Wireless communication method for applying ai / ML functionalities and related devices
The proposed wireless communication method using AI/ML capabilities in user equipment and base stations addresses radio condition degradation in 5G/6G networks by enhancing data collection and prediction, improving mobility management and reducing service interruptions.
Patent Information
- Application Number
- PCT/CN2025/108262
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-11-01
- Filing Date
- 2025-07-11
- Publication Date
- 2026-01-15
AI Technical Summary
In high-speed mobile 5G/6G cellular communication networks, user equipment experiences radio condition degradation due to the Doppler effect and fast fading, leading to inefficient mobility management and frequent cell switching, causing service interruptions. AI/ML technologies are needed to enhance data collection and prediction for improved system performance.
A wireless communication method is proposed that enables user equipment and base stations to exchange AI/ML capability information, perform data collection based on life cycle management frameworks, and report predicted data for model training and inference, facilitating enhanced mobility and network management.
The method improves communication performance by enabling efficient data collection and prediction, reducing service interruptions and enhancing mobility management in high-speed mobile networks.
Smart Images

Figure CN2025108262_15012026_PF_FP_ABST
Abstract
Description
WIRELESS COMMUNICATION METHOD FOR APPLYING AI / ML FUNCTIONALITIES AND RELATED DEVICESTECHNICAL FIELD
[0001] The present disclosure relates to the field of wireless communications, and more particularly, to a wireless communication method for applying AI / ML functionalities and related devices.BACKGROUND ART
[0002] Communication systems and networks have developed towards being a broadband and mobile system. In cellular wireless communication systems developed by the Third Generation Partnership Project (3GPP) , user equipment (UE) is connected by a wireless link to a next generation radio access network (NG-RAN) . The NG-RAN includes a set of base stations which provide wireless links to the UEs located in cells covered by the base station, and the interface (s) to a core network (CN) which provides overall network control. The 3GPP has developed the so-called 4G or Long-Term Evolution (LTE) system, namely, an Evolved Universal Mobile Telecommunication System Territorial Radio Access Network (E-UTRAN) , for a mobile access network where one or more macro-cells are supported by a base station known as an eNodeB or eNB (evolved NodeB) . More recently, evolved from LTE, the so-called 5G or new radio (NR) systems where one or more cells are supported by a base station known as a gNB.
[0003] As 5G or NR system is becoming pervasive across industries and geographical areas, handling more advanced services and applications requiring very high data rate (e.g., XR) , networks are being denser, use more antennas, larger bandwidths, and more frequency bands. Beyond 5G (B5G) or 6G will be able to use higher frequencies than 5G and provide substantially higher capacity and much lower latency. Artificial intelligence and machine learning (AI / ML) and AI / ML prediction are expected to facilitate large improvements in the presence technology of new radio and mobility awareness.
[0004] In high-speed mobile 5G / 6G and further release cellular communication networks, the UE may experience degradation in radio conditions due to the doppler effect and fast fading. The traditional mobility management procedure (e.g., handover, location update) is not performed smoothly enough to adapt to large channel variations in extreme mobility scenarios. Furthermore, the usage of the higher frequency in the cell will cause smaller coverage and it will further cause frequent handover or cell switching or ping-pong effects. A non-negligible service interruption due to the frequent change of cell would be aggravated. AI / ML enables new capabilities and drives advancements in technology. Therefore, there is a need to develop a new data collection procedure for AI / ML prediction. Furthermore, AI / ML technologies have to be enhanced in new radio, while AI / ML predictions provide proper insights into future events for optimal system performance.SUMMARY
[0005] An object of the present disclosure is to propose a wireless communication method for applying AI / ML functionalities and related devices (such as a user equipment (UE) and / or a base station (BS) ) , which can solve issues in the prior art, enhancing data collection procedure, realizing measurement result / event prediction and inference, and / or provide a good communication performance.
[0006] In a first aspect of the present disclosure, provided is a wireless communication method for applying AI / ML functionality, executable by a user equipment (UE) , the method including sending, by the UE in a network, artificial intelligence (AI) / machine learning (ML) capability information to the network, wherein the AI / ML capability information indicates AI / ML functionality at UE side; receiving from the network a data collection configuration configured based on the AI / ML functionality in accordance with a life cycle management (LCM) framework; performing data collection based on the data collection configuration and the AI / ML functionality to obtain collected data and predicted data; and reporting the predicted data to the network for AI / ML model training, management, and / or inference.
[0007] In a second aspect of the present disclosure, provided is a wireless communication method for applying AI / ML functionalities, executable by a base station (BS) , the method including receiving from a user equipment (UE) in a network artificial intelligence (AI) / machine learning (ML) capability information, wherein the AI / ML capability information indicates AI / ML functionality at UE side; transmitting to the UE a data collection configuration configured based on the AI / ML functionality in accordance with a life cycle management (LCM) framework; and receiving from the UE collected data and predicted data, which are obtained by performing data collection based on the data collection configuration and the AI / ML functionality by the UE, for AI / ML model training, management, and / or inference.
[0008] In a third aspect of the present disclosure, provided is a wireless communication method for applying AI / ML functionality, executable by a user equipment (UE) , the method including receiving from the network a measurement configuration including configurations of actual measurement and predicted measurement to start data collection; performing predicted measurement by using an artificial intelligence (AI) / machine learning (ML) based on the measurement configuration; and sending a predicted measurement report to the network for AI / ML model training, management, and inference.
[0009] In a fourth aspect of the present disclosure, provided is a wireless communication method for applying AI / ML functionalities, executable by a base station (BS) , the method including transmitting to a user equipment (UE) in a network a measurement configuration including configurations of actual measurement and predicted measurement for the UE to start data collection; and receiving from the UE a predicted measurement report for AI / ML model training, management, and inference, wherein the predicted measurement report is obtained by performing predicted measurement by using an artificial intelligence (AI) / machine learning (ML) based on the measurement configuration by the UE.
