code phase deviation data or code deviation data to compensate for one or more of the following conditions or states: (1) A first state in which the time-to-width lane offset is caused by an observed change in the received signal-to-noise ratio or carrier-to-noise density ratio of the encoded L1C/a or L2C (or L2C or L5) satellite signal of the satellite transmitter; (2) A second state in which the time-varying narrow lane bias is caused by an observed change in the received signal-to-noise ratio or carrier-to-noise density ratio of the encoded L1C/a or L2C (or L2C or L5) satellite signal of the satellite transmitter; (3) A third state in which the time-to-wide lane offset and the narrow lane offset are caused by thermally-related inter-frequency clock offsets (e.g., different hardware signal paths in the satellite, or different thermal environments of different transmitter signal paths within the satellite).

The embodiment of the method of fig. 23 is similar to the embodiment of the method of fig. 22, except that step S824 is replaced with step S825 in fig. 23. Like reference numerals in fig. 22 and 23, or in any set of drawings in this regard, refer to like elements or features.

In step S825, the low delay clock module 52, the correction manager 40, or the electronic data processor 20 of the data processing center 118 provides correction signals including widelane ambiguities, time-widelane ambiguities and narrow-lane ambiguities, and time-varying narrow-lane ambiguities and code-deviations (or code-phase-deviations) for each satellite (of the corresponding GNSS constellation) within reception range of the mobile receiver 12 for one or more epochs. Generally, as a practical matter, the code bias (of the pseudorandom noise code encoding any carrier of the GNSS signal component) is related or can be related to the phase bias (e.g., of the satellite clock bias component) of the corresponding carrier phase of the GNSS signal component, e.g., due to or associated with a change in the elastic power or the transmit power of the corresponding carrier phase of the GNSS signal component; thus, in this disclosure, the term "code phase bias" may be used to indicate such correlation.

Further, in alternative embodiments, step S825 may replace step S824 in any of the methods or flowcharts in the present disclosure.

The embodiment of the method of fig. 24 is similar to the embodiment of the method of fig. 22, except that step S820 is replaced with step S821 in fig. 24. Like reference numerals in fig. 22 and 24, or in any set of drawings in this regard, refer to like elements or features.

In step S821, the electronic data processor (e.g., 20), the measurement preprocessing module 36, or one or more of its filters (400, 37, 501) determine a wide-lane fixed ambiguity for each satellite and a corresponding time-varying wide-lane bias based on the adaptive estimates responsive to the tuned dynamic noise within the supplemental wide-lane bias prediction filter 504 for that satellite: (A) The wide-lane homodyne filter provides a wide-lane residual error for a complementary wide-lane deviation prediction filter of a given satellite; (B) Wherein the supplemental wide-lane bias prediction filter provides tuning dynamic noise (e.g., refined, tuning dynamic noise) to the wide-lane homodyne filter of a given satellite based on the wide-lane residual provided for the respective satellite.

Further, in alternative embodiments, step S821 may replace step S820 in any method or flow chart in the present disclosure.

In an alternative embodiment of step S821, the determination of the wide lane fix ambiguity and wide lane bias for each satellite includes:

(A) Establishing a pair of wide-lane homodyne filters and a complementary wide-lane bias prediction filter for each satellite;

(B) Providing, by the wide-lane homodyne filter 501, wide-lane residuals to a pair of corresponding supplemental wide-lane bias prediction filters for a given satellite; and

(C) Tuning dynamic noise is provided to the wide-lane homodyne filter for a given satellite based on the wide-lane residual provided for the respective satellite by supplementing the wide-lane bias prediction filter (e.g., 504), wherein the pair of WL ZD filters 501 and the supplemental WL bias prediction filter (e.g., 504) for the given satellite include software instructions and/or kalman filter logic structures associated with the preprocessing module 36 within the data processing center.

The embodiment of the method of fig. 25 is similar to the embodiment of the method of fig. 22, except that step S822 is replaced with step S823 in fig. 25. Like reference numerals in fig. 22 and 25, or in any set of drawings in this regard, refer to like elements or features.

