Granted Patent B2

US 12,366,451 · App. 17/785,342 · Granted Jul 22, 2025

Inventors: Sholeh Yasini (Sundbyberg, SE); Torbjörn Wigren (Uppsala, SE); Richard Wirén (Vantaa, FI); Juhani Kauppi (Espoo, FI)

Assignee: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

G01C21/16H04B7/18504

Abstract

Claims (48)

1. A method of monitoring a kinematic state of an unmanned aerial vehicle, UAV, the method comprising:

obtaining one or more predicted pathlosses between a UAV and one or more base stations at a first time instance, wherein the predicted pathlosses are determined using an estimate of a kinematic state of the UAV at the first time instance and one or more pathloss models developed using a machine-learning process;

obtaining one or more measurements of a pathloss between each of the one or more base stations and the UAV at the first time instance; and

re-determining the estimate of the kinematic state of the UAV at the first time instance based on the one or more predicted pathlosses and the one or more measurements of the pathloss.

2. The method of claim 1 , wherein each of the one or more pathloss models corresponds to a respective base station of the one or more base stations.

3. The method of claim 2 , wherein the UAV is a first UAV and each of the one or more pathloss models is developed using training data for its respective base station, the training data comprising:

a plurality of position measurements for at least one second UAV, and

a plurality of pathloss measurements of at least one pathloss between the at least one second UAV and the respective base station, each of the plurality of pathloss measurements being associated with a respective position measurement in the plurality of position measurements.

4. The method of claim 3 , wherein each of the plurality of position measurements comprise one or more of the following:

a distance between the at least one second UAV and the respective base station;

an altitude of the at least one second UAV; and

latitude and longitude of the at least one second UAV.

5. The method of claim 3 , wherein obtaining the one or more predicted pathlosses comprises:

obtaining the estimate of the kinematic state of the first UAV at the first time instance; and

inputting the estimate of the kinematic state of the first UAV at the first time instance into the one or more pathloss models to determine the one or more predicted pathlosses.

6. The method of claim 3 , wherein the one or more measurements of the pathloss between the one or more base stations and the first UAV comprise one or more measurements performed by the first UAV on one or more signals transmitted by the one or more base stations.

7. The method of claim 3 , wherein re-determining the estimate of the kinematic state of the first UAV at the first time instance comprises:

inputting state estimation data into an extended Kalman filter, to re-determine the estimate of the kinematic state at the first time instance, the state estimation data comprising the one or more predicted pathlosses and the one or more measurements of the pathloss.

8. The method of claim 7 , wherein the first UAV is modelled as being in one of a plurality of movement modes at the first time instance, each of the plurality of movement modes being associated with a respective extended Kalman filter in a plurality of extended Kalman filters, the method comprising:

inputting the state estimation data into the plurality of extended Kalman filters; and

combining one or more outputs of the plurality of extended Kalman filters to re-determine the estimate of the kinematic state of the first UAV at the first time instance according to an interacting-multiple-model filtering process.

9. The method of claim 8 , wherein the plurality of movement modes comprise one or more of the following:

a three-dimensional constant velocity movement Wiener process;

a three-dimensional constant acceleration movement Wiener process; and

a three-dimensional constant position Wiener process.

10. The method of claim 7 , wherein the state estimation data further comprises:

one or more travel time measurements for signals transmitted between the one or more base stations and the first UAV at the first time instance; and

one or more predicted signal travel times for signals transmitted between the first UAV and the one or more base stations at the first time instance, wherein the one or more predicted signal travel times are determined using the estimate of a kinematic state of the first UAV at the first time instance.

11. The method of claim 7 , wherein the state estimation data further comprises:

one or more measurements of a carrier frequency offset for signals transmitted between the first UAV and the one or more base stations at the first time instance, wherein the carrier frequency offset is indicative of a velocity of the first UAV; and

one or more predicted carrier frequency offsets for signals transmitted between the first UAV and the one or more base stations at the first time instance, wherein the one or more predicted carrier frequency offsets are determined using the estimate of a kinematic state of the first UAV at the first time instance.

12. The method of claim 3 , comprising:

performing said method for the first UAV at each of a plurality of time instances to monitor a trajectory of the first UAV.

13. The method of claim 1 , wherein the machine-learning process is a random-forest process or a least-squares process.

14. An apparatus configured to monitor a kinematic state of an unmanned aerial vehicle, UAV, the apparatus comprising:

processing circuitry;

memory coupled with the processing circuitry, wherein the memory includes instructions that when executed by the processing circuitry causes the apparatus to perform operations comprising:

obtain one or more predicted pathlosses between a UAV and one or more base stations at a first time instance, wherein the predicted pathlosses are determined using an estimate of a kinematic state of the UAV at the first time instance and one or more pathloss models developed using a machine-learning process;

obtain one or more measurements of a pathloss between each of the one or more base stations and the UAV at the first time instance; and

re-determine the estimate of the kinematic state of the UAV at the first time instance based on the one or more predicted pathlosses and the one or more measurements of the pathloss.

15. The apparatus of claim 14 , wherein each of the one or more pathloss models corresponds to a respective base station of the one or more base stations.

16. The apparatus according to claim 15 , wherein the UAV is a first UAV and each of the one or more pathloss models is developed using training data for the one or more respective base stations corresponding to the one or more pathloss models, the training data comprising:

a plurality of position measurements for at least one second UAV, and

a plurality of pathloss measurements of at least one pathloss between the at least one second UAV and the respective base station, each of the plurality of pathloss measurements being associated with a respective position measurement in the plurality of position measurements.

17. The apparatus according to claim 16 , wherein obtaining the one or more predicted pathlosses comprises:

obtaining the estimate of the kinematic state of the first UAV at the first time instance; and inputting the estimate of the kinematic state of the first UAV at the first time instance into the one or more pathloss models to determine the one or more predicted pathlosses.

18. The apparatus according to claim 14 further comprising a computer program, wherein the computer program is contained on a carrier comprising a non-transitory storage medium and the carrier comprises one of an electronic signal, optical signal, radio signal or machine-readable storage medium.

19. The computer program according to claim 18 , wherein the computer program is stored on a computer program product comprising non-transitory machine-readable media.

Assignments (3)

Continuity (1)

References Cited (23)