[0010] In a fifth aspect of the present disclosure, a user equipment includes a memory and a processor coupled to the memory, the processor configured to call and run program instructions stored in a memory, to execute the method of any of the first aspect and the third aspect.
[0011] In a sixth aspect of the present disclosure, a base station includes a memory and a processor coupled to the memory, the processor configured to call and run program instructions stored in a memory, to execute the method of any of the second aspect and the fourth aspect.
[0012] In a seventh aspect of the present disclosure, a non-transitory machine-readable storage medium has stored thereon instructions that, when executed by a computer, cause the computer to perform any of the above methods.
[0013] In an eighth aspect of the present disclosure, a chip includes a processor, configured to call and run a computer program stored in a memory, to cause a device in which the chip is installed to execute any of the above methods.
[0014] In a ninth aspect of the present disclosure, a non-transitory computer readable storage medium, in which a computer program is stored, causes a computer to execute any of the above methods.
[0015] In a tenth aspect of the present disclosure, a computer program product includes a computer program, and the computer program causes a computer to execute any of the above methods.
[0016] In an eleventh aspect of the present disclosure, a computer program causes a computer to execute any of the above methods.DESCRIPTION OF DRAWINGS
[0017] In order to more clearly illustrate the embodiments of the present disclosure or related art, the following figures will be described in the embodiments are briefly introduced. It is obvious that the drawings are merely some embodiments of the present disclosure, a person having ordinary skill in this field can obtain other figures according to these figures without paying the premise.
[0018] FIG. 1 (a) is a schematic diagram illustrating a communication controlling system according to an embodiment of the present disclosure.
[0019] FIG. 1 (b) is a block diagram of a user equipment and a base station of wireless communication in a communication controlling system according to an embodiment of the present disclosure.
[0020] FIG. 2 is a schematic diagram illustrating radio protocol architecture within gNB and UE for AI / ML.
[0021] FIG. 3 is a schematic diagram illustrating a gNB further including a centralized unit (CU) and a plurality of distributed unit (DUs) .
[0022] FIG. 4 is a schematic diagram illustrating a general functional AI / ML framework for RAN intelligence according to an embodiment of the present disclosure.
[0023] FIG. 5 is a flowchart of a data collection and reporting procedure according to an embodiment of the present disclosure.
[0024] FIG. 6 is a schematic diagram illustrating an example of an allowed prediction range according to an embodiment of the present disclosure.
[0025] FIG. 7 is a flowchart of a data collection configuration / reconfiguration procedure according to an embodiment of the present disclosure.
[0026] FIG. 8 is a schematic diagram illustrating an association between condition identifier and model identifier according to an embodiment of the present disclosure.
[0027] FIG. 9 is a flowchart of a wireless communication method for applying AI / ML functionalities by a user equipment according to an embodiment of the present disclosure.
[0028] FIG. 10 is a flowchart of a wireless communication method for applying AI / ML functionalities by a base station according to an embodiment of the present disclosure.
[0029] FIG. 11 is a flowchart of a wireless communication method for applying AI / ML functionalities by a user equipment according to another embodiment of the present disclosure.
[0030] FIG. 12 is a flowchart of a wireless communication method for applying AI / ML functionalities by a base station according to another embodiment of the present disclosure.
[0031] FIG. 13 is a flowchart of data collection and prediction for NW-initiated Handover according to a first embodiment of the present disclosure.
[0032] FIG. 14 is a flowchart of data collection and prediction for UE-initiated Handover according to a second embodiment of the present disclosure.
[0033] FIG. 15 is a flowchart of data collection and prediction for MBS transmission / handover according to a third embodiment of the present disclosure.
[0034] FIG. 16 is a flowchart of data collection and prediction for UE-initiated RAN-based MBS notification area update according to a fourth embodiment of the present disclosure.
[0035] FIG. 17 is a flowchart of data collection and prediction for paging according to a fifth embodiment of the present disclosure.
[0036] FIG. 18 is a flowchart of data collection and prediction for NW-initiated conditional Handover according to a sixth embodiment of the present disclosure.
[0037] FIG. 19 is a flowchart of data collection and prediction for NES according to a seventh embodiment of the present disclosure.
[0038] FIG. 20 is a flowchart of data collection and prediction for NW-initiated load balancing or SON according to an eighth embodiment of the present disclosure.DETAILED DESCRIPTION OF EMBODIMENTS
[0039] Embodiments of the disclosure are described in detail with the technical matters, structural features, achieved objects, and effects with reference to the accompanying drawings as follows. Specifically, the terminologies in the embodiments of the present disclosure are merely for describing the purpose of the certain embodiment, but not to limit the disclosure.
[0040] In this document, the term " / " should be interpreted to indicate "and / or. " A combination such as “at least one of A, B, or C, ” “one or more of A, B, or C, ” “at least one of A, B, and C, ” “one or more of A, B, and C, ” or “A, B, and / or C” may be A only, B only, C only, A and B, A and 30 C, B and C, or A and B and C, where any combination may contain one or more members of A, B, or C.
[0041] A schematic view and a functional block diagram of a communication controlling system 1 (e.g., 5GS, 6GS) according to the present disclosure are shown in FIG. 1 (a) and FIG. 1 (b) respectively. The communication controlling system 1 includes a user equipment 10 and a base station 20. The user equipment 10 and the base station 20 may communicate with each other either wirelessly or in a wired way. The base station 20 and a next generation core network 30 may also communicate with each other either wirelessly or in a wired way. When the communication controlling system 1 complies with the New Radio (NR) standard of the 3rd Generation Partnership Project (3GPP) , the next generation core network (5GCN) 30 is a backend serving network system and may include an Access and Mobility Management Function (AMF) , User Plane Function (UPF) , and a Session Management Function (SMF) .