In step S823, an electronic data processor (e.g., 20); the orbit solution module 38 or a filter (404, 39, 506) thereof; or the clock solution module 44 or its filter (408, 43, 506) determines the slot fix ambiguity, the satellite slow clock solution, and the time-varying slot bias for each satellite based on adaptive estimates of adaptive estimates responsive to tuned dynamic noise within the slot bias/code phase bias filter 512 for the corresponding satellite: (A) Wherein the narrow lane homodyne filter provides refraction corrected code residuals to a corresponding narrow lane filter/code bias filter for a given satellite; (B) Wherein the narrow-lane filter/code bias filter provides tuning dynamic noise (e.g., refined tuning dynamic noise) to the narrow-lane homodyne filter of a given satellite based on the refraction-corrected code residuals provided for the respective satellite.

Further, in alternative embodiments, step S823 may replace step S822 in any method or flowchart in the present disclosure.

In an alternative embodiment of step S823, the determination of the lane-fixing ambiguity and lane-bias for each satellite includes:

(A) Establishing a pair of narrow lane homodyne filters 506 and a supplemental narrow lane bias prediction filter for each satellite;

(B) Providing refraction-corrected code residuals to a pair of corresponding narrow lane filters/code bias filters (e.g., 512) for a given satellite by narrow lane homodyne filter 506;

(C) Tuning dynamic noise is provided to the narrow lane homodyne filter for a given satellite by a narrow lane filter/code bias filter (e.g., 512) based on the refraction-corrected code residuals provided for the respective satellite, wherein the pair of NL ZD filter 506 and the narrow lane bias filter/code bias filter (e.g., supplemental NL bias prediction filter 512) for each satellite include software instructions and/or kalman filter logic structures associated with the orbit solution module 38 or clock solution module 44 within the data processing center for the given satellite.

The embodiment of the method of fig. 26 is similar to the embodiment of the method of fig. 22, except that step S826 is added in fig. 26. Like reference numerals in fig. 22 and 26, or in any set of drawings in this regard, refer to like elements or features.

In step S826, one or more of the electronic data processor (e.g., 20), the measurement preprocessing module 36, or the filters thereof (400, 37, 501, 504) determine whether the next wide lane bias correction for the given satellite is operating in the first Wide Lane (WL) filter mode or the second WL filter mode at the next epoch (or a subsequent transient period): (A) Wherein the first WL filter mode (of one or more of WL filters 400, 501, 504) is within a corresponding first WL threshold (e.g., a first limited deviation) associated with a previous mean of WL deviations from a series of previous consecutive epochs (prior to a next epoch); or (B) wherein a second WL filter pattern (of one or more of WL filters 400, 501, 504) has a corresponding second WL threshold (e.g., a second limited offset) associated with the previous mean that is greater than the first WL threshold (e.g., a first limited offset) for each corresponding satellite.

Further, in alternative embodiments, step S826 may be added in any method or flowchart of the present disclosure.

The embodiment of the method of fig. 27 is similar to the embodiment of the method of fig. 22, except that step S828 is added in fig. 27. Like reference numerals in fig. 22 and 27, or in any set of drawings in this regard, refer to like elements or features.

In step S828, an electronic data processor (e.g., 20); the orbit solution module 38 or a filter (404, 39, 506) thereof; or the clock solution module 44 or its filter (408, 43, 506, 512) determines whether the next lane offset correction for a given satellite is operating in the first lane (NL) filter mode or the second NL filter mode at the next epoch (or a transient period thereafter): (A) Wherein the first NL filtering mode (of the one or more NL filters 408, 43, 506, 512) is within a corresponding first NL threshold (e.g., a first limited offset) related to the previous average of NL offsets from a series of previous consecutive epochs (prior to the next epoch); or (B) wherein a second NL filter mode (of the one or more NL filters 408, 43, 506, 512) has a corresponding second NL threshold (e.g., a second limited offset) related to the previous mean, the second NL threshold being greater than the first NL threshold (e.g., the first limited offset) for each corresponding satellite.

Further, in alternative embodiments, step S828 may be added to any method or flowchart of the present disclosure, either alone or cumulatively with step S826.

The embodiment of the method of fig. 28 is similar to the embodiment of the method of fig. 22, except that steps S830 and S832 are added in fig. 21. Like reference numerals in fig. 22 and 28, or any set of drawings in this regard, refer to like elements or features.

In step S830, for each satellite within the reception range, a supplemental WL bias prediction filter determines a derived metric (T) based on the WL residual and its variance _WL ) Or is configured to determine a derived metric (T) based on the WL residual and its variance _WL )。

In step S832, if the first WL tuning dynamic noise is greater than the derived metric, one or more of the supplemental WL bias prediction filters 504 provide or are configured to provide a second WL tuning dynamic noise (as a transition or a next setting for a next epoch) to the WL ZD filter 501 to replace the first WL tuning dynamic noise of the WL ZD filter 501 (as a current setting for a current epoch), wherein the second WL tuning dynamic noise is greater than the first WL tuning dynamic noise, and wherein the WL residual and corresponding resolved WL ambiguity associated with the WL bias are related to a predetermined satellite.

