Method and apparatus for user equipment positioning estimation based on artificial intelligence

Abstract

The present application discloses a method, in which new AI-based positioning related functions are introduced in NG-RAN nodes to enable the UE to select between AI and non-AI positioning models for UE position estimation and / or enable the UE to select a suitable AI model among different AI positioning models according to an indication provided by the UE. The method also defines measurement criteria for selecting AI or non-AI models for UE positioning estimation and defines related signaling procedures to enable proper interaction between different NR positioning related nodes.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of wireless communication systems, and more specifically, to methods and apparatuses for direct positioning estimation based on artificial intelligence / machine learning (AI / ML). More specifically, the present application relates to enhancing radio access network (RAN) signaling and procedures in new radio (NR) specifications to support AI-based direct positioning methods, thereby improving the accuracy of user equipment (UE) position estimation. BACKGROUND

[0002] At the 3GPP RAN#94 meeting, a new study item (SI) on artificial intelligence / machine learning (AI / ML) for NR air interface was approved, with the main goal of exploring better support for AI / ML algorithms by enhancing air interface functionalities, thereby improving performance and / or reducing complexity / overhead. The study item aims to study several carefully selected application scenarios, such as CSI feedback, beam management, and positioning accuracy enhancement, and identify areas where AI / ML can improve the performance of air interface functionalities.

[0003] Regarding the application scenario of positioning enhancement, it was agreed at the 3GPP RAN1#109 meeting to study two AI / ML-based positioning method options in NR Release-18 specifications. The first is a direct AI / ML positioning method, where an AI / ML model will replace existing positioning methods to provide the final UE position estimate. The second is an indirect AI / ML positioning method or AI-aided method, where an AI model is used to assist existing positioning methods to improve the accuracy of UE position estimation. In addition, it was decided at the meeting to further study the potential specification impact of AI / ML positioning methods and provide relevant inputs, including aspects of AI / ML model indication / configuration, such as assistance signaling and procedures, including model configuration, model activation / deactivation, model recovery / termination, and model selection, etc.

[0004] For the direct AI / ML positioning method, it is proposed in multiple contributions submitted in 3GPP RAN1#109 meeting that this option can face some technical challenges. These challenges include: due to the introduction of new types of positioning measurements, it can lead to increased signaling overhead or complexity issues; the generalization ability of AI models when changing the UE scenario; backward compatibility issues with existing NR positioning methods; high requirements for processing power or computing resources in the entity deploying direct AI / ML positioning. Therefore, corresponding device, method, process or signaling enhancement measures are needed to effectively integrate such direct AI / ML positioning model in the NR specification, while considering the above technical challenges. SUMMARY

[0005] The purpose of the present application is to propose a method and device based on a direct artificial intelligence / machine learning (AI / ML) positioning model for user equipment (UE) position estimation.

[0006] In a first aspect of the present application, a method for user equipment (UE) position estimation based on a direct AI / ML positioning model is provided, which is executed by a communication network system, comprising: deploying a direct AI positioning related function or information element in a radio access network (RAN) entity, an access and mobility management function (AMF) and / or a location management function (LMF) entity; the direct AI positioning related function allows the node responsible for UE position estimation (such as UE or LMF) to select the appropriate AI / ML model between different AI models, and / or to make adaptive selection between direct AI positioning method and non-AI positioning method based on a set of parameters or configuration in the communication network system entity; the parameter set or indication includes at least one of the following parameters: performance error of existing AI or non-AI positioning model; processing capability limit of AI positioning model deployment entity (such as UE type information); required positioning accuracy level of application (i.e. UE position estimation requirement requested by application); processing load, overhead or computing resource availability state of AI model deployment entity.

[0007] In a second aspect of the present application, a communication network system is provided, comprising: a memory; a transceiver; a processor coupled with the memory and the transceiver; wherein the processor is configured to execute the above method.

[0008] In a third aspect of the present application, a non-transitory machine-readable storage medium having instructions stored thereon, when executed by a computer, cause the computer to perform the above method.

[0009] In a fourth aspect of the present application, a chip comprising a processor configured to invoke and run a computer program stored in a memory to cause a device installed with the chip to perform the above method.

[0010] In a fifth aspect of the present application, a computer-readable storage medium having a computer program stored therein, wherein the computer program causes a computer to perform the above method.

[0011] In a sixth aspect of the present application, a computer program product comprising a computer program, wherein the computer program causes a computer to perform the above method.

[0012] In a seventh aspect of the present application, a computer program, wherein the computer program causes a computer to perform the above method. BRIEF DESCRIPTION OF DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present application or related art, the following will describe various embodiments and briefly introduce the drawings. Obviously, these drawings are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without departing from the scope of the present application.

[0014] Figure 1 is a schematic diagram of an example of an NR positioning system architecture with different AI positioning model locations and collaboration options according to embodiments of the present application.

[0015] Figure 2 is a schematic diagram of an example of a direct AI positioning and AI assisted positioning method according to embodiments of the present application.

[0016] Figure 3 is a schematic diagram of an example of an NR positioning system architecture in which direct AI related positioning functions are located at gNB according to embodiments of the present application.

[0017] Figure 4 is a schematic diagram of an example of an NR positioning system architecture in which direct AI positioning related functions are located at AMF / LMF according to embodiments of the present application.

[0018] Figure 5 is a schematic diagram of an example of an NR positioning system architecture in which direct AI models are located at UE and AI positioning related functions are located at gNB or AMF / LMF in the case of no UE-network collaboration according to embodiments of the present application.

[0019] Figure 6 is a schematic diagram of an example of a NR positioning system architecture assuming no UE-network collaboration, according to embodiments shown herein, in which the direct AI positioning model is located at the gNB side, while the AI positioning related functions are located at the gNB or AMF / LMF side.