[0042] The user equipment 10 may be an RRC_IDLE / RRC_INACTIVE / RRC_CONNECTED AI / ML-capable UE or non-AI / ML UE while the base station 20 may be an intelligent NG-RAN node or legacy NG-RAN node, but the present disclosure is not limited to this. Both the AI / ML-capable UE and intelligent NG-RAN node support Artificial Intelligence (AI) / Machine Learning (ML) capabilities and techniques whereas AI / ML is not supported by the non-AI / ML UE and legacy NG-RAN node.
[0043] The user equipment 10 includes at least one transceiver 12 and a processor 14, which are electrically connected with each other. The base station 20 includes a transceiver 22 and a processor 24, which are electrically connected with each other. The transceiver 12 of the user equipment 10 is configured to transmit a signal to the base station 20 (and receive a signal from the base station 20) and the processor 24 of the base station 20 processes the signal, the transceiver 22 of the base station 20 is configured to transmit a signal to the user equipment 10 (and receive a signal from the user equipment 10) and the processor 14 of the user equipment 10 processes the signal. In this way, the user equipment 10 communicates with the base station 20 each other.
[0044] The radio protocol architecture within the base station (gNB) and UE is shown in FIG. 2, which includes Radio Resource Control (RRC) , Service Data Adaptation Protocol (SDAP) , Packet Data Convergence Protocol (PDCP) , Radio Link Control (RLC) , Medium Access Control (MAC) . RAN functional split is supported for new radio and further generation mobile networks. The new radio protocol architecture within the base station (gNB) and the UE for RAN functional split is shown in FIG. 3, which includes SDAP, RRC, PDCP, RLC, MAC. In case of RAN functional split, the gNB further includes a centralized unit (CU) and a plurality of distributed unit (DUs) as shown in FIG. 3. The protocol stack of CU includes an RRC layer, a SDAP layer, and a PDCP layer, while the protocol stack of DU includes an RLC layer, a MAC layer, and a PHY layer. The F1 interface between the CU and DU is established between the PDCP layer of the protocol stack and the RLC layer of the protocol stack.
[0045] To support the efficient operation of AI / ML techniques in new radio and next generation mobile networks, a unified AI / ML framework and data flows for AI-enabled NG-RAN (e.g., RAN intelligence) corresponding to various use cases is shown in FIG. 4. The unified AI / ML framework (e.g., Life Cycle Management framework) consists of Data Collection, Model Training, Management, Inference, and Model Storage. A data collection procedure performed by the UE / Network (e.g., gNB, LMF, or OAM) is used to produce the training data, monitoring data, and inference data for AI / ML model training, inference, and data analytics and management. The AI / ML model can be stored at UE / network side when the trained / updated / transferred model is derived. AI / ML prediction can provide configuration into future events for optimal system performance. Examples of input data for the data collection procedure may include L1 / L3 measurements from UEs or different network entities, feedback from AI / ML management, output from an AI / ML inference / prediction model. The UE and RAN intelligence can use AI / ML prediction to facilitate enhanced mobility, multicast and broadcast service, load balancing, self-organizing network, and network energy saving applications, etc.
[0046] An AI / ML model is associated with a specific configuration (s) / condition (s) indicated by UE capability and additional condition (s) (e.g., applicability related information, scenarios, monitoring data) . UE capabilities are related to collect / log the training / monitoring / inference data at the UE side. Additional conditions refer to any additional data that are used for model training, management, and inference but that may be not a part of UE capability. The additional conditions can be divided into two categories: network-sided additional conditions and UE-sided additional conditions. For the model training and inference, the network and UEs can collect network-sided additional conditions and UE-sided additional conditions jointly so that the model can be trained and inferenced under the additional conditions. In some cases, transferring UE capabilities and network-sided / UE-sided additional conditions can facilitate the network / UE to determine whether / which the AI / ML function / function or feature group (FG) / functionality is applicable or not. Therefore, the network can configure AI / ML function / FG / functionalities via RRC / LPP message to the UE while the UE reports its applicable AI / ML function / FG / functionalities via RRC / UAI / LPP message based on UE capability and additional conditions. Both proactive reporting and reactive reporting of UE-sided additional conditions are supported to allow UE to report UE-sided additional conditions with or without network configured candidate information.
[0047] Regarding Life Cycle Management (LCM) for network-sided model, the model training, inference, and management are performed at the network side. The instruction / signaling for model monitoring / management are negotiated internally within the network entities. The UE may not be involved in any decision making for network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) , except being configured to provide the required measurements and data collection. The network would configure the radio measurements and the related reporting via radio signaling (e.g., RRC Reconfiguration) to enable data collection as a basis for AI / ML model training, management, and inference. The following are the types of data collection for network-sided model.
[0048] (1) gNB-centric data collection: based on UE request or applicable functionality, the gNB can configure the UE with the enhanced data collection configuration (e.g., an adaptive measurement configuration, an adaptive CSI measurement configuration) to start / stop the data collection procedure. A L1 / L2 / L3-based solution can be a suitable candidate for the gNB-centric data collection. The network should transmit the enhanced data collection configuration to the UE for the related data collection (e.g., measurement, UE interests) and reporting. The associated measurement results (i.e., RSRP / RSRQ / SINR / RSSI) may be immediate or logged and should be reported by the UE for network-sided data collection. In addition, the extending CSI report is used to allow reporting measurements of more beams (e.g., more than 4) for model training and management. Accordingly, the UE can report the collected data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) signaling as shown in FIG. 5.
