User-aided signal line-of-sight (LOS) machine learning classifier
Machine learning techniques can be used to mitigate multipath in a GNSS receiver that includes a first trained model that provides extra path length (EPL) corrections in the GNSS receiver. The first trained model can be updated using an updated and trained model from one or more assistance servers that are in communication with the GNSS receiver. The GNSS receiver can provide, for a particular computed position and time, extracted features from received GNSS signals to the one or more assistance servers. The assistance servers can then use the extracted features and a source of true EPL corrections (e.g., from a 3D building map database for the particular computed position and time) to train a server model. The server model, once trained to a desired level of accuracy, can be transmitted to the GNSS receiver to replace the first trained model.
1. A method of operating a global navigation satellite system (GNSS) receiver, the method comprising:
receiving GNSS signals from GNSS satellites (SVs) at a first time and determining one or more pseudoranges to one or more GNSS SVs;
extracting, from the received GNSS signals, features for use as inputs for a first trained model that provides extra path length (EPL) corrections;
computing an output from the first trained model based on the extracted features;
computing a first position from the received GNSS signals, the first position based on the output from the first trained model and the one or more pseudoranges;
transmitting, to one or more first servers, the extracted features for use in obtaining an updated and trained model based on the extracted features, the computed first position and its associated first time;
receiving data representing the updated and trained model; and
storing the data representing the updated and trained model as a second trained model for use in providing EPL corrections when computing positions for the GNSS receiver.
2. The method as in claim 1 , wherein the method further comprises:
requesting a set of EPL corrections based on the first position and the first time.
3. The method as in claim 2 , wherein the method further comprises:
receiving, in response to the requesting, the set of EPL corrections based on the first position and the first time.
4. The method as in claim 3 , wherein the method further comprises:
transmitting the set of EPL corrections for use in obtaining the updated and trained model.
5. The method as in claim 4 , wherein the GNSS receiver discontinues use of the first trained model after storing the second trained model.
6. The method as in claim 1 , wherein the method further comprises:
receiving a further updated model based on a region in which the GNSS receiver operates, the further updated model trained for the region.
7. The method as in claim 1 , wherein each of the one or more pseudoranges is corrected by subtracting a corresponding EPL correction from a measured pseudorange, and wherein the measured pseudorange is measured in a delay locked loop in the GNSS receiver and wherein the corresponding EPL correction is derived based on outputs at the delay locked loop (DLL) which is a pseudorandom noise (PRN) tracking loop to track pseudoranges in the GNSS receiver and wherein the corresponding EPL correction corrects for multipath reflections of GNSS signals in an urban canyon that surrounds the GNSS receiver.
8. The method as in claim 7 , wherein the features comprise: (1) correlation vector features from a set of successive correlation vectors from correlators, in the GNSS receiver, for GNSS signals from a GNSS SV; (2) a mutual probability distribution function (pdf) data of relative delay and relative amplitude from each of the correlators; and (3) other features relating to the GNSS receiver or GNSS SV.
9. The method as in claim 8 , wherein the correlation vector features comprise one or more of: (1) relative amplitude of local maximum points in a correlation vector; (2) relative delay of peaks in a correlation vector; (3) a width of a strongest peak in a correlation vector; or (4) a number of strong peaks in a correlation vector.
10. The method as in claim 9 , wherein the other features comprise one or more of: (1) an elevation of a GNSS SV; (2) a signal to noise ratio (SNR) of a measured pseudorange; (3) a type of antenna used to receive GNSS signals; (4) a tracking mode of the GNSS receiver; or (5) a tracking loop configuration of the GNSS receiver.
11. A data processing system that comprises a global navigation satellite system (GNSS) receiver, the data processing system comprising:
an antenna to receive RF GNSS signals;
an RF receiving section coupled to the antenna;
an analog to digital (A/D) converter coupled to the RF receiving section to digitize received GNSS signals;
a GNSS processing system coupled to the A/D converter, the GNSS processing system to determine one or more pseudoranges to one or more GNSS SVs and to extract, from the received GNSS signals, features for use as inputs for a first trained model that provides extra path length (EPL) corrections; and the GNSS processing system to compute an output from the first trained model based on the extracted features; and the GNSS processing system to compute a first position from the received GNSS signals, the first position based on the output from the first trained model and the one or more pseudoranges; and
a set of one or more communication transceivers coupled to the GNSS processing system, the set of one or more communication transceivers to transmit, to one or more first servers, the extracted features for use in obtaining an updated and trained model based on the extracted features, the computed first position and its associated first time; and the set of one or more communication transceivers to receive data representing the updated and trained model; and the GNSS processing system to store the data representing the updated and trained model as a second trained model for use in providing EPL corrections when computing positions for the GNSS receiver.
12. The data processing system as in claim 11 , wherein the set of one or more communication transceivers request a set of EPL corrections based on the first position and the first time, and the set of one or more communication transceivers receive, in response to the request, the set of EPL corrections based on the first position and the first time; and the set of one or more communication transceivers transmit the set of EPL corrections for use in obtaining the updated and trained model.
13. A non-transitory machine readable medium storing executable program instructions which when executed by a data processing system cause the data processing system to perform a method, the method comprising:
receiving global navigation satellite system (GNSS) signals from GNSS satellites (SVs) at a first time and determining one or more pseudoranges to one or more GNSS SVs;
extracting, from the received GNSS signals, features for use as inputs for a first trained model that provides extra path length (EPL) corrections;
computing an output from the first trained model based on the extracted features;
computing a first position from the received GNSS signals, the first position based on the output from the first trained model and the one or more pseudoranges;
transmitting, to one or more first servers, the extracted features for use in obtaining an updated and trained model based on the extracted features, the computed first position and its associated first time;
receiving data representing the updated and trained model; and
storing the data representing the updated and trained model as a second trained model for use in providing EPL corrections when computing positions for the GNSS receiver.
14. The non-transitory machine readable medium as in claim 13 , wherein the GNSS receiver discontinues use of the first trained model after storing the second trained model.
15. The non-transitory machine readable medium as in claim 14 , wherein the method further comprises:
receiving a further updated model based on a region in which the GNSS receiver operates, the further updated model trained for the region.
16. The non-transitory machine readable medium as in claim 14 , wherein the features comprise: (1) correlation vector features from a set of successive correlation vectors from correlators, in the GNSS receiver, for GNSS signals from a GNSS SV; (2) a mutual probability distribution function (pdf) data of relative delay and relative amplitude from each of the correlators; and (3) other features relating to the GNSS receiver or GNSS SV.
17. The non-transitory machine readable medium as in claim 16 , wherein the correlation vector features comprise one or more of: (1) relative amplitude of local maximum points in a correlation vector; (2) relative delay of peaks in a correlation vector; (3) a width of a strongest peak in a correlation vector; or (4) a number of strong peaks in a correlation vector.
18. The non-transitory machine readable medium as in claim 17 , wherein the other features comprise one or more of: (1) an elevation of a GNSS SV; (2) a signal to noise ratio (SNR) of a measured pseudorange; (3) a type of antenna used to receive GNSS signals; (4) a tracking mode of the GNSS receiver; or (5) a tracking loop configuration of the GNSS receiver.