Method and apparatus for a device for a wireless communication system

Abstract

A method, for example a computer-implemented method, for a device for a wireless communication system, the method comprising: receiving (300), e.g., from at least one further device, first information (I-1) characterizing inference results of a positioning procedure based on artificial intelligence, AI, e.g., machine learning, ML, associated with the at least one further device, determining (302), based at least on the first information (I-1), second information (I-2) characterizing reference data, e.g., ground truth, for the positioning procedure.

Description

[0001] R.415400

[0002] - 1 -

[0003] Method and Apparatus for a Device for a Wireless Communication System

[0004] Technical Field

[0005] The disclosure relates to a method for a device for a wireless communication system.

[0006] The disclosure further relates to an apparatus for a device for a wireless communication system.

[0007] Summary

[0008] Some examples relate to a method, for example a computer-implemented method, for a device for a wireless communication system, the method comprising: receiving, e.g., from at least one further device, first information characterizing inference results of a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML, associated with (e.g., executed by) the at least one further device, determining, based at least on the first information, second information characterizing reference data, e.g., ground truth, for the positioning procedure. In some examples, this may enable to efficiently determine reference data that can be used, e.g., for assessing a performance of the positioning procedure, and / or for training the positioning procedure or a respective Al, e.g., ML, model.

[0009] In some examples, the wireless communications system may, e.g., be a wireless, for example cellular, communications system, which is for example based on and / or adheres at least partially to at least one third generation partnership project, 3GPP, radio standard such as 4G (fourth generation), 5G (fifth generation), or 6G (sixth generation), or to another other radio access technology. R.415400

[0010] - 2 -

[0011] In some examples, the device for the wireless communication system may, e.g., be an entity or device, respectively, for a location management function, LMF, e.g., for and / or of the wireless communication system, e.g., an LMF device or LMF entity.

[0012] In some examples, the at least one further device may, e.g., be a terminal device, e.g., user equipment, e.g., UE.

[0013] In some examples, the at least one further device may, e.g., be network device, e.g., base station, e.g., gNB.

[0014] In some examples, the method further comprises: determining the second information based on the first information and based on historical data, e.g., historical data associated at least with the first information. In some examples, this may enable to provide a particularly precise determination of the second information.

[0015] In some examples, the inference results are based on at least one of: a) measurements, e.g., channel measurements, of the at least one further device, or b) measurements, e.g., channel measurements, of one or more position reference units, PRU. In other words, in some examples, the at least one further device may use at least one of its, e.g., own, measurements, e.g., channels measurements, or measurements, e.g., channel measurements, of one or more position reference units for performing the inference using the positioning procedure.

[0016] In some examples, the method comprises: determining, based on at least one of a) the first information or b) the second information, third information characterizing at least one performance metric associated with the positioning procedure. In some examples, this may enable to assess a performance of the positioning procedure.

[0017] In some examples, the at least one performance metric associated with the positioning procedure comprises at least one of: a) a positioning error, e.g., characterizing a difference between a predicted position as, e.g., inferred by the positioning procedure and a reference, e.g., ground truth, position as R.415400

[0018] - 3 - characterized by the reference data, or b) an accuracy, e.g., characterizing a number or percentage of position predictions that fall within a predetermined error threshold compared to the respective reference, e.g., ground truth, position, or c) a reliability, e.g., characterizing a consistency of position predictions, e.g., over a predetermined time interval.

[0019] In some examples, the method comprises: determining, based at least on the third information, fourth information characterizing one or more adjustments of the positioning procedure, e.g., for the at least one further device, sending the fourth information to the at least one further device. In some examples, this enables to provide feedback to the at least one further device, e.g., for improving at least one aspect of the positioning procedure, such as, e.g., a model, e.g., an Al- or ML-based model, for the positioning procedure.

[0020] In some examples, the method comprises: receiving measurements, e.g., channel measurements, of (the) one or more position reference units, sending the measurements, e.g., channel measurements,, to the at least one further device. In some examples, this enables the at least one further device to provide inference results based on the measurements of the one or more PRUs.

[0021] Some examples relate to an apparatus for performing the method according to the disclosure, wherein for example the apparatus is configured to perform the method according to the disclosure.

[0022] Some examples relate to a device for a wireless communication system, comprising at least one apparatus according to the disclosure, wherein for example the device is a device or entity for, e.g., of, a location management function, LMF, e.g., an LMF entity.

[0023] Some examples relate to a method, for example a computer-implemented method, for a device for a wireless communication system, the method comprising: determining first information characterizing inference results of a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML, associated with the device, transmitting the first information to at least one further device. In some examples, the device may, e.g., be a terminal device, e.g., UE. In some examples, the at least one further device may, e.g., be a R.415400

[0024] - 4 - device or entity for, e.g., of, a location management function, LMF, e.g., an LMF entity. In some examples, this enables the UE to provide the LMF entity with inference results of the position procedure associated with, e.g., performed by, the UE.