[0020] Figure 7 is a schematic diagram of an example of a NR positioning system architecture assuming no UE-network collaboration, according to embodiments shown herein, in which the direct AI model is located at the LMF / AMF, while the AI positioning related functions are located at the gNB or AMF / LMF side.

[0021] Figure 8 is a schematic diagram of an example of a NR positioning system architecture under the collaboration option of AI model transmission from gNB to UE, according to embodiments shown herein.

[0022] Figure 9 is a schematic diagram of an example of a NR positioning system architecture under the collaboration option of AI model transmission from LMF to UE, according to embodiments shown herein.

[0023] Figure 10 is a schematic diagram of an example of providing AI model indication in the RRC signaling response of the UE to gNB’s position measurement indication, according to embodiments shown herein.

[0024] Figure 11 is a schematic diagram of an example of providing AI model indication in the NR-PPa signaling request of the LMF to gNB sending positioning information response or update message, according to embodiments shown herein.

[0025] Figure 12 is a schematic diagram of an example of RRC signaling configuration 0 / 1 of gNB to UE exchanging model transmission related information, according to embodiments shown herein.

[0026] Figure 13 is a schematic diagram of an example of providing model transmission related information in the NR-PPa signaling response of gNB to LMF sending positioning information response or update message, according to embodiments shown herein.

[0027] Figure 14 is a block diagram of a communication network system, according to embodiments shown herein.

[0028] Figure 15 is a flowchart of a method for user equipment (UE) location estimation based on direct AI / ML positioning model, according to embodiments shown herein.

[0029] Figure 16 is a block diagram of a wireless communication system, according to embodiments shown herein. DETAILED DESCRIPTION

[0030] The technical content, structural features, purposes and effects of the present application are described in detail below in combination with the drawings. Specifically, the terms in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not a limitation on the present application.

[0031] At the 3GPP RAN#94 meeting, a new study item (SI) on AI / ML for NR air interface was approved, with the main goal of exploring better support for AI / ML algorithms by enhancing air interface functions to improve performance and / or reduce complexity / overhead. The goal of this study item is to focus on the study of several carefully selected application scenarios, such as channel state information (CSI) feedback, beam management, and positioning accuracy enhancement, to identify areas where AI / ML can improve the performance of air interface functions, and to study a general AI / ML framework, including the functional requirements of AI / ML architecture and the specification impact required to implement AI / ML techniques to improve air interface performance. Regarding NR AI positioning accuracy enhancement, it was agreed at the 3GPP RAN1#109 meeting to further study the deployment location of different AI positioning models, such as: AI models located on one side or a single entity (UE, gNB or LMF), i.e. both training and inference are performed on a single end; some AI models are deployed on one side, i.e. training and inference are mainly performed on the network or UE side, but additional signaling or process optimization between the two ends is required, and may be combined with the existing signaling framework. In addition, it was agreed at the meeting to consider different levels of network and UE cooperation, such as: no cooperation; signaling-based cooperation (no model transmission); signaling-based cooperation (with model transmission) (see Figure 1 for details). It was also decided to further study the applicability of AI / ML model generalization in NR positioning.

[0032] In addition to the above agreement, it was agreed at the RAN1#109 meeting to further study the following AI / ML-based positioning enhancement techniques, including: a) direct AI positioning: using AI positioning methods to replace existing positioning methods, and using new types of measurement data and / or measurement reports for UE position estimation, including: signal amplitude or power; time of arrival (TOA); angle of arrival (AoA); channel impulse response (CIR); beam index; transmission and reception point (TRP) index; reflection order of signals received from multiple TRPs (see Figure 2b) Indirect or AI-assisted positioning: AI is used as an auxiliary means to optimize existing positioning methods to improve the accuracy of UE position estimation. By extracting intermediate features from measurement data and / or measurement reports, such as: downlink-reference signal timing difference (DL-RSTD); downlink-angle of departure (DL-AoD); uplink-relative time of arrival (UL-RTOA); uplink sounding reference signal-reference signal received power (UL SRS-RSRP); downlink positioning reference signal RSRP (DL PRS-RSRP); UE receive-transmit time difference (UERx-Tx Time Difference); gNB receive-transmit time difference (gNB Rx-Tx Time Difference); uplink angle of arrival (UL-AOA), angle of departure (AoD), or zenith angle of arrival (ZOA) (calculated per path); then, based on the extracted intermediate features, the final UE position is estimated (see Figure 2

[0033] In this report, we focus on direct AI-based positioning methods, as this approach may require the introduction of new measurement methods and / or additional information exchange or procedures, which can increase system overhead and / or system complexity, thus reducing network efficiency. In addition, this approach also has the following problems: since the training of AI / ML models is usually highly dependent on specific scenarios or geographical distribution, when the scenario changes, the adaptability of the model may be limited. Direct AI positioning methods may be difficult to be compatible with current non-AI positioning standards. The computational requirements of direct AI positioning models may exceed the processing capacity or computing resources of entities such as UEs, gNBs, or LMFs. Therefore, it is currently unclear how to integrate direct AI / ML positioning models into NR specifications while balancing the following factors: signaling overhead, system complexity, generalization ability of AI models, backward compatibility issues, processing capacity or computing resource availability of AI model running entities, and achievable positioning accuracy level in UE position estimation for direct AI positioning methods. In view of the above technical challenges, some embodiments of the present application propose a method for effectively integrating direct AI positioning models while taking into account the above technical factors.

