Interaction between AI-based positioning technology and traditional positioning technology

Abstract

A user equipment (UE) configured to determine that the UE is capable of performing a first positioning scheme, determine that the UE is capable of performing a second positioning scheme, select one of the first positioning scheme and the second positioning scheme to be used by the UE to perform a positioning operation, and determine that the UE is capable of performing the first positioning scheme and the second positioning scheme. And calculating a location of the UE using the one of the first positioning scheme and the second positioning scheme.

Description

Technical Field

[0001] The present application relates generally to wireless communications, and specifically to the interaction between AI-based positioning technology and traditional positioning technology. Background Art

[0002] 5G New Radio (NR) introduces many radio access network (RAN) and core network (CN) enhancements as well as an enhanced security architecture. Artificial intelligence (AI) and / or machine learning (ML) processes, e.g., deep learning neural networks, may be used to enhance the operation of the air interface. Use cases for AI / ML for the air interface include: channel state information (CSI) feedback enhancements (e.g., overhead reduction, improved accuracy, prediction); beam management (e.g., beam prediction in the time and / or spatial domain for overhead and latency reduction, beam selection accuracy improvement); and positioning accuracy enhancements for different scenarios, including, for example, those with heavy non-line-of-sight (NLOS) conditions. Summary of the invention

[0003] Some exemplary implementation aspects relate to a processor of a user equipment (UE) configured to perform operations including determining that the UE is capable of performing a first positioning scheme, determining that the UE is capable of performing a second positioning scheme, selecting one of the first positioning scheme and the second positioning scheme for the UE to use to perform positioning operations, and calculating a position of the UE using the one of the first positioning scheme and the second positioning scheme.

[0004] Other exemplary embodiments are directed to a user equipment (UE) having: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform operations. The operations include: determining that the UE is capable of performing a first positioning scheme, determining that the UE is capable of performing a second positioning scheme, selecting one of the first positioning scheme and the second positioning scheme for the UE to use to perform positioning operations, and calculating a position of the UE using the one of the first positioning scheme and the second positioning scheme.

[0005] Still further exemplary embodiments relate to a processor of a user equipment (UE) configured to perform operations including storing one or more trained models and performing an artificial intelligence (AI) based positioning method using the one or more trained models to determine a location of the UE.

[0006] Another exemplary embodiment relates to a user equipment (UE) having: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform operations. The operations include storing one or more trained models and performing an artificial intelligence (AI)-based positioning method using the one or more trained models to determine a location of the UE. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 Network arrangements according to various exemplary embodiments are shown.

[0008] Figure 2 An exemplary UE according to various exemplary embodiments is shown.

[0009] Figure 3 A diagram illustrating the interaction between a traditional positioning method and an AI-based positioning method according to various exemplary embodiments is shown.

[0010] Figure 4a A diagram is shown in which Doppler estimates inform a neural network (NN) update rate for an AI-based positioning scheme, according to various exemplary embodiments.

[0011] Figure 4b A diagram is shown in which Doppler estimation is used to select one of "n" trained NNs, according to various exemplary embodiments.

[0012] Figure 4c A diagram is shown in which an AI model identifying Line-of-Site (LOS) / Non-Line-of-Site (NLOS) conditions informs the execution of traditional positioning techniques according to various exemplary embodiments.

[0013] Figure 4d A diagram is shown in which a selector AI identifies LOS / NLOS parameters of a UE relative to a positioning node and selects an AI-based positioning solution or a conventional positioning solution according to various exemplary embodiments.

[0014] Figure 4e A diagram is shown in which a selector AI recognizes a scene and selects a NN or a traditional positioning solution according to various exemplary embodiments.

[0015] Figure 4f A diagram is shown in which the update rate AI is an input to the position updates of the AI ​​model according to various exemplary embodiments.

[0016] Figure 4g A diagram is shown in which a selector AI selects the number of next generation Node Bs (gNBs) as input to an AI-based positioning scheme according to various exemplary embodiments.

[0017] Figure 5 A more generalized model in which multiple estimator stages may be used to determine position is shown according to various exemplary embodiments.

[0018] Figure 6 A diagram of a multi-input based convolutional NN (CNN) is shown according to various exemplary embodiments. DETAILED DESCRIPTION

[0019] The exemplary embodiments may be further understood with reference to the following description and the associated drawings, in which similar elements have the same reference numerals. The exemplary embodiments describe how traditional positioning schemes can interact with positioning schemes based on artificial intelligence (AI) and / or machine learning (ML) in wireless networks. In addition, the exemplary embodiments also describe exemplary ways of selecting between traditional positioning schemes or AI-based positioning schemes and using AI-based positioning schemes to enhance traditional positioning schemes or using traditional positioning schemes to enhance AI-based positioning schemes.

[0020] The exemplary aspects are described with reference to a UE. However, the use of a UE is provided for illustration purposes. The exemplary aspects can be utilized with any electronic component that can establish a connection with a network and is configured with hardware, software, and / or firmware for exchanging information and data with the network. Therefore, a UE as described herein is used to represent any electronic component that can access a wireless network and perform an AI-based positioning scheme and / or a traditional positioning scheme to locate the UE.

[0021] The exemplary aspects are described with reference to the network being a 5G New Radio (NR) network and the base station being a next generation Node B (gNB). However, for illustrative purposes, the use of a 5G NR network and a gNB is provided. The exemplary aspects may be applicable to any type of network utilizing similar functionality. For example, some positioning methods as described herein may be RAT-independent.

[0022] The exemplary embodiments are further described with respect to location estimation based on artificial intelligence (AI) and / or machine learning (ML). Depending on the UE and network implementation, any number of different AI / ML models may be used. For example, a deep learning neural network may be used. In addition, various types of models may use different types of channel response data for training the models, as well as different types / densities of RS for the inference phase of location estimation. Therefore, for illustrative purposes, references to any particular AI-based location estimation model are provided. The exemplary aspects may be applied to any type of AI-based location estimation model using a training phase and an inference phase that may be performed at a UE and / or a core network element (e.g., a location management function (LMF)).