[0025] In some examples, the determining comprises at least one of: a) determining the first information based on measurements, e.g., channel measurements, of the device, e.g., UE, or b) determining the first information based on measurements, e.g., channel measurements, of one or more position reference units, PRU, wherein, e.g., the UE receives the PRU measurements and performs inference based on the received PRU measurements. In other words, in some examples, the method comprises: receiving, e.g., by the UE, measurements, e.g., channel measurements, from one or more position reference units, determining, by the UE, the first information at least based on the received measurements, e.g., channel measurements.

[0026] In some examples, the method comprises: receiving information, e.g., fourth information, characterizing one or more adjustments of the positioning procedure from the at least one further device, and, optionally, adjusting the positioning procedure based at least on the fourth information. In some examples, the device, e.g., UE, may receive the fourth information from an LMF entity, e.g., the at least one further device, which may have determined the fourth information, e.g., based at least on the first information as provided by the device, e.g., UE. In some examples, this may enable to provide a feedback mechanism, e.g., in the sense of a closed loop control, e.g., for improving at least one aspect of the positioning procedure associated with, e.g., performed by, the device, e.g., UE.

[0027] Some examples relate to an apparatus for performing the method according to the disclosure, e.g., according to at least one of the claims 10 to 13, wherein for example the apparatus is configured to perform the method according to the disclosure.

[0028] Some examples relate to a device for a wireless communication system, comprising at least one apparatus according to the disclosure, wherein for example the device is a terminal device or a network device, e.g., gNB. R.415400

[0029] - 5 -

[0030] Some examples relate to a computer program comprising instructions which, when the program is executed by a computer and / or an apparatus and / or a device and / or an entity, cause the computer and / or the apparatus and / or the device to perform the method(s) according to the disclosure.

[0031] Some examples relate to a computer-readable storage medium comprising instructions which, when executed by a computer and / or an apparatus and / or a device and / or an entity, cause the computer and / or the apparatus and / or the device to perform the method(s) according to the disclosure.

[0032] Some examples relate to a data carrier signal carrying and / or characterizing the computer program according to the disclosure.

[0033] Some examples relate to a use of the method according to the disclosure and / or of the apparatus according to the disclosure and / or of the device according to the disclosure and / or of the computer program according to the disclosure and / or of the computer-readable storage medium according to the disclosure and / or of the data carrier signal according to the disclosure for at least one of: a) determining reference data, e.g., ground truth, for a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML, or b) monitoring a performance of a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML, or c) enhancing a reliability of a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML.

[0034] Brief Description of Some Example Figures

[0035] Some example embodiments will now be described with reference to the accompanying drawings, in which:

[0036] Fig. 1 schematically depicts a simplified flow-chart,

[0037] Fig. 2 schematically depicts a simplified block diagram,

[0038] Fig. 3 schematically depicts a simplified flow-chart,

[0039] Fig. 4 schematically depicts a simplified flow-chart, R.415400

[0040] - 6 -

[0041] Fig. 5 schematically depicts a simplified flow-chart,

[0042] Fig. 6 schematically depicts a simplified flow-chart,

[0043] Fig. 7 schematically depicts a simplified flow-chart,

[0044] Fig. 8 schematically depicts a simplified flow-chart,

[0045] Fig. 9 schematically depicts a simplified block diagram,

[0046] Fig. 10 schematically depicts aspects of use.

[0047] Some examples, see, for example, Fig. 1, 2, relate to a method, for example a computer-implemented method, for a device 10 (Fig. 2) for a wireless communication system 1 , the method comprising: receiving 300 (Fig. 1), e.g., from at least one further device 20, first information 1-1 characterizing inference results INF-RES of a positioning procedure POS-PROC based on artificial intelligence, Al, e.g., machine learning, ML, associated with (e.g., executed by) the at least one further device 20, determining 302, based at least on the first information 1-1 , second information I-2 characterizing reference data RD, e.g., ground truth, e.g., ground truth labels, for the positioning procedure POS-PROC. In some examples, this may enable to efficiently determine reference data RD that can be used, e.g., for assessing a performance of the positioning procedure POS-PROC, and / or for training the positioning procedure POS-PROC or a respective Al, e.g., ML, model.

[0048] In some examples, Fig.2, the wireless communications system 1 may, e.g., be a wireless, for example cellular, communications system, which is for example based on and / or adheres at least partially to at least one third generation partnership project, 3GPP, radio standard such as 4G (fourth generation), 5G (fifth generation), or 6G (sixth generation), or to another other radio access technology.

[0049] In some examples, Fig. 2, the device 10 for the wireless communication system 1 may, e.g., be an entity or device, respectively, for a location management R.415400

[0050] - 7 - function LMF, e.g., for and / or of the wireless communication system 1 , e.g., an LMF device or LMF entity 10.

[0051] In some examples, Fig. 2, the at least one further device 20 may, e.g., be a terminal device, e.g., user equipment, e.g., UE.

[0052] In some examples, the at least one further device 20 may, e.g., be network device, e.g., base station, e.g., gNB.

[0053] In some examples, Fig. 1 , the method further comprises: determining 302a the second information I-2 based on the first information 1-1 and based on historical data, e.g., historical data associated at least with the first information 1-1. In some examples, this may enable to provide a particularly precise determination of the second information I-2.