[0034] ​To balance the above technical factors and the positioning accuracy that can be achieved by direct AI-based positioning estimation models, the present application proposes to introduce direct AI positioning related functions or information elements in existing NR specifications, such as deployed in gNB and / or AMF / LMF entities. The role of this function is to act as a management entity for AI / ML positioning models, responsible for the indication / configuration of AI / ML models, including: activation / deactivation of AI models; recovery / termination of models; selection of models; transmission of models; management applicable to different model deployment entities and collaboration levels, regardless of where the model is deployed (such as UE, gNB or LMF) or the degree of collaboration required between these entities. The main goal of the direct AI positioning related function is to allow the node responsible for UE position estimation (such as UE or LMF) to select the appropriate model between different AI / ML positioning models; allow adaptive selection between direct AI / ML positioning methods and non-AI positioning methods to achieve UE position estimation. The selection of AI models can be based on existing parameters or configurations, determined by the node responsible for positioning estimation (such as UE, gNB or LMF); explicit or implicit indication, through signaling message transmission from the final position estimation node. The parameters or indications of AI model selection can include at least one of the following factors: performance error of existing AI or non-AI positioning models; processing capability limit of AI positioning model deployment entity (gNB, UE, LMF) (such as UE type information); required positioning accuracy level of application (i.e. accuracy requirement of UE position estimation); processing load, overhead or computing resource availability status of AI positioning model deployment entity (such as gNB, UE or LMF).

[0035] To support different direct AI / ML positioning model position options and different collaboration levels, the present application discloses or defines different levels of procedures and interaction signaling between UE, LMF and AI direct positioning related functions at gNB or LMF. Detailed information of these signaling and procedures is provided in the following embodiments.

[0036] Embodiment 1: Signaling and procedures considering direct AI positioning function located at gNB.

[0037] Considering that the direct AI positioning function is located at gNB, the procedures and interactions between AI direct positioning related functions at UE, LMF and gNB can be as shown in Figure 3 and described as follows:

[0038] The UE or core network (CN) entity (such as 5G location server) sends a positioning or location service request for the UE to the entity responsible for the final position estimation or calculation (i.e. location management function LMF).

[0039] The LMF entity forwards or initiates a location estimation procedure request (e.g. through NR-PPa signaling request) to the NG-RAN node or gNB, which can contain explicit or implicit indication of at least one AI-related positioning parameter as mentioned above.

[0040] Based on this indication or internal NG-RAN node configuration, the node can request the UE (e.g. through RRC signaling configuration) to provide location information, including explicit indication of enabling or supporting AI-based positioning methods, or implicit indication based on the above-mentioned parameters (or requiring new measurements to support AI models at gNB or LMF), or indicating whether it has AI positioning models or needs model transfer from LMF or gNB.

[0041] The UE can respond to the gNB request through a signaling configuration 0 response message containing explicit or implicit indication of AI model support and / or model transfer information.

[0042] Based on the indication signaling configuration 0 response message from the UE, or the indication provided by LMF to gNB, the gNB or AI-based related positioning function at gNB can decide to provide / configure one of the following through signaling configuration 1: positioning reference signal (PRS) or positioning sounding reference signal (SRS) measurement gap configuration, measurement object and / or measurement indication configuration to support AI or non-AI based positioning methods; or, the gNB can provide both types of configurations according to the indication, in order to guarantee other requirements (such as guaranteeing certain positioning KPI, reducing signaling overhead, or ensuring fallback to non-AI model in order to allow selection between different AI positioning models when UE scenario or condition changes). In addition, the gNB can also configure to transfer AI positioning inference model information / configuration to the UE, or transfer inference input data to the UE before transferring AI-related measurement object, indication or gap configuration. In another scheme, the gNB can transfer AI / non-AI related information to LMF through the interface connecting NG-RAN node and AMF (i.e. NR-PPa signaling response), as well as AI / non-AI related measurement data received from UE, for final position calculation. Optionally, the gNB can exchange AI / non-AI positioning support information related to UE with another RAN node through X2 / Xn interface, to support non-AI or AI positioning methods at this node.

[0043] Based on the measurement gap configured by the NG-RAN node, the UE can perform AI / non-AI related measurements as described below and report them back to the gNB.

[0044] Direct AI is based on measurements, such as measuring the amplitude, time of arrival (TOA), angle of arrival (AoA), channel impulse response (CIR), beam index, TRP index, power, and / or signal reflection order of the received signal from a gNB or multiple gNBs.

[0045] Alternatively, based on non-AI measurements, such as measuring downlink RSTD and uplink RTOA, downlink PRS-RSRP and uplink SRS-RSRP, UE transceiver time difference (Rx-Tx) and gNB transceiver time difference (Rx-Tx), and uplink AoA, ZoA or AoD values on each path.

[0046] After receiving the signaling configuration 1 from the gNB, the UE can send a response message containing the measurement report to the gNB or LMF through the signaling response 1 for the final position estimation; or if the AI model has been deployed at the UE end, or transmitted to the UE from the LMF or gNB, the UE can perform local position estimation or inference.

[0047] Embodiment 2: Signaling and procedures considering that the direct AI positioning function is located at the LMF:

[0048] When the direct AI positioning function is located at the LMF, the procedures and interactions between the UE, LMF, and AI direct positioning related functions at the LMF can be as shown in ( Figure 4 ) and described as follows:

[0049] The UE or core network (CN) entity (such as the 5G location server) sends a positioning or location service request for the UE to the entity responsible for the final position estimation or calculation (i.e., the location management function LMF).

[0050] The LMF entity forwards or initiates a position estimation process request to the gNB (e.g., through NR-PPa signaling request), or can optionally send a request to the UE (e.g., through non-access stratum (NAS) signaling). The request can contain a request or query about supporting or activating / deactivating the direct AI positioning method on the RAN entity (e.g., gNB or UE).

[0051] Based on the request of the LMF entity, the gNB can request the UE (e.g., through RRC signaling configuration 0) to provide location information, including explicit indication about enabling or supporting the direct AI based positioning method, or implicit indication based on the above parameters; whether to provide new measurement data to support the direct AI model at the gNB or LMF; whether the UE has an AI positioning model, or whether the model needs to be transmitted from the LMF or gNB.

[0052] The UE can respond to the gNB's request by providing a Signaling Configuration 0 Response message containing explicit or implicit indication about AI model support / activation / deactivation and / or model transfer information.

[0053] Based on the Signaling Configuration 0 Response message provided by the UE and the indication provided by the LMF to the gNB, the gNB can decide to provide / configure, through Signaling Configuration 1 : Positioning Reference Signal (PRS) or Positioning Sounding Reference Signal (SRS) measurement gap configuration; Measurement Object and / or Measurement Indication configuration for AI or non-AI based positioning methods; or, the gNB can provide both types of configurations in combination to guarantee other requirements (e.g. guarantee positioning KPIs, reduce signaling overhead, ensure fallback to non-AI model when conditions change, or allow selection between different AI positioning models). In addition, the gNB can indicate the LMF to transfer AI positioning inference model information / configuration, or transfer inference input data to the UE before transferring AI related measurement objects, indications or gap configurations. In another approach, the gNB can transfer AI / non-AI related information to the LMF through the interface connecting the NG-RAN node and the AMF (i.e. NR-PPa Signaling Response), as well as AI / non-AI related measurement data received from the UE, to support model transfer and / or final position computation. Optionally, the gNB can exchange AI / non-AI positioning support information related to the UE with another RAN node through X2 / Xn interface to support non-AI or AI positioning methods on that node.

[0054] Based on the measurement gap configured by the NG-RAN node, the UE can perform AI / non-AI related measurements as described below and report them back to the gNB, or provide them to the LMF through Non-Access Stratum (NAS) signaling.

[0055] Direct AI based on measurements, such as measuring the amplitude, time of arrival (TOA), angle of arrival (AoA), channel impulse response (CIR), beam index, TRP index, power, and / or signal reflection order of signals received from the gNB or multiple gNBs.

[0056] Non-AI based on measurements, such as measuring downlink RSTD and uplink RTOA, downlink PRS-RSRP and uplink SRS-RSRP, UE receive-transmit time difference (Rx-Tx) and gNB receive-transmit time difference (Rx-Tx), uplink AoA, ZoA, or AoD values on each path.

[0057] After receiving the signaling configuration 1 from gNB, the UE can send a response message containing the measurement report to gNB or LMF through signaling response 1 for final position estimation; or if the AI model is deployed at UE side, or transmitted to UE from LMF or gNB, the UE can perform local position estimation or inference.

[0058] Embodiment 3: No UE-network cooperation, direct AI positioning model at UE side.

[0059] This embodiment is applicable to scenarios with high positioning accuracy requirements and AI model deployed at UE side, as shown in Figure 5 In this case, the direct AI related functions can be located at gNB or LMF. In this scenario, gNB instructs UE to perform measurements and provides an indication to UE to perform one of the following operations: run AI model to estimate position and return the estimated position result; optionally, provide parameters required for position calculation to LMF; or directly provide measurement data to LMF for LMF to calculate the final UE position. Subsequently, LMF provides the final position estimation result to the entity requesting UE position.

[0060] Embodiment 4: No UE-network cooperation, direct AI positioning model at gNB side.

[0061] This embodiment is also applicable to scenarios with high positioning accuracy requirements and AI model deployed at gNB side, as shown in Figure 6 In this case, the direct AI related functions can be located at gNB or LMF. In this scenario, gNB or LMF instructs UE to perform measurements and reports the measurement results back to gNB. Based on the measurement results, gNB can perform one of the following operations: run AI model for position inference and provide the final position estimation result to LMF; or directly provide measurement data to LMF for LMF to calculate / estimate the final position by traditional positioning methods. Subsequently, LMF provides the final position estimation result to the entity requesting UE position.

[0062] Embodiment 5: No UE-network cooperation, direct AI positioning model at LMF side.

[0063] This embodiment is also applicable to scenarios with high positioning accuracy requirements and AI model deployed at LMF side, as shown in Figure 7 In this case, the direct AI related functions can be located at gNB or LMF. In this scenario, gNB or LMF instructs UE to perform measurements and reports the measurement results back to gNB, and then gNB transmits the measurement data and an indication about AI / non-AI model usage to LMF. According to the indication, LMF can perform AI or non-AI positioning calculation. Subsequently, UE provides the final position estimation result to the entity requesting UE position.

[0064] Embodiment 6: Collaboration option AI model transfer from gNB to UE.

[0065] This embodiment is applicable to scenarios where the positioning accuracy requirement is high, but the UE itself does not have an AI model, but supports AI models, and the delay allows the download of AI models, as shown in Figure 8 In this case, the gNB can instruct the UE to provide AI model transfer related information. Subsequently, the gNB transfers the required direct AI positioning AI model to the UE and configures the UE to perform AI based measurement gap / object measurement. Then, the gNB instructs the UE to perform the measurement and / or calculate the required position, or provide AI based data to the LMF for final position estimation.

[0066] Embodiment 7: Collaboration option AI model transfer from network to UE.

[0067] This embodiment is applicable to scenarios where the positioning accuracy requirement is high, but the UE itself does not have an AI model, but supports AI models, and the delay allows the download of AI models, as shown in Figure 9 In this case, the gNB can instruct the UE to provide AI model transfer related information to the gNB or directly to the LMF through NAS signaling. Subsequently, the gNB requests the LMF to transfer the required direct AI positioning AI model to the UE. In addition, the gNB also configures the UE to perform AI based measurement gap / object measurement and instructs the UE to perform the measurement and calculate the final UE position, and then provides the final UE position estimation result to the LMF.