[0049] In FIG. 5, an initial enhanced data collection configuration is optional configured from the network based on the pre-configured parameters or history-data collection. An optional AI / ML capability enquiry is transmitted by network to initiate a UE for reporting the supported AI / ML functionalities. In some cases, the network may configure a list of AI / ML capable cells and non-AI / ML capable cells for UE reporting. The UE should report its AI / ML capability (e.g., supported functionality, model structure, applicable functionality and conditions, additional condition) via UE-assistance (e.g., UECapability report, LPP) and / or real-time report framework (e.g., UEAssitanceInformation (UAI) , NeedForGap) to indicate the supported / applicable functionality. After model management, the network may configure the AI / ML-enabled feature for the UE. In some cases, the network may configure AI / ML function / FG / functionalities via RRC / LPP message which may include network-sided additional condition to the UE. The enhanced data collection configuration would be configured for the UE based on the AI / ML capability and UE-sided / network-sided additional condition (s) . The UE AI / ML capability and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . Then the UE reports the collected data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) signaling. In some cases, the UE may report the applicable functionality in response to the request by network-sided additional condition. In further cases, when the at least one reporting criterion is met, the UE can send the predicted data report for AI / ML model training and inference.
[0050] (2) OAM-centric data collection: the OAM provides the enhanced data collection configuration (e.g., an adaptive measurement configuration, an adaptive logged measurement configuration via the gNB) needed for the UE to start / stop the data collection procedure. The Minimization of Drive Tests (MDT) -based solution can be a suitable candidate for the OAM-centric data collection. The immediate MDT and logged MDT can be supported based on the state of UE. For example, the immediate MDT can be works for the RRC_CONNECTED UEs whereas the logged MDT can be enabled for power saving (e.g., RRC_IDLE or RRC_INACTIVE) UEs. Immediate MDT is operated upon the existing RRC / RRM reporting procedure. Logged MDT is used to reduce the reporting overhead from the UE / network.
[0051] Considering model training, inference, and management are performed at the network side, real-time data collection may not have the stringent requirements while the signaling overhead of data collection might be a concern. Due to the hardware limitation of UEs, a mass of continuously data reporting would extremely increase the signaling overhead and UE power consumption, especially for the long-term model monitoring. For the sake of signaling overhead reduction and power saving, the UE should support intelligent measurement and report based on the enhanced data collection configuration. The network may configure / request the UE to perform intelligent measurement and to report the periodical / event-based collected / predicted data when the reporting condition (s) is met (e.g., based on SS / PBCH block (s) , CSI-RS framework) . The intelligent measurement and report are performed while the UE is in RRC_IDLE / RRC_INACTIVE / RRC_CONNECTED. During UE configuration phase, the network initiates the enhanced data collection configuration to the UE by means of dedicated signaling (e.g., the intelligent measurement configuration message, OtherConfig, using RRCReconfiguration or RRCResume) . The initial data collection configuration is configured based on the pre-configured parameters or history-data collection. For AI / ML model training and inference, the enhanced data collection configuration (e.g., intelligent measurement configuration, CSI measurement configuration) may include the following parameters to provide measurement and report without increasing signaling overhead and power consumption: - Measurement objects: According to AI / ML model training, management, and inference for UE location / UE trajectory, the network can provide the information of measurement objects on which the UE shall perform the measurements. The information indicates the frequency / time / space and subcarrier spacing of reference signals to be measured. To avoid service interruption, the network may configure a list of AI / ML capable cells and non-AI / ML capable cells for UE measurement / prediction and data collection. In some cases, the network may provide the predicted candidate cell / beam list for UE measurement, and provide the predicted excluded cell / beam list for UE skipping measurements. There are two categories for measurement prediction of measurement object: (1) direct prediction and (2) indirect prediction. In direct measurement prediction, the actual RRM beam / cell-level measurement result (e.g., RSRP) could be a part of the model inputs. On the other hand, the predicted beam / cell-level measurement data of the measurement objects is used for the indirect measurement prediction. In some cases, the associated estimated information (e.g., UE trajectory, movement trends) can be also considered as input of the model for the indirect measurement prediction. The output of the inference / prediction model would be the predicted measurement data. - Reporting configuration: According to AI / ML model training, management, and inference for UE location / UE trajectory, the network can provide the at least one criterion (e.g., triggering threshold, allowed predicting threshold, preconfigured event, etc. ) that triggers the UE to measure / predict or send a measurement / predicted report. This can either be periodical or event triggered. The network may configure at least one of triggering threshold (s) and an allowed predicting threshold for intelligent measurement and report. The reporting type may be a prediction result in addition to the actual beam / cell-level measurement results (SS / PBCH block-based or CSI-based) . It means that the UE may send a predicted measurement report based on the triggering threshold and / or allowed predicting threshold. For example, an adaptive RSRP / RSSI threshold is one type of allowed predicting threshold that triggers the UE to report the predicted measurement if necessary. In some cases, an adaptive RSRP / RSSI change threshold is the other type of allowed predicting threshold that can be used for the UE to determine whether to transmit the (predicted) measurement report for AI / ML model training and inference. The adaptive RSRP / RSSI change threshold is a criterion of the difference between two measurements (e.g., two predicted measurements or an actual measurement and an allowed predicted measurement) . For example, if the UE is configured with an adaptive RSRP change threshold for AI / ML mobility prediction, the first actual measurement (e.g., RSRP1) and the second predicted measurement (e.g., RSRP2) should be performed within a time duration. When the triggering threshold is met and the difference of RSRP1 and RSRP2 is met on the adaptive RSRP change threshold which is one of the allowed predicting thresholds, the UE can send the predicted measurement reports to the network for AI / ML prediction. FIG. 6 shows an example of allowed prediction range that can be configured for the UE to perform predicted measurement and to report the measurements when the at least one reporting criterion is met. The UE may report the first measured data within time duration D1 and then predict the measurement during a period P. If the triggering threshold is met and the difference of two measurements is met on the adaptive RSRP change threshold CT, the UE can send the predicted measurement reports to the network for AI / ML prediction. At least the UE should send an actual measurement report at timing T2 which is the end of period P if no other predicted measurement result meets the at least one reporting criterion (e.g., the UE is stable and needs periodic reporting) . In some cases, the UE reports the actual and / or predicted measurement for the same cell / beam (e.g., cell identifier, beam identifier) once the at least one reporting criterion is met (e.g., event-triggered reporting) . In some other cases, the UE would not report the predicted measurement since the at least one reporting criterion and / or the adaptive RSRP change threshold