[0023] An exemplary embodiment is further described with respect to a location management function (LMF) of a 5G core network (5GC) that supports location determination for a target UE. As will be further described below, in the exemplary aspects described herein, the LMF may perform operations that facilitate AI-based estimation of the UE location, the operations comprising: performing either or both of an inference phase and a training phase of a positioning NN for locating the target UE; configuring a RAN node and / or UE to transmit / receive a reference signal (RS) for channel response acquisition; and transmitting / receiving position estimation and / or channel response estimation feedback to / from the target UE. It should be understood that the LMF is not required to be in the core network. For example, the LMF may reside on a separate server connected to the 5GC or may reside within the 5G RAN.

[0024] In addition, throughout the specification, a base station (e.g., a gNB) may be referred to as a "serving cell". A gNB acting as a serving cell is a cell to which the UE is currently connected, for example, the UE may be in a radio resource control (RRC) connected state with the gNB and may actively exchange data and / or control information with the base station. A gNB may also be referred to as a "positioning gNB", "positioning node", or "positioning cell". A gNB acting as a positioning cell is a base station that helps locate the UE, for example, transmitting a positioning reference signal (PRS) to the UE to help locate the UE. A gNB may simultaneously act as a serving cell and a positioning cell relative to the UE, or may only act as a positioning cell for the UE. In addition, throughout the specification, a gNB may be referred to as a "neighbor cell" or "neighboring cell". According to the present disclosure, a neighboring cell may not act as a serving cell for the UE, however certain signals may be exchanged between the neighboring cell and the UE without entering an RRC connected state. One or more neighboring cells may act as additional positioning gNBs to help locate the UE.

[0025] A serving cell or a neighboring cell acting as a positioning gNB may include one or more transmission and reception points (TRPs), for example, a first TRP and a second TRP. One or more of the TRPs located at a specific positioning gNB may be used in the exemplary positioning method and may be referred to as a "positioning TRP". Multiple positioning TRPs may be located at a single positioning gNB.

[0026] In addition, throughout the specification, various positioning methods are generally described as traditional positioning schemes or AI-based positioning schemes. These positioning methods may include downlink (DL)-based positioning, uplink (UL)-based positioning, or combined DL+UL-based positioning, including but not limited to: Assisted Global Navigation Satellite System (A-GNSS) (GPS); Wireless Local Area Network (WLAN); Terrestrial Beacon System (TBS); Downlink Time Difference of Arrival (DL-TDOA); DL Angle of Departure (DL-AoD); and Multiple Round Trip Time (Multiple RTT). In some positioning methods, a positioning reference signal (PRS) is transmitted to a UE from each of a plurality of network nodes (i.e., positioning transmission and reception points (TRPs)) on the DL, so that the DL arrival timing of the corresponding PRS at the UE can be determined, and the position of the UE can be determined based thereon. In addition, other types of reference signals (including channel state information reference signals (CSI-RS) and / or synchronization signal blocks (SSBs), such as demodulation RS (DMRS)) can be used on the DL to position the UE. Uplink (UL) reference signals (including sounding reference signals (SRS)) may also be used to locate the UE. Those skilled in the art will appreciate that these reference signals may also be used for purposes other than locating the UE. Therefore, the RS described herein is not limited to any particular type of reference signal.

[0027] Furthermore, throughout the specification, it will be understood that AI-based positioning schemes can use the same inputs as traditional positioning schemes, however, the inputs are fed into a model (e.g., NN) that analyzes the inputs to output a position, rather than strictly applying mathematical rules or formulas as in traditional positioning schemes. Furthermore, those skilled in the art will appreciate that the channel over which the reference signal is transmitted typically has a delay spread across time, and therefore can be modeled with multiple taps at different delays. Therefore, when the term tap is used with reference to various inputs, it should be understood that taps refer to this concept.

[0028] Figure 1 An exemplary network arrangement 100 according to various exemplary embodiments is shown. The exemplary network arrangement 100 includes a user equipment (UE) 110. Those skilled in the art will appreciate that a UE may be any type of electronic component configured to communicate via a network, such as a mobile phone, a tablet, a smart phone, a phablet, an embedded device, a wearable device, a Cat-M device, a Cat-M1 device, an MTC device, an eMTC device, other types of Internet of Things (IoT) devices, etc. It should also be understood that an actual network arrangement may include any number of UEs used by any number of users. Therefore, for purposes of illustration, only an example with a single UE 110 is provided.

[0029] UE 110 may communicate directly with one or more networks. In the example of network configuration 100, the networks with which UE 110 may wirelessly communicate are 5G NR radio access network (5G NR-RAN) 120, LTE radio access network (LTE-RAN) 122, and wireless local area network (WLAN) 124. Therefore, UE 110 may include a 5G NR chipset that communicates with 5G NR-RAN 120, an LTE chipset that communicates with LTE-RAN 122, and an ISM chipset that communicates with WLAN 124. However, UE 110 may also communicate with other types of networks (e.g., traditional cellular networks), and UE 110 may also communicate with the network via a wired connection. Referring to the exemplary aspect, UE 110 may establish a connection with 5G NR-RAN 122.

[0030] 5G NR-RAN 120 and LTE-RAN 122 may be parts of cellular networks that may be deployed by cellular providers (e.g., Verizon, AT&T, T-Mobile, etc.). These networks 120, 122 may include, for example, cells or base stations (NodeB, eNodeB, HeNB, eNBS, gNB, gNodeB, macrocell base stations, microcell base stations, small cell base stations, femtocell base stations, etc.) configured to send and receive traffic from UEs equipped with appropriate cellular chipsets. WLAN 124 may include any type of wireless local area network (WiFi, hotspot, IEEE 802.11x network, etc.).