[0068] Embodiment 8: Signaling messages and instructions.

[0069] In the above method, the UE provides explicit or implicit indication to the NG-RAN node over the air interface, i.e. RRC signaling response 0 message, which is an RRC signaling message sent by the UE to the gNB over uplink logical channel. The RRC signaling response 0 message can optionally belong to one of the following categories: first category of messages, e.g. uplink information transfer RRC message, location measurement indication RRC message, UE assistance information RRC message. Second category of messages, e.g. UE information response RRC message, UE positioning assistance information RRC message (new RRC message defined for AI positioning integration). The RRC signaling response 0 message can contain the following: explicit AI positioning model indication information, including: assistance information, e.g. model error, processing load and / or accuracy level indication, for the gNB or LMF to decide whether to activate or deactivate the AI positioning model (this information is represented in the AI-ModelAssistanceInfo IE field). Direct request for activation or deactivation of AI model (this information is represented in the AI-ModelDirectIndication IE field). AI inference input transfer or AI positioning model transfer request information, including: whether AI model inference input transfer (InferenceInput IE) is needed, whether AI positioning model transfer (ModeTransferInfo IE) is needed, if model transfer is needed, the UE can indicate the need and provide the model ID to the gNB. Measurement type indication information, including: which type of measurement the gNB needs to configure for the UE to support AI positioning method or non-AI based positioning method. The related information is contained in the Non-AIMeasurementInfoList-rel-18 IE and AI-MeasurementInfoList-rel-18 IE. Based on the measurement type indication, the gNB can perform the following operations: configure measurement gap, configure measurement parameters, configure one or more measurement objects to support AI or non-AI based positioning method. In addition, for examples of explicit or implicit indication provided in the location measurement indication RRC message, please refer to Figure 10 .

[0070] In the above method, the LMF / AMF entity of the core network provides explicit or implicit indication to the NG-RAN through NR Positioning Protocol A (NR-PPa) signaling messages. The NR-PPa signaling request message can be one of the following types: Positioning Information Request / Update message; Measurement Initiation Request message; Positioning Activation Request message; a new NR-PPa signaling message defined for exchanging AI / non-AI positioning method related information between LMF / AMF and NG-RAN nodes. The NR-PPa signaling message can contain the following: explicit AI positioning model indication information, including: assistance information such as model error, processing load and / or accuracy level indication for gNB or LMF to decide whether to activate or deactivate AI positioning model (this information is represented in the AI-ModelAssistanceInfo IE field). Directly request activation or deactivation of AI model (this information is represented in the AI-ModelDirectIndication IE field). AI model inference input transmission or AI positioning model transmission request information, including: whether AI model inference input transmission (InferenceInput IE) is needed, whether AI positioning model transmission (ModeTransferInfo IE) is needed, if model transmission is needed, UE can indicate the need (AImodelTransfer IE) and provide model ID to gNB (AI-Model-Id IE). Examples of explicit or implicit indication provided in the positioning information request NR-PPa message sent by LMF to NG-RAN nodes can be referred to Figure 11 .

[0071] In the above method, the NG-RAN node transmits AI model or measurement gap configuration RRC signaling configuration 0 or RRC signaling configuration 1 to the UE, which can take the following forms: RRC setup; RRC reconfiguration; RRC resume; RRC release; RRC reestablishment; RRC logged measurement configuration message; a new RRC message defined for exchanging AI positioning related information between gNB and UE. The RRC signaling configuration 0 message can contain the following: AI positioning activation and deactivation information (this information is represented in the AI-Pos-ModelSupport IE field). AI model inference input configuration (this information is represented in the ModelInferenceInput IE field). AI positioning model transmission configuration information, including: asking UE whether AI model transmission from gNB is needed; providing a set of AI models that can be selected by UE (this information is represented in the ModelTransferInfo IE field). Examples of RRC signaling configuration 0 messages carrying AI positioning related information sent by gNB to UE can be referred to Figure 12 .

[0072] In the above method, the NR-PPa signaling response or update message can be used to transmit AI / non-AI model transfer information from gNB to LMF, which can be one of the following types: positioning information response; positioning information update message; new NR-PPa signaling message defined for exchanging AI / non-AI positioning method related information between NG-RAN node and LMF. The NR-PPa signaling response or update message can contain the following contents: AI positioning activation and deactivation information (this information is represented in the AI-Pos-ModelSupport IE field). AI model inference input configuration (this information is represented in the ModelInferenceInput IE field). AI positioning model transfer configuration information, including: asking whether AI model needs to be transferred from gNB to specified UE; providing a set of AI models for UE to choose (this information is represented in the ModelTransferInfo IE field). For examples of NR-PPa signaling response or update messages carrying AI positioning related information sent by LMF to gNB or UE, please refer to Figure 13 .

[0073] Figure 14 It is illustrated that in some embodiments, one or more user equipment (UE) 10, RAN node (e.g., gNB) 20 and network node (e.g., LMF or AMF) 30 communicate in a communication network system 40. According to embodiments of the present application, the communication network system 40 includes one or more UEs 10, RAN nodes 20 and network nodes 30 (e.g., LMF or AMF). The one or more UEs 10 can include a memory 12, a transceiver 13 and a processor 11, wherein the processor 11 is connected to the memory 12 and the transceiver 13. The RAN node 20 can include a memory 22, a transceiver 23 and a processor 21, wherein the processor 21 is connected to the memory 22 and the transceiver 23. The network node 30 can include a memory 32, a transceiver 33 and a processor 31, wherein the processor 31 is connected to the memory 32 and the transceiver 33. The processor 11, 21 or 31 can be configured to perform the proposed functions, processes and / or methods described in the present specification. Layers of the radio interface protocol can be implemented in the processor 11, 21 or 31. The memory 12, 22 or 32 is operatively coupled with the processor 11, 21 or 31 and stores various information to enable the processor 11, 21 or 31 to operate. The transceiver 13, 23 or 33 is operatively coupled with the processor 11, 21 or 31 and is responsible for transmitting and / or receiving wireless signals.