is not met. In some further cases, the network may enlarge the measurement gap (i.e., reduce the number of actual measurement time instances) for the actual measurement while the predicted measurement can be used to compensate for the reduced actual measurement. In an ideal case, if the predicted measurement can be derived consistently between the UE side and network side, the predicted measurement may not be transmitted by the UE even if the at least one reporting criterion and / or the adaptive RSRP change threshold is met. The measurement reduction rate in temporal domain is associated with the skipped actual measurement time instances within the total measurement time instances. The higher measurement reduction rate gains the better resource efficiency. It means that the allowed prediction range can be reconfigured for the ideal case based on the long-term output of AI / ML model training and performance feedback so that the signaling overhead reduction and UE power saving is achieved. - Quantity configuration: According to AI / ML model training, management, and inference, the network can provide the quantity to filter the event evaluation and related measurement report. - Measurement identifier: According to AI / ML model training, management, and inference, the network can provide a list of predicting measurement identifiers where each predicted measurement identifier is associated with at least one measurement object and report configuration. The measurement identifier can be included in the intelligent measurement report that is a predicted report, providing as a reference to the network for AI / ML model training and inference. - Measurement gaps: According to AI / ML model training, management, and inference, the network can provide a time duration that the UE may use to perform predicted measurements. The time duration may be started upon the reception of enhanced data collection configuration. The predicted measurements can be checked with the allowed predicting threshold within the measurement gap, when the at least one reporting criterion is met, the UE can send the predicted measurement report for AI / ML model training and inference. At least the UE should send a data report at the end of time duration if no other predicted data meets the at least one reporting criterion during the time duration as shown in FIG. 7. In some cases, measurement gap reduction may be considered to provide available resource gain to the UE. Based on the correlation of the predicted measurement / previous measurement reports, the number of measurement gaps is adaptive configured for the next data collection interval.
[0052] Upon the reception of data collection from UE (s) for network-sided model training, inference, and management, the network can check the performance metrics based on AI / ML model output and the actual measured results to determine the functionality management (e.g. selection, (de) activation, switching, fallback, etc. ) . The network can finetune / reconfigure the enhanced data collection configuration according to AI / ML model output and performance monitoring feedback as shown in FIG. 7. A release / removal / addition / modification operation for enhanced data collection configuration in the UE is realized when the configuration is overwritten or by configuration in case that an adaptive measurement timer stops or is expired. The enhanced data collection configuration can be used to facilitate AI / ML prediction for enhanced mobility, multicast and broadcast service, load balancing, self-organizing network, and network energy saving applications, etc.
[0053] A UE-sided model refers to an AI / ML model that inference is performed entirely at the UE side while two-sided model refers to an AI / ML model that inference is performed jointly across the UE and the network, e.g., the first part of inference is performed by the UE and then the remaining part is performed by the gNB. The AI / ML model / functionality should be transferred from the network to the UE for supporting AI / ML-based approach at UE side. In UE-sided model or UE part of two-sided model where the model is transferred from the network to the UE, UE initially reports its AI / ML capability (e.g., supported functionality, model structure, applicable functionality and conditions, additional condition) via UE-assistance (e.g., UECapability report, LPP) and / or real-time report framework (e.g., UEAssitanceInformation (UAI) , NeedForGap) to indicate the applicable functionality. In some cases, the real-time report framework includes an indication from the UE to request AI / ML configuration (e.g., measurement) for data collection. Then at least one of AI / ML configurations / models is transferred from the network to the UE. The network may configure the AI / ML-enabled feature for the UE. The UE may apply the associated actions when the AI / ML-enabled feature / FG / functionality and the applicable condition (s) are met based on the AI / ML model. In some embodiments, the applicable conditions are either determined by UE-sided model or configured together with the transferred model from the network for the model selection at UE side. For example, there is a condition identifier associated with an additional condition. The UE collects / predicts the data and trains the AI / ML model based on the condition identifier. And then the UE takes the condition identifier into account and reports a model identifier associated with an AI / ML model to the network. It should be noted that there is an association between the condition identifier and model identifier. The association of condition identifier and model identifier is pre-configured and aligned between the network side and UE side as shown in FIG. 8.
[0054] As a result of the enhanced data collection, the UE may report the radio resource management (RRM) measurement result and / or measurement event to make as the reference for mobility and / or resource management (e.g., handover decision, network energy saving) . In some cases, the measurement event predictions can be performed based on the intelligent measurement result. There are two categories for measurement event prediction: (1) direct prediction and (2) indirect prediction. In direct measurement event prediction, the actual RRM beam / cell-level measurement result (e.g., RSRP) could be a part of the model inputs. Some of the associated actual UE information (e.g., moving speed of UE) can be also considered as input of the model. On the other hand, the predicted beam / cell-level measurement data of the serving cell and neighboring cells is used for the indirect measurement event prediction. Some of the associated estimated information (e.g., UE trajectory, movement trends) can be also considered as input of the model. The output of the inference / prediction model would be the predicted measurement data and / or the probability of measurement event occurrence within a time duration. Since multiple measurement thresholds / offsets and new events are predefined / preconfigured in the system for different mobility scenarios and features, the UE can report the predicted measurement data / event when the at least one reporting criterion is met. In some cases (e.g., load balancing, SON) , the network can transmit the prediction / request to next generation core network (NGC) and / or AI / ML server based on UE report and network self-measurement. According to the output of the inference / prediction model and UE report, the UE and / or the network can determine whether to perform the enhanced mobility, multicast and broadcast service, load balancing, self-organizing network, and network energy saving applications. The network can provide predicted configuration and action into the future events to optimize system performance. In one embodiment of the present disclosure, the direct / indirect measurement event prediction can be used to provide the predicted configuration and report from network or UEs. In some cases, the direct / indirect measurement event prediction is not only used for UE mobility enhancement but also network resource management. For example, when an indirect measurement event Ai is predicted with a high probability of occurrence (e.g., over a configured threshold) , the event Ai is used to trigger an early conditional handover procedure based on the intelligent measurement and report. In an example, the event Ai is predicted to trigger NES associated procedure.