[0031] UE 110 may be connected to 5G NR-RAN via at least one of next generation nodeB (gNB) 120A and / or gNB 120B. Reference to two gNBs 120A, 120B is for illustrative purposes only. The exemplary aspects may be applied to any suitable number of gNBs.

[0032] In addition to the networks 120, 122, and 124, the network arrangement 100 also includes a cellular core network 130, the Internet 140, an IP multimedia subsystem (IMS) 150, and a network service backbone 160. The cellular core network 130 (e.g., the 5GC of the 5G NR network) can be viewed as an interconnected collection of components that manage the operation and traffic of the cellular network. The cellular core network 130 also manages the traffic flowing between the cellular network and the Internet 140. As will be further described below, the core network 130 may include a location management function (LMF) 131 to support location determination of the UE.

[0033] The IMS 150 may be generally described as an architecture for delivering multimedia services to the UE 110 using the IP protocol. The IMS 150 may communicate with the cellular core network 130 and the Internet 140 to provide multimedia services to the UE 110. The network service backbone 160 communicates directly or indirectly with the Internet 140 and the cellular core network 130. The network service backbone 160 may be generally described as a set of components (e.g., servers, network storage arrangements, etc.) that implement a set of services that may be used to extend the functionality of the UE 110 to communicate with various networks.

[0034] Figure 2 An exemplary UE 110 is shown according to various exemplary embodiments. Figure 1 100 is used to describe the UE 110. The UE 110 may represent any electronic device and may include a processor 205, a memory arrangement 210, a display device 215, an input / output (I / O) device 220, a transceiver 225, and other components 230. The other components 230 may include, for example, an audio input device, an audio output device, a battery providing a limited power source, a data acquisition device, a port for electrically connecting the UE 110 to other electronic devices, a sensor for detecting the status of the UE 110, etc. In addition, the UE 110 may be configured to access the SNPN.

[0035] The processor 205 may be configured to execute a plurality of engines of the UE 110. For example, the engines may include a positioning engine 235 for performing operations including those related to positioning. Each of these operations will be described in more detail below.

[0036] The above-described engine as an application (e.g., a program) executed by the processor 205 is merely exemplary. The functionality associated with the engine may also be represented as an independently integrated component of the UE 110, or may be a modular component coupled to the UE 110, such as an integrated circuit with or without firmware. For example, an integrated circuit may include an input circuit for receiving a signal and a processing circuit for processing the signal and other information. The engine may also be embodied as an application or an independent application. In addition, in some UEs, the functions described for the processor 205 are split between two or more processors such as a baseband processor and an application processor. The exemplary aspects may be implemented in any of these or other configurations of the UE.

[0037] The memory 210 may be a hardware component configured to store data related to operations performed by the UE 110. The display device 215 may be a hardware component configured to display data to a user, and the I / O device 220 may be a hardware component that enables a user to perform input. The display device 215 and the I / O device 220 may be separate components or may be integrated together (such as a touch screen). The transceiver 225 may be a hardware component configured to establish a connection with the 5G-NR RAN 120, the LTE RAN 122, etc. Therefore, the transceiver 225 may operate on a variety of different frequencies or channels (e.g., a continuous frequency group).

[0038] An exemplary network base station, in this case gNB 120A, may represent a serving cell for UE 110. gNB 120A may represent any access node belonging to a 5G NR network that UE 110 may use to establish a connection and manage network operations. gNB 120A may include a processor, a memory arrangement, an input / output (I / O) device, a transceiver, and other components. Other components may include, for example, an audio input device, an audio output device, a battery, a data acquisition device, a port for electrically connecting gNB 120A to other electronic devices, and the like. The functionality associated with the processor of gNB 120A may also be represented as a separate integrated component of gNB 120A, or may be a modular component coupled to gNB 120A, such as an integrated circuit with or without firmware. For example, an integrated circuit may include input circuits for receiving signals and processing circuits for processing signals and other information. In addition, in some gNBs, the functions described for the processor are split between multiple processors (e.g., a baseband processor, an application processor, etc.). The exemplary aspects may be implemented in any of these or other configurations of the gNB.

[0039] The memory may be a hardware component configured to store data related to operations performed by the UE 110, 112. The I / O device may be a hardware component or port that enables a user to interact with the gNB 120A. The transceiver may be a hardware component configured to exchange data with the UE 110 and any other UE in the system 100. The transceiver may operate on a variety of different frequencies or channels (e.g., a set of contiguous frequencies). Thus, the transceiver may include one or more components (e.g., a radio component) to enable data exchange with various networks and UEs.

[0040] According to various exemplary embodiments, the interaction between traditional positioning methods and AI-based positioning methods is described. In previous NR versions, many traditional positioning technologies were proposed, including RAT-related positioning technologies (e.g., time difference of arrival (TDOA), angle of arrival (AOA), etc.) and RAT-independent positioning technologies (e.g., global navigation satellite system (GNSS)). These traditional technologies can interact with AI-based positioning in a variety of ways. The exemplary embodiments may define interactions for UEs.

[0041] In one aspect of the interaction between the traditional positioning method and the AI-based positioning method, the UE will not perform the traditional positioning scheme and the AI-based positioning scheme at the same time. In one option, the UE does not expect to be configured with two schemes at the same time. In another option, the UE can be configured with two schemes at the same time, but can switch between the schemes semi-statically or dynamically. In this option, a new UE capability can be introduced to report whether the UE supports such concurrent configuration.

[0042] On the other hand of the interaction between traditional positioning methods and AI-based positioning methods, the UE may be configured with both traditional positioning technology and AI-based positioning technology, and may switch between the two in specific scenarios. For example, in certain cases such as high Doppler or non-line of sight (NLOS) scenarios, the UE may fall back to the traditional positioning solution and default to the AI-based positioning solution outside of these predefined scenarios. In some exemplary embodiments, when a scenario that triggers switching is detected, the LMF may semi-statically or dynamically configure switching between positioning solution types. In other exemplary embodiments, when a scenario that triggers switching is detected, the UE may autonomously switch between positioning types.

[0043] In another aspect of the interaction between traditional positioning methods and AI-based positioning methods, traditional positioning services can be used to calibrate AI-based positioning, and AI-based positioning can be used to calibrate traditional positioning. For example, the calibration position can be determined based on: (a) any of the RAT-independent technologies (e.g., GNSS) defined in 3GPP Release 16 and Release 17 or (b) any of the RAT-dependent technologies (e.g., TDOA, AOA). The calibration error can be used as an input to the AI ​​model update rate decision. For example, if the calibration error is greater than the calibration error threshold for a period of time greater than the calibration period, an AI model update can be requested.

[0044] In another aspect of the interaction between the traditional positioning method and the AI-based positioning method, the NN inference input RS parameters (e.g., periodicity, RS density) for the traditional positioning method and / or the AI-based positioning method may be modified based on the AI-based positioning service. In one example, the AI ​​model may be able to predict that based on the current performance, reducing the density or periodicity of the measurement signal and / or the measurement feedback will fall within the position tolerance requirement. In another example, if the error between the two methods exceeds a threshold, the periodicity / density may be increased. In yet another example, the PRS / P-SRS may be turned on / off. In yet another example, the output of the AI ​​model may include accuracy or error estimates (in addition to the position). This output may be used to determine the RS parameters (periodicity, density). The RS parameters may be changed semi-statically (RRC configured) or dynamically based on the error estimate. In one scenario, the UE may request multiple reports or inputs (position, NN input) for both the AI-based method and the traditional method. The request may be turned on / off for a specific method as needed. In some exemplary embodiments, there may be multiple predefined sets of RSs with various characteristics (e.g., periodicity, density, etc.). The AI ​​model may be used to select a specific one of the sets that should be used in the current UE environment.

[0045] In an additional aspect of the interaction between traditional positioning methods and AI-based positioning methods, a hybrid service of AI-based positioning and traditional positioning can be used. In one example, the UE can detect a line-of-sight (LOS) or non-line-of-sight (NLOS) condition between a gNB and the UE (or between each of multiple gNBs and the UE). For a LOS link, a traditional positioning method can be used to identify the distance (time of arrival, TOA) between the UE and the gNB. For a NLOS link, an AI-based positioning method can be used to identify the distance (TOA) between the UE and the gNB.

[0046] Figure 3 A diagram 300 is shown illustrating the interaction between an exemplary traditional positioning method and an AI-based positioning method according to various exemplary embodiments. As shown in the exemplary aspects described above, various aspects of this interaction are included in the diagram 300. However, it should be understood that the various features shown in the diagram 300 may be used alone or in combination with other features shown in the diagram 300.

[0047] Box 305 shows the training of the AI ​​model. Box 305 may represent the initial training of the AI ​​model or the retraining of the model after a calibration phase, e.g., an AI training update.

[0048] Block 310 shows the inference phase of the AI ​​model. Block 310 includes AI-based position determination based on a neural network (NN) inference input (e.g., channel response parameters for a reference signal (RS)). The RS may be an uplink (UL) RS (e.g., a positioning SRS (P-SRS)) or a downlink (UL) RS (e.g., a PRS) with a certain periodicity and density.

[0049] Block 315 shows a conventional (RAT-based or RAT-independent) positioning determination, and block 320 shows a position determination from a positioning reference unit (where the position is known).

[0050] Block 325 shows a comparison of the position determinations from the AI ​​model, conventional approaches, and / or positioning reference units to determine a positioning error (eg, the difference between positions determined by different positioning methods) in block 330. This positioning error can then be used for various purposes.

[0051] For example, if the error exceeds a threshold, the AI ​​model may be retrained (calibrated), for example, using different training data, in block 310. As described above, there may also be a time period associated with the error, for example, an error exceeding a threshold for a predetermined time period may trigger retraining.

[0052] In another example, based on the positioning error and input from the legacy estimator 340 (such as a Doppler estimator), the selector 350 of the UE may select a positioning method to be applied to the output position. For example, if the error exceeds a threshold and the Doppler estimator indicates that the UE is operating in a high Doppler environment, the selector 350 may select a position determined by a legacy approach.

[0053] In another example, the positioning error may be used by the UE to indicate to the network the RS type and / or measurement feedback selection to be applied for positioning services, as shown in block 360. Figure 3 As shown, this selection shown in block 360 may be based on positioning error, or may be based on the AI ​​positioning output without taking positioning error into account.

[0054] According to further aspects of the exemplary embodiments, multiple inputs and / or input types may be used to perform AI-based positioning. AI-based positioning may be based on raw high-density data, such as channel state information (CSI), channel impulse response (CIR), and / or positioning features, such as layer 1 reference signal received power (L1-RSRP), power delay profile, beam index, NLOS, and Doppler. The interaction between these multiple inputs may affect the AI ​​architecture in various ways.

[0055] In one aspect of multi-input AI-based positioning, traditional estimation techniques may inform AI-based modeling. Figure 4aA diagram 400 is shown in which a Doppler estimate 405 informs a neural network (NN) update rate 410 for an AI-based positioning solution, according to various exemplary embodiments. Thus, in this example, a traditional velocity-related technique (Doppler) is used as an input to decide how often the AI ​​model (NN) should be updated.

[0056] In another example, a Doppler estimator may be used to select a NN from a plurality of trained NNs. Figure 4b A diagram 420 is shown in which a Doppler estimate 405 is used to select one of "n" trained NNs 425, according to various exemplary embodiments. In this example, there are three (3) trained NNs 425. Each of the NNs 425 may be trained based on a different scenario (e.g., mobility states such as stationary, slow moving, fast moving, etc.) for which the NN 425 is designed. The Doppler estimate 405 may be used to select the best NN 425 for a particular mobility scenario. For example, if the Doppler estimate 405 indicates that the UE is moving at a high velocity rate, the Doppler estimate 405 may be used to select a NN 425 trained for a high mobility scenario. Again, in this example, traditional velocity related techniques (Doppler) are used as input to decide which NN should be used for positioning purposes.

[0057] On the other hand of multi-input AI-based positioning, AI-based modeling can inform traditional estimation techniques. Figure 4c A diagram 430 is shown in which an AI model 435 that identifies line of site (LOS) / line of non-site (NLOS) conditions informs the execution of a traditional positioning technique 440 according to various exemplary embodiments. In this example, the traditional positioning method is a positioning method based on TDOA. However, this is merely exemplary, as other types of traditional positioning methods may also be supplemented with AI information. As will be appreciated by those skilled in the art, TDOA is based on the UE receiving reference signals (RS) from multiple transmission and reception points (TRPs). The TRPs may be different base stations, different cells of the same base station co-located, different cells of the same base station geographically separated, and the like. That is, the TRP may be any device capable of transmitting RS for the purposes of the positioning techniques described herein. Receiving RS for the purposes of TDOA (and other positioning techniques) depends on tight synchronization and LOS. By understanding the LOS / NLOS characteristics of different TRPs with respect to the UE, the UE may (a) identify the correct TRP to be used for TDOA estimation, (b) correct synchronization errors, or (c) select the correct traditional positioning technique to be used (e.g., using a technique different from TDOA). Therefore, in this example, the AI ​​model that outputs LOS / NLOS information is used as input to improve traditional positioning techniques.

[0058] Figure 4dA diagram 450 is shown in which a selector AI 455 identifies LOS / NLOS parameters of a UE relative to a positioning node (e.g., a TRP) and selects an AI-based positioning solution or a traditional positioning solution according to various exemplary embodiments. As described above, the positioning technique may depend on the LOS / NLOS of various TRPs for which the UE will measure or send RS. Therefore, an AI model 455 that can be used to identify specific LOS / NLOS scenarios for each TRP can be used as an input to make a basic selection as to whether the UE should perform an AI-based positioning solution 460 or a traditional positioning solution 465.

[0059] In another aspect of multi-input AI-based positioning, multiple AI models can be linked together, where the output of one AI model comprises the input of another AI model. That is, each AI model can be trained using different types of information, and the position output (or some other output) of the first model in the model can be used as an input to the second model in the model trained using different information. By using multiple models, the UE can refine the determined position.

[0060] Figure 4e A diagram 470 is shown in which the selector AI 455 identifies a scene and selects either the NN 425 or the traditional positioning solution 440 according to various exemplary embodiments. It should be understood that although the NN 425 has Figure 4b , but this does not mean that they need to be the same NN. Similarly, the conventional positioning scheme 440 has the same reference numerals as Figure 4c Although the TDOA positioning scheme 440 in the conventional positioning scheme 440 is given the same reference numeral, this does not mean that the conventional positioning scheme 440 must be a TDOA technique. In this example, each NN 425 can be trained based on different scenarios such as mobility, LOS / NLOS, etc. The selector AI model 455 can be used to select the NN 425 based on the environment the UE is experiencing, or even the conventional positioning scheme 440 if no NN 425 matches the current scenario. In addition, as discussed above, the selector AI 455 can select more than one of the NN 425 and / or the conventional positioning scheme 440, so that the output of one of the techniques can be used as an input to another technique.

[0061] Figure 4fA diagram 480 is shown in which update rate AI 485 is an input to the location update of AI model 410 according to various exemplary embodiments. Update rate AI 485 can provide an estimate of the confidence of the location. For example, the input to update rate AI 485 can include the location determined by AI model 410 and the location determined by one or more traditional positioning techniques or positioning reference units. Those skilled in the art will understand that for the purpose of positioning calibration, the network owner (or other entity) can deploy positioning reference units at known locations. Update rate AI 485 can be trained using various techniques to understand the relationship between these various inputs. Update rate AI 485 can run a model to determine whether it is necessary to retrain (update) AI model 410 based on the confidence that update rate AI 485 has in the location output of AI model 485.

[0062] Figure 4g A diagram 490 is shown in which a selector AI 495 selects a number of next generation node Bs (gNBs) as input to an AI-based positioning scheme 460 according to various exemplary embodiments. As described above, a gNB may be a base station that transmits and / or receives RS for the purpose of a positioning scheme. The selector AI 495 may be used to select the number and / or type of gNBs for the AI-based positioning scheme 460. For example, the selector AI 495 may identify a gNB for input into an AI-based positioning model. Data from a specified gNB is used as input to the AI-based positioning scheme 460. This input may include a channel impulse response (CIR) input that may be selected based on an AI-based LOS / NLOS classification, a candidate selected based on an L1-RSRP value relative to an AI-generated threshold, a candidate based on an optimal number of gNBs for positioning accuracy, and the like. The selector AI 495 may consider each of these factors and indicate to the AI-based positioning scheme 460 which gNBs and which inputs to use to determine the location.

[0063] Figure 5 A more generalized model 500 is shown in which multiple estimator stages may be used to determine position according to various exemplary embodiments. Each stage may consist of a mixture of AI models or traditional estimators (e.g., Doppler estimator 405, update rate AI 485, etc.). The selection of a stage may indicate the probability of selecting the next stage, e.g., NLOS means AI locator. The selector may be local to each stage or global. The positioning method may be as follows: Figure 5 The various ways shown implement these different stages to produce position output 510.