[0074] The processors 11, 21, or 31 can include application-specific integrated circuit (ASIC), other chip sets, logic circuit, and / or a data processing device. The memories 12, 22, or 32 can include read-only memory (ROM), random access memory (RAM), flash memory, storage card, storage medium, and / or other storage devices. The transceivers 13, 23, or 33 can include baseband circuitry for processing radio frequency signals. When the embodiments are implemented in software, the techniques described herein can be implemented using a suitably-programmed processor 11, 21, or 31. The processor 11, 21, or 31 can be programmed using software instructions, which can refer to applications 14, 24, or 34. The software instructions can be read into the memory 12, 22, or 32 from another computer-readable medium or from another device via the transceivers 13, 23, or 33. The software instructions can be stored in the memory 12, 22, or 32 until needed for execution, after which the software instructions can be executed by the processor 11, 21, or 31. The memory 12, 22, or 32 can also be, or include, non-transitory machine-readable media. The transceivers 13, 23, or 33 can be configured to transmit and / or receive radio frequency signals using the antennas 15, 25, or 35. The transceivers 13, 23, or 33 can be implemented as a single transceiver or as multiple transceivers. The transceivers 13, 23, or 33 can be communicatively coupled to the processors 11, 21, or 31 via the bus 16, 26, or 36. The transceivers 13, 23, or 33 can be configured to transmit and / or receive signals under the control of the processors 11, 21, or 31. The antennas 15, 25, or 35 can be implemented as single antennas or as multiple antennas. The antennas 15, 25, or 35 can be communicatively coupled to the transceivers 13, 23, or 33 via the bus 16, 26, or 36. The antennas 15, 25, or 35 can be configured to transmit and / or receive signals under the control of the processors 11, 21, or 31.

[0075] Figure 15 is a flowchart of a method for user equipment (UE) position estimation based on direct AI / ML positioning model according to embodiments of the present application. Figure 15 It is explained that in some embodiments, the method comprises the following steps: step 1502: using direct AI positioning related functions or information elements in radio access network (RAN) entities, access and mobility management function (AMF) and / or location management function (LMF) entities. Step 1504: performing direct AI positioning related functions so that the node responsible for UE position estimation (such as UE or LMF) can select the appropriate model from different AI / ML models; adaptively select between direct AI positioning method and non-AI positioning method based on a set of parameters or configuration of communication network system entities; or adaptively select between AI positioning method and non-AI positioning method based on explicit or implicit indication transmitted by the final position estimation node through signaling messages. The parameters or indications include at least one of the following parameters: performance error of existing AI or non-AI positioning model; processing capability limit of AI positioning model deployment entity (such as UE type information); required positioning accuracy level of application (i.e. accuracy requirement indicated by the application requesting UE position estimation); processing load, overhead or computing resource availability status of AI model deployment entity.

[0076] In summary, some embodiments of the present application provide a method that relies on the introduction of a new AI-based positioning related function in the NG-RAN or LMF node that allows the UE to adaptively select between AI and non-AI positioning models, and defines within this function the metrics that can be used to select an AI or non-AI positioning model for position estimation, while defining the related signaling procedures to ensure the correct interaction between the UE, NG-RAN and LMF entities. The main advantages of the method provided by some embodiments of the present application include: the new method introduces a mechanism that makes the selection of AI or non-AI positioning methods configurable, enabling the best balance between the level of positioning accuracy and network efficiency, signaling overhead and complexity that can be brought by the adoption of AI models. The new method provides flexibility to support both AI and non-AI methods and to enable selection between different AI models, or to adaptively combine these methods to support each UE to select the appropriate model (e.g., select the best model with less error in case of multiple models), and to fall back to non-AI models when needed (e.g., when the scenario changes cause the model performance error). The new method helps to avoid the unnecessary transmission of AI model related information over multiple network interfaces. For example, when the AI positioning model is located at the UE or LMF, AI model related information needs to be transmitted over the interface between the UE and the NG-RAN node and the interface between the NG-RAN node and the LMF, while this method can effectively avoid this situation. The new method reuses the existing NR positioning signaling and procedures, thus minimizing the specification impact brought by the integration of AI in positioning.

[0077] Figure 16 is a block diagram of a wireless communication system in accordance with embodiments shown by the present application. The embodiments described herein can be implemented into the system using any suitably configured hardware and / or software. Figure 16 The system 700 is shown to include radio frequency (RF) circuitry 710, baseband circuitry 720, application circuitry 730, a memory / storage 740, a display 750, a camera 760, a sensor 770, and an input / output (I / O) interface 780, which are at least partially coupled together via one or more buses 790, as shown. The application circuitry 730 can include circuitry such as, but not limited to, one or more single-core or multi-core processors. The processors can include any combination of general-purpose processors and dedicated processors (e.g., graphics processors, application processors). The processors can be coupled with memory / storage and configured to execute instructions stored in the memory / storage to enable various applications and / or operating systems to run on the system.

[0078] While the present application has been described in connection with what is considered the most practical and preferred embodiments, it is recognized that the application is not limited to the disclosed embodiments, but is intended to cover any arrangements which fall within the scope of the appended claims interpreted in their broadest manner.

Claims

1. A method for direct artificial intelligence / machine learning, AI / ML, positioning model for user equipment, UE, position estimation, performed by a communication network system, characterized by, Comprise: Methods for UE position estimation based on direct AI / ML based positioning models, employing direct AI positioning related functions or information elements in wireless access network RAN entities, access and mobility management function AMF and / or location management function LMF entities, wherein the direct AI positioning related functions allow the node responsible for UE position estimation to select a suitable AI / ML model among different AI models and / or to make an adaptive selection between direct AI positioning methods or non-AI positioning methods based on a set of parameters or a configuration of the communication network system entity, and / or based on an explicit or implicit indication passed through signaling messages from the node responsible for the final position estimation; wherein the set of parameters or the indication comprises at least one of the following parameters: performance error of existing AI or non-AI positioning models, processing capability limit of the entity deploying the AI positioning model, application required UE position estimation accuracy level indication, and / or processing load, overhead or computing resource availability status of the entity deploying the AI model.

2. The method of claim 1, wherein, Also comprise employing different levels of procedures and interactions between the UE, the LMF entity and the related AI direct positioning functions to support different direct AI positioning options and different levels of cooperation.

3. The method of claim 2, wherein, Wherein, Employing the different levels of procedures and interactions between the UE, the LMF entity and the related AI direct positioning functions comprises considering procedures to place the direct AI positioning functions on a gNB.

4. The method of claim 3, wherein, Considering procedures to place the direct AI positioning functions on the gNB comprises: The UE or core network CN sends a positioning or location service request of the UE to the LMF entity; The LMF entity forwards or initiates a location estimation procedure request to a next generation wireless access network NG-RAN node or the gNB, wherein the location estimation procedure request contains an explicit or implicit indication related to at least one AI positioning parameter; Based on the explicit or implicit indication or internal configuration of the NG-RAN node, the NG-RAN node requests the UE to provide location information, including: an explicit indication about enabling or supporting AI positioning methods, an implicit indication based on at least one AI positioning parameter, or providing new measurement data to support AI models of the gNB or the LMF entity, or indicating whether the UE is equipped with AI positioning models or needs to acquire models from the LMF entity or the gNB; The UE provides a signaling configuration 0 response message containing AI model support and / or model transmission information in response to the gNB request; and The LMF entity or the gNB provides the UE with an AI model or a set of AI models based on the UE's capability and the AI model support indication. Based on the signaling configuration 0 response message from the UE or the indication provided by the LMF entity to the gNB, the gNB or the AI related positioning function on the gNB decides whether to provide / configure the positioning reference signal (PRS) or positioning sounding reference signal (SRS) measurement gap configuration, measurement object and / or measurement indication configuration by signaling configuration 1 to support AI or non-AI positioning methods, or select to combine both configurations according to the indication, to ensure other requirements including positioning KPI, reduce signaling overhead, or guarantee fallback to non-AI model when conditions change, or allow selection between different AI positioning models; Based on the measurement gap configured by the NG-RAN node, the UE performs AI / non-AI related measurements and reports the AI / non-AI related measurement results to the gNB; Upon receiving the signaling configuration 1 from the gNB, the UE sends a response containing measurement report to the gNB or the LMF entity by signaling response 1 for final position estimation, or in the case that AI model is deployed at UE side or transferred from the LMF entity or the gNB to the UE side, the UE performs local position estimation or inference.

5. The method of claim 4, wherein, The gNB is configured to transmit AI positioning inference model information / configuration or inference input data to the UE before transmitting AI measurement object, measurement indication or measurement gap configuration.

6. The method of claim 4, wherein, The gNB transmits the AI / non-AI related information and AI / non-AI measurement data provided by the UE to the LMF entity through the interface connecting the NG-RAN node and the AMF for final position calculation.

7. The method of claim 6, wherein, The AI / non-AI positioning measurements include: direct AI measurements including measuring the amplitude, time of arrival (TOA), angle of arrival (AoA), channel impulse response (CIR), beam index, TRP index, power and / or reflection order of the signals received from the gNB or multiple gNBs; or non-AI measurements including measuring DL-RSTD and UL-RTOA, UE receive-transmit time difference and gNB receive-transmit time difference, and UL-AOA and ZOA values of each path.

8. The method of claim 6, wherein, The gNB exchanges AI / non-AI positioning support information related to the UE with another RAN node through X2 / Xn interface to support non-AI or AI positioning methods of the other node.

9. The method of claim 2, wherein, Different levels of procedures and interactions are adopted between the UE, the LMF entity and the related AI direct positioning function, including procedures considering the direct AI positioning function being located at the LMF entity.

10. The method of claim 9, wherein, Procedures considering the direct AI positioning function being located at the LMF entity include: the UE or core network (CN) sends a positioning or location service request of the UE to the LMF entity; the LMF entity forwards or initiates a location estimation procedure request to the gNB or the UE, wherein the location estimation procedure request contains a request or query about supporting or activating or deactivating direct AI positioning methods at the RAN entity; Based on the request from the LMF entity, the gNB configures the UE to provide location information by RRC signaling configuration 0 request including explicit indication about enabling or supporting direct AI positioning method or implicit indication based on at least one AI positioning parameter, or requiring to provide new measurements for supporting the gNB or the LMF entity AI model, or indicating whether owning AI positioning model or requiring model transfer from the LMF entity or the gNB; The UE responds to the gNB's request by signaling configuration 0 response message containing explicit or implicit indication about AI model support and / or model transfer information; Based on the indication in the signaling configuration 0 response message from the UE or the indication provided by the LMF entity to the gNB, the gNB decides to provide / configure positioning reference signal, PRS, or positioning sounding reference signal, SRS, measurement gap configuration, measurement object and / or measurement indication configuration by signaling configuration 1 to support AI or non-AI based positioning method, or selectively combine both types of configuration according to the indication to ensure including positioning KPI or reducing signaling overhead, or guarantee fallback to non-AI model when condition changes, or allow selection among different AI positioning models; Based on the measurement gap configured by the NG-RAN node, the UE performs AI / non-AI related measurements and reports the AI / non-AI related measurement results back to the gNB; Upon receiving the signaling configuration 1 from the gNB, the UE responds to the gNB or the LMF entity by signaling response 1 providing response containing measurement report for final location estimation; or in case of AI model deployed at the UE or transferred to the UE from the LMF entity or the gNB, the UE locally performs location estimation or inference.