[0055] In an aspect, an enhanced data collection procedure is proposed in this disclosure for AI / ML model training and inference. An intelligent measurement configuration of data collection is proposed to configure measurements for multiple datasets and use cases (e.g., mobility, MBS, positioning, NES) . The collected / predicted data are expected to facilitate the AI / ML operations including UE-side, network-side, two-side model to predict UE mobility and network management for reducing interruption time and improving system throughput.
[0056] FIG. 9 illustrates a wireless communication method 100 for applying AI / ML functionalities by a UE according to an embodiment of the present disclosure. The method 100 is executable by a user equipment (UE) in a network and includes the following.
[0057] In Step 110, the UE sends artificial intelligence (AI) / machine learning (ML) capability information to the network, wherein the AI / ML capability information indicates AI / ML functionality at UE side.
[0058] In an example, the UE may receive from the network an AI / ML capability enquiry used to initiate the UE for reporting the AI / ML capability information. Upon reception of the AI / ML capability enquiry, the AI / ML capability information is sent from the UE to the network to indicate the AI / ML functionality. Furthermore, before the receiving the AI / ML capability enquiry, the UE may receive an initial data collection configuration configured based on pre-configured parameters or history-data collection. In an example, the UE may report applicable AI / ML functionality in response to a request made by network-sided additional condition. For example, the applicable AI / ML functionality may be reported to the network via UE assistance information (UAI) or radio resource control (RRC) message based on UE capability and additional condition.
[0059] In Step 120, the UE receives from the network a data collection configuration configured based on the AI / ML functionality in accordance with a life cycle management (LCM) framework.
[0060] The data collection configuration may include a measurement configuration to enable predicted measurement related to the AI / ML model training, management, and / or inference. The data collection configuration may include the measurement objects, reporting configuration, quantity configuration, measurement identifier and measurement gaps as described above. More specifically, the data collection configuration may include an information of measurement objects on which the UE should perform measurements / predictions. The information of measurement objects may indicate resource of reference signals to be measured / predicted. The resource of reference signals may include at least one of frequency, time, space, and subcarrier spacing to be measured / predicted. The information of measurement objects may include a list of AI / ML capable cells and non-AI / ML capable cells, a candidate cell / beam list, and / or an excluded cell / beam list. For the most cases, the data collection configuration may include an intelligent measurement configuration to enable online / offline data collection.
[0061] In an example, the data collection configuration may be configured for the UE based on the AI / ML functionality and further based on UE-sided / network-sided additional condition related to the AI / ML model training, management, and / or inference. In an example, the data collection may be performed based on the data collection configuration, the AI / ML functionality and the network-sided additional condition.
[0062] In addition, the measurement gap is a time duration the UE uses to perform predicted measurement, and a predicted measurement is checked with an allowed predicting threshold within the measurement gap. The measurement gap may be started upon the reception of data collection configuration. A data report may be sent to the network at an end of the measurement gap if no other predicted data meets at least one reporting criterion during the measurement gap. The measurement gap may be enlarged to reduce a number of actual measurement time instances for the actual measurement.
[0063] Additionally, the data collection configuration may be finetuned / reconfigured according to AI / ML model output and performance monitoring feedback. The network may transmit an updated data collection configuration to the UE according to the AI / ML model output. Additionally, the UE may preform a release / removal / addition / modification operation for the data collection configuration when the data collection configuration is overwritten or by configuration in case that an adaptive measurement timer stops or is expired.
[0064] In Step 130, the UE performs data collection based on the data collection configuration and the AI / ML functionality to obtain collected data and predicted data.
[0065] In an example, the UE may receive from the network-sided additional condition, and the data collection may be performed based on the data collection configuration, the AI / ML functionality and the network-sided additional condition.
[0066] In Step 140, the UE reports the predicted data to the network for AI / ML model training, management, and / or inference.
[0067] In an example, the UE may perform prediction based on the collected data to obtain the predicted data and report the predicted data to the network for the AI / ML model training, management, and / or inference. The UE may perform direct measurement prediction based on the data collection configuration and an actual measurement result to obtain the predicted data. Alternatively, the UE may perform indirect measurement prediction based on the data collection configuration and a predicted measurement result to obtain the predicted data.
[0068] For example, the UE may report periodical and / or event-based predicted data when reporting condition is met. A predicted data report may be reported to the network based on at least one of a triggering threshold and an allowed predicting threshold. The allowed predicting threshold may include a reference signal received power (RSRP) or received signal strength indicator (RSSI) threshold or a RSRP or RSSI change threshold. The RSRP or RSSI change threshold may be a criterion of a difference between two measurements. The two measurements are two predicted measurements, or an actual measurement and a predicted measurement. In an example, the predicted data report may be sent to the network when the triggering threshold is met and the difference between a first actual measurement and a second predicted measurement is met on the RSRP or RSSI change threshold. More specifically, the UE may report a first actual measurement result to the network within a first time duration, predicts the measurement to obtain a predicted measurement result within a second time duration that is after the first time duration, and reports a second actual measurement result at an end of the second time duration if no other predicted measurement result meets at least one reporting criterion. The predicted measurement result may be reported to the network during the second time duration when the at least one reporting criterion is met.
[0069] In an example, the UE may train an AI / ML model based on a condition identifier and report a model identifier associated with the AI / ML model to the network, wherein an association exists between the condition identifier and model identifier. In an example, the UE may collect data based on a condition identifier associated with an additional condition and report a model identifier associated with an AI / ML model to the network, wherein an association exists between the condition identifier and model identifier.
[0070] In an example, the data collection may include L1 and / or L3 measurements from the UE or feedback from AI / ML management. In an example, the UE may report to the network an immediate data or a logged data regarding signal strength measurement. In an example, the UE may report to the network an immediate minimization of drive tests (MDT) or a logged MDT according to a state of the UE.
[0071] In an example, the data collection configuration may include a measurement configuration, which configures measurements or resources for any one of mobility, multicast-broadcast service (MBS) , load balancing, self-organizing network (SON) , and network energy saving (NES) . For the mobility, in response to a predicted handover command from a serving network, the UE may perform measurement on a predicted target network. A predicted measurement result on the predicted target network may be reported when an allowed predicting threshold or a reporting triggering threshold is met. An actual measurement result may be reported at an end of a time duration if no other predicted measurement result meets a reporting triggering threshold. If connecting to the predicted target network is failed, the UE may feedback a prediction error to the serving network for fine-tuning the data collection configuration. For the mobility, based on the predicted data, the UE may transmit a predicted handover request to initiate L1 or L2 based handover.
[0072] With the proposed method 100 illustrated above, data collection procedure is enhanced for AI / ML model training, management and / or inference.
[0073] Referring to FIG. 10, the embodiment of the present disclosure further provides a wireless communication method 200 for applying AI / ML functionalities, executable by a base station (BS) , the method 200 including receiving from a user equipment (UE) in a network artificial intelligence (AI) / machine learning (ML) capability information, wherein the AI / ML capability information indicates AI / ML functionality at UE side (Step 210) ; transmitting to the UE a data collection configuration configured based on the AI / ML functionality in accordance with a life cycle management (LCM) framework (Step 220) ; and receiving from the UE collected data and predicted data, which are obtained by performing data collection based on the data collection configuration and the AI / ML functionality by the UE, for AI / ML model training, management, and / or inference (Step 230) . Details can be referred to the method 100 of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0074] In another aspect, intelligent measurement and / or measurement event prediction are proposed in this disclosure. Using AI / ML predictions including UE-side, network-side, two-side model is expected to facilitate UE mobility and network resource management for increasing service continuation and system throughput.
[0075] FIG. 11 illustrates a wireless communication method 300 for applying AI / ML functionalities by a UE according to an embodiment of the present disclosure. The method 300 is executable by a user equipment (UE) in a network and includes the following.
[0076] In Step 310, the UE receives from the network a measurement configuration including configurations of actual measurement and predicted measurement to start data collection.
[0077] In Step 320, the UE performs predicted measurement by using an artificial intelligence (AI) / machine learning (ML) based on the measurement configuration.
[0078] In Step 330, the UE sends a predicted measurement report to the network for AI / ML model training, management, and inference. The predicted measurement report may be sent to the network when the at least one reporting criterion is met. The predicted measurement report may include a predicted measurement result and / or a predicted measurement event. The predicted measurement report may be sent to the network via 3GPP signaling including any one of non-access stratum (NAS) , radio resource control (RRC) , medium access control control element (MAC-CE) , uplink control information (UCI) , downlink control information (DCI) , layer 1 (L1) signaling, location positioning protocol (LPP) , and UE assistance information (UAI) .
[0079] As to the predicted measurement result, the predicted measurement result sent in the predicted measurement report may be within an allowed predicting threshold. A predicted measurement result is reported to the network only when the predicted measurement result meets the allowed predicting threshold. For example, the predicted measurement result may be cell-level or beam-level and includes predicted reference signal received power (RSRP) , received signal strength indicator (RSSI) , RSRP change and / or RSSI change. The allowed predicting threshold may include an adaptive RSRP or RSSI threshold or an adaptive RSRP or RSSI change threshold. The adaptive RSRP or RSSI change threshold is a criterion of a difference between two measurements, and the two measurements are two predicted measurements, or an actual measurement and a predicted measurement. The predicted RSRP or RSSI change that meets an adaptive RSRP or RSSI change threshold is reported to the network in the predicted measurement report. For example, the UE may report an actual measurement result to the network within a first time duration, predict the measurement to obtain a predicted measurement result within a second time duration that is after the first time duration, and report at least one of the actual measurement result and the predicted measurement result at an end of the second time duration when at least one reporting criterion is met.
[0080] For example, the measurement configuration may include a measurement gap that is a time duration the UE uses to perform the predicted measurement, and a predicted measurement result is checked with an allowed predicting threshold within the measurement gap. A data report may be sent to the network at an end of the measurement gap if no other predicted data meets at least one reporting criterion during the measurement gap. The measurement gap may be enlarged to reduce a number of actual measurement time instances for actual measurements. A measurement reduction rate may be associated with skipped actual measurement time instances within total measurement time instances.
[0081] In addition, in a case that an allowed predicting threshold is not set or configured, an actual or predicted measurement result may be reported to the network when a triggering threshold for the reporting is met. At least one criterion that triggers the UE to perform measurement or send a measurement report may be provided by the network, and whether the UE performs the actual or predicted measurement or report an actual or predicted measurement result may be determined by the UE based on the at least one criterion. At least one of a triggering threshold and an allowed predicting threshold may be configured for the UE for intelligent measurement and / or reporting. For example, a measurement triggering threshold may be configured for the UE to determine whether to perform the actual or predicted measurement, while a report triggering threshold may be configured for the UE to determine whether to report an actual or predicted measurement result. In another case, a triggering threshold may be configured for the UE to determine whether to perform the actual or predicted measurement and report an actual or predicted measurement result. In some cases, the predicted measurement may not be transmitted by the UE even if at least one reporting criterion and / or an adaptive RSRP change threshold is met.