Further, in alternative embodiments, steps S830 and S832 may be added to any method or flowchart of the present disclosure.

In another alternative embodiment, one or more supplemental WL bias prediction filters 504 (e.g., or electronic data processor 20 of the data processing center) determine or are configured to determine a derived metric (T _WL ). In addition, or separately from step S832, if the first WL tuned dynamic noise is equal to or less than the derived metric, then the WL bias prediction filter 504 is supplemented withIs provided to WL ZD filter 501 or is configured to provide first WL tuned dynamic noise (as a maintenance of the next setting of the next epoch) to WL ZD filter 501 to maintain first WL tuned dynamic noise (as a current setting of the current epoch) of WL ZD filter 501, and wherein WL residuals associated with WL bias and corresponding resolved WL ambiguities are related to predetermined satellites.

The embodiment of the method of fig. 29 is similar to the embodiment of the method of fig. 22, except that steps S834 and S836 are added in fig. 29. Like reference numerals in fig. 22 and 29, or in any set of drawings in this regard, refer to like elements or features.

In step S834, for each satellite within reception range, one or more supplemental NL bias filters/code phase filters 512 determine a derived metric (T) based on the refraction-corrected code residuals (e.g., RC code phase bias residuals) and their variances _NL ) Or configured to determine a derived metric (T) based on the refraction-corrected code residuals (e.g., RC code phase bias residuals) and their variances _NL ). The RC code residuals or RC code phase bias residuals may be referred to individually or collectively as NL residuals. In certain illustrative embodiments, a supplemental NL bias prediction filter (e.g., a kalman filter or an extended kalman filter) may be used to implement the NL bias filter/code phase filter.

Step S836 may be performed in accordance with various techniques or modifications that may be applied singly or cumulatively in combination with step S834 mentioned above.

Under the first technique for performing step S836, if the first NL tuning dynamic noise is greater than the derived metric, one or more of the NL bias filters/code phase filters 512 provide or are configured to provide a second NL tuning dynamic noise (as a transition or a next setting for a next epoch) to the NL ZD filter 506 to replace the first NL tuning dynamic noise of the NL ZD filter 506 (as a current setting for a current epoch), wherein the second NL tuning dynamic noise is greater than the first NL tuning dynamic noise, and wherein the RC code residuals associated with NL bias and corresponding resolved NL ambiguities are related to predetermined satellites.

Under the second technique for performing step S836, if the first NL tuning dynamic noise is less than or equal to the derived metric, then one or more NL bias filters/code phase filters 512 are provided to the NL ZD filter 506 or are configured to provide the first NL tuning dynamic noise (as the next setting for the next epoch) to the NL ZD filter 506 to maintain the first NL tuning dynamic noise of the NL ZD filter 506 (as the current setting for the current epoch), wherein the RC code residuals associated with NL bias and corresponding resolved NL ambiguities are related to predetermined satellites.

Under a third technique for performing step S836, if the first NL tuning dynamic noise is greater than the derived metric, one or more NL bias filters/code phase filters 512 (e.g., to the NL ZD filter 506, to the low delay clock module, or to the correction module 52) are provided or configured to provide code phase biases (e.g., refraction corrected code phase biases) for incorporation into (or enhancement) correction signals, wherein the RC code residuals associated with the NL biases and corresponding resolved NL ambiguities are associated with predetermined satellites. By selectively not combining or enhancing correction signals with code phase bias under the third technique, wireless bandwidth or satellite channel bandwidth may be saved, such as for increasing maximum wireless communication channel (e.g., physical or virtual) throughput or maximum transmission rate of other correction data.

Under a fourth technique for performing step S836, if the first NL tuning dynamic noise is less than or equal to the derived metric, then one or more NL bias filters/code phase filters 512 provide or optionally preserve (withhold) (or are configured to provide or optionally preserve, for example, to the NL ZD filter 506, to the low delay clock module, or to the correction module 52) code phase bias (e.g., refraction corrected code phase bias) that is not incorporated into the correction signal, wherein the RC code residuals associated with the NL bias and corresponding resolved NL ambiguities are related to a predetermined satellite.

Further, in alternative embodiments, steps S834 and S836 may be added in any method or flowchart of the present disclosure, either alone or in addition to steps S830 and S832.

The embodiment of the method of fig. 30 is similar to the embodiment of the method of fig. 22, except that steps S840 and S842 are added in fig. 30. Like reference numerals in fig. 22 and 30, or in any set of drawings in this regard, refer to like elements or features.

In step S840, one or more of the NL bias filter/code phase filters 512 determines whether the estimated refraction-corrected code phase bias meets or exceeds a code phase bias threshold. In addition, the master NL ZD filter 506 applies satellite NL bias dynamic noise 510 associated with code phase biases that meets or exceeds code phase bias thresholds to determine resolved or fixed NL ambiguities and associated NL biases and associated code biases or code phase biases.

In step S842, if the code phase deviation meets or exceeds the code phase deviation threshold, the data processor 20, the low delay clock module 52, or the correction manager 40 incorporates the refraction-corrected code deviation or code phase deviation into the correction signal.

In alternative embodiments, steps S840 and S842 may be modified collectively as follows: first, in an alternative step S840, if the code bias meets or exceeds the code bias threshold, the data processor 20, the low delay clock module 52, or the correction manager 40 incorporates the refraction-corrected code bias into the correction signal. Next, in an alternative step S842, if the code bias meets or exceeds the code bias threshold, the data processor 20, the low delay clock module 52, or the correction manager 40 incorporates the refraction-corrected code bias or code phase bias into the correction signal.

The foregoing description, for purposes of explanation, has been described with reference to specific embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated.

Claims

1. A method for providing a global satellite differential correction signal, wherein the method comprises an improvement comprising:

determining a wide-lane fixed ambiguity for each satellite and a corresponding time-varying wide-lane bias based on adaptive estimates responsive to tuning dynamic noise in a supplemental wide-lane bias prediction filter for each satellite as a current setting of a corresponding current epoch;

determining a slot fixed ambiguity, a satellite slow clock solution, and a time-varying slot bias for each satellite based on an adaptive estimate of an adaptive estimate responsive to another tuned dynamic noise within the slot bias/code phase bias filter for the corresponding current epoch; and

Providing a correction signal for each satellite comprising the widelane ambiguity, the time-varying widelane bias, and the narrow elane ambiguity and the time-varying narrow elane bias, wherein the correction signal further comprises a code bias or code phase bias for one or more epochs for each satellite.

2. The method of claim 1, further comprising:

estimating test statistics based on a ratio of wide-lane bias residuals for respective satellites to square roots of variances of the wide-lane bias residuals; and

determining whether the estimated test statistic is greater than a specified threshold of dynamic noise for the corresponding satellite; and

if the estimated test statistic is determined to be greater than a specified threshold of the dynamic noise for the corresponding satellite, the tuned dynamic noise is increased.

3. The method of claim 2, wherein the tuned dynamic noise of the wide lane departure prediction filter increases from a first tuned dynamic noise range to a second tuned dynamic noise range.

4. The method of claim 1, further comprising:

estimating test statistics based on a ratio of a narrow lane bias residual of a respective satellite to a square root of a variance of the narrow lane bias residual; and

if the estimated test statistic is determined to be greater than the specified threshold for the dynamic noise for the corresponding satellite, the tuned dynamic noise is increased.

5. The method of claim 4, wherein the tuned dynamic noise of the narrow lane departure prediction filter increases from a first tuned dynamic noise range to a second tuned dynamic noise range for the respective satellite.

6. The method of claim 1, wherein determining, in a preprocessing module of a clock and orbit estimation module within a data processing center, the wide lane fix ambiguity and wide lane bias for each satellite estimated by a pair of wide lane homodyne filters and the supplemental wide lane bias prediction filter for each satellite further comprises:

providing, by the wide-lane homodyne filter, wide-lane residuals to a pair of corresponding supplemental wide-lane bias prediction filters for a given satellite;

the tuning dynamic noise is provided to the wide-lane homodyne filter for the given satellite based on the wide-lane residuals provided for the respective satellite by the supplemental wide-lane bias prediction filter.

7. The method of claim 1, wherein determining the lane fix ambiguity and lane bias for each satellite estimated by a pair of lane homodyne filters and the lane bias filter/code bias filter for each satellite in a clock solution module of a clock and orbit estimation module within a data processing center further comprises:

providing refraction corrected code residuals by the narrow lane homodyne filter to a pair of corresponding narrow lane filters/code bias filters for a given satellite;

the tuning dynamic noise is provided to the narrow lane homodyne filter of the given satellite based on the refraction-corrected code residuals provided for the respective satellite by the narrow lane filter/code bias filter.

8. The method of claim 1, further comprising:

determining whether to operate in a first wide-lane filtering mode in which a next wide-lane offset correction for a given satellite for a next epoch has a corresponding first WL threshold associated with a previous mean of WL offsets from a series of previous consecutive epochs, or a second wide-lane filtering mode in which a next wide-lane offset correction for the given satellite for the next epoch or a transient period thereafter has a corresponding second WL threshold associated with the previous mean that is greater than the first WL threshold (or exceeds the limit of the first WL threshold) for each corresponding satellite.

9. The method of claim 1, further comprising:

determining whether to operate in a first narrow lane filtering mode in which a next narrow lane offset correction for a given satellite for a next epoch has a corresponding first NL threshold associated with a previous mean of NL offsets from a series of previous consecutive epochs, or a second narrow lane filtering mode in which a next narrow lane offset correction for the given satellite for the next epoch or a transient period thereafter has a corresponding second NL threshold associated with the previous mean that is greater than the first NL threshold (or exceeds the limit of the first NL threshold) for each corresponding satellite.

10. The method of claim 1, further comprising:

for each satellite in the reception range, determining a derived metric (T) based on the WL residual and its variance by supplementing the WL bias prediction filter (or an electronic data processor of the data processing center) _WL )；

Providing a second WL tuned dynamic noise (as a transition or as a next setting for a next epoch) to the WL ZD filter by the supplemental WL deviation prediction filter to replace the first WL tuned dynamic noise of the WL ZD filter (as a current setting for a current epoch) if the first WL tuned dynamic noise is greater than the derived metric, wherein the second WL tuned dynamic noise is greater than the first WL tuned dynamic noise, and wherein the WL residual and corresponding resolved WL ambiguity associated with WL deviation are correlated with a predetermined satellite.

11. The method of claim 1, further comprising:

for each satellite in the reception range, determining a derived metric (T) based on the refraction-corrected code residuals and their variances by supplementing an NL bias filter/code phase filter _NL )；

Providing a second NL tuning dynamic noise (as a transition or as a next setting for a next epoch) to the NL ZD filter by the NL bias filter/code phase filter if the first NL tuning dynamic noise is greater than the derived metric to replace the first NL tuning dynamic noise of the NL ZD filter (as a current setting for a current epoch), wherein the second NL tuning dynamic noise is greater than the first NL tuning dynamic noise, and wherein the RC code residuals associated with NL bias and corresponding resolved NL ambiguities are associated with predetermined satellites.

12. The method of claim 1, further comprising:

for each satellite in the reception range, determining a derived metric (T) based on the WL residual and its variance by supplementing a WL bias prediction filter _WL )；

Providing, by the supplemental WL-bias prediction filter, a first WL-tuned dynamic noise (as a maintenance of a next setting for a next epoch) to the WL ZD filter if the first WL-tuned dynamic noise is equal to or less than the derived metric, to maintain the first WL-tuned dynamic noise of the WL ZD filter (as a current setting for a current epoch), and wherein the WL residual and corresponding resolved WL ambiguity associated with WL bias are related to a predetermined satellite.

13. The method of claim 1, further comprising:

for each satellite in the reception range, determining a derived metric (T) based on the refraction-corrected code residuals (RC code residuals) and their variances by supplementing NL bias filters/code phase filters _NL )；

Providing a first NL tuning dynamic noise (as the next setting of the next epoch) to the NL ZD filter through the NL bias filter/code phase filter if the first NL tuning dynamic noise is less than or equal to the derived metric to maintain the first NL tuning dynamic noise of the NL ZD filter (as the current setting of the current epoch), wherein the RC code residuals associated with NL bias and corresponding resolved NL ambiguities are related to a predetermined satellite.

14. The method of claim 1, further comprising:

for each satellite in the reception range, determining a derived metric (T) based on the refraction-corrected code residuals (e.g., RC code phase residuals) and their variances by supplementing NL bias filter/code phase filter _NL )；

Providing the code phase bias to the NL ZD filter by the NL bias filter/code phase filter and incorporating the code phase bias into the correction signal if the first NL tuning dynamic noise is greater than the derived metric, wherein the RC code residual and corresponding resolved NL ambiguity associated with NL bias are correlated with a predetermined satellite.

15. The method of claim 1, wherein determining the WL bias further comprises:

the satellite WL bias applicable to the first column WL ZD filter for each corresponding satellite is set by the following stepsIs described in (1) tuning dynamic noise:

determining, for each satellite, a variance of wide-lane bias residuals input at a corresponding measurement time (t) in a plurality of supplemental WL bias prediction filters of the supplemental WL bias prediction filtersAnd

based on the sum of T (T _WL ) Defined tuning dynamic noise consistent with dynamic noise constraints (determined separately and independently for the WL kalman filter of the respective satellite)To estimate a second variance or a next variance associated with a sampling time (t+1) after said measurement time>

16. The method of claim 1, wherein determining the NL bias further comprises:

the satellite NL bias applicable to the first column NLZD filter for each corresponding satellite is set by the following stepsOr satellite RC code phaseDeviation->Is described in (1) tuning dynamic noise:

determining, for each satellite, in a plurality of said NL bias filters/code phase filters, the variance of the narrow lane bias residual inputted at the corresponding measurement time (t)And

based on the sum of T (T _NL ) Defined tuned NL dynamic noise consistent with dynamic noise constraints (determined separately and independently for said NL kalman filter of the respective satellite)Or->To estimate a second variance or a next variance associated with a sampling time (t+1) after said measurement time>

17. A system for providing a global satellite differential correction signal, wherein the system comprises:

a wide-lane bias filtering system configured to determine a wide-lane fixed ambiguity for each satellite and a corresponding time-varying wide-lane bias based on adaptive estimates responsive to tuning dynamic noise within a supplemental wide-lane bias prediction filter for each satellite;

a narrow lane bias filtering system configured to determine a narrow lane fixed ambiguity, a satellite slow clock solution, and a time-varying narrow lane bias for each satellite based on adaptive estimates of adaptive estimates responsive to tuning dynamic noise within a narrow lane bias/code phase bias filter for the corresponding satellite, wherein the narrow lane bias/code phase bias is configured to estimate a code bias or a code phase bias; and

A correction data estimator configured to provide a correction signal for each satellite comprising the widelane ambiguity, the time-varying widelane bias, and the narrow-lane ambiguity and the time-varying narrow-lane bias, wherein the correction signal further comprises the code bias or the code phase bias for one or more epochs for each satellite.

18. The system of claim 17, wherein the wide lane offset filtering system further comprises:

a pair of wide-lane homodyne filters and a supplemental wide-lane bias prediction filter for each satellite;

the wide lane homodyne filter is configured to provide a plurality of wide lane residuals to a pair of corresponding supplemental wide lane bias prediction filters for a given satellite;

the supplemental wide-lane bias prediction filter is configured to provide the tuning dynamic noise to the wide-lane homodyne filter of the given satellite based on a wide-lane residual provided for the respective satellite.

19. The system of claim 17, wherein the narrow lane departure filtering system further comprises:

a pair of narrow lane homodyne filters and a narrow lane bias filter/code bias filter for each satellite;

the narrow lane homodyne filter is configured to provide a plurality of refraction-corrected code residuals to a pair of corresponding narrow lane filter/code bias filters for a given satellite;

The narrow lane filter/code bias filter is configured to provide the tuning dynamic noise to the narrow lane homodyne filter of the given satellite based on refraction-corrected code residuals provided for the respective satellite.

20. The system of claim 17, wherein the wide lane offset filtering system is configured to determine the time-wide lane offset according to the following equation:

wherein:

is a satellite wide-lane bias (one for each satellite for all receivers); />Is the L1 satellite code bias; />Is the L2 satellite code bias; />Is the L1 satellite phase bias; />Is the L2 satellite phase offset; />Is the L1 frequency;is the L2 frequency and wherein both the satellite wide lane offset and the receiver wide lane offset are not constant over time.

21. The system of claim 17, wherein the narrow lane departure filterThe wave system is configured to determine the time-varying narrow lane code bias according to the following equation

Wherein:

is the narrow lane L1 code bias for receiver r and satellite s; />Is the narrow lane L2 code bias for receiver r and satellite s; />Is the L1 frequency; />Is the L2 frequency; />(e.g.)>) Represents a generic term for any given frequency; b (B) _r，NL Is the receiver r narrow lane code bias (one for each receiver and constellation for all satellites in view); />Is satellite s narrow lane code deviation; and->Is an error in the refraction-corrected narrow lane code bias, such as an unmodeled error.