11. The method of claim 2, wherein, Different levels of procedures and interactions are employed among the UE, the LMF entity and the related AI direct positioning function including procedures considering no UE-network cooperation and direct AI positioning model located at the UE.

12. The method of claim 11, wherein, Procedures considering no UE-network cooperation and direct AI positioning model located at the UE include: gNB instructs the UE to perform measurements and provides indication to the UE to run model to estimate location and return estimated location, or to provide position calculation parameters to the LMF entity, or to provide pure measurement data to the LMF entity to calculate final location of the UE, and the LMF entity provides final location estimation result to the entity requesting the UE location.

13. The method of claim 2, wherein, Different levels of procedures and interactions are employed among the UE, the LMF entity and the related AI direct positioning function including procedures considering no UE-network cooperation and direct AI positioning model located at the gNB. Different levels of procedures and interactions are employed among the UE, the LMF entity and the related AI direct positioning function including procedures considering no UE-network cooperation and direct AI positioning model located at the gNB.

14. The method of claim 13, wherein, The procedure considering no UE-network cooperation and the direct AI positioning model located at the gNB includes that the gNB or the LMF entity instructs the UE to perform measurements and report the measurement results back to the gNB, based on the measurement results, the gNB runs an AI model for position inference, and provides a final position estimate to the LMF entity, or provides pure measurement data to the LMF entity to calculate / estimate a final position using a traditional positioning procedure, and provides the final position estimate to an entity requesting the UE position.

15. The method of claim 2, wherein, The procedures and interactions at different levels among the UE, the LMF entity and the related AI direct positioning function include a procedure considering no UE-network cooperation and the direct AI positioning model located at the LMF entity.

16. The method of claim 15, wherein, The procedure considering no UE-network cooperation and the direct AI positioning model located at the LMF entity includes that the gNB or the LMF entity instructs the UE to perform measurements and report the measurement results back to the gNB, based on the measurement results, the gNB provides measurement data and an indication of using an AI / non-AI model to the LMF entity, the LMF entity performs AI or non-AI based position calculation according to the indication, and provides a final position estimate to an entity requesting the UE position.

17. The method of claim 2, wherein, The procedures and interactions at different levels among the UE, the LMF entity and the related AI direct positioning function include a procedure considering a cooperation option of transmitting an AI model from a gNB to the UE.

18. The method of claim 17, wherein, The procedure considering a cooperation option of transmitting the AI model from the gNB to the UE includes that the gNB instructs the UE to provide information related to AI model transmission to the gNB, transmits a required direct AI positioning AI model to the UE, configures the UE to perform AI based measurement gap / object measurement, instructs the UE to perform measurement and / or calculate a required position, and provides AI based measurement results to the LMF entity for final position estimation.

19. The method of claim 2, wherein, The procedures and interactions at different levels among the UE, the LMF entity and the related AI direct positioning function include a procedure considering a cooperation option of transmitting an AI model from a network to the UE.

20. The method of claim 19, wherein, The procedure considering a cooperation option of transmitting the AI model from the network to the UE includes that the gNB instructs the UE to provide information related to AI model transmission to the gNB or directly to the LMF entity through NAS signaling, requests the LMF entity to transmit a required direct AI positioning AI model to the UE, configures the UE to perform AI based measurement gap / object measurement, instructs the UE to perform measurement and calculate a final UE position, and provides the final UE position estimate to the LMF entity.

21. The method of claim 3, wherein, The explicit or implicit indication provided by the UE to the NG-RAN node over the air interface is a RRC signaling configuration response message sent by the UE to the gNB on an uplink logical channel.

22. The method of claim 21, wherein, The RRC message includes a first type of message containing an uplink information transfer message, a location measurement indication, or UE assistance information; or the RRC message includes a second type of message containing a UE information response, UE positioning assistance information, or a new RRC message defined for AI positioning integration.

23. The method of claim 3, wherein, The explicit or implicit indication provided by the LMF entity of the core network CN to the NG-RAN through a new radio positioning protocol, NR-PPa, signaling request is a location information transfer / request procedure / signaling / message.

24. The method of claim 23, wherein, The NR-PPa signaling request message is a positioning information request / update message, a measurement initiation request message, a positioning activation request message, or a new NR-PPa signaling message defined for exchanging AI / non-AI positioning method related information between the LMF entity and the NG-RAN node.

25. The method of claim 3, wherein, The explicit indication in the message is a direct request to configure measurement gaps and / or provide configuration to indicate support for AI or non-AI positioning methods, or to provide assistance information indicating the same, or as an input for inference or training.

26. The method of claim 3, wherein, The RRC signaling configuration 0 / 1 for AI model or measurement gap transfer is an RRC setup, RRC reconfiguration, RRC resume, RRC release, RRC reestablishment, RRC logged measurement configuration message, or a newly defined RRC message.

27. The method of claim 3, wherein, The NR-PPa signaling message for AI / non-AI model transfer information from the gNB to the LMF entity is a positioning information response message, a positioning information update message, or a newly defined NR-PPa signaling message sent from the gNB to the LMF entity for exchanging AI / non-AI positioning method related information between the NG-RAN node and the LMF entity.

28. A communication network system, characterized by comprises: a memory; a transceiver; and a processor coupled with the memory and the transceiver; wherein the processor is configured to perform the method of any one of claims 1 to 27.

Images