[0082] As to the predicted measurement event, in an example, the predicted measurement event may be obtained by direct measurement event prediction, where an actual measurement result is a part of model inputs. In the direct measurement event prediction, UE mobility information is also a part of the model inputs. In another example, the predicted measurement event may be obtained by indirect measurement event prediction, where a predicted measurement result is used for the indirect measurement event prediction. In the indirect measurement event prediction, estimated information is used as an input of a model. In addition, an output of the AI / ML model may be a probability of measurement event occurrence within a time duration. In a scenario, the predicted measurement event is used to provide predicted configuration and / or reporting. In another scenario, the predicted measurement event is used for UE mobility enhancement and / or network resource management. According to an output of the AI / ML model and / or the predicted measurement report sent to the network, whether to carry out at least one of the following applications may be determined: enhanced mobility, multicast and broadcast service, load balancing, self-organizing network, and network energy saving.
[0083] With the proposed method 300 illustrated above, intelligent measurement and / or measurement event prediction is realized for AI / ML model training, management and / or inference.
[0084] Referring to FIG. 12, the embodiment of the present disclosure further provides a wireless communication method 400 for applying AI / ML functionalities, executable by a base station (BS) , the method 400 including transmitting to a user equipment (UE) in a network a measurement configuration including configurations of actual measurement and predicted measurement for the UE to start data collection (Step 410) ; and receiving from the UE a predicted measurement report for AI / ML model training, management, and inference, wherein the predicted measurement report is obtained by performing predicted measurement by using an artificial intelligence (AI) / machine learning (ML) based on the measurement configuration by the UE (Step 420) . Details can be referred to the method 300 of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0085] In one embodiment of the present disclosure is as shown in FIG. 13, which depicts implementation scenarios of AI / ML-capable network triggered handover between the UE 10 and the base station 20 according to the present disclosure.
[0086] In FIG. 13, an initial enhanced data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection.
[0087] The UE AI / ML capability and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent measurement configuration to enable online / offline data collection and measurement reporting without increasing signaling overhead and power consumption. The UE may determine whether to send an actual or a predicted measurement / data report based on the reporting triggering threshold and / or allowed predicting threshold as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., CSI measurement configuration, MDT configuration) depending on the data required by AI / ML model training / management / inference. L1 / L2-triggered mobility (LTM) and L3-based mobility are both considered. The UE may report the collected / predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) .
[0088] Based on data collection and AI / ML prediction for AI / ML model training / management / inference, the serving NG-RAN may establish UE path switching and PDU session with the predicted target NG-RAN and NGC / AIML server in advance and may transmit a predicted handover command (e.g., RRCReconfiguration) to trigger a seamless handover procedure. Upon the reception of predicted handover command, the UE should measure with the predicted target NG-RAN to check whether the predicted target NG-RAN is qualified or not. If the predicted target NG-RAN can be a target NG-RAN to provide service for the UE (e.g., RACH complete) , a UE connection with the predicted target NG-RAN would be established for the seamless data transmission and reception. On the other hand, if the predicted target NG-RAN cannot provide service for the UE (e.g., RACH failure) , the UE would feedback the prediction error to the serving NG-RAN for fine-tuning AI / ML prediction.
[0089] In one embodiment of the present disclosure is as shown in FIG. 14, which depicts implementation scenarios of AI / ML-capable UE triggered handover between the UE 10 and the base station 20 according to the present disclosure.
[0090] In FIG. 14, an initial enhanced data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection.
[0091] The UE AI / ML capability and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent measurement configuration to enable online / offline data collection and measurement reporting without increasing signaling overhead and power consumption. The UE may determine whether to send an actual or a predicted measurement / data report based on the reporting triggering threshold and / or allowed predicting threshold as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., CSI measurement configuration, MDT configuration) depending on the data required by AI / ML model training / management / inference. L1 / L2-triggered mobility (LTM) and L3-based mobility are both considered. The UE may report the collected / predicted data or predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) .
[0092] Based on data collection and AI / ML prediction for AI / ML model training / management / inference, the UE may trigger a predicted L1 / L2 handover. If the predicted handover request is accepted by the serving NG-RAN, the serving NG-RAN may establish UE path switching and PDU session with the predicted target NG-RAN and NGC / AIML server. And then the serving NG-RAN or predicted target NG-RAN may transmit a handover command to the UE in response to the predicted handover request. Upon the reception of handover command, the UE should measure with the predicted target NG-RAN to check whether the predicted target NG-RAN is qualified or not. If the predicted target NG-RAN can be a target NG-RAN to provide service for the UE (e.g., RACH complete) , a UE connection with the predicted target NG-RAN would be established for the seamless data transmission and reception. On the other hand, if the predicted target NG-RAN cannot provide service for the UE (e.g., RACH failure) , the UE would feedback the prediction error to the serving NG-RAN for fine-tuning AI / ML prediction.
[0093] In one embodiment of the present disclosure is as shown in FIG. 15, which depicts implementation scenarios of AI / ML-capable MBS transmission between the UE 10 and the base station 20 according to the present disclosure.
[0094] In FIG. 15, an initial enhanced data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection.
[0095] The UE capability (e.g., MBS, AI / ML) and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent interests report configuration to enable online / offline data collection and reporting without increasing signaling overhead and power consumption. The UE may determine whether to send an actual or a predicted data (e.g., predicted MBS interest) based on the reporting triggering threshold and / or allowed predicting threshold as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., MBS indication, frequency preference) depending on the data required by AI / ML model training / management / inference. The UE may report the collected / predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) .
[0096] Based on data collection and AI / ML prediction for AI / ML model training / management / inference, when the serving NG-RAN predicts there is at least one UE interested in receiving MBS, a predicted MBS subscription for the UE may be established with the NGC / AIML server. The serving NG-RAN may determine whether to transmit MCCH information and MBS for the UE via PTP or PTM based on the data collection from UE, MBS transmission threshold, and AI / ML prediction.
[0097] In one embodiment of the present disclosure is as shown in FIG. 16, which depicts implementation scenarios of AI / ