Methods and apparatuses for supporting a life cycle management (LCM) of an artificial intelligence (AI) / machine learning (ML) model or functionality supported at a user equipment (UE)
By aligning Life Cycle Management approaches for AI/ML models through signaling exchanges between UE and network nodes, the proposed solution addresses the challenges of managing AI/ML models in wireless communication systems, improving stability and efficiency.
Patent Information
- Application Number
- PCT/CN2024/110648
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-08-08
- Publication Date
- 2025-06-12
AI Technical Summary
Current wireless communication systems lack efficient methods for managing the lifecycle of artificial intelligence (AI)/machine learning (ML) models deployed on user equipment (UE), leading to potential collisions in LCM approaches and frequent activation/deactivation issues.
The proposed solution involves designing signaling exchanges between UE and network nodes to align Life Cycle Management (LCM) approaches for AI/ML models, including stopping applicability-related information transmission upon deactivation and allowing fallback to non-AI/ML behaviors when performance degrades.
This approach ensures seamless alignment of LCM approaches between UE and network nodes, reducing collisions and frequent activations/deactivations, thereby enhancing the stability and efficiency of AI/ML model management in wireless communication systems.
Smart Images

Figure CN2024110648_12062025_PF_FP_ABST
Abstract
Description
METHODS AND APPARATUSES FOR SUPPORTING A LIFE CYCLE MANAGEMENT (LCM) OF AN ARTIFICIAL INTELLIGENCE (AI) / MACHINE LEARNING (ML) MODEL OR FUNCTIONALITY SUPPORTED AT A USER EQUIPMENT (UE)TECHNICAL FIELD
[0001] The present disclosure relates to wireless communications, and more specifically to methods and apparatuses for supporting an a life cycle management (LCM) of an artificial intelligence (AI) / machine learning (ML) model or functionality supported at a user equipment (UE) .BACKGROUND
[0002] A wireless communications system may include one or multiple network communication devices, such as base stations, which may support wireless communications for one or multiple user communication devices, which may be otherwise known as user equipment (UE) , or other suitable terminology. The wireless communications system may support wireless communications with one or multiple user communication devices by utilizing resources of the wireless communication system (e.g. time-domain resources (e.g. symbols, slots, subframes, frames, or the like) or frequency-domain resources (e.g. subcarriers, carriers, or the like) . Additionally, the wireless communications system may support wireless communications across various radio access technologies including third generation (3G) radio access technology, fourth generation (4G) radio access technology, fifth generation (5G) radio access technology, among other suitable radio access technologies beyond 5G (e.g. sixth generation (6G) ) .SUMMARY
[0003] An article "a" before an element is unrestricted and understood to refer to "at least one" of those elements or "one or more" of those elements. The terms "a, " "at least one, " "one or more, " and "at least one of one or more" may be interchangeable. As used herein, including in the claims, "or" as used in a list of items (e.g. a list of items prefaced by a phrase such as "at least one of" or "one or more of" or "one or both of" ) indicates an inclusive list such that, for example, a list of at least one of A, B, or C means A or B or C or AB or AC or BC or ABC (i.e., A and B and C) . Also, as used herein, the phrase "based on" shall not be construed as a reference to a closed set of conditions. For example, an example step that is described as "based on condition A" may be based on both a condition A and a condition B without departing from the scope of the present disclosure. In other words, as used herein, the phrase "based on" shall be construed in the same manner as the phrase "based at least in part on. Further, as used herein, including in the claims, a "set" may include one or more elements.
[0004] Some implementations of the present disclosure provide a user equipment (UE) . The UE includes at least one memory; and at least one processor coupled to the at least one memory and configured to cause the UE to: transmit capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; receive a first configuration regarding applicability-related information for the one or more AI / ML functionalities; and determine whether to transmit the applicability-related information and determine one or more elements included in the applicability-related information according to the first configuration.
[0005] In some implementations of the UE described herein, the capability information is transmitted to a network node via a radio resource control (RRC) UE capability report message or via a long term evolution (LTE) positioning protocol (LPP) provide capabilities message.
[0006] In some implementations of the UE described herein, the capability information indicates at least one of the following: the one or more AI / ML functionalities that can be supported by the UE; one or more life cycle management (LCM) approaches that can be supported by the UE for the one or more AI / ML functionalities; or at least one LCM approach adopted by the UE for each of the one or more AI / ML functionalities.
[0007] In some implementations of the UE described herein, the one or more AI / ML functionalities that can be supported by the UE and the one or more LCM approaches that can be supported by the UE have one-to-one mapping relationship or one-to-multiple or multiple-to-one mapping relationship.
[0008] In some implementations of the UE described herein, the one or more LCM approaches that can be supported by the UE include at least one of the following: an approach for monitoring the one or more AI / ML functionalities; an approach for determining applicability of the one or more AI / ML functionalities; an approach for making an activation decision for the one or more AI / ML functionalities; or an approach for making a deactivation decision for the one or more AI / ML functionalities.
[0009] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to: determine the at least one LCM approach adopted by the UE; or receive, from the network node, a message including: one or more LCM approaches to be adopted by the UE; and a configuration regarding the one or more LCM approaches to be adopted by the UE.
[0010] In some implementations of the UE described herein, the message is an RRC reconfiguration message or an LPP request location information message.
[0011] In some implementations of the UE described herein, the one or more elements included in the applicability-related information indicates at least one of the following: whether the one or more AI / ML functionalities are applicable; whether the one or more AI / ML functionalities are applicable at the UE and informing that a network node needs to further determine applicability of the one or more AI / ML functionalities; or informing the network node to determine the applicability of the one or more AI / ML functionalities.
[0012] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to: receive, from a network node, additional condition information associated with the applicability-related information; and determine whether the one or more AI / ML functionalities are applicable based on the additional condition information.
[0013] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to transmit the applicability-related information to a network node, and wherein the applicability-related information is carried in at least one of the following: a long term evolution (LTE) positioning protocol (LPP) provide capabilities message; a radio resource control (RRC) UE assistance information message; an RRC reconfiguration complete message; or an RRC resume complete message.
[0014] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to receive an RRC other configuration message including information indicating the UE to use the RRC UE assistance information message to carry the applicability-related information.
[0015] In some implementations of the UE described herein, if at least one AI / ML functionality within the one or more AI / ML functionalities has been activated or has been reported to be applicable, the at least one processor is configured to cause the UE to: stop transmitting the applicability-related information for the at least one AI / ML functionality for at least one of a time period or a pre-defined condition; and resume transmitting the applicability-related information for the at least one AI / ML functionality after the at least one of the time period or the pre-defined condition.
[0016] In some implementations of the UE described herein, transmitting the applicability-related information for the at least one AI / ML functionality is stopped in response to at least one of the following: the at least one AI / ML functionality is deactivated by a decision from the network node; the at least one AI / ML functionality is deactivated by a decision of the UE; the at least one AI / ML functionality is deactivated by the decision of the UE, and a deactivation result of the at least one AI / ML functionality is reported by the UE to the network node; the at least one AI / ML functionality is deactivated by the decision of the UE, the deactivation result of the at least one AI / ML functionality is reported by the UE to the network node, and a confirmation is received by the UE from the network node; or the at least one AI / ML functionality is determined by the UE to be non-applicable.
[0017] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to consider that the at least one AI / ML functionality is to be non-applicable in response to at least one of the following: the at least one AI / ML functionality is deactivated by the decision from the network node; the at least one AI / ML functionality is deactivated by the decision of the UE; the at least one AI / ML functionality is deactivated by the decision of the UE, and the deactivation result of the at least one AI / ML functionality is reported by the UE to the network node; or the at least one AI / ML functionality is deactivated by the decision of the UE, the deactivation result of the at least one AI / ML functionality is reported by the UE to the network node, and a confirmation is received by the UE from the network node.
[0018] In some implementations of the UE described herein, the first configuration includes a second configuration regarding the at least one of the time period or the pre- defined condition; or the at least one processor is configured to cause the UE to receive an RRC other configuration message including the second configuration.
[0019] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to: stop determining whether the at least one AI / ML functionality is applicable during the time period; and resume determining whether the at least one AI / ML functionality is applicable after the time period.
[0020] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to receive an AI / ML activation or deactivation decision from a network node.
[0021] In some implementations of the UE described herein, the AI / ML activation or deactivation decision is carried in at least one of the following: a long term evolution (LTE) positioning protocol (LPP) request location information message; a radio resource control (RRC) reconfiguration message; or a medium access control (MAC) control element (CE) .
[0022] In some implementations of the UE described herein, if at least one AI / ML functionality within the one or more AI / ML functionalities has been activated, the at least one processor is configured to cause the UE to: determine whether the at least one AI / ML functionality is degrading or whether the at least one AI / ML functionality is applicable; and fallback to a non-AI / ML behavior or a default AI / ML method, if the at least one AI / ML functionality is degrading or if the at least one AI / ML functionality is not applicable, and if it is upon a network node’s final decision to deactivate the at least one AI / ML functionality.
[0023] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to receive, from the network node, a configuration indicating that the UE can fallback in case that the at least one AI / ML functionality is degrading or in case that the at least one AI / ML functionality is not applicable without the network node confirming a deactivation of the at least one AI / ML functionality.
[0024] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to transmit, to the network node, information indicating that the UE will fallback to the non-AI / ML behavior or the default AI / ML method.
[0025] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to: measure a measurement result; and transmit the measurement result to the network node.
[0026] In some implementations of the UE described herein, the at least one processor is configured to cause the UE to: perform a location estimation for the UE using a non-AI / ML method; and transmit a location of the UE estimated by the location estimation to the network node.
[0027] In some implementations of the UE described herein, the location of the UE is transmitted in a long term evolution (LTE) positioning protocol (LPP) provide location information message.
[0028] In some implementations of the UE described herein, the network node is at least one of a location management function (LMF) or a radio access network (RAN) node.
[0029] Some implementations of the present disclosure provide a processor for wireless communication, comprising at least one controller coupled with at least one memory and configured to cause the processor to: transmit capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; receive a first configuration regarding applicability-related information for the one or more AI / ML functionalities; and determine whether to transmit the applicability-related information and determine one or more elements included in the applicability-related information according to the first configuration.
[0030] Some implementations of the present disclosure provide a method performed by user equipment (UE) . The method includes: transmitting capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; receiving a first configuration regarding applicability-related information for the one or more AI / ML functionalities; and determining whether to transmit the applicability-related information and determining one or more elements included in the applicability-related information according to the first configuration.
[0031] Some implementations of the present disclosure provide a network node. The network node includes at least one memory; and at least one processor coupled to the at least one memory and configured to cause the network node to: receive, from a user equipment (UE) , capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; and transmit, to the UE, a first configuration regarding applicability-related information for the one or more AI / ML functionalities.
[0032] In some implementations of the network node described herein, the capability information is received via a radio resource control (RRC) UE capability report message or via a long term evolution (LTE) positioning protocol (LPP) provide capabilities message.
[0033] In some implementations of the network node described herein, the capability information indicates at least one of the following: the one or more AI / ML functionalities that can be supported by the UE; one or more life cycle management (LCM) approaches that can be supported by the UE for the one or more AI / ML functionalities; or at least one LCM approach adopted by the UE for each of the one or more AI / ML functionalities.
[0034] In some implementations of the network node described herein, the one or more AI / ML functionalities that can be supported by the UE and the one or more LCM approaches that can be supported by the UE have one-to-one mapping relationship or one-to-multiple or multiple-to-one mapping relationship.
[0035] In some implementations of the network node described herein, the one or more LCM approaches that can be supported by the UE include at least one of the following: an approach for monitoring the one or more AI / ML functionalities; an approach for determining applicability of the one or more AI / ML functionalities; an approach for making an activation decision for the one or more AI / ML functionalities; or an approach for making a deactivation decision for the one or more AI / ML functionalities.
[0036] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to transmit, to the UE, a message including: one or more LCM approaches to be adopted by the UE; and a configuration regarding the one or more LCM approaches to be adopted by the UE.
[0037] In some implementations of the network node described herein, the message is an RRC reconfiguration message or an LPP request location information message.
[0038] In some implementations of the network node described herein, the applicability-related information indicates at least one of the following: whether the one or more AI / ML functionalities are applicable; whether the one or more AI / ML functionalities are applicable at the UE and informing that the network node needs to further determine applicability of the one or more AI / ML functionalities; or informing the network node to determine the applicability of the one or more AI / ML functionalities.
[0039] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to: if the applicability-related information indicates whether the one or more AI / ML functionalities are applicable, consider that the UE has fully determined the applicability of the one or more AI / ML functionalities; if the applicability-related information indicates whether the one or more AI / ML functionalities are applicable at the UE and informing that the network node needs to further determine applicability of the one or more AI / ML functionalities, consider that the UE has partially determined the applicability of the one or more AI / ML functionalities and further determine a final applicability of the one or more AI / ML functionalities considering a condition at the network node; or if the applicability-related information indicates informing the network node to determine the applicability of the one or more AI / ML functionalities, determine the applicability of the one or more AI / ML functionalities considering the condition at the network node.
[0040] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to receive the applicability-related information from the UE, and wherein the applicability-related information is carried in at least one of the following: a long term evolution (LTE) positioning protocol (LPP) provide capabilities message; a radio resource control (RRC) UE assistance information message; an RRC reconfiguration complete message; or an RRC resume complete message.
[0041] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to transmit, to the UE, an RRC other configuration message including information indicating the UE to use an RRC UE assistance information message to carry the applicability-related information.
[0042] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to transmit additional condition information associated with the applicability-related information to the UE, and wherein the additional condition information is used by the UE to determine whether the one or more AI / ML functionalities are applicable.
[0043] In some implementations of the network node described herein, the first configuration includes a second configuration regarding at least one of a time period or a pre-defined condition, wherein the at least one of the time period or the pre-defined condition can be used by the UE to stop transmitting the applicability-related information for the one or more AI / ML functionalities; or the at least one processor is configured to cause the network node to transmit, to the UE, an RRC other configuration message including the second configuration.
[0044] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to transmit an AI / ML activation or deactivation decision to the UE.
[0045] In some implementations of the network node described herein, the AI / ML activation or deactivation decision is carried in at least one of the following: a long term evolution (LTE) positioning protocol (LPP) request location information message; a radio resource control (RRC) reconfiguration message; or a medium access control (MAC) control element (CE) .
[0046] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to transmit, to the UE, a configuration indicating that the UE can fallback in case that at least one AI / ML functionality is degrading or in case that the at least one AI / ML functionality is not applicable without the network node confirming a deactivation of the at least one AI / ML functionality.
[0047] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to receive, from the UE, information indicating that the UE will fallback to a non-AI / ML behavior or a default AI / ML method.
[0048] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to receive a measurement result measured by the UE from the UE.
[0049] In some implementations of the network node described herein, the at least one processor is configured to cause the network node to receive, from the UE, a location of the UE estimated by the UE performing a location estimation using a non-AI / ML method.
[0050] In some implementations of the network node described herein, the location of the UE is transmitted in a long term evolution (LTE) positioning protocol (LPP) Provide location information message.
[0051] In some implementations of the network node described herein, the network node is at least one of a location management function (LMF) or a radio access network (RAN) node.
[0052] Some implementations of the present disclosure provide a processor for wireless communication, comprising at least one controller coupled with at least one memory and configured to cause the processor to: receive, from a user equipment (UE) , capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; and transmit, to the UE, a first configuration regarding applicability-related information for the one or more AI / ML functionalities.
[0053] Some implementations of the present disclosure provide a method performed by a network node. The method includes: receiving, from a user equipment (UE) , capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; and transmitting, to the UE, a first configuration regarding applicability-related information for the one or more AI / ML functionalities.BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Figure 1 illustrates an example of a wireless communications system 100 in accordance with aspects of the present disclosure.
[0055] Figure 2 illustrates an example of a user equipment (UE) 200 in accordance with aspects of the present disclosure.
[0056] Figure 3 illustrates an example of a processor 300 in accordance with aspects of the present disclosure.
[0057] Figure 4 illustrates an example of a network equipment (NE) 400 in accordance with aspects of the present disclosure.
[0058] Figures 5 and 6 illustrate flowcharts of methods related to an AI / ML model or functionality in accordance with aspects of the present disclosure.
[0059] Figures 7A and 7B illustrate schematic diagrams of aligning an LCM approach for an AI / ML model or functionality in accordance with aspects of the present disclosure.
[0060] Figures 8 and 9 illustrate schematic diagrams of transmitting an applicability-related report for an AI / ML model or functionality in accordance with aspects of the present disclosure.DETAILED DESCRIPTION
[0061] In general, 3GPP has been discussing using AI / ML technology to optimize access stratum operation, an AI / ML model or functionality can be deployed at a UE's side to support the following use cases:
[0062] (1) Channel State Information (CSI) feedback compression
[0063] - An AI / ML model or functionality can be used to compress a CSI report at the UE's side, and another AI / ML model or functionality can be used to decompress a CSI report at the RAN node's side.
[0064] (2) CSI temporal domain prediction
[0065] - An AI / ML model or functionality can be used to predict a future measurement result of one set of CSI-RS based on a historical measurement result of another set of CSI-RS.
[0066] (3) Beam spatial domain prediction
[0067] - An AI / ML model or functionality can be used to predict a measurement result (e.g., layer-1 (L1) beam level reference signal received power (RSRP) , or a best beam index) of one set of beams based on a measurement result of another set of beams.
[0068] (4) Beam temporal domain prediction
[0069] - An AI / ML model or functionality can be used to predict a future measurement result (e.g., L1 beam level RSRP, or a best beam index) of one set of beams based on a historical measurement result another set of beams.
[0070] (5) AI / ML assisted positioning estimation
[0071] - An AI / ML model or functionality can be used to improve the accuracy of an intermediate result (e.g., line of sight (LOS) or non-line of sight (NLOS) ) which will be further used to determine the UE's location.
[0072] (6) Direction positioning
[0073] - An AI / ML model or functionality can be used to directly determine the UE's location
[0074] (7) RRM temporal domain prediction
[0075] - Predict the measurement result (e.g., layer-3 (L3) beam or cell level RSPR, or a best beam or cell index) in the future for one set of cell or beam using a measurement result in the past of another set of cell or beam.
[0076] (8) RRM spatial domain prediction
[0077] - Predict a measurement result (e.g., L3 beam or cell level RSPR, or best beam or cell index) for one set of cell or beam using a measurement result of another set of cell or beam at the same time instance, wherein these two sets of cell or beam are of the same frequency. The associated cells could be collocated.
[0078] (9) Radio resource management (RRM) frequency domain prediction
[0079] - Predict a measurement result (e.g., L3 beam or cell level RSPR, or best beam or cell index) for one set of cell or beam using a measurement result of another set of cell or beam at the same time instance, wherein these two sets of cell or beam are of different frequencies.
[0080] An AI / ML model or functionality deployed at a UE may also be named as "a UE-sided AI / ML model or functionality" or the like. A UE-sided AI / ML model or functionality can be applicable and activated for inference only if a UE is under a circumstance (e.g., a current network configuration, the UE's side condition, a network's side condition) that is consistent with the circumstance under which the AI / ML model or functionality is trained.
[0081] Besides, after the UE's sided AI / ML model or functionality is activated, a UE or a network (NW) should keep monitoring the performance of the AI / ML model or functionality and deactivate the AI / ML model or functionality if the performance of the AI / ML model or functionality is degrading.
[0082] So far, some possible options with respect to an AI / ML model or functionality include: "#1 Monitoring, " "#2 Applicability Determination, " and "#3 Activation or Deactivation Decision" as below. "#3 Activation or Deactivation Decision" may include "#3 Activation Decision" or "#3 Deactivation Decision. " For example:
[0083] #1 Monitoring an AI / ML model or functionality may include the following two LCM approaches:
[0084] · #1a NW side monitoring
[0085] · NW calculates AI / ML model or functionality performance metrics (e.g., throughput, CSI overhead, squared generalized cosine similarity (SGCS) , accuracy) .
[0086] · A UE calculates AI / ML model or functionality performance metrics (e.g., throughput, CSI overhead, SGCS, accuracy) and report the metrics to NW.
[0087] · #1b UE side monitoring
[0088] · A UE calculates AI / ML model or functionality performance metrics (e.g., throughput, CSI overhead, SGCS, accuracy) and determines whether the AI / ML model or functionality performance is degrading.
[0089] #2 Applicability Determination of an AI / ML model or functionality may include the following four LCM approaches:
[0090] · #2a Applicability determination is made by a UE
[0091] · #2b Applicability determination is made by NW
[0092] · #2c Applicability determination is made by a UE based on initial filtering by NW
[0093] · #2d Applicability determination is made by NW based on initial filtering by a UE #3 Activation or Deactivation Decision for an AI / ML model or functionality may include the following three LCM approaches:
[0094] · #3a NW makes an activation or deactivation decision for an AI / ML model or functionality, and the NW initiates AI / ML management
[0095] · #3b a UE autonomously makes an activation or deactivation decision for an AI / ML model or functionality, and, and such decision for decision is reported to NW
[0096] · #3c NW makes an activation or deactivation decision for an AI / ML model or functionality, and a UE initiates AI / ML management
[0097] For instance, regarding LCM approach #2a "Applicability Determination made by UE, " a UE decides the applicable AI / ML models or functionalities based on UE-side additional conditions (internally known by the UE) and NW-side additional conditions (if provided) . In a special case for LCM approach #2a, a network node (e.g. a RAN node or an LMF) sending NW-side additional conditions to the UE may be in advance for applicable AI / ML model or functionality determination.
[0098] Regarding LCM approach #2b "Applicability Determination made by NW, " a network node (e.g. a RAN node or an LMF) decides the applicable AI / ML models or functionalities based on UE-side additional conditions (if required) and NW-side additional conditions. In LCM approach #2b, a UE may need to report its UE-side additional conditions to NW via UE assistance information (UAI) . For proactive reporting, the network node only needs to provide network configurations to functionalities with available models at the UE's side. Therefore, in LCM approach #2b, the UE may only need to report UE-side additional conditions of available models, while the network node doesn’ t need to know the exact model used by the UE. However, the UE may also need to include related NW-side additional condition of the corresponding reported UE-considered applicable functionalities transmitted from the UE to the network node.
[0099] Regarding LCM approach #2c "Applicability Determination made by UE based on initial filtering by NW" , a joint decision is made by a UE and NW. In LCM approach #2c, UE-side additional conditions are known by a UE internally, and NW-side additional conditions are known at NW-side internally. First of all, a network node (e.g. a RAN node or an LMF) decides NW-considered applicable AI / ML models or functionalities based on NW-side additional conditions. The network node reports NW-considered applicable AI / ML models or functionalities to a UE. The UE further checks its NW-side additional conditions by implementation and decides the final applicable AI / ML models or functionalities by providing the corresponding configuration to the network node. However, the network node may also need to include UE-side additional conditions of the corresponding reported NW-considered applicable AI / ML models or functionalities which is transmitted from the network node to the UE.
[0100] Regarding LCM approach #2d "Applicability Determination made by NW based on initial filtering by UE" , a joint decision is made by a UE and NW. In LCM approach #2d, UE-side additional conditions are known by a UE internally, and NW-side additional conditions are known at NW-side internally. First of all, a UE decides UE-considered applicable AI / ML models or functionalities based on UE-side additional conditions (e.g. the UE's speed, scenario, hardware capabilities, model availability, etc. ) . The UE reports UE-considered applicable AI / ML models or functionalities to a network node (e.g. a RAN node or an LMF) . The network node further checks its NW-side additional conditions by implementation and decides the final applicable AI / ML models or functionalities by providing the corresponding configuration to the UE. However, the UE may also need to include related NW-side additional conditions of the corresponding reported UE-considered applicable AI / ML models or functionalities which is transmitted from the UE to the network node.
[0101] Regarding LCM approach #3a "NW makes a decision, NW initiates AI / ML management, " an LCM decision is taken and initiated by a network node (e.g. a RAN node or an LMF) . In particular, the network node sends configurations (e.g. measurement, reporting) to a UE. The UE sends the performance or assistance information (e.g. measurements) to the network node. The network node may perform management and then send a management instruction to the UE. The management instruction may be a result of AI / ML model or functionality performance monitoring at the network node. The management instruction may include information about the AI / ML model or functionality.
[0102] Regarding LCM approach #3b "UE autonomously makes a decision, the decision is reported to NW, " an LCM decision can autonomously be taken by a UE. In particular, a network node (e.g. a RAN node or an LMF) sends configurations on decision reporting to a UE. The UE may perform management and then may be configured to send a management decision report to the network node upon performing an AI / ML model or functionality management decision.
[0103] Regarding LCM approach #3c "NW makes a decision, UE initiates AI / ML management, " an LCM decision is taken by a network node (e.g. a RAN node or an LMF) but where a request is initiated by a UE. In particular, the network node sends configurations (e.g. measurement, reporting) to a UE. The UE may perform management and then send a management request to the network node. Then, the network node may send a management instruction to the UE. The management Request may be a result of AI / ML model or functionality monitoring at the UE. In response to the management request, the network may send a management instruction to the UE. The management request may include information about the AI / ML model or functionality. The network node may accept or reject the management request from the UE. The management request may include information related to AI / ML model or functionality performance metrics. The management instruction may include information about the AI / ML model or functionality.
[0104] In general, depending on the demand of a particular AI / ML model or functionality, different one or more LCM approaches could be taken. In an example, for some AI / ML model or functionality, a UE can perform monitoring (i.e. #1b as described above) and determine the applicability by itself (i.e. #2a as described above) and thus can suggest or decide the activation or deactivation (i.e. #3b or #3c as described above) . In another example, for some AI / ML model or functionality, the applicability of the AI / ML model or functionality is determined by a UE and NW jointly (i.e. #2c or #2d as described above) , and the activation or deactivation decision is made by the NW (i.e. #3a as described above) . In an additional example, the NW could perform monitoring (i.e. #1a as described above) and decides to deactivate the AI / ML model or functionality (i.e. #3a as described above) if performance of the AI / ML model or functionality is degrading. The following exemplary table summarized different possible combinations of LCM approaches.
[0105] Currently, details of supporting an LCM of a UE-sided AI / ML model or functionality have not been discussed. For instance, to support an LCM approach of a UE-sided AI / ML model or functionality, the following issues need to be solved:
[0106] - Issue #1: For an LCM approach of a particular UE-sided AI / ML model or functionality, a UE and NW have to align which LCM approach will be used for "#1 Monitoring, " "#2 Applicability Determination, " and "#3 Activation or Deactivation Decision. " Otherwise, collision could happen.
[0107] - Issue #2: In case that an activation or deactivation decision for an AI / ML model or functionality is made by NW (e.g. #3a or #3c) , it could cause too frequent AI / ML model or functionality activation or deactivation if a UE determines the same AI / ML model or functionality is applicable again and then sends related information to the NW immediately after the NW sends the deactivation decision or command to the UE.
[0108] - Issue #3: In case of a combination of LCM approaches of #1b, #2a or #2c or #2d, and #3a or #3c, even if a UE could monitor the performance of a UE-sided AI / ML model or functionality and determine the applicability of the UE-sided AI / ML model or functionality, the UE can deactivate the UE-sided AI / ML model or functionality only if NW has confirmed. On the other hand, before receiving the confirmation from the NW, some the UE's behavior (e.g., the UE's measurement) could fallback to a conventional or legacy manner (e.g. a non-AI / ML behavior or a default AI / ML method) already.
[0109] For example, if an AI / ML model or functionality is related to a temporal or spatial domain beam prediction, a UE may skip an actual measurement for some beam instances if the prediction result is available, and the UE may report the prediction result to NW. However, if the UE realizes that the accuracy of such beam prediction result degrades, before the NW confirms to deactivate the AI / ML model or functionality, the UE could perform the actual measurement on the beam instances which could be skipped, and the UE could report the measurement result instead of the prediction result (e.g., in spatial domain prediction, it could be fine for the NW's operation even if the reported result is the actual UE measurement instead of the prediction result, e.g., the reported best beam index is determined by the UE measuring all related beam instances) .
[0110] The present disclosure aims to solve the above issues and provides methods and apparatuses for supporting an LCM of a UE-sided AI / ML model or functionality. In particular, some embodiments of the present disclosure provide a solution to design signaling exchange for a UE and NW to align one or more LCM approaches to be used for a UE-sided AI / ML model or functionality, e.g. with regard to "#1 Monitoring, " "#2 Applicability Determination, " and "#3 Activation or Deactivation Decision" as described above.
[0111] In some embodiments of the present disclosure, in one or more LCM approaches, within a time period, a UE will stop reporting any applicability-related information for a UE-sided AI / ML model or functionality in case that the same UE-sided AI / ML model is deactivated by NW.
[0112] In some embodiments of the present disclosure, in one or more LCM approaches, even if NW has not deactivated a UE-sided AI / ML model or functionality yet, a UE may fallback to a non-AI / ML behavior or a default AI / ML method, if the UE determines that the performance of the UE-sided AI / ML model or functionality is degrading or the UE-side AI / ML model or functionality is no longer applicable.
[0113] In the embodiments of the present disclosure, an AI / ML model or functionality supported at a UE may also be named as "an AI / ML model or functionality deployed at a UE, " "a UE-sided AI / ML model or functionality" or the like. An AI / ML model or functionality may also be named as "an AI or ML model or functionality, " "an AI model, " "an AI functionality" or the like. An LCM approach may also be named as "an LCM method, " "an LCM manner, " "an LCM" or the like. More details of the embodiments of the present disclosure will be illustrated in the following text in combination with the appended drawings.
[0114] Figure 1 illustrates an example of a wireless communications system 100 in accordance with aspects of the present disclosure. The wireless communications system 100 may include one or more NE 102, one or more UE 104, and a core network (CN) 106. The wireless communications system 100 may support various radio access technologies. In some implementations, the wireless communications system 100 may be a 4G network, such as an LTE network or an LTE-Advanced (LTE-A) network. In some other implementations, the wireless communications system 100 may be a NR network, such as a 5G network, a 5G-Advanced (5G-A) network, or a 5G ultrawideband (5G-UWB) network. In other implementations, the wireless communications system 100 may be a combination of a 4G network and a 5G network, or other suitable radio access technology including Institute of Electrical and Electronics Engineers (IEEE) 802.11 (Wi-Fi) , IEEE 802.16 (WiMAX) , IEEE 802.20. The wireless communications system 100 may support radio access technologies beyond 5G, for example, 6G. Additionally, the wireless communications system 100 may support technologies, such as time division multiple access (TDMA) , frequency division multiple access (FDMA) , or code division multiple access (CDMA) , etc.
[0115] The one or more NE 102 may be dispersed throughout a geographic region to form the wireless communications system 100. One or more of the NE 102 described herein may be or include or may be referred to as a network node, a base station, a network element, a network function, a network entity, a radio access network (RAN) , a NodeB, an eNodeB (eNB) , a next-generation NodeB (gNB) , or other suitable terminology. An NE 102 and a UE 104 may communicate via a communication link, which may be a wireless or wired connection. For example, an NE 102 and a UE 104 may perform wireless communication (e.g. receive signaling, transmit signaling) over a Uu interface.
[0116] An NE 102 may provide a geographic coverage area for which the NE 102 may support services for one or more UEs 104 within the geographic coverage area. For example, an NE 102 and a UE 104 may support wireless communication of signals related to services (e.g. voice, video, packet data, messaging, broadcast, etc. ) according to one or multiple radio access technologies. In some implementations, an NE 102 may be moveable, for example, a satellite associated with a non-terrestrial network (NTN) . In some implementations, different geographic coverage areas associated with the same or different radio access technologies may overlap, but the different geographic coverage areas may be associated with different NE 102.
[0117] The one or more UE 104 may be dispersed throughout a geographic region of the wireless communications system 100. A UE 104 may include or may be referred to as a remote unit, a mobile device, a wireless device, a remote device, a subscriber device, a transmitter device, a receiver device, or some other suitable terminology. In some implementations, the UE 104 may be referred to as a unit, a station, a terminal, or a client, among other examples. Additionally, or alternatively, the UE 104 may be referred to as an Internet-of-Things (IoT) device, an Internet-of-Everything (IoE) device, or machine-type communication (MTC) device, among other examples.
[0118] A UE 104 may be able to support wireless communication directly with other UEs 104 over a communication link. For example, a UE 104 may support wireless communication directly with another UE 104 over a device-to-device (D2D) communication link. In some implementations, such as vehicle-to-vehicle (V2V) deployments, vehicle-to-everything (V2X) deployments, or cellular-V2X deployments, the communication link may be referred to as a sidelink. For example, a UE 104 may support wireless communication directly with another UE 104 over a PC5 interface.
[0119] An NE 102 may support communications with the CN 106, or with another NE 102, or both. For example, an NE 102 may interface with other NE 102 or the CN 106 through one or more backhaul links (e.g. S1, N2, or network interface) . In some implementations, the NE 102 may communicate with each other directly. In some other implementations, the NE 102 may communicate with each other or indirectly (e.g. via the CN 106. In some implementations, one or more NE 102 may include subcomponents, such as an access network entity, which may be an example of an access node controller (ANC) . An ANC may communicate with the one or more UEs 104 through one or more other access network transmission entities, which may be referred to as a radio heads, smart radio heads, or transmission-reception points (TRPs) .
[0120] The CN 106 may support user authentication, access authorization, tracking, connectivity, and other access, routing, or mobility functions. The CN 106 may be an evolved packet core (EPC) , or a 5G core (5GC) , which may include a control plane entity that manages access and mobility (e.g. a mobility management entity (MME) , an access and mobility management functions (AMF) ) and a user plane entity that routes packets or interconnects to external networks (e.g. a serving gateway (S-GW) , a Packet Data Network (PDN) gateway (P-GW) , or a user plane function (UPF) ) . In some implementations, the control plane entity may manage non-access stratum (NAS) functions, such as mobility, authentication, and bearer management (e.g. data bearers, signal bearers, etc. ) for the one or more UEs 104 served by the one or more NE 102 associated with the CN 106.
[0121] The CN 106 may communicate with a packet data network over one or more backhaul links (e.g. via an S1, N2, or another network interface) . The packet data network may include an application server. In some implementations, one or more UEs 104 may communicate with the application server. A UE 104 may establish a session (e.g. a protocol data unit (PDU) session, or the like) with the CN 106 via an NE 102. The CN 106 may route traffic (e.g. control information, data, and the like) between the UE 104 and the application server using the established session (e.g. the established PDU session) . The PDU session may be an example of a logical connection between the UE 104 and the CN 106 (e.g. one or more network functions of the CN 106) .
[0122] In the wireless communications system 100, the NEs 102 and the UEs 104 may use resources of the wireless communications system 100 (e.g. time resources (e.g. symbols, slots, subframes, frames, or the like) or frequency resources (e.g. subcarriers, carriers) ) to perform various operations (e.g. wireless communications) . In some implementations, the NEs 102 and the UEs 104 may support different resource structures. For example, the NEs 102 and the UEs 104 may support different frame structures. In some implementations, such as in 4G, the NEs 102 and the UEs 104 may support a single frame structure. In some other implementations, such as in 5G and among other suitable radio access technologies, the NEs 102 and the UEs 104 may support various frame structures (i.e., multiple frame structures) . The NEs 102 and the UEs 104 may support various frame structures based on one or more numerologies.
[0123] One or more numerologies may be supported in the wireless communications system 100, and a numerology may include a subcarrier spacing and a cyclic prefix. A first numerology (e.g. μ=0) may be associated with a first subcarrier spacing (e.g. 15 kHz) and a normal cyclic prefix. In some implementations, the first numerology (e.g. μ=0) associated with the first subcarrier spacing (e.g. 15 kHz) may utilize one slot per subframe. A second numerology (e.g. μ=1) may be associated with a second subcarrier spacing (e.g. 30 kHz) and a normal cyclic prefix. A third numerology (e.g. μ=2) may be associated with a third subcarrier spacing (e.g. 60 kHz) and a normal cyclic prefix or an extended cyclic prefix. A fourth numerology (e.g. μ=3) may be associated with a fourth subcarrier spacing (e.g. 120 kHz) and a normal cyclic prefix. A fifth numerology (e.g. μ=4) may be associated with a fifth subcarrier spacing (e.g. 240 kHz) and a normal cyclic prefix.
[0124] A time interval of a resource (e.g. a communication resource) may be organized according to frames (also referred to as radio frames) . Each frame may have a duration, for example, a 10 millisecond (ms) duration. In some implementations, each frame may include multiple subframes. For example, each frame may include 10 subframes, and each subframe may have a duration, for example, a 1 ms duration. In some implementations, each frame may have the same duration. In some implementations, each subframe of a frame may have the same duration.
[0125] Additionally or alternatively, a time interval of a resource (e.g. a communication resource) may be organized according to slots. For example, a subframe may include a number (e.g. quantity) of slots. The number of slots in each subframe may also depend on the one or more numerologies supported in the wireless communications system 100. For instance, the first, second, third, fourth, and fifth numerologies (i.e., μ=0, μ=1, μ=2, μ=3, μ=4) associated with respective subcarrier spacings of 15 kHz, 30 kHz, 60 kHz, 120 kHz, and 240 kHz may utilize a single slot per subframe, two slots per subframe, four slots per subframe, eight slots per subframe, and 16 slots per subframe, respectively. Each slot may include a number (e.g. quantity) of symbols (e.g. OFDM symbols) . In some implementations, the number (e.g. quantity) of slots for a subframe may depend on a numerology. For a normal cyclic prefix, a slot may include 14 symbols. For an extended cyclic prefix (e.g. applicable for 60 kHz subcarrier spacing) , a slot may include 12 symbols. The relationship between the number of symbols per slot, the number of slots per subframe, and the number of slots per frame for a normal cyclic prefix and an extended cyclic prefix may depend on a numerology. It should be understood that reference to a first numerology (e.g. μ=0) associated with a first subcarrier spacing (e.g. 15 kHz) may be used interchangeably between subframes and slots.
[0126] In the wireless communications system 100, an electromagnetic (EM) spectrum may be split, based on frequency or wavelength, into various classes, frequency bands, frequency channels, etc. By way of example, the wireless communications system 100 may support one or multiple operating frequency bands, such as frequency range designations FR1 (410 MHz –7.125 GHz) , FR2 (24.25 GHz –52.6 GHz) , FR3 (7.125 GHz –24.25 GHz) , FR4 (52.6 GHz –114.25 GHz) , FR4a or FR4-1 (52.6 GHz –71 GHz) , and FR5 (114.25 GHz –300 GHz) . In some implementations, the NEs 102 and the UEs 104 may perform wireless communications over one or more of the operating frequency bands. In some implementations, FR1 may be used by the NEs 102 and the UEs 104, among other equipment or devices for cellular communications traffic (e.g. control information, data) . In some implementations, FR2 may be used by the NEs 102 and the UEs 104, among other equipment or devices for short-range, high data rate capabilities.
[0127] FR1 may be associated with one or multiple numerologies (e.g. at least three numerologies) . For example, FR1 may be associated with a first numerology (e.g. μ=0) , which includes 15 kHz subcarrier spacing; a second numerology (e.g. μ=1) , which includes 30 kHz subcarrier spacing; and a third numerology (e.g. μ=2) , which includes 60 kHz subcarrier spacing. FR2 may be associated with one or multiple numerologies (e.g. at least 2 numerologies) . For example, FR2 may be associated with a third numerology (e.g. μ=2) , which includes 60 kHz subcarrier spacing; and a fourth numerology (e.g. μ=3) , which includes 120 kHz subcarrier spacing.
[0128] Figure 2 illustrates an example of a UE 200 in accordance with aspects of the present disclosure. The UE 200 may include a processor 202, a memory 204, a controller 206, and a transceiver 208. The processor 202, the memory 204, the controller 206, or the transceiver 208, or various combinations thereof or various components thereof may be examples of means for performing various aspects of the present disclosure as described herein. These components may be coupled (e.g. operatively, communicatively, functionally, electronically, electrically) via one or more interfaces.
[0129] The processor 202, the memory 204, the controller 206, or the transceiver 208, or various combinations or components thereof may be implemented in hardware (e.g. circuitry) . The hardware may include a processor, a digital signal processor (DSP) , an application-specific integrated circuit (ASIC) , or other programmable logic device, or any combination thereof configured as or otherwise supporting a means for performing the functions described in the present disclosure.
[0130] The processor 202 may include an intelligent hardware device (e.g. a general-purpose processor, a DSP, a CPU, an ASIC, an FPGA, or any combination thereof) . In some implementations, the processor 202 may be configured to operate the memory 204. In some other implementations, the memory 204 may be integrated into the processor 202. The processor 202 may be configured to execute computer-readable instructions stored in the memory 204 to cause the UE 200 to perform various functions of the present disclosure.
[0131] The memory 204 may include volatile or non-volatile memory. The memory 204 may store computer-readable, computer-executable code including instructions when executed by the processor 202 cause the UE 200 to perform various functions described herein. The code may be stored in a non-transitory computer-readable medium such the memory 204 or another type of memory. Computer-readable media includes both non-transitory computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A non-transitory storage medium may be any available medium that may be accessed by a general-purpose or special-purpose computer.
[0132] In some implementations, the processor 202 and the memory 204 coupled with the processor 202 may be configured to cause the UE 200 to perform one or more of the functions described herein (e.g. executing, by the processor 202, instructions stored in the memory 204) . For example, the processor 202 may support wireless communication at the UE 200 in accordance with examples as disclosed with respect to Figure 5. The UE 200 may be configured to support: a means for transmitting capability information related to one or more AI / ML functionalities supported at the UE; a means for receiving a configuration regarding applicability-related information for the one or more AI / ML functionalities; and a means for determining whether to transmit the applicability-related information and determining one or more elements included in the applicability-related information according to the configuration.
[0133] The controller 206 may manage input and output signals for the UE 200. The controller 206 may also manage peripherals not integrated into the UE 200. In some implementations, the controller 206 may utilize an operating system such as or other operating systems. In some implementations, the controller 206 may be implemented as part of the processor 202.
[0134] In some implementations, the UE 200 may include at least one transceiver 208. In some other implementations, the UE 200 may have more than one transceiver 208. The transceiver 208 may represent a wireless transceiver. The transceiver 208 may include one or more receiver chains 210, one or more transmitter chains 212, or a combination thereof. The means for receiving abovementioned in the processor 202 or the means for transmitting in the processor 202 may be implemented via at least one transceiver 208.
[0135] A receiver chain 210 may be configured to receive signals (e.g. control information, data, packets) over a wireless medium. For example, the receiver chain 210 may include one or more antennas for receive the signal over the air or wireless medium. The receiver chain 210 may include at least one amplifier (e.g. a low-noise amplifier (LNA) ) configured to amplify the received signal. The receiver chain 210 may include at least one demodulator configured to demodulate the receive signal and obtain the transmitted data by reversing the modulation technique applied during transmission of the signal. The receiver chain 210 may include at least one decoder for decoding the processing the demodulated signal to receive the transmitted data.
[0136] A transmitter chain 212 may be configured to generate and transmit signals (e.g. control information, data, packets) . The transmitter chain 212 may include at least one modulator for modulating data onto a carrier signal, preparing the signal for transmission over a wireless medium. The at least one modulator may be configured to support one or more techniques such as amplitude modulation (AM) , frequency modulation (FM) , or digital modulation schemes like phase-shift keying (PSK) or quadrature amplitude modulation (QAM) . The transmitter chain 212 may also include at least one power amplifier configured to amplify the modulated signal to an appropriate power level suitable for transmission over the wireless medium. The transmitter chain 212 may also include one or more antennas for transmitting the amplified signal into the air or wireless medium.
[0137] Figure 3 illustrates an example of a processor 300 in accordance with aspects of the present disclosure. The processor 300 may be an example of a processor configured to perform various operations in accordance with examples as described herein. The processor 300 may include a controller 302 configured to perform various operations in accordance with examples as described herein. The processor 300 may optionally include at least one memory 304, which may be, for example, an L1 / L2 / L3 cache. Additionally, or alternatively, the processor 300 may optionally include one or more arithmetic-logic units (ALUs) 306. One or more of these components may be in electronic communication or otherwise coupled (e.g. operatively, communicatively, functionally, electronically, electrically) via one or more interfaces (e.g. buses) .
[0138] The processor 300 may be a processor chipset and include a protocol stack (e.g. a software stack) executed by the processor chipset to perform various operations (e.g. receiving, obtaining, retrieving, transmitting, outputting, forwarding, storing, determining, identifying, accessing, writing, reading) in accordance with examples as described herein. The processor chipset may include one or more cores, one or more caches (e.g. memory local to or included in the processor chipset (e.g. the processor 300) or other memory (e.g. random access memory (RAM) , read-only memory (ROM) , dynamic RAM (DRAM) , synchronous dynamic RAM (SDRAM) , static RAM (SRAM) , ferroelectric RAM (FeRAM) , magnetic RAM (MRAM) , resistive RAM (RRAM) , flash memory, phase change memory (PCM) , and others) .
[0139] The controller 302 may be configured to manage and coordinate various operations (e.g. signaling, receiving, obtaining, retrieving, transmitting, outputting, forwarding, storing, determining, identifying, accessing, writing, reading) of the processor 300 to cause the processor 300 to support various operations in accordance with examples as described herein. For example, the controller 302 may operate as a control unit of the processor 300, generating control signals that manage the operation of various components of the processor 300. These control signals include enabling or disabling functional units, selecting data paths, initiating memory access, and coordinating timing of operations.
[0140] The controller 302 may be configured to fetch (e.g. obtain, retrieve, receive) instructions from the memory 304 and determine subsequent instruction (s) to be executed to cause the processor 300 to support various operations in accordance with examples as described herein. The controller 302 may be configured to track memory address of instructions associated with the memory 304. The controller 302 may be configured to decode instructions to determine the operation to be performed and the operands involved. For example, the controller 302 may be configured to interpret the instruction and determine control signals to be output to other components of the processor 300 to cause the processor 300 to support various operations in accordance with examples as described herein. Additionally, or alternatively, the controller 302 may be configured to manage flow of data within the processor 300. The controller 302 may be configured to control transfer of data between registers, arithmetic logic units (ALUs) , and other functional units of the processor 300.
[0141] The memory 304 may include one or more caches (e.g. memory local to or included in the processor 300 or other memory, such RAM, ROM, DRAM, SDRAM, SRAM, MRAM, flash memory, etc. In some implementations, the memory 304 may reside within or on a processor chipset (e.g. local to the processor 300) . In some other implementations, the memory 304 may reside external to the processor chipset (e.g. remote to the processor 300) .
[0142] The memory 304 may store computer-readable, computer-executable code including instructions that, when executed by the processor 300, cause the processor 300 to perform various functions described herein. The code may be stored in a non-transitory computer-readable medium such as system memory or another type of memory. The controller 302 and / or the processor 300 may be configured to execute computer-readable instructions stored in the memory 304 to cause the processor 300 to perform various functions. For example, the processor 300 and / or the controller 302 may be coupled with or to the memory 304, the processor 300, the controller 302, and the memory 304 may be configured to perform various functions described herein. In some examples, the processor 300 may include multiple processors and the memory 304 may include multiple memories. One or more of the multiple processors may be coupled with one or more of the multiple memories, which may, individually or collectively, be configured to perform various functions herein.
[0143] The one or more ALUs 306 may be configured to support various operations in accordance with examples as described herein. In some implementations, the one or more ALUs 306 may reside within or on a processor chipset (e.g. the processor 300) . In some other implementations, the one or more ALUs 306 may reside external to the processor chipset (e.g. the processor 300) . One or more ALUs 306 may perform one or more computations such as addition, subtraction, multiplication, and division on data. For example, one or more ALUs 306 may receive input operands and an operation code, which determines an operation to be executed. One or more ALUs 306 be configured with a variety of logical and arithmetic circuits, including adders, subtractors, shifters, and logic gates, to process and manipulate the data according to the operation. Additionally, or alternatively, the one or more ALUs 306 may support logical operations such as AND, OR, exclusive-OR (XOR) , not-OR (NOR) , and not-AND (NAND) , enabling the one or more ALUs 306 to handle conditional operations, comparisons, and bitwise operations.
[0144] The processor 300 may support wireless communication in accordance with examples as described with respect to Figures 5-9 as described below.
[0145] In some implementations, the processor 300 may be configured to support a means for performing operations of a UE as described with respect to Figure 5. The processor 300 may be configured to or operable to support: a means for transmitting capability information related to one or more AI / ML functionalities supported at the UE; a means for receiving a configuration regarding applicability-related information for the one or more AI / ML functionalities; and a means for determining whether to transmit the applicability-related information and determining one or more elements included in the applicability-related information according to the configuration.
[0146] In some implementations, the processor 300 may be configured to support a means for performing operations of a network node as described with respect to Figure 6. The processor 300 may be configured to or operable to support: a means for receiving, from a UE, capability information related to one or more AI / ML functionalities supported at the UE; and a means for transmitting, to the UE, a configuration regarding applicability-related information for the one or more AI / ML functionalities.
[0147] It should be appreciated by persons skilled in the art that the components in exemplary processor 300 may be changed, for example, some of the components in exemplary processor 300 may be omitted or modified or new component (s) may be added to exemplary processor 300, without departing from the spirit and scope of the disclosure. For example, in some embodiments, the processor 300 may not include the ALUs 306.
[0148] Figure 4 illustrates an example of a NE 400 in accordance with aspects of the present disclosure. The NE 400 may include a processor 402, a memory 404, a controller 406, and a transceiver 408. The processor 402, the memory 404, the controller 406, or the transceiver 408, or various combinations thereof or various components thereof may be examples of means for performing various aspects of the present disclosure as described herein. These components may be coupled (e.g. operatively, communicatively, functionally, electronically, electrically) via one or more interfaces.
[0149] The processor 402, the memory 404, the controller 406, or the transceiver 408, or various combinations or components thereof may be implemented in hardware (e.g. circuitry) . The hardware may include a processor, a digital signal processor (DSP) , an application-specific integrated circuit (ASIC) , or other programmable logic device, or any combination thereof configured as or otherwise supporting a means for performing the functions described in the present disclosure.
[0150] The processor 402 may include an intelligent hardware device (e.g. a general-purpose processor, a DSP, a CPU, an ASIC, an FPGA, or any combination thereof) . In some implementations, the processor 402 may be configured to operate the memory 404. In some other implementations, the memory 404 may be integrated into the processor 402. The processor 402 may be configured to execute computer-readable instructions stored in the memory 404 to cause the NE 400 to perform various functions of the present disclosure.
[0151] The memory 404 may include volatile or non-volatile memory. The memory 404 may store computer-readable, computer-executable code including instructions when executed by the processor 402 cause the NE 400 to perform various functions described herein. The code may be stored in a non-transitory computer-readable medium such the memory 404 or another type of memory. Computer-readable media includes both non-transitory computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A non-transitory storage medium may be any available medium that may be accessed by a general-purpose or special-purpose computer.
[0152] In some implementations, the processor 402 and the memory 404 coupled with the processor 402 may be configured to cause the NE 400 to perform one or more of the functions described herein (e.g. executing, by the processor 402, instructions stored in the memory 404) . For example, the processor 402 may support wireless communication at the NE 400 in accordance with examples as disclosed herein. For example, the NE 400 may be configured to support a means for performing the operations as described with respect to Figures 6-9 as described below.
[0153] In some implementations, the NE 400 may be a network node as described with respect to Figure 6. The NE 400 may be configured to support: a means for receiving, from a UE, capability information related to one or more AI / ML functionalities supported at the UE;and a means for transmitting, to the UE, a configuration regarding applicability-related information for the one or more AI / ML functionalities.
[0154] The controller 406 may manage input and output signals for the NE 400. The controller 406 may also manage peripherals not integrated into the NE 400. In some implementations, the controller 406 may utilize an operating system such as or other operating systems. In some implementations, the controller 406 may be implemented as part of the processor 402.
[0155] In some implementations, the NE 400 may include at least one transceiver 408. In some other implementations, the NE 400 may have more than one transceiver 408. The transceiver 408 may represent a wireless transceiver. The transceiver 408 may include one or more receiver chains 410, one or more transmitter chains 412, or a combination thereof. The means for receiving or the means for transmitting abovementioned in the processor 402 may be implemented via at least one transceiver 408.
[0156] A receiver chain 410 may be configured to receive signals (e.g. control information, data, packets) over a wireless medium. For example, the receiver chain 410 may include one or more antennas for receive the signal over the air or wireless medium. The receiver chain 410 may include at least one amplifier (e.g. a low-noise amplifier (LNA) ) configured to amplify the received signal. The receiver chain 410 may include at least one demodulator configured to demodulate the receive signal and obtain the transmitted data by reversing the modulation technique applied during transmission of the signal. The receiver chain 410 may include at least one decoder for decoding the processing the demodulated signal to receive the transmitted data.
[0157] A transmitter chain 412 may be configured to generate and transmit signals (e.g. control information, data, packets) . The transmitter chain 412 may include at least one modulator for modulating data onto a carrier signal, preparing the signal for transmission over a wireless medium. The at least one modulator may be configured to support one or more techniques such as amplitude modulation (AM) , frequency modulation (FM) , or digital modulation schemes like phase-shift keying (PSK) or quadrature amplitude modulation (QAM) . The transmitter chain 412 may also include at least one power amplifier configured to amplify the modulated signal to an appropriate power level suitable for transmission over the wireless medium. The transmitter chain 412 may also include one or more antennas for transmitting the amplified signal into the air or wireless medium.
[0158] It should be appreciated by persons skilled in the art that the components in exemplary NE 400 may be changed, for example, some of the components in exemplary NE 400 may be omitted or modified or new component (s) may be added to exemplary NE 400, without departing from the spirit and scope of the disclosure. For example, in some embodiments, the NE 400 may not include the controller 406.
[0159] Figure 5 illustrates a flowchart of a method related to an AI / ML model or functionality in accordance with aspects of the present disclosure. The operations of the method may be implemented by a UE as described herein. In some implementations, the UE may execute a set of instructions to control the function elements of the UE to perform the described functions. In some implementations, aspects of operations 502, 504, and 506 may be performed by UE 200 as described with reference to Figure 2. Each of operations 502, 504, and 506 may be performed in accordance with examples as described herein. Specific examples are described in the embodiments of Figures 7A-9 as follows.
[0160] At 502, the method may include transmitting, by a UE (denoted as UE #1) , capability information related to one or more AI / ML functionalities (denoted as AI / ML set #1) supported at UE #1. An AI / ML functionality may also be named as "an AI / ML model or functionality, " "an AI or ML model or functionality, " "an AI model, " "an AI functionality" or the like.
[0161] In some implementations, a network node may be at least one of a RAN node or an LMF. The capability information may be transmitted at 502 to the network node via an RRC UE capability report message (e.g. an RRC UECapabilityReport message to the RAN node) or via an LPP provide capabilities message (e.g. an LPP ProvideCapabilities message to the LMF) .
[0162] For instance, the capability information may indicate at least one of the following:
[0163] (1) AI / ML set #1 that can be supported by UE #1.
[0164] (2) one or more LCM approaches (denoted as LCM set #1) that can be supported by UE #1 for AI / ML set #1; or
[0165] (3) at least one LCM approach adopted by UE #1 for each AI / ML functionality of AI / ML set #1, i.e. at least one LCM approach adopted by UE #1 for a supported UE-sided AIML model or functionality.
[0166] In some embodiments, AI / ML set #1 and LCM set #1 have one-to-one mapping relationship, or have one-to-multiple or multiple-to-one mapping relationship. That is, each UE-sided supported AI / ML functionality may correspond to one or more LCM approaches. Each LCM approach may be adopted to one or more supported UE-sided AIML models or functionalities.
[0167] In some implementations, LCM set #1 may include at least one of the following:
[0168] (1) an approach for monitoring AI / ML set #1, e.g. "#1 Monitoring" as described above;
[0169] (2) an approach for determining applicability of AI / ML set #1, e.g. "#2 Applicability Determination" as described above;
[0170] (3) an approach for making an activation decision for AI / ML set #1, e.g. "#3 Activation Decision" as described above; or
[0171] (4) an approach for making a deactivation decision for AI / ML set #1, e.g. "#3 Deactivation Decision" as described above.
[0172] In some implementations of the method, UE #1 may determine the at least one LCM approach adopted by UE #1. In some other implementations of the method, UE #1 may receive, from the network node, a message including: (1) one or more LCM approaches to be adopted by UE #1; and (2) a configuration regarding the one or more LCM approaches to be adopted by UE #1. For example, the message may be an RRC reconfiguration message or an LPP request location information message.
[0173] At 504, the method may include receiving, by UE #1, a configuration (denoted as configuration #1) regarding applicability-related information for AI / ML set #1. At 506, the method may include determining whether to transmit the applicability-related information and determining one or more elements included in the applicability-related information by UE #1 according to configuration #1.
[0174] In some implementations, the one or more elements included the applicability-related information may indicate at least one of the following:
[0175] (1) whether AI / ML set #1 are applicable;
[0176] (2) whether AI / ML set #1 are applicable at UE #1, and informing that a network node needs to further determine applicability of AI / ML set #1; or
[0177] (3) informing the network node to determine the applicability of AI / ML set #1.
[0178] In some implementations of the method, UE #1 may receive, from the network node, additional condition information associated with the applicability-related information. UE #1 may determine whether AI / ML set #1 are applicable based on the received additional condition information. For example, the additional condition can be the network node's side associated identifier (ID) that represents a specific deployment or condition of the network node, such as an antenna location, height, an angle and etc.
[0179] In some implementations of the method, UE #1 may transmit the applicability-related information to the network node, e.g. after UE #1 determines to transmit the applicability-related information at 506. For instance, applicability-related information may be carried in at least one of the following messages:
[0180] (1) an LPP provide capabilities message;
[0181] (2) an RRC UE assistance information message, e.g. an RRC UEAssistingInfomation message. In an implementation, UE #1 may receive an RRC other configuration message (e.g. an RRC OtherConfig message) including information which indicates UE #1 to use the RRC UE assistance information message to carry the applicability-related information;
[0182] (3) an RRC reconfiguration complete message, e.g. an RRCReconfigurationComplete message; or
[0183] (4) an RRC resume complete message, e.g. an RRCResumeComplete message.
[0184] In some implementations, if at least one AI / ML functionality (denoted as AI / ML #x) within AI / ML set #1 has been activated or has been reported to be applicable, UE #1 may stop transmitting the applicability-related information for AI / ML #x for a time period and / or a pre-defined condition, e.g. a prohibit timer is running. UE #1 may resume transmitting the applicability-related information for AI / ML #x after the time period and / or the pre-defined condition. Specific examples are described in the embodiments of Figures 8 and 9 as follows.
[0185] In some implementations, configuration #1 received by UE #1 at 504 includes a configuration regarding the time period and / or the pre-defined condition. In some other implementations, UE #1 may receive an RRC other configuration message (e.g. an RRC OtherConfig message) including the configuration regarding the time period and / or the pre-defined condition.
[0186] In some implementations, UE #1 may stop transmitting the applicability-related information for AI / ML #x in at least one of the following cases:
[0187] (1) AI / ML #x is deactivated by a decision (or command) from the network node, e.g. LCM approach #3a or #3c as described above;
[0188] (2) AI / ML #x is deactivated by a decision of UE #1, e.g. LCM approach #3b as described above;
[0189] (3) AI / ML #x is deactivated by the decision of UE #1, and a deactivation result of AI / ML #x is reported by UE #1 to the network node, e.g. LCM approach #3b as described above;
[0190] (4) AI / ML #x is deactivated by the decision of UE #1, the deactivation result of AI / ML #x is reported by UE #1 to the network node, and a confirmation is received by UE #1 from the network node, e.g. LCM approach #3b as described above; or
[0191] (5) AI / ML #x is determined by UE #1 to be non-applicable, e.g. LCM approach #2a or #2c or #2d as described above. For example, UE #1 may consider that AI / ML #x is to be non-applicable in at least one of the following cases:
[0192] a) AI / ML #x is deactivated by the decision from the network node, e.g. LCM approach #3a or #3c as described above;
[0193] b) AI / ML #x is deactivated by the decision of UE #1, e.g. LCM approach #3b as described above;
[0194] c) AI / ML #x is deactivated by the decision of UE #1, and the deactivation result of AI / ML #x is reported by UE #1 to the network node, e.g. LCM approach #3b as described above; or
[0195] d) AI / ML #x is deactivated by the decision of UE #1, the deactivation result of AI / ML #x is reported by UE #1 to the network node, and a confirmation is received by UE #1 from the network node, e.g. LCM approach #3b as described above.
[0196] In some implementations of the method, UE #1 may stop determining whether AI / ML #x is applicable during the time period (e.g. the prohibit timer is running) , and resume determining whether AI / ML #x is applicable after the time period (e.g. the prohibit timer is expired) .
[0197] In some implementations of the method, UE #1 may receive an AI / ML activation or deactivation decision (or command) from a network node. For example, the AI / ML activation or deactivation decision (or command) may be carried in: an LPP request location information message; an RRC reconfiguration message; and / or a MAC CE.
[0198] In some implementations of the method, if at least one AI / ML functionality (denoted as AI / ML #y) within AI / ML set #1 has been activated, UE #1 may determine whether AI / ML #y is degrading or whether AI / ML #y is applicable, e.g. under the current circumstance (e.g., current NW configuration, UE side condition, and / or NW side condition) . If AI / ML #y is degrading or if AI / ML #y is not applicable (e.g. in case of LCM approach #1b and "#2a or #2c or #2d" ) , and if it is upon a network node’s final decision to deactivate AI / ML #y (e.g. in case of LCM approach #1b and "#3a or #3c" ) , UE #1 may fallback to a non-AI / ML behavior or a default AI / ML method.
[0199] For example, UE #1 may determine that AI / ML #y is degrading or that AI / ML #y is not applicable in at least one of the following cases:
[0200] (1) an accuracy of AI / ML #y is below a threshold;
[0201] (2) a measured L1 / L3 beam level reference signal received power (RSRP) of AI / ML #y is above or below a threshold; or
[0202] (3) a measured L3 cell level RSRP of AI / ML #y is above or below a threshold.
[0203] In an implementation, UE #1 may receive, from the network node, a configuration indicating that UE #1 can fallback in case that AI / ML #y is degrading or in case that AI / ML #y is not applicable without the network node confirming a deactivation of AI / ML #y.
[0204] In some embodiments, UE #1 may transmit, to the network node, information indicating that UE #1 will fallback to the non-AI / ML behavior or the default AI / ML method. In some other embodiments, UE #1 may not indicate anything to the network node, and UE #1 will autonomously fallback to the non-AI / ML behavior or the default AI / ML method.
[0205] In some implementations of the method, UE #1 may measure a measurement result (e.g., a beam level L1 / L3 RSRP measurement result or a cell level L3 RSRP measurement result) , and transmit the measurement result to the network node. In some implementations, UE #1 is configured to perform a beam prediction (e.g. in temporal or spatial domain) or a radio resource management (RRM) prediction using AI / ML set #1, and transmit a prediction result of the beam prediction or the RRM prediction (e.g., predicted beam / cell level L1 / L3 RSRP, predicted best beam index, predicted best cell index etc. ) to the network node. In an implementation, the prediction result and the measurement result are transmitted in a same report, e.g. an uplink control information (UCI) CSI report (in case of L1 result) or a RRC measurement report (in case of L3 result) . In another implementation, the prediction result and the measurement result are transmitted in separate reports.
[0206] In some implementations of the method, UE #1 may perform a location estimation for UE #1 using a non-AI / ML method; and transmit an estimated location of UE #1 to the network node. In some implementations, UE #1 is configured to perform a location prediction using AI / ML set #1, and transmit a predicted location of UE #1 to the network node. In an implementation, "the location of UE #1 estimated by the location estimation using the non-AI / ML method" and / or "the location of UE #1 predicted by the location prediction using AI / ML set #1" are transmitted in an LPP provide location information message.
[0207] Figure 6 illustrates another flowchart of a method related to an AI / ML model or functionality in accordance with aspects of the present disclosure. The operations of the method may be implemented by a network node (e.g. a RAN node or an LMF) as described herein. In some implementations, the network node may execute a set of instructions to control the function elements of the network node to perform the described functions. In some implementations, aspects of operations 602 and 604 may be performed by NE 400 as described with reference to Figure 4. Each of operations 602 and 604 may be performed in accordance with examples as described herein. Specific examples are described in the embodiments of Figures 7A-9 as follows.
[0208] At 602, the method may include receiving, by a network node (e.g. a RAN node and / or an LMF) from a UE (e.g. UE #1 as described in the embodiments of Figure 5) , capability information related to one or more AI / ML functionalities (e.g. AI / ML set #1 as described in the embodiments of Figure 5) supported at the UE. The capability information received at 602 may include the same or similar elements as those in the capability information transmitted by UE #1 at operation 502 as described in the embodiments of Figure 5. For example, the capability information is received via an RRC UE capability report message or via an LPP provide capabilities message.
[0209] At 604, the method may include transmitting, by the network node to the UE, a configuration (e.g. configuration #1 as described in the embodiments of Figure 5) regarding applicability-related information for the one or more AI / ML functionalities. The configuration may include the same or similar elements as those in configuration #1 received by UE #1 at operation 504 as described in the embodiments of Figure 5.
[0210] In some implementations of the method, the network node may transmit a message (e.g. an RRC reconfiguration message or an LPP request location information message) including the following information to the UE:
[0211] (1) one or more LCM approaches (e.g. LCM set #1 as described in the embodiments of Figure 5) to be adopted by the UE; and
[0212] (2) a configuration regarding the one or more LCM approaches to be adopted by the UE.
[0213] In some implementations of the method, the network node may receive the applicability-related information from the UE, which is carried in:
[0214] (1) an LPP provide capabilities message, e.g. an LPP ProvideCapabilities to LMF message;
[0215] (2) an RRC UE assistance information message, e.g. an RRC UEAssistanceInformation message;
[0216] (3) an RRC reconfiguration complete message, e.g. an RRCReconfigurationComplete message; and / or
[0217] (4) an RRC resume complete message, e.g. an RRCResumeComplete message.
[0218] In some implementations of the method, the network node may transmit, to the UE, an RRC other configuration message (e.g. an RRC OtherConfig message) including information indicating the UE to use an RRC UE assistance information message (e.g. the RRC UEAssistanceInformation message) to carry the applicability-related information. The applicability-related information may include the same or similar elements as those in the applicability-related information as described in the embodiments of Figure 5.
[0219] In some implementations, if the applicability-related information indicates whether AI / ML set #1 are applicable, the network node may consider that the UE has fully determined the applicability of AI / ML set #1. If the applicability-related information indicates whether AI / ML set #1 are applicable at the UE and informing that the network node needs to further determine applicability of AI / ML set #1, the network node may consider that the UE has partially determined the applicability of AI / ML set #1 and further determine a final applicability of AI / ML set #1 considering a condition at the network node. If the applicability-related information indicates informing the network node to determine the applicability of AI / ML set #1, the network node may determine the applicability of AI / ML set #1 considering the condition at the network node.
[0220] In some implementations of the method, the network node may transmit additional condition information associated with the applicability-related information to the UE, and the additional condition information is used by the UE to determine whether AI / ML set #1 are applicable. For example, the additional condition can be the network node's side associated ID that represents a specific deployment or condition of the network node, such as an antenna location, height, an angle and etc.
[0221] In some implementations, the configuration transmitted by the network node at 604 includes a configuration regarding a time period and / or a pre-defined condition, which can be used by the UE to stop transmitting the applicability-related information for at least one within AI / ML set #1 (e.g. AI / ML #x as described in the embodiments of Figure 5) . In some other implementations, the network node may transmit, to the UE, an RRC other configuration message (e.g. an RRC OtherConfig message) including the configuration regarding a time period and / or a pre-defined condition.
[0222] In some implementations of the method, the network node may transmit an AI / ML activation or deactivation decision (or command) to the UE, e.g. which may be carried in an LPP request location information message, an RRC reconfiguration message, and / or a MAC CE.
[0223] In some implementations of the method, the network node may transmit, to the UE, a configuration indicating that the UE can fallback in case that at least one AI / ML functionality (e.g. AI / ML #y as described in the embodiments of Figure 5) is degrading or in case that the at least one AI / ML functionality is not applicable without the network node confirming a deactivation of the at least one AI / ML functionality.
[0224] In some implementations of the method, the network node may receive, from the UE, information indicating that the UE will fallback to a non-AI / ML behavior or a default AI / ML method. In some other implementations, the UE may not indicate anything to the network node, and the UE will autonomously fallback to the non-AI / ML behavior or the default AI / ML method.
[0225] In some implementations of the method, the network node may receive a measurement result (e.g., a beam level L1 / L3 RSRP measurement result or a cell level L3 RSRP measurement result) measured by the UE from the UE. In some implementations the network node may receive a prediction result of the beam prediction or the RRM prediction (e.g., predicted beam / cell level L1 / L3 RSRP, predicted best beam index, predicted best cell index etc. ) predicted by the UE performing a beam prediction (e.g. in temporal or spatial domain) or an RRM prediction using AI / ML set #1. In an implementation, the measurement result and the prediction result are transmitted in a same report, e.g. an uplink control information (UCI) CSI report (in case of L1 result) or a RRC measurement report (in case of L3 result) . In another implementation, the measurement result and the prediction result are transmitted in separate reports.
[0226] In some implementations of the method, the network node may receive, from the UE, a location of the UE estimated by the UE performing a location estimation using a non-AI / ML method. In some implementations, the network node may receive, from the UE, a location of the UE predicted by the UE performing a location prediction using AI / ML set #1. In an implementation, the estimated location of the UE and / or the predicted location of the UE are transmitted in an LPP provide location information message.
[0227] It should be noted that the method described in Figure 5 or Figure 6 describes possible implementations, and that the operations and the steps may be rearranged or otherwise eliminated or modified and that other implementations are possible, without departing from the spirit and scope of the disclosure.
[0228] The following text describes specific embodiments of the flowcharts as shown and illustrated above, i.e. Embodiment 1, Embodiment 2 and Embodiment 3 as below.
[0229] Embodiment 1 (to align an LCM approach between a UE and NW)
[0230] In Embodiment 1, for the sake of a UE-sided AI / ML model or functionality, a UE and NW needs to determine which one or more LCM approaches for some or all of "#1 Monitoring, " "#2 Applicability Determination" and "#3 Activation or Deactivation Decision" is going to be adopted for one or more UE-sided AI / ML models or functionalities. An LCM approach will be adopted for each aspect of "#1 Monitoring, " "#2 Applicability Determination" and "#3 Activation or Deactivation Decision, " respectively. That is, determining the one or more LCM approaches includes respectively selecting #1a or #1b of "#1 Monitoring, " selecting one from #2a, #2b, #2c or #2d of "#2 Applicability Determination" and / or selecting one from #3a, #3b or #3c of "#3 Activation or Deactivation Decision" as described above.
[0231] In an example, for some AI / ML model or functionality, a UE can perform monitoring (i.e. #1b) and determine the applicability of the AI / ML model or functionality by itself (i.e. #2a) and thus can suggest or decide the activation or deactivation (i.e. #3b or #3c) ; for the AI / ML model or functionality. In another example, the applicability of the AI / ML model or functionality is determined by the UE and NW jointly (i.e. #2c or #2d) and the activation or deactivation decision is made by the NW (i.e. #3a) . In an additional example, the NW could perform monitoring (i.e. #1a) and decide to deactivate the AI / ML model or functionality (i.e. #3a) if performance of the AI / ML model or functionality is degrading.
[0232] Embodiment 1 may include two embodiments according to different cases, i.e. Embodiment 1.1 and Embodiment 1.2 as below.
[0233] In Embodiment 1.1, it is a UE that determines one or more LCM approaches and informs a network node (aRAN node and / or an LMF) , and the network node will act accordingly.
[0234] Figure 7A illustrates a schematic diagram of aligning an LCM approach for an AI / ML model or functionality in accordance with aspects of the present disclosure. Details described in all other embodiments of the present disclosure are applicable for the embodiments shown in Figure 7A. Figure 7A describes a specific example of Embodiment 1.1.
[0235] In an example of Embodiment 1.1, a UE may inform a network node (aRAN node or an LMF) about supported one or more UE-sided AI / ML models or functionalities of the UE. The UE may also indicate, to the network node, one or more LCM approaches adopted for the supported one or more UE-sided AI / ML models or functionalities.
[0236] In particular, as shown in Figure 7A, at 711, a UE transmits a message to a network node (aRAN node or an LMF) , to inform the network node about the supported one or more UE-sided AI / ML models or functionalities of the UE. For example, the UE transmits an RRC UECapabilityReport message to a RAN node or an LPP ProvideCapabilities message to an LMF at 711. In some embodiments, the UE also indicates the adopted LCM approaches at 711, with regard to one or more of "#1 Monitoring, " "#2 Applicability Determination" and "#3 Activation or Deactivation Decision" for each supported UE-sided AI / ML model or functionality. There could be one-to-one or one-to-multiple or multiple-to-one mapping between the LCM approaches and the supported one or more UE-sided AI / ML models or functionalities.
[0237] In another example of Embodiment 1.1, the UE transmits applicability-related information to the network node, which may explicitly or implicitly indicate the adopted LCM approaches for the supported one or more UE-sided AI / ML models or functionalities.
[0238] As shown in Figure 7A, at 713, the UE transmits applicability-related information to the network node (e.g., via an RRC UEAssistanceInformation message to a RAN node or LPP ProvideCapabilities to an LMF or any other new defined RRC message) . The UE may implicitly indicate the adopted LCM approach at 713 in the following cases according to different embodiments, and the network node may act differently. For example:
[0239] (1) If the applicability-related information reported from the UE to the network node at 713 includes "information indicating whether the supported one or more UE-sided AI / ML models or functionalities are applicable or not, " the network node will consider that the UE has (fully) determined the applicability of the supported one or more UE-sided AI / ML models or functionalities (i.e. adopting LCM approach #2a or #2c)
[0240] a) In some implementations, the UE may also indicate to distinguish LCM approach #2a or #2c. For instance, in the message at 713, the UE may indicate that the applicability-related information is "applicable both UE-sided and NW-sided information" (i.e. adopting LCM approach #2a) , or the UE may indicate that the applicability-related information is "applicable UE-sided information only" and the network node's further determination on the applicability of the supported one or more UE-sided AI / ML models or functionalities is needed (i.e. adopting LCM approach #2c) .
[0241] (2) If the applicability-related information reported from the UE to the network node at 713 includes "information indicating whether the supported UE-sided AI / ML models or functionalities are applicable or not from the UE's side" and "information indicating that the network node's further determination on the applicability of the supported UE-sided AI / ML models or functionalities is needed, " the network node will consider that the UE has (partially) determined the applicability considering the UE's side condition (i.e. adopting LCM approach #2d) , and then the network node will further determine the final applicability considering the network node's side condition.
[0242] (3) If the applicability-related information reported from the UE to the network node at 713 includes "information indicating the network node to determine the applicability of the supported UE-sided AI / ML models or functionalities, " the network node will need to determine the applicability considering the network node's side condition (i.e. adopting LCM approach #2b) .
[0243] In another example for Embodiment 1.1, the network node may provide the network node's side additional condition to the UE. For example, the additional condition can be the network node's side associated ID that represents a specific deployment or condition of the network node, such as an antenna location, height, an angle and etc. In particular, at 712 (optional) , the network node's side additional condition is provided to the UE, e.g. in system information block (SIB) or in an RRC Reconfiguration message to OtherConfig IE. Then, the UE will adopt LCM approach #2a. After that, in the UEAssistanceInformation message at 713 which indicates the applicability of the one or more UE-sided AI / ML models or functionalities, the UE may include information indicating whether the one or more UE-sided AI / ML models or functionalities are applicable or not.
[0244] Otherwise, if the network node's side additional condition is not provided to the UE (i.e. operation 712 is not performed) , the UE may adopt LCM approach #2b or #2d. Then, the UEAssistanceInformation message at 713, which indicates the applicability of the one or more UE-sided AI / ML models or functionalities, may include "information for the network node to determine the applicability of the one or more UE-sided AI / ML models or functionalities. " For example, the network node may determine the associated network node configuration or the network node's side additional condition under which the one or more UE-sided AI / ML models or functionalities are applicable.
[0245] In Embodiment 1.2, it is NW that determines a combination of LCM approaches and informs a UE, and the UE will act accordingly.
[0246] Figure 7B illustrates a schematic diagram of aligning an LCM approach for an AI / ML model or functionality in accordance with aspects of the present disclosure. Details described in all other embodiments of the present disclosure are applicable for the embodiments shown in Figure 7B. Figure 7B describes a specific example of Embodiment 1.2.
[0247] In one example of Embodiment 1.2, a UE may inform a network node (aRAN node or an LMF) about supported one or more UE-sided AI / ML models or functionalities of the UE. The UE may also indicate, to the network node, supported one or more LCM approaches for the supported one or more UE-sided AI / ML models or functionalities. Then, it is upon the network node to further decide which one or more LCM approaches to be adopted.
[0248] In particular, as shown in Figure 7B, at 701, a UE transmits a message to a network node (aRAN node or an LMF) , to inform the network node about the supported one or more UE-sided AI / ML models or functionalities of the UE. For example, the UE transmits an RRC UECapabilityReport message to the RAN node or an LPP ProvideCapabilities message to the LMF at 701. In some embodiments, the UE also indicates the supported LCM approaches at 701, with regard to one or more of "#1 Monitoring, " "#2 Applicability Determination" and "#3 Activation or Deactivation Decision" for each supported UE-sided AI / ML model or functionality. There could be one-to-one or one-to-multiple or multiple-to-one mapping between the supported one or more LCM approaches and the supported one or more UE-sided AI / ML models or functionalities.
[0249] At 702, the network node may transmit a message (e.g. an RRC RRCReconfiguration message from the RAN node or an LPP RequestLocationInformation message from the LMF) to the UE, to indicate one or more LCM approaches that are going to used and provide the related configuration as well.
[0250] At 703 (optional) , the network node may transmit a message (e.g. an RRC OtherConfig message) to the UE, to indicate the UE to use an RRC UEAssistanceInformation message to carry at least one of the following information:
[0251] (1) information indicating whether the one or more UE-sided AIML models or functionalities are applicable or not (e.g. #2a or #2c) , e.g. the applicability-related information of the one or more UE-sided AIML models or functionalities;
[0252] (2) information indicating whether the one or more UE-sided AIML models or functionalities are applicable or not from the UE's side. In this case, the network node's further determination on the applicability of the one or more UE-sided AIML models or functionalities is needed (e.g. #2d) ; or
[0253] (3) information indicating the network node to determine the applicability of the one or more UE-sided AIML models or functionalities (e.g. #2b) .
[0254] At 704 (optional) , the UE may transmit the RRC UEAssistanceInformation message to the network node, which carries at least one of the above information, e.g. the applicability-related information of the one or more UE-sided AIML models or functionalities.
[0255] Embodiment 2 provides embodiments to avoid an unnecessary applicability-related report.
[0256] Figure 8 illustrates a schematic diagram of transmitting an applicability-related report for an AI / ML model or functionality in accordance with aspects of the present disclosure. Details described in all other embodiments of the present disclosure are applicable for the embodiments shown in Figure 8. Figure 8 describes a specific example of Embodiment 2.
[0257] In Embodiment 2, if a UE is configured to report applicability-related information of a UE-sided AI / ML model or functionality to a network node (aRAN node or an LMF) , the applicability-related information can be one the following:
[0258] (1) information indicating whether the UE-sided AI / ML model or functionality is applicable or not, if the UE is involved in the applicability determination (e.g. LCM approach #2a or #2c or #2d) ;
[0259] (2) information indicating the network node to determine the applicability of the UE-sided AI / ML model or functionality (e.g. LCM approach #2b) , e.g. the associated network node configuration or the network node's side additional condition under which it is applicable.
[0260] In particular, as shown in Figure 8, at 801, a UE may be configured to report applicability-related information of one or more UE-sided AI / ML models or functionalities to a network node (aRAN node or an LMF) in an RRC message, e.g. via an applicability report. When the network node is a RAN node, the applicability report can be in the form of an RRC UEAssistingInfomation message or an RRCReconfigurationComplete message or an RRCResumeComplete message. When the network node is an LMF, the applicability report can be an LPP ProvideCapabilities messages.
[0261] In Embodiment 2, if one or more particular UE-sided AI / ML models or functionalities have been activated or have been reported to be applicable before and now, upon either one of the following cases, the UE may stop reporting the applicability-related information for such particular UE-sided AI / ML models or functionalities for a time period (e.g. T1 as shown in Figure 8) and / or some pre-defined conditions:
[0262] (1) the particular UE-sided AI / ML models or functionalities are deactivated by the network node's decision or command (e.g. LCM approach #3a or #3c) ;
[0263] (2) the particular UE-sided AI / ML models or functionalities are deactivated by the UE decision and reported to the network node (e.g. LCM approach #3b) ;
[0264] (3) the particular UE-sided AI / ML models or functionalities are deactivated by the UE's decision, the UE reports to the network node that the particular UE-sided AI / ML models or functionalities are deactivated, and the UE receives a confirmation from the network node (e.g. LCM approach #3b) ; or
[0265] (4) the particular UE-sided AI / ML models or functionalities are determined to be not applicable by the UE (e.g. LCM approach #2a or #2c or #2d) .
[0266] Then, the UE may resume reporting the applicability-related information for the particular UE-sided AI / ML models or functionalities after the time period and / or the pre-defined conditions.
[0267] In Embodiment 2, the UE may consider that the at least one UE-sided AI / ML model or functionality to be non-applicable, if the at least one UE-sided AI / ML model or functionality is deactivated by the network node's decision or command (e.g. LCM approach #3a or #3c) , or if the at least one UE-sided AI / ML model or functionality is deactivated by the UE's decision and the UE's decision is reported to the network node (e.g. LCM approach #3b) .
[0268] In Embodiment 2, when the network node is a RAN node, the AI / ML activation or deactivation decision or command can be conveyed in an RRC RRCReconfiguration message or a new MAC CE. When the network node is an LMF, the AI / ML activation or deactivation decision or command can be conveyed in an LPP RequestLocationInformation message as well.
[0269] In one example of Embodiment 2, the network node will configure the UE with a prohibit timer (e.g. T1 as shown in Figure 8) and the UE is configured to report the applicability-related information of the one or more UE-sided AI / ML models or functionalities. The prohibit timer may be triggered running in at least one of the following cases:
[0270] - If at least one UE-sided AI / ML model or functionality has been activated and now deactivated by the network node's decision or command (e.g. LCM approach #3a or #3c) ,
[0271] - if at least one UE-sided AI / ML model or functionality has been activated and now deactivated by the UE's decision (e.g. LCM approach #3b) ,
[0272] - if at least one UE-sided AI / ML model or functionality has been determined to be applicable and now not applicable determined by the UE (e.g. LCM approach #2a or #2c or #2d) , or
[0273] - if at least one UE-sided AI / ML model or functionality has been determined to be applicable and now not applicable determined by the UE (e.g. LCM approach #2a or #2c or #2d) , and relevant UAI is triggered to be transmitted.
[0274] When the prohibit timer (e.g. T1 as shown in Figure 8) is running, the UE will stop reporting the applicability-related information for that particular UE-sided AI / ML model or functionality to the network node, and the UE may resume reporting the applicability-related information after the prohibit timer expires.
[0275] In another example of Embodiment 2, the prohibit timer (e.g. T1 as shown in Figure 8) can be possibly configured in an RRC OtherConfig message. When the network node configures the UE to report the applicability information related to one or more UE-sided AI / ML models or functionalities, the network node could also configure a length of an associated prohibit timer. For example, at 800 (optional) of Figure 8, the network node may transmit an RRC OtherConfig message to configure the UE to report the applicability information and configure a length of an associated prohibit timer (e.g. T1 as shown in Figure 8) .
[0276] In another example of Embodiment 2, in LCM approach #2a or #2c or #2d, the UE will stop determining whether the UE-sided AI / ML model or functionality is applicable or not (e.g., based on the UE's side condition, the network node's configuration, and / or the network node's side condition) during the time period (e.g. T1 as shown in Figure 8) , and the UE may resume determining whether the UE-sided AI / ML model or functionality is applicable or not after the time period.
[0277] For example, as shown in Figure 8, at 802, the network node may transmit an activation decision or command to the UE to activate at least one UE-sided AI / ML model or functionality, e.g. in an RRC message or a MAC CE. At 803, the network node may perform monitoring. At 804, the network node may decide to deactivate the at least one UE-sided AI / ML model or functionality. At 805, the network node may transmit a deactivation decision or command (e.g. LCM approach #3a or #3c) to the UE to deactivate the at least one UE-sided AI / ML model or functionality, e.g. in an RRC message or a MAC CE, and prohibit timer T1 is triggered running. At 806, during T1 as shown in Figure 8, the UE stops any applicability report related to the at least one UE-sided AI / ML model or functionality. At 807, after the prohibit timer T1 expires, the UE resumes determining the applicability of the one or more UE-sided AI / ML models or functionalities and resumes an applicability report related to the one or more UE-sided AI / ML models or functionalities. At 808, the UE transmits the applicability report related to the one or more UE-sided AI / ML models or functionalities to the network node.
[0278] In an example of Embodiment 2, the applicability report is allowed to be transmitted by the UE to the network node (e.g. at 808) only if one or more contents of applicability report are changed. More specifically, if the UE-sided AI / ML models or functionalities have been reported already (e.g. at 801) , the same contents will not be reported again (e.g. at 808) . For example, the one or more contents could be:
[0279] (1) information indicating whether the UE-sided AI / ML model or functionality is applicable or not, if UE is involved in the applicability determination (e.g. LCM approach #2a or #2c or #2d) , or
[0280] (2) information indicating the network node to determine the applicability of the UE-sided AI / ML model or functionality (e.g. LCM approach #2b) , e.g., the associated NW configuration or NW side additional condition under which it is applicable.
[0281] Embodiment 3 provides embodiments in which a UE may fallback to a non-AI / ML behavior or a default AI / ML method even if a UE-sided AI / ML model or functionality is activated by a network node.
[0282] In Embodiment 3, in case of LCM approach #1b and "#2a or #2c or #2d, " a UE is able to determine whether a UE-sided AI / ML model or functionality performance is degrading or whether the UE-sided AI / ML model or functionality is applicable under a current circumstance (e.g., a current network node configuration, the UE's side condition, the network node's side condition) . On the other hand, in case of LCM approach #3a or #3c, it is upon the network node’s final decision to deactivate an activated UE-sided AI / ML model or functionality. In such scenario, i.e., in case of one or more LCM approaches of #1b, "#2a or #2c or #2d, " and / or "#3a or #3c, " the UE can (partially) fallback to a non-AI / ML behavior or a default AI / ML method, even if the network node has not formally deactivated the UE-sided AI / ML model or functionality. In some embodiment, whether the UE can behave like above could be upon the network node’s configuration.
[0283] In one example of Embodiment 3, it is upon the network node to configure if the UE can fallback if the UE detects the performance degrading or non-applicable without the network node's confirming regarding the deactivation. For instance, when the network node wants to activate the UE-sided AI / ML model or functionality, the network node configures the UE whether to perform fallback to a non-AI / ML behavior or a default AI / ML method immediately after the UE indicates "the performance degrading or non-applicability" to the network node. The UE's determination of performance degrading or non-applicable could also be upon the network node configuration, e.g., if the accuracy is below a threshold, the measured L1 / L3 beam level RSRP is above or below a threshold, or the measured L3 cell level RSRP is above or below a threshold.
[0284] In one example of Embodiment 3, when the UE indicates "the performance degrading or non-applicability" to the network node, the UE will also indicate to the network node that from now, when reporting, the UE will fallback to a non-AI / ML behavior or a default AI / ML method instead of (or along side) the current activated AI / ML model or functionality. In another example of Embodiment 3, the UE may not indicate anything to the network node, and the UE will autonomously fallback to the non-AI / ML behavior.
[0285] In one example of Embodiment 3, the UE uses a UE-sided AI / ML model or functionality for performing a temporal or spatial domain beam prediction or an RRM prediction. After a prediction result (e.g., predicted beam or cell level L1 / L3 RSRP, a predicted best beam index, or a predicted best cell index etc. ) becomes available, the UE will report the prediction result to the network node. Besides, according to the network node's configuration (e.g., the network node configures the UE to predict one or more certain beam instances and report the prediction result) , the UE may skip measuring the one or more certain beam instances that already have available prediction results, so as to save the UE's measurement effort. However, in one or more LCM approaches of #1b, "#2a or #2c or #2d, " and / or "#3a or #3c, " the UE may indicate "the performance degrading or non-applicability" to the network node and wait for the network node to formally deactivate the UE-sided AI / ML model or functionality. In the meanwhile, since the UE is aware that the prediction result is not enough accurate (e.g. after sending the applicability report to the network node to indicate "the performance degrading or non-applicability" ) , instead of skipping, the UE will perform a measurement and also report an actual measurement result even if the relevant prediction result is available or has been reported. In an example, in the report which is used to convey the prediction result, the UE will report to the network node with the actual measurement result, e.g., a beam level L1 / L3 RSRP measurement result or a cell level L3 RSRP measurement result. In another example, both the actual measurement result and the available prediction result will be sent to the network node either in the same report or in separate reports respectively.
[0286] In one example of Embodiment 3, the UE uses a UE-sided AI / ML model or functionality for direct positioning. In one or more LCM approaches of #1b, "#2a or #2c or #2d, " and / or "#3a or #3c, " the UE may indicate "the performance degrading or non-applicability" to an LMF and wait for the LMF to formally deactivate the UE-sided AI / ML model or functionality. In the meanwhile, since the UE is aware that the prediction result is not enough accurate, (e.g. after sending the applicability report to the LMF indicating "the performance degrading or non-applicability" ) , the UE will use a legacy non-AI / ML method to estimate the UE's location and send the estimated UE's location to the LMF. In an example, both the UE's location estimated by the legacy non-AI / ML method and the available UE's location predicted by the AI / ML method will be sent to the LMF.
[0287] In Embodiment 3, when the network node is a RAN node, the applicability report can be in the form of an RRC UEAssistingInfomation message or an RRCReocnfiguraionComplete message or an RRCResumeComplete message. The AI / ML activation or deactivation command can be conveyed in an RRC RRCReconfiguration message or a new MAC CE. The prediction or measurement result can be carried in a UCI CSI report (in case of L1 result) or an RRC measurement report (in case of L3 result) .
[0288] In Embodiment 3, when the network node is an LMF, the applicability report can be in the form of an LPP ProvideCapabilities messages. The AI / ML activation or deactivation command can be conveyed in an LPP RequestLocationInformation message as well. The UE's location prediction result can be carried in an LPP ProvideLocationInformation message.
[0289] Figure 9 illustrates a schematic diagram of transmitting an applicability-related report for an AI / ML model or functionality in accordance with aspects of the present disclosure. Details described in all other embodiments of the present disclosure are applicable for the embodiments shown in Figure 9. Figure 9 describes a specific example of Embodiment 3.
[0290] As shown in Figure 9, at 901, a UE reports applicability-related information of one or more UE-sided AI / ML models or functionalities to a network node (aRAN node or an LMF) in an RRC message, e.g. via an applicability report. When the network node is a RAN node, the applicability report can be in the form of an RRC UEAssistingInfomation message or an RRCReconfigurationComplete message or an RRCResumeComplete message. When the network node is an LMF, the applicability report can be an LPP ProvideCapabilities messages.
[0291] A 902, the network node may transmit an activation decision or command to the UE to activate at least one UE-sided AI / ML model or functionality, e.g. in an RRC message or a MAC CE. At 903, the UE starts a beam prediction using the at least one UE-sided AI / ML model or functionality. At 904, the UE transmits a beam prediction result in a CSI report in UCI.
[0292] At 905, the UE may determine that the performance of the at least one UE-sided AI / ML model or functionality is degrading or determine that the at least one UE-sided AI / ML model or functionality not applicable any more. At 906, the UE transmits information which indicates "the performance degrading or non-applicability" of the at least one UE-sided AI / ML model or functionality to the network node, e.g. via an applicability report. In the applicability report at 906, the UE may suggest the network node to deactivate the at least one UE-sided AI / ML model or functionality.
[0293] At 907, the UE fallbacks to a beam measurement using a non-AI / ML method. At 908, the UE transmits a beam measurement result in a CSI report in UCI. At 909, the network node may transmit a deactivation decision or command to the UE to deactivate the at least one UE-sided AI / ML model or functionality, e.g. in an RRC message or a MAC CE. As shown in Figure 9, T2 is a time gap between the UE transmitting the information indicating "the performance degrading or non-applicability" to the network node and the UE receiving a formal deactivation decision or command.
[0294] The description herein is provided to enable a person having ordinary skill in the art to make or use the disclosure. Various modifications to the disclosure will be apparent to a person having ordinary skill in the art, and the generic principles defined herein may be applied to other variations without departing from the scope of the disclosure. Thus, the disclosure is not limited to the examples and designs described herein but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein.
Claims
1.A user equipment (UE) , comprising:at least one memory; andat least one processor coupled to the at least one memory and configured to cause the UE to:transmit capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE;receive a first configuration regarding applicability-related information for the one or more AI / ML functionalities; anddetermine whether to transmit the applicability-related information and determine one or more elements included in the applicability-related information according to the first configuration.2.The UE of Claim 1, wherein the capability information indicates at least one of the following:the one or more AI / ML functionalities that can be supported by the UE;one or more life cycle management (LCM) approaches that can be supported by the UE for the one or more AI / ML functionalities; orat least one LCM approach adopted by the UE for each of the one or more AI / ML functionalities.3.The UE of Claim 2, wherein the one or more AI / ML functionalities that can be supported by the UE and the one or more LCM approaches that can be supported by the UE have one-to-one mapping relationship or one-to-multiple or multiple-to-one mapping relationship.4.The UE of Claim 2, wherein the one or more LCM approaches that can be supported by the UE include at least one of the following:an approach for monitoring the one or more AI / ML functionalities;an approach for determining applicability of the one or more AI / ML functionalities;an approach for making an activation decision for the one or more AI / ML functionalities; oran approach for making a deactivation decision for the one or more AI / ML functionalities.5.The UE of Claim 2, wherein the at least one processor is configured to cause the UE to:determine the at least one LCM approach adopted by the UE; orreceive, from the network node, a message including:one or more LCM approaches to be adopted by the UE; anda configuration regarding the one or more LCM approaches to be adopted by the UE.6.The UE of Claim 1, wherein the one or more elements included in the applicability-related information indicate at least one of the following:whether the one or more AI / ML functionalities are applicable;whether the one or more AI / ML functionalities are applicable at the UE and informing that a network node needs to further determine applicability of the one or more AI / ML functionalities; orinforming the network node to determine the applicability of the one or more AI / ML functionalities.7.The UE of Claim 1, wherein the at least one processor is configured to cause the UE to transmit the applicability-related information to a network node, and wherein the applicability-related information is carried in at least one of the following:a long term evolution (LTE) positioning protocol (LPP) provide capabilities message;a radio resource control (RRC) UE assistance information message;an RRC reconfiguration complete message; oran RRC resume complete message.8.The UE of Claim 7, wherein the at least one processor is configured to cause the UE to receive an RRC other configuration message including information indicating the UE to use the RRC UE assistance information message to carry the applicability-related information.9.The UE of Claim 1 or Claim 7, wherein if at least one AI / ML functionality within the one or more AI / ML functionalities has been activated or has been reported to be applicable, the at least one processor is configured to cause the UE to:stop transmitting the applicability-related information for the at least one AI / ML functionality for at least one of a time period or a pre-defined condition; andresume transmitting the applicability-related information for the at least one AI / ML functionality after the at least one of the time period or the pre-defined condition.10.The UE of Claim 9, wherein transmitting the applicability-related information for the at least one AI / ML functionality is stopped in response to at least one of the following:the at least one AI / ML functionality is deactivated by a decision from the network node;the at least one AI / ML functionality is deactivated by a decision of the UE;the at least one AI / ML functionality is deactivated by the decision of the UE, and a deactivation result of the at least one AI / ML functionality is reported by the UE to the network node;the at least one AI / ML functionality is deactivated by the decision of the UE, the deactivation result of the at least one AI / ML functionality is reported by the UE to the network node, and a confirmation is received by the UE from the network node; orthe at least one AI / ML functionality is determined by the UE to be non-applicable.11.The UE of Claim 10, wherein the at least one processor is configured to cause the UE to consider that the at least one AI / ML functionality is to be non-applicable in response to at least one of the following:the at least one AI / ML functionality is deactivated by the decision from the network node;the at least one AI / ML functionality is deactivated by the decision of the UE;the at least one AI / ML functionality is deactivated by the decision of the UE, and the deactivation result of the at least one AI / ML functionality is reported by the UE to the network node; orthe at least one AI / ML functionality is deactivated by the decision of the UE, the deactivation result of the at least one AI / ML functionality is reported by the UE to the network node, and a confirmation is received by the UE from the network node.12.The UE of Claim 10, wherein:the first configuration includes a second configuration regarding the at least one of the time period or the pre-defined condition; orthe at least one processor is configured to cause the UE to receive an RRC other configuration message including the second configuration.13.The UE of Claim 10, wherein the at least one processor is configured to cause the UE to:stop determining whether the at least one AI / ML functionality is applicable during the time period; andresume determining whether the at least one AI / ML functionality is applicable after the time period.14.The UE of Claim 1, wherein if at least one AI / ML functionality within the one or more AI / ML functionalities has been activated, the at least one processor is configured to cause the UE to:determine whether the at least one AI / ML functionality is degrading or whether the at least one AI / ML functionality is applicable; andfallback to a non-AI / ML behavior or a default AI / ML method, if the at least one AI / ML functionality is degrading or if the at least one AI / ML functionality is not applicable, and if it is upon a network node’s final decision to deactivate the at least one AI / ML functionality.15.The UE of Claim 14, wherein the at least one processor is configured to cause the UE to receive, from the network node, a configuration indicating that the UE can fallback in case that the at least one AI / ML functionality is degrading or in case that the at least one AI / ML functionality is not applicable without the network node confirming a deactivation of the at least one AI / ML functionality.16.The UE of Claim 14, wherein the at least one processor is configured to cause the UE to transmit, to the network node, information indicating that the UE will fallback to the non-AI / ML behavior or the default AI / ML method.17.The UE of Claim 14, wherein the at least one processor is configured to cause the UE to:measure a measurement result; andtransmit the measurement result to the network node.18.A network node, comprising:at least one memory; andat least one processor coupled to the at least one memory and configured to cause the network node to:receive, from a user equipment (UE) , capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE; andtransmit, to the UE, a first configuration regarding applicability-related information for the one or more AI / ML functionalities.19.A processor for wireless communication, comprising:at least one controller coupled with at least one memory and configured to cause the processor to:transmit capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE;receive a first configuration regarding applicability-related information for the one or more AI / ML functionalities; anddetermine whether to transmit the applicability-related information and determine one or more elements included in the applicability-related information according to the first configuration.20.A method performed by a user equipment (UE) , comprising:transmitting capability information related to one or more artificial intelligence (AI) / machine learning (ML) functionalities supported at the UE;receiving a first configuration regarding applicability-related information for the one or more AI / ML functionalities; anddetermining whether to transmit the applicability-related information and determining one or more elements included in the applicability-related information according to the first configuration.
Citation Information
Patent Citations
Method and apparatus for using artificial intelligence / machine learning model in wireless communication network
US20240214840A1
ML model generalization and specification
WO2024041595A1
Model management method and apparatus, and communication system
WO2024092766A1