Methods and apparatus and machine-readable media relating to airborne wireless devices
A method performed by a node in a wireless communication network. The method comprises obtaining measurement data for one or more measurements performed, by a wireless device in the wireless communication network, on at least one first radio signal transmitted by one or more base stations in the wireless communication network. The method further comprises determining whether or not the wireless device is airborne by inputting the measurement data to a model developed using a machine learning algorithm.
1. A method performed by a node in a wireless communication network, the method comprising:
obtaining measurement data for one or more measurements performed, by a wireless device in the wireless communication network, on at least one first radio signal transmitted by a serving base station and one or more neighbouring base stations in the wireless communication network; and
determining whether or not the wireless device is airborne by inputting the measurement data to a model developed using a machine learning algorithm, wherein the model is developed using training data comprising:
measurement data obtained from at least one airborne wireless device and at least one grounded wireless device on the at least one first radio signal transmitted by the serving base station and one or more neighbouring base stations; and
identification information for the measurement data, indicating which of the measurement data are from the at least one airborne wireless device and which of the measurement data are from the at least one grounded wireless device.
2. The method according to claim 1 , wherein determining whether or not the wireless device is airborne further comprises:
obtaining, from the model, a classification score indicating a likelihood that the wireless device is airborne.
3. The method according to claim 2 , wherein determining whether or not the wireless device is airborne further comprises:
comparing the classification score to a first threshold to determine whether or not the wireless device is airborne.
4. The method according to claim 1 , wherein the machine learning algorithm is a logistic regression algorithm.
5. The method according to claim 1 , wherein the measurement data comprise a number of cells detected by the wireless device.
6. The method according to claim 1 , wherein the measurement data comprise one or more of: a received power of the at least one first radio signal and a received signal quality of the at least one first radio signal.
7. The method according to claim 1 , wherein the measurement data comprise a variance in received signal power of the serving base station.
8. The method according to claim 1 , wherein the measurement data comprise a difference between a received signal power of the serving base station and a received signal power of the one or more neighbouring base stations.
9. The method according to claim 8 , wherein the received signal power of the one or more neighbouring base stations is a maximum received signal power.
10. The method according to claim 1 , further comprising:
obtaining a carrier frequency offset for a second radio signal transmitted between the wireless device and one of the serving base station and one or more neighbouring base stations, wherein the carrier frequency offset is indicative of a velocity of the wireless device; and
determining whether or not the wireless device is an unmanned aerial vehicle using the carrier frequency offset.
11. The method according to claim 10 , wherein determining whether or not the wireless device is an unmanned aerial vehicle using the carrier frequency offset comprises:
comparing the carrier frequency offset to a second threshold to determine whether or not the wireless device is an unmanned aerial vehicle, wherein the second threshold is specific to a carrier frequency of the second radio signal.
12. The method according to claim 1 , wherein the node is the serving base station, or the node is a core network node in the wireless communication network.
13. A node in a wireless communication network comprising:
processing circuitry; and
a storage medium storing instructions which, when executed by the processing circuitry, cause the node to:
obtain measurement data for one or more measurements performed, by a wireless device in the wireless communication network, on at least one first radio signal transmitted by a serving base station and one or more neighbouring base stations in the wireless communication network; and
determine whether or not the wireless device is airborne by inputting the measurement data to a model developed using a machine learning algorithm, wherein the model is developed using training data comprising:
measurement data obtained from at least one airborne wireless device and at least one grounded wireless device on the at least one first radio signal transmitted by the serving base station and one or more neighbouring base stations; and
identification information for the measurement data, indicating which of the measurement data are from the at least one airborne wireless device and which of the measurement data are from the at least one grounded wireless device.
14. The node according to claim 13 , wherein to determine whether or not the wireless device is airborne further comprises the node to obtain, from the model, a classification score indicating a likelihood that the wireless device is airborne.
15. The node according to claim 14 , wherein to determine whether or not the wireless device is airborne further comprises the node to compare the classification score to a first threshold to determine whether or not the wireless device is airborne.
16. The node according to claim 13 , wherein the machine learning algorithm is a logistic regression algorithm.
17. A non-transitory machine-readable storage medium storing instructions which, when executed by processing circuitry of a node in a wireless communication network, cause the node to perform operations comprising:
obtaining measurement data for one or more measurements performed, by a wireless device in the wireless communication network, on at least one first radio signal transmitted by a serving base station and one or more neighbouring base stations in the wireless communication network; and
determining whether or not the wireless device is airborne by inputting the measurement data to a model developed using a machine learning algorithm, wherein the model is developed using training data comprising:
measurement data obtained from at least one airborne wireless device and at least one grounded wireless device on the at least one first radio signal transmitted by the serving base station and one or more neighbouring base stations; and
identification information for the measurement data, indicating which of the measurement data are from the at least one airborne wireless device and which of the measurement data are from the at least one grounded wireless device.