[0054] In some examples, Fig. 2, the inference results INF-RES are based on at least one of: a) measurements, e.g., channel measurements, 20-CM of the at least one further device 20, or b) measurements, e.g., channel measurements, PRU- CM of one or more position reference units PRU. In other words, in some examples, the at least one further device 20 may use at least one of its, e.g., own, measurements, e.g., channels measurements, 20-CM or measurements, e.g., channel measurements, PRLI-CM of one or more position reference units PRU for performing the inference using the positioning procedure POS-PROC.

[0055] In some examples, Fig. 1 , the method comprises: determining 304, based on at least one of a) the first information 1-1 or b) the second information I-2, third information I-3 characterizing at least one performance metric PM-POS-PROC associated with the positioning procedure POS-PROC. In some examples, this may enable to assess a performance of the positioning procedure POS-PROC.

[0056] In some examples, Fig. 2, the at least one performance metric PM-POS-PROC associated with the positioning procedure POS-PROC comprises at least one of: a) a positioning error PM-1 , e.g., characterizing a difference between a predicted position as, e.g., inferred by the positioning procedure and a reference, e.g., ground truth, position as characterized by the reference data RD, or b) an accuracy PM-2, e.g., characterizing a number or percentage of position R.415400

[0057] - 8 - predictions that fall within a predetermined error threshold compared to the respective reference, e.g., ground truth, position, or c) a reliability PM-3, e.g., characterizing a consistency of position predictions, e.g., over a predetermined time interval. In some examples, alternatively or additionally to the abovementioned aspects PM-1 , PM-2, PM-3, one or more further aspects may be provided as the at least one performance metric PM-POS-PROC or as a part of the at least one performance metric PM-POS-PROC.

[0058] In some examples, Fig. 3, the method comprises: determining 310, based at least on the third information, fourth information I-4 characterizing one or more adjustments of the positioning procedure POS-PROC, e.g., for the at least one further device 20, sending 312 the fourth information I-4 to the at least one further device 20. In some examples, this enables to provide feedback to the at least one further device 20, e.g., for improving at least one aspect of the positioning procedure POS-PROC, such as, e.g., a model, e.g., an Al- or ML-based model, for the positioning procedure POS-PROC.

[0059] In some examples, Fig. 4, the method comprises: receiving 320 measurements, e.g., channel measurements, PRI-CM of one or more position reference units PRU, sending 322 the measurements, e.g., channel measurements, PRLI-CM, to the at least one further device 20. In some examples, this enables the at least one further device 20 to provide inference results INF-RES based on the measurements PRLI-CM of the one or more PRUs.

[0060] Some examples, Fig. 2, relate to an apparatus 100 for performing the method according to the disclosure, wherein for example the apparatus 100 is configured to perform the method according to the disclosure.

[0061] Some examples, Fig. 2, relate to a device 10 for a wireless communication system 1 , comprising at least one apparatus 100 according to the disclosure, wherein for example the device 10 is a device or entity for, e.g., of, a location management function LMF, e.g., an LMF entity.

[0062] Some examples, Fig. 5, relate to a method, for example a computer-implemented method, for a device 20 for a wireless communication system 1 , the method comprising: determining 350 first information 1-1 characterizing inference results R.415400

[0063] - 9 -

[0064] INF-RES of a positioning procedure POS-PROC based on artificial intelligence, Al, e.g., machine learning, ML, associated with, e.g., executed by, the device 20, transmitting 352 the first information 1-1 to at least one further device 10. In some examples, the device 20 may, e.g., be a terminal device, e.g., UE. In some examples, the at least one further device 10 may, e.g., be a device or entity for, e.g., of, a location management function LMF, e.g., an LMF entity. In some examples, Fig. 2, this enables the UE 20 to provide the LMF entity 10 with inference results INF-RES of the position procedure POS-PROC associated with, e.g., performed by, the UE 20.

[0065] In some examples, Fig. 5, the determining 350 comprises at least one of: a) determining 350a the first information 1-1 based on measurements, e.g., channel measurements, 20-CM of the device, e.g., UE, 20 or b) determining 350b the first information 1-1 based on measurements, e.g., channel measurements, PRU-CM of one or more position reference units PRU, wherein, e.g., the UE 20 receives the PRU measurements PRU-CM and performs inference based on the received PRU measurements. In other words, in some examples, Fig. 6, the method comprises: receiving 360, e.g., by the UE 20, measurements, e.g., channel measurements, PRU-CM from one or more position reference units PRU, determining 362, by the UE 20, the first information 1-1 at least based on the received measurements, e.g., channel measurements.

[0066] In some examples, Fig. 7, the method comprises: receiving 370 information, e.g., fourth information I-4, characterizing one or more adjustments of the positioning procedure POS-PROC from the at least one further device 10, and, optionally, adjusting 372 the positioning procedure POS-PROC (e.g., at least one aspect of the positioning procedure POS-PROC) based at least on the fourth information I- 4. In some examples, the device, e.g., UE, 20 may receive the fourth information I-4 from an LMF entity, e.g., the at least one further device 10, which may have determined the fourth information I-4, e.g., based at least on the first information 1-1 as provided (see, for example, block 300 of Fig. 1) by the device, e.g., UE 20. In some examples, this may enable to provide a feedback mechanism, e.g., in the sense of a closed loop control, e.g., for improving at least one aspect of the positioning procedure POS-PROC associated with, e.g., performed by, the device, e.g., UE 20. R.415400

[0067] - 10 -

[0068] Some examples, Fig. 2, relate to an apparatus 200 for performing the method according to the disclosure, e.g., according to at least one of the claims 10 to 13, wherein for example the apparatus 200 is configured to perform the method according to the disclosure.

[0069] Some examples, Fig. 2, relate to a device 20 for a wireless communication system, comprising at least one apparatus 200 according to the disclosure, wherein for example the device 20 is a terminal device or a network device, e.g., gNB.

[0070] Fig. 8 schematically depicts a simplified flow-chart according to some examples. Element E1 symbolizes collecting data by one or more devices 10 (e.g., LMF entity), 20 (e.g., UE or gNB), ... of the wireless communication system 1. In some examples, the data collection E1 may comprise at least one of: a) performing and / or receiving measurements (e.g., channel measurements) 20-CM, PRLI-CM, or b) performing positioning (e.g., obtaining the inference results INF-RES (Fig. 2), e.g., using the positioning procedure POS-PROC, and / or collecting data associated with the positioning procedure POS-PROC, or c) collecting one or more, e.g., other, operational parameters of the wireless communication system 1 or at least one component associated with the wireless communication system 1 or arranged within an environment of the wireless communication system 1.

[0071] Elements E2, E5 symbolize a determination, e.g., generation, of reference data RD, e.g., ground truth data, according to different variants.

[0072] As an example, element E2 symbolizes reference data generation on an LMF side (e.g., based on measurements as provided by the device 20 to the device 10) and sending, e.g., transmission, of the reference data so generated from the LMF device 10 to the device, e.g., UE, 20.

[0073] As a further example, element E5 symbolizes reference data generation on an LMF side, e.g., at the LMF entity 10 (see Fig. 2, e.g., based on measurements and / or first information as provided by the device 20 to the device 20, see, for example, the receiving block 300 of Fig. 1), e.g., without sending, e.g., transmission, of the reference data so generated from the LMF 10 to the device, e.g., UE, 20. Thus, in other words, in some examples, in element E5, the R.415400

[0074] - 11 - reference data RD is generated on the LMF side, e.g., for further use on the LMF side, e.g., by the LMF device 10.

[0075] Elements E3, E6 symbolize a determination of at least one performance metric, see, for example block PM-POS-PROC of Fig. 2 or the optional block 304 of Fig. 1. As an example, element E3 of Fig. 8 symbolizes the determination of the at least one performance metric on a UE side (e.g., device 20 of Fig. 2), according to some examples. By contrast, element E6 of Fig. 8 symbolizes the determination of the at least one performance metric on the LMF side, e.g., by the LMF device 10, according to some other examples.

[0076] Elements E4, E7 symbolize respective feedback actions. According to some examples, element E4 symbolizes a feedback, e.g., of the at least one performance metric, as, e.g., obtained according to element E3, to a model associated with the positioning procedure POS-PROC, e.g., on the UE side, e.g., device 20, also see arrow a1 of Fig. 8. According to some other examples, element E7 symbolizes a feedback, e.g., of the at least one performance metric, as, e.g., obtained according to element E6, to a model associated with the positioning procedure POS-PROC, e.g., on the UE side, e.g., device 20, also see arrow a2 of Fig. 8.

[0077] Some examples, Fig. 9, relate to a configuration or apparatus 400 for performing at least one aspect of the disclosure.

[0078] In some examples, Fig. 9, the apparatus 400 comprises at least one calculating unit, e.g. processor or "computer", 402 and at least one memory unit 404 associated with (i.e. , usable by) the at least one calculating unit 402 for at least temporarily storing a computer program PRG and / or data DAT, wherein the computer program PRG is e.g. configured to at least temporarily control an operation of the apparatus 400 and / or the device 10, 20, e.g. an execution of a method according to the disclosure.

[0079] In some examples, Fig. 9, the at least one calculating unit 402 comprises at least one core (see the dashed rectangle) for executing the computer program PRG or at least parts thereof, e.g. for executing the method according to the disclosure or at least one or more steps thereof. R.415400

[0080] - 12 -

[0081] According to some examples, the at least one calculating unit 402 may comprise at least one of the following elements: a microprocessor, a microcontroller, a digital signal processor (DSP), a programmable logic element (e.g., FPGA, field programmable gate array), an ASIC (application specific integrated circuit), hardware circuitry, a tensor processor, a graphics processing unit (GPU). According to further preferred embodiments, any combination of two or more of these elements is also possible.

[0082] According to some examples, the memory unit 404 comprises at least one of the following elements: a volatile memory 404a, e.g., a random-access memory (RAM), a non-volatile memory 404b, e.g., a Flash-EEPROM.

[0083] In some examples, Fig. 9, the computer program PRG is at least temporarily stored in the non-volatile memory 404b. Data DAT (e.g. associated with the first information 1-1 and / or the second information I-2 and / or the third information I-3, and the like), which may, e.g. be used for performing the method according to the disclosure, may at least temporarily be stored in the RAM 404a.

[0084] In some examples, Fig. 9, an optional computer-readable storage medium SM comprising instructions, e.g. in the form of a or the computer program PRG, may be provided. As an example, the storage medium SM may comprise or represent a digital storage medium such as a semiconductor memory device (e.g., solid state drive, SSD) and / or a magnetic storage medium such as a disk or harddisk drive (HDD) and / or an optical storage medium such as a compact disc (CD) or DVD (digital versatile disc) or the like.

[0085] In some examples, Fig. 9, the configuration or apparatus 400 may comprise an optional data interface 406, e.g. for bidirectional data exchange with at least one further device (not shown). As an example, by means of the data interface 406, a data carrier signal DCS may be exchanged, e.g. with an external device, for example via a wired or a wireless data transmission medium, e.g. over a (virtual) private computer network and / or a public computer network such as e.g. the Internet. R.415400

[0086] - 13 -

[0087] Some examples, Fig. 9, relate to a computer program PRG comprising instructions which, when the program is executed by a computer 402, cause the computer 402 to perform the method according to the disclosure.

[0088] In some examples, Fig. 2, at least one of the devices 10, 20 or apparatus 100, 200 may comprise a configuration identical or at least similar to the configuration or apparatus 400 of Fig. 9.

[0089] In the following, further aspects and examples are disclosed that, in some examples, may be combined with each other and / or with at least one of the aspects and examples disclosed above, e.g., to form further aspects and examples illustrating the disclosure.

[0090] In some examples, one or more information elements and / or messages may be provided, e.g., for exchanging at least one of the following elements: a) the first information 1-1, or b) the second information I-2, or c) the third information I-2, or d) the fourth information I-4, or e) the measurements PRLI-CM, e.g., between at least some of the devices or entities 10, 20, LMF.

[0091] In some examples, the principle according to the disclosure enables to address at least some challenges of at least some conventional AI / ML-based positioning systems, e.g., regarding accurately monitoring a model performance, e.g., due to a difficulty of generating reliable ground truth labels. In some examples, e.g., in this context, ground truth labels correspond to a true position, that is the output of a positioning system. In some examples, the ground truth labels may, e.g., be used for training at least one AI-, e.g., ML-based model of and / or for the positioning procedure POS-PROC.

[0092] In some examples, the principle according to the disclosure enables to enhance an accuracy and reliability of performance metrics, ultimately leading to better AI / ML model performance associated with the positioning procedure POS-PROC, e.g., for providing a positioning system, e.g., for and / or using at least some components 10, 20, LMF of the wireless communication system 1.

[0093] In some examples, the principle according to the disclosure enables to provide a robust framework, e.g., for providing, e.g., generating, and / or utilizing reference R.415400

[0094] - 14 - data RD, e.g., ground truth labels, e.g., to monitor AI / ML model performance associated with at least one positioning procedure POS-PROC, e.g., in positioning systems. By implementing one or more aspects of the disclosure, more accurate and reliable performance metrics may be achieved in some examples, thus, e.g., improving an overall quality and reliability of positioning services.

[0095] In some examples, the principle according to the disclosure enables to provide different options for providing, e.g., generating, reference data RD, e.g., ground truth labels, either at a) a UE, e.g., target UE, 20, or b) at an LMF or LMF entity 10 or LMF side (see, for example, device 10).

[0096] In some examples, the principle according to the disclosure enables to provide mechanisms for an accurate calculation and / or efficient reporting of performance metrics.

[0097] In some examples, the principle according to the disclosure enables to ensure flexibility, e.g., enabling an integration of various different options or methods for ground truth label generation and / or metric calculation.

[0098] In some examples, the principle according to the disclosure enables to provide, e.g., generate, ground truth labels, e.g., at a UE side ("Aspect A"), e.g., at the device 20, e.g., a target UE, e.g., for the positioning procedure POS-PROC, and / or at an LMF side, e.g., at the device 10 or at the location management function LMF ("Aspect B").

[0099] In some examples, in the target UE side-approach (Aspect A), the location management function LMF (and / or device 10) may provide ground truth labels to the target UE 10, e.g., based on "measurements" (e.g., characterized by the first information 1-1 , e.g., the inference results INF-RES), e.g., sent from the target UE 20 (and / or gNB, e.g., if the device 20 is a gNB).

[0100] Alternatively, in some examples, in the LMF side-approach (Aspect B), the device 20, e.g., target UE, may send the inference results INF-RES (e.g., of a positioning procedure POS-PROC (e.g., in some examples, based on Al and / or ML, or, in some other examples, not based on Al and / or ML) to the LMF side, R.415400

[0101] - 15 - e.g., to at least one of the device 10 or the location management function LMF, wherein at least one of element 10, LMF may then generate the reference data RD, e.g., ground truth labels, based on the inference results INF-RES, and may, e.g., calculate one or more performance metrics PM-POS-PROC (Fig. 2).

[0102] In some examples, the AI / ML-based positioning procedure POS-PROC may, e.g., be trained on data obtained by one or more other positioning methods, e.g., legacy positioning methods, such as llu positioning and / or sidelink positioning.

[0103] In some examples, the principle according to the disclosure enables to provide advanced algorithms, e.g., based on mean square error or root mean square error between the AI / ML output and the ground truth or between quantities derived from the output and the ground truth, such as the Euclidean or Manhattan distance, e.g., for calculating performance metrics PM-POS-PROC such as positioning error, accuracy, and latency, e.g., using the generated ground truth labels. In some examples, these metrics may provide a comprehensive evaluation of the AI / ML model's performance, guiding improvements and adjustments.

[0104] In some examples, the principle according to the disclosure enables to provide and / or use optimized signaling protocols, e.g., including new message types and / or information elements and / or headers, e.g., extended headers, e.g., to ensure efficient data transmission, e.g., with minimal signaling overhead.

[0105] In some examples, the principle according to the disclosure enables to provide tools, e.g., for performance monitoring, e.g., for continuous performance monitoring, which may, e.g., be implemented to assess a performance of at least one AI / ML model (e.g., for the positioning procedure POS-PROC), e.g., in realtime. In some examples, one or more feedback loops may be established (e.g., using the fourth information I-4, see, for example, blocks 310, 312 of Fig. 3), e.g., to refine and improve one or more algorithms, e.g., based on performance data.

[0106] In some examples, the principle according to the disclosure may at least temporarily address and / or improve at least one of the following aspects: a) enhance accuracy, e.g., by providing reliable ground truth labels, e.g., leading to precise performance metrics and improved tuning of AI / ML models, or b) enable R.415400

[0107] - 16 - timely adjustments, e.g., based on accurate monitoring, e.g., enhancing the reliability and performance of the models, or c) flexibility and / or adaptability, e.g., to support multiple methods for ground truth generation and metric calculation (e.g., on UE side, e.g., side of device 20, and / or on LMF side, e.g., at element(s) 10, LMF), thus, e.g., ensuring versatility and robustness.

[0108] In some examples, the principle according to the disclosure may contribute to improve an operational efficiency, e.g., through optimized signaling protocols (e.g., comprising at least one of efficient data packing, or selective data transmission, or compression, or adaptive signaling depending on network conditions, or edge processing) that may reduce signaling overhead, freeing up network resources. In some examples, the scalable design ensures that the system can handle increasing amounts of data and complexity, maintaining longterm effectiveness. Furthermore, by maintaining compatibility with existing 3GPP standards, the invention facilitates widespread adoption and interoperability across different networks and devices.

[0109] In some examples, the principle according to the disclosure enables to provide one or more of the following aspects related to providing, e.g., generating, the reference data RD, e.g., as ground truth labels: a) information on a ground truth label, e.g., of the target UE 20, is generated by the location management function LMF (and / or device 10), and, e.g., provided to the target UE 20, e.g., by sending the information from any of the entities 10, LMF to the device 20. As an example, the target UE 20 (and / or gNB, in some examples) may send measurements (e.g., legacy measurements and / or inference results INF-RES) to the LMF (or device 10), which may, e.g., derive the ground truth label information based at least on these measurements / inference results. b) Position calculation assistance data may be provided, e.g., from LMF entity 10 to the target UE 20. c) PRU measurements and / or corresponding PRU location(s) may be sent, e.g., via the LMF device 10, to the target UE 20, e.g., reusing aspects of a 3GPP R.415400

[0110] - 17 -

[0111] Rel-18 assistance data transfer framework, e.g., from LMF device 10 to the target UE 20. d) PRU measurements (e.g., PRLI-CM and, optionally, corresponding PRU location, e.g., if not known to the UE 20) may be sent from at least one PRU to the target UE 20. In some examples, this may, e.g., be implemented in a manner transparent to specifications, e.g., if the PRU sends information to the target UE 20 in a proprietary method.

[0112] In the following, one or more aspects of performance metric calculation according to further examples are disclosed.

[0113] In some examples, the performance metric calculation may involve several steps, e.g., to ensure accurate and reliable monitoring of AI / ML model performance, e.g., in positioning systems. In some examples, a calculation process may, e.g., be implemented either at the target UE side (see Aspect A as mentioned above) or at the LMF side (see Aspect B as mentioned above), e.g., with each option or aspect offering unique advantages.

[0114] In some examples, a performance metric calculation related to Aspect A (e.g., in the sense of a "target UE side monitoring metric calculation") may comprise one or more of the following aspects:

[0115] • Ground Truth Label Generation:

[0116] • In some examples, the LMF device 10 may generate ground truth labels based on measurements I inference results INF-RES (see, for example, the first information 1-1) received from the target UE and / or gNB 20. In some examples, these measurements may include at least one of signal strength, time of arrival, angle of arrival, etc.

[0117] • In some examples, the generated ground truth labels may then be sent to the target UE 20, providing a reference for performance evaluation.

[0118] • Metric Calculation at Target UE:

[0119] • In some examples, the target UE 20 may use the received ground truth labels, e.g., to calculate performance metrics such as positioning error, accuracy, and reliability.

[0120] • Example Metrics:

[0121] • Positioning Error: The difference between the predicted position (inferred by the AI / ML model) and the ground truth position. R.415400

[0122] - 18 -

[0123] • Accuracy: The percentage of predictions that fall within a certain error threshold compared to the ground truth.

[0124] • Reliability: The consistency of the positioning predictions over time.

[0125] • Reporting to LMF:

[0126] • In some examples, the target UE 20 may send the calculated performance metrics to the device 10 and / or the LMF, e.g., for further analysis and optimization.

[0127] In some examples, a performance metric calculation related to Aspect B (e.g., in the sense of a " LMF Side Monitoring Metric Calculation") may comprise one or more of the following aspects:

[0128] • Inference Result Transmission:

[0129] • In some examples, the target UE 20 may perform initial positioning using its AI / ML model (e.g., the positioning procedure POS-PROC) and may send the inference results INF-RES (i.e. , the model output corresponding to the target UE’s channel measurement, as, e.g., characterized by the first information 1-1) to the LMF device 10.

[0130] • Alternatively, in some examples, a PRU’s channel measurement CM-PRU may be sent, e.g., via the LMF device 10, e.g., to the target UE 20, and the inference result INF-RES corresponding to PRU’s channel measurement CM-PRU may then be sent by the target UE 20 to the LMF device 10.

[0131] • Ground Truth Label Generation at LMF:

[0132] • In some examples, the LMF device 10 may generate ground truth labels based on the received measurements and, optionally, based on historical data.

[0133] • Metric Calculation at LMF:

[0134] • In some examples, the LMF device 10 may use the ground truth labels and the received inference results INF-RES, e.g., to calculate performance metrics PM-POS-PROC.

[0135] • In some examples, example performance metrics PM-POS-PROC comprise at least one of:

[0136] • Positioning Error: Calculated by comparing the AI / ML model’s predicted positions with the ground truth positions.

[0137] • Model Confidence: Evaluates the confidence level of the AI / ML model’s predictions based on the variance of prediction errors.

[0138] • Latency: Measures the time taken from receiving the measurements to generating the positioning output.

[0139] • Feedback to Target UE: R.415400

[0140] - 19 -

[0141] • In some examples, the LMF device 10 may provide feedback to the target UE 20, e.g., based on the calculated metrics, e.g., suggesting adjustments to improve model performance, e.g., in the form of the fourth information I-4, see, for example, block 312 of Fig. 3.

[0142] In some examples, the principle according to the disclosure enables to attain at least some of the following benefits: a) Improved accuracy: More accurate performance metrics as obtained by using the principle according to the disclosure may lead to better tuning of AI / ML models, thus, e.g., resulting in enhanced positioning accuracy. b) Enhanced reliability: Reliable ground truth labels as obtained by using the principle according to the disclosure may enable to improve an overall reliability of positioning services, providing consistent and dependable performance. c) Scalability: The flexible framework as may be provided by using the principle according to the disclosure enables to adapt to future advancements and to integrate new methods for performance monitoring, ensuring long-term relevance and effectiveness.

[0143] Some examples, Fig. 10, relate to a use 500 of the method according to the disclosure and / or of the apparatus 100, 200 according to the disclosure and / or of the device 10, 20 according to the disclosure and / or of the computer program PRG according to the disclosure and / or of the computer-readable storage medium SM according to the disclosure and / or of the data carrier signal DCS according to the disclosure for at least one of: a) determining 501 reference data, e.g., ground truth, for a positioning procedure POS-PROC based on artificial intelligence, Al, e.g., machine learning, ML, or b) monitoring 502 a performance of a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML, or c) enhancing 503 a reliability of a positioning procedure based on artificial intelligence, Al, e.g., machine learning, ML.

Claims

R.415400- 20 -Claims1. A method, for example a computer-implemented method, for a device (10) for a wireless communication system (1), the method comprising:- receiving (300), e.g., from at least one further device (20), first information (1-1) characterizing inference results (INF-RES) of a positioning procedure (POS-PROC) based on artificial intelligence, Al, e.g., machine learning, ML, associated with the at least one further device (20),- determining (302), based at least on the first information (1-1), second information (I-2) characterizing reference data (RD), e.g., ground truth, for the positioning procedure (POS-PROC).

2. The method according to claim 1 , comprising:- determining (302a) the second information (I-2) based on the first information (1-1) and based on historical data.

3. The method according to at least one of the preceding claims, wherein the inference results (INF-RES) are based on at least one of: a) measurements, e.g., channel measurements, (20-CM) of the at least one further device (20), or b) measurements, e.g., channel measurements, (PRLI-CM) of one or more position reference units (PRU).

4. The method according to at least one of the preceding claims, comprising:- determining (304), based on at least one of a) the first information (1-1) or b) the second information (I-2), third information (I-3) characterizing at least one performance metric (PM-POS-PROC) associated with the positioning procedure (POS- PROC).

5. The method according to claim 4, wherein the at least one performance metric (PM-POS-PROC) associated with the positioning procedure (POS- PROC) comprises at least one of:R.415400- 21 - a) a positioning error (PM-1), e.g., characterizing a difference between a predicted position as, e.g., inferred by the positioning procedure (POS- PROC) and a reference, e.g., ground truth, position as characterized by the reference data (RD), or b) an accuracy (PM-2), e.g., characterizing a number or percentage of position predictions that fall within a predetermined error threshold compared to the respective reference, e.g., ground truth, position, or c) a reliability (PM-3), e.g., characterizing a consistency of position predictions, e.g., over a predetermined time interval.

6. The method according to any of the claims 4 to 5, comprising:- determining (310), based at least on the third information (I-3), fourth information (I-4) characterizing one or more adjustments of the positioning procedure (POS-PROC), e.g., for the at least one further device (20), sending (312) the fourth information (I-4) to the at least one further device (20).

7. The method according to at least one of the preceding claims, comprising:- receiving (320) measurements, e.g., channel measurements, (PRLI-CM) of (the) one or more position reference units (PRU),- sending (322) the measurements, e.g., channel measurements, (PRU- CM), to the at least one further device (20).

8. An apparatus (100) for performing the method according to at least one of the preceding claims, wherein for example the apparatus (100) is configured to perform the method according to at least one of the preceding claims.

9. A device (10) for a wireless communication system (1), comprising at least one apparatus (100) according to claim 8, wherein for example the device (10) is a device or entity for, e.g., of, a location management function (LMF).

10. A method, for example a computer-implemented method, for a device (20) for a wireless communication system (1), the method comprising:- determining (350) first information (1-1) characterizing inference results (INF-RES) of a positioning procedure (POS-PROC) based on artificial intelligence, Al, e.g., machine learning, ML, associated with the deviceR.415400- 22 -- transmitting (352) the first information (1-1) to at least one further device (10).

11. The method according to claim 10, wherein the determining (350) comprises at least one of: a) determining (350a) the first information (1-1) based on measurements, e.g., channel measurements, (20-CM) of the device (20), or b) determining (350b) the first information (1-1) based on measurements, e.g., channel measurements, (PRLI-CM) of one or more position reference units (PRU).

12. The method according to any of the claims 10 to 11 , comprising:- receiving (360) measurements, e.g., channel measurements, (PRLI-CM) from one or more position reference units (PRU),- determining (362) the first information (1-1) at least based on the received measurements, e.g., channel measurements, (PRU-CM).

13. The method according to any of the claims 10 to 12, comprising:- receiving (370) information, e.g., fourth information (I-4), characterizing one or more adjustments of the positioning procedure (POS-PROC) from the at least one further device (10), and,- optionally, adjusting (372) the positioning procedure (POS-PROC) based at least on the fourth information (I-4).

14. An apparatus (200) for performing the method according to at least one of the claims 10 to 13, wherein for example the apparatus (200) is configured to perform the method according to at least one of the claims 10 to 13.

15. A device (20) for a wireless communication system (1), comprising at least one apparatus (200) according to claim 14, wherein for example the device (20) is a terminal device or a network device, e.g., gNB.

16. A computer program (PRG) comprising instructions which,- when the program (PRG) is executed by a computer (402) and / or the apparatus (100) according to claim 8 and / or the device (10) according to claim 9, cause the computer (402) and / or the apparatus (100) and / or theR.415400- 23 - device (10) to perform the method according to at least one of the claims 1 to 7, and / or- when the program (PRG) is executed by a computer (402) and / or the apparatus (200) according to claim 14 and / or the device (20) according to claim 15, cause the computer (402) and / or the apparatus (200) and / or the device (20) to perform the method according to at least one of the claims 10 to 13.

17. A computer-readable storage medium (SM) comprising instructions (PRG) which,- when executed by a computer (402) and / or the apparatus (100) according to claim 8 and / or the device (10) according to claim 9, cause the computer (402) and / or the apparatus (100) and / or the device (10) to perform the method according to at least one of the claims 1 to 7, and / or- when by a computer (402) and / or the apparatus (200) according to claim 14 and / or the device (20) according to claim 15, cause the computer (402) and / or the apparatus (200) and / or the device (20) to perform the method according to at least one of the claims 10 to 13.

18. A data carrier signal (DCS) carrying and / or characterizing the computer program (PRG) of claim 16.

19. A use (500) of the method according to any of the claims 1 to 7 or 10 to 13 and / or of the apparatus (100; 200) according to claim 8 or 14 and / or of the device (10, 20) of any of the claims 9 or 15 and / or of the computer program (PRG) according to claim 16 and / or of the computer-readable storage medium (SM) according to claim 17 and / or of the data carrier signal (DCS) according to claim 18 for at least one of: a) determining (501) reference data (RD), e.g., ground truth, for a positioning procedure (POS-PROC) based on artificial intelligence, Al, e.g., machine learning, ML, or b) monitoring (502) a performance of a positioning procedure (POS-PROC) based on artificial intelligence, Al, e.g., machine learning, ML, or c) enhancing (503) a reliability of a positioning procedure (POS-PROC) based on artificial intelligence, Al, e.g., machine learning, ML.

Images