[0064] In additional aspects of multi-input based AI localization, a multi-input based convolutional NN (CNN) can be used with different types (and amounts) of inputs to each branch. In a first example, it can be considered that the CIR is more difficult to obtain with the required accuracy compared to the L1-RSRP and has more overhead. Therefore, for frequency range 2 (FR2), the beam index can also be more accurate in cases of low mobility and non-narrow beams. Thus, the CNN can implement rules such as for N gNBs, when M < N gNBs, use the CIR, when P ≤ N gNBs, use a simpler metric (e.g., L1-RSRP), and when Q ≤ N gNBs, use the beam index. In other examples, a multi-input NN with different classes of inputs can be used. In yet another example, in addition to the position, the output of the AI model can also include an accuracy metric or an error estimate.

[0065] Figure 6 FIG. 600 shows a multi-input based convolutional NN (CNN) according to various exemplary embodiments. In this example, there are three (3) NNs, a CIR NN 610, an L1-RSRP NN 620, and a beam NN 630. These NNs 610 to 630 can include fully connected layers that serve as inputs to the connection layer. As described above, the connection layer 640 can provide a position output based on the outputs of the three NNs 610 to 630. However, as Figure 6 shown, there can be additional layers 650 and 660 that can be used to process the inputs to determine the output position.

[0066] The various examples of traditional localization schemes interacting with AI and / or machine learning (ML) based localization schemes in a wireless network have been described above. Some exemplary use cases and their impact on the network will be provided below.

[0067] As described above, in some exemplary embodiments, the UE can implement only one localization scheme. In one example, the UE can implement an AI-only based localization scheme, where the UE position is the output of an AI model. In this example, the CIR and / or L1-RSRP can be inputs to the NN that generates the UE position output. Thus, in this example, the UE can perform CIR estimation or L1-RSRP measurement on multiple gNBs and can also provide corresponding feedback (e.g., to the LMF).

[0068] In another example, the UE may implement two types of positioning, where the AI-based positioning scheme does not output the UE location, but instead outputs information that may be useful to traditional positioning schemes. For example, the output of the AI ​​model may be the probability that a particular tap has LOS or NLOS with the UE. This input may be used by traditional positioning schemes (e.g., TDOA-based positioning) to determine which gNBs to use for positioning purposes. In another example, the output of the AI ​​model may be a time of arrival (TOA) estimate for each of the departure gNBs. This may be used as an input to TDOA-based positioning. In such an example, the UE may signal TOA instead of TDOA when performing positioning, for example, the UE may signal TOA instead of signaling LMF with TDOA information between two or more gNBs.

[0069] As described above, inputs to the AI ​​model may include CIR, L1-RSRP, power delay profile (PDP), beam index, etc. These inputs may be generated from NN inference input collection processes from multiple gNBs (e.g., any or all of the gNBs operating as positioning gNBs for the UE). These inference input collection processes may include pre-processing, signaling, measurements, and feedback related to the RS and corresponding NN inference inputs.

[0070] In some exemplary embodiments, AI may be used to optimize feedback for CIR, PDP, and L1-RSRP using an inferred entity (e.g., an LMF that may indicate a degree of optimization). To provide some examples, for CIR, the optimized feedback may include the timing and amplitude of each tap. For PDP, the optimized feedback may include the timing and energy amplitude of each tap. For L1-RSRP, the optimized feedback may not include the timing and amplitude of all taps. In addition, feedback may be switched between types for different gNBs, or there may be a modified PDP that includes the timing of a set of taps and the sum of the amplitudes of the set of taps.

[0071] Example

[0072] In a first embodiment, a method performed by a user equipment (UE) includes: determining that the UE is capable of performing a first positioning scheme, determining that the UE is capable of performing a second positioning scheme, selecting one of the first positioning scheme and the second positioning scheme for the UE to use to perform a positioning operation, and using one of the first positioning scheme and the second positioning scheme to calculate the position of the UE.

[0073] In a second embodiment, according to the method of the first embodiment, the method further includes sending a capability message to the network indicating that the UE supports the first positioning scheme and the second positioning scheme.

[0074] In a third embodiment, a method is described according to the first embodiment, wherein the first positioning scheme is one of the following: a global navigation satellite system (GNSS) (GPS), a wireless local area network (WLAN) positioning method, a terrestrial beacon system (TBS), a downlink time difference of arrival (DL-TDOA) positioning method, an uplink angle of departure (UL-AoD) positioning method, a DL angle of arrival (DL-AoA) positioning method, or a multiple round trip time (multi-RTT) positioning method.

[0075] In a fourth embodiment, according to the method of the first embodiment, the second positioning scheme is an artificial intelligence (AI) based positioning method using one or more trained models.

[0076] In a fifth embodiment, according to the method of the first embodiment, determining one of the first positioning solution and the second positioning solution is based at least on an environment in which the UE is operating.

[0077] In a sixth embodiment, the method according to the fifth embodiment, wherein the environment includes Doppler shift or non-line of sight (NLOS) with a base station that the UE is experiencing.

[0078] In a seventh embodiment, according to the method of the first embodiment, the method further comprises receiving a message from a location and management function (LMF) of a network to which the UE is connected, wherein determining one of the first positioning solution and the second positioning solution is based at least on the message.

[0079] In an eighth embodiment, according to the method of the first embodiment, the method further includes calibrating one of the first positioning scheme or the second positioning scheme using positioning data generated by the other of the first positioning scheme or the second positioning scheme.

[0080] In a ninth embodiment, according to the method described in the first embodiment, the method also includes: using the other of the first positioning scheme and the second positioning scheme to determine the position of the UE; comparing the positions determined by the first positioning scheme and the second positioning scheme; and performing additional operations based on the comparison of the positions determined by the first positioning scheme and the second positioning scheme.

[0081] In a tenth embodiment, the method according to the ninth embodiment, wherein the additional operation includes: when the comparison indicates that the difference between the positions determined by the first positioning scheme and the second positioning scheme is greater than a threshold for more than a predetermined time period, updating the AI ​​model for one of the first positioning scheme and the second positioning scheme.

[0082] In an eleventh embodiment, the method according to the ninth embodiment, wherein the further operation includes sending a message to a network indicating one of a reference signal (RS) parameter or measurement feedback applied to one of the first positioning scheme and the second positioning scheme.

[0083] In a twelfth embodiment, the method according to the eleventh embodiment, wherein the RS parameter includes one of a periodicity or a density of the RS signal for one of the first positioning scheme or the second positioning scheme.

[0084] In a thirteenth embodiment, a method is described according to the eleventh embodiment, wherein the RS signal includes one of a positioning reference signal (PRS) or a positioning sounding reference signal (P-SRS), and the RS parameter includes one of the following: turning the RS signal on or off or selecting one of a plurality of defined sets of RS signals.

[0085] In a fourteenth embodiment, the method of the ninth embodiment, wherein the operation includes selecting a position determined by the first positioning scheme or the second positioning scheme based at least on input from another estimator.

[0086] In a fifteenth embodiment, the method according to the first embodiment, wherein the first positioning scheme and the second positioning scheme include the UE receiving and measuring signals from one or more base stations, and wherein the determination is performed for each of the one or more base stations, and wherein the determination for each base station is based at least on whether the UE has a line of sight (LOS) or non-line of sight (NLOS) with the base station.

[0087] In a sixteenth embodiment, a processor of a user equipment (UE) is configured to perform any operation of the operations described in the first to fifteenth embodiments.

[0088] In a seventeenth embodiment, a user equipment (UE) includes: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform any of the operations described in accordance with the first to fifteenth embodiments.

[0089] In an eighteenth embodiment, a method performed by a user equipment (UE) includes storing one or more trained models, and performing an artificial intelligence (AI)-based positioning method using the one or more trained models to determine a location of the UE.

[0090] In a nineteenth embodiment, according to the method of the eighteenth embodiment, the method further includes: determining a speed of the UE based on Doppler estimation; and determining whether to update one or more trained models based on the speed.

[0091] In a twentieth embodiment, according to the method of the eighteenth embodiment, the method further includes: determining a velocity of the UE based on Doppler estimation; and selecting one of the one or more models to determine the position based on the velocity.

[0092] In a twenty-first embodiment, a method according to the eighteenth embodiment, wherein more than one trained model is used to determine the position, wherein the output of a first model is used as an input to a second model.

[0093] In a twenty-second embodiment, the method according to the eighteenth embodiment, wherein a trained model of the one or more trained models is used to determine the location, wherein the one of the trained models is based at least on the environment in which the UE is operating.

[0094] In a twenty-third embodiment, the method of the eighteenth embodiment is performed, wherein a trained model of the one or more trained models is used to determine the position, wherein the trained model of the trained models is based on input from the other of the first positioning scheme and the second positioning scheme.

[0095] In a twenty-fourth embodiment, a method is described according to the eighteenth embodiment, wherein one of the first positioning scheme and the second positioning scheme includes the UE receiving and measuring signals from one or more base stations, wherein one or more models are used to select a subset of one or more base stations to perform one of the first positioning scheme and the second positioning scheme.

[0096] In a twenty-fifth embodiment, a method according to the eighteenth embodiment is provided, wherein one of the first positioning scheme and the second positioning scheme includes inputting one of a channel input response (CIR), a layer 1 reference signal received power (L1-RSRP), or a beam index into one or more models, wherein one of the CIR, L1-RSRP, or a beam index is selected based at least on the number of base stations from which the UE is receiving a reference signal (RS) when performing one of the first positioning scheme and the second positioning scheme.

[0097] In a twenty-sixth embodiment, a method is described according to the eighteenth embodiment, wherein one of the first positioning scheme and the second positioning scheme includes a radio access technology (RAT)-based positioning method, the RAT-based positioning method includes the UE receiving and measuring signals from one or more base stations, and the other of the first positioning scheme and the second positioning scheme includes an AI model estimating line-of-sight (LOS) or non-line-of-sight (NLOS) information for each of the base stations.

[0098] In a twenty-seventh embodiment, the method according to the twenty-sixth embodiment, wherein the UE selects a subset of one or more base stations for performing the RAT-based positioning method based at least on LOS or NLOS information for each of the base stations.

[0099] In a twenty-eighth embodiment, according to the method of the twenty-sixth embodiment, wherein a synchronization error of a position determined by one of the first positioning scheme and the second positioning scheme is corrected based on at least LOS or NLOS information for each of the base stations.

[0100] In a twenty-ninth embodiment, a method is described according to the twenty-sixth embodiment, wherein the RAT-based positioning method includes more than one positioning method, and the UE selects each RAT-based positioning method in the RAT-based positioning method based at least on LOS or NLOS information for each base station in the base stations.

[0101] In a thirtieth embodiment, a processor of a user equipment (UE) is configured to perform any operation of the operations described in accordance with the eighteenth to twenty-ninth embodiments.

[0102] In a thirty-first embodiment, a user equipment (UE) includes: a transceiver configured to communicate with a network; and a processor communicatively coupled to the transceiver and configured to perform any of the operations described in embodiments 18 to 29.

[0103] Those skilled in the art will appreciate that the exemplary embodiments described above may be implemented with any suitable software configuration or hardware configuration or combination thereof. Exemplary hardware platforms for implementing the exemplary embodiments may include, for example, Intel x86-based platforms with compatible operating systems, Windows OS, Mac platforms and MAC OS, mobile devices with operating systems such as iOS, Android, etc. The exemplary embodiments of the above methods may be embodied as a program comprising lines of code stored on a non-transitory computer-readable storage medium, which, when compiled, may be executed on a processor or microprocessor.

[0104] Although this patent application describes various combinations of various embodiments each having different features, those skilled in the art will understand that any feature of an embodiment may be combined with features of other embodiments in any manner not publicly denied or with features that are not functionally or logically inconsistent with the operation or function of the device of the embodiments disclosed in the present invention.

[0105] It is understood that the use of personally identifiable information should be subject to privacy policies and practices that are generally recognized to meet or exceed industry or government requirements for maintaining user privacy. Specifically, personally identifiable information data should be managed and processed to minimize the risk of unintentional or unauthorized access or use, and the nature of the authorized use should be clearly stated to users.

[0106] It will be apparent to those skilled in the art that various modifications may be made to the present disclosure without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is intended to cover modifications and variations of the present disclosure, provided that these modifications and variations are within the scope of the appended claims and their equivalents.

Claims

1. A processor of a user equipment (UE), the processor being configured to perform operations comprising: Determining that the UE is capable of executing a first positioning scheme; Determining that the UE is capable of executing a second positioning scheme; selecting one of the first positioning scheme and the second positioning scheme to be used by the UE to perform a positioning operation; as well as The position of the UE is calculated using the one of the first positioning scheme and the second positioning scheme.

2. The processor of claim 1 , wherein the operations further comprise: A capability message indicating that the UE supports the first positioning scheme and the second positioning scheme is sent to a network.

3. The processor of claim 1 , wherein the first positioning scheme is one of: a global navigation satellite system (GNSS) (GPS), a wireless local area network (WLAN) positioning method, a terrestrial beacon system (TBS), a downlink time difference of arrival (DL-TDOA) positioning method, an uplink angle of departure (UL-AoD) positioning method, a DL angle of arrival (DL-AoA) positioning method, or a multiple round trip time (multi-RTT) positioning method. 4 . The processor of claim 1 , wherein the second positioning scheme is an artificial intelligence (AI)-based positioning method using one or more trained models. 5 . The processor of claim 1 , wherein determining one of the first positioning scheme and the second positioning scheme is based at least on an environment in which the UE is operating.

6. The processor of claim 1, wherein the operations further comprise: A message is received from a location and management function (LMF) of a network to which the UE is connected, wherein determining one of the first positioning solution and the second positioning solution is based at least on the message.

7. The processor of claim 1, wherein the operations further comprise: One of the first positioning solution or the second positioning solution is calibrated using positioning data generated by the other of the first positioning solution or the second positioning solution.

8. The processor of claim 1, wherein the operations further comprise: determining the position of the UE using the other of the first positioning scheme and the second positioning scheme; comparing the positions determined by the first positioning solution and the second positioning solution; and Further operations are performed based on a comparison of the positions determined by the first positioning solution and the second positioning solution.

9. The processor of claim 8, wherein the further operations comprise: When the comparison indicates that the difference between the positions determined by the first positioning solution and the second positioning solution is greater than a threshold for more than a predetermined period of time, an AI model is updated for the one of the first positioning solution and the second positioning solution.

10. The processor of claim 8, wherein the further operations comprise: (i) sending a message to a network indicating one of reference signal (RS) parameters or measurement feedback that should be used for the one of the first positioning scheme and the second positioning scheme, or (ii) selecting the position determined by the first positioning scheme or the second positioning scheme based at least on input from another estimator.

11. The processor of claim 1 , wherein the first positioning scheme and the second positioning scheme include the UE receiving and measuring signals from one or more base stations, and wherein the determination is performed for each of the one or more base stations, and wherein the determination for each base station is based at least on whether the UE has line of sight (LOS) or non-line of sight (NLOS) with the base station.

12. A processor of a user equipment (UE), the processor being configured to perform operations comprising: storing one or more trained models; and An artificial intelligence (AI) based positioning method using the one or more trained models is performed to determine the location of the UE.

13. The processor of claim 12, wherein the operations further comprise: determining a velocity of the UE based on a Doppler estimate; as well as A determination is made based on the speed whether to update the one or more trained models.

14. The processor of claim 12, wherein the operations further comprise: determining a velocity of the UE based on a Doppler estimate; as well as Selecting one of one or more models to determine the position based on the velocity.

15. The processor of claim 12, wherein more than one trained model is used to determine the position, wherein an output of a first model is used as an input to a second model.

16. The processor of claim 12, wherein a trained model of the one or more trained models is used to determine the location, wherein the one of the trained models is based at least on an environment in which the UE is operating.

17. The processor of claim 12, wherein one of the one or more trained models is used to determine the location, wherein the one of the trained models is based on input from the other of a first positioning scheme and a second positioning scheme.

18. The processor of claim 12, wherein one of the first positioning scheme and the second positioning scheme comprises the UE receiving and measuring signals from one or more base stations, wherein one or more models are used to select a subset of the one or more base stations to perform the one of the first positioning scheme and the second positioning scheme.

19. The processor of claim 12, wherein one of the first positioning scheme and the second positioning scheme comprises inputting one of a channel input response (CIR), a layer 1 reference signal received power (L1-RSRP), or a beam index into one or more models, wherein the CIR, the L1-RSRP, or the beam index is selected based at least on the number of base stations from which the UE is receiving reference signals (RSs) when performing the first positioning scheme and the second positioning scheme.

20. The processor of claim 12, wherein one of the first positioning scheme and the second positioning scheme comprises a radio access technology (RAT)-based positioning method, the RAT-based positioning method comprising the UE receiving and measuring signals from one or more base stations, and the other of the first positioning scheme and the second positioning scheme comprises an AI model estimating line-of-sight (LOS) or non-line-of-sight (NLOS) information for each of the base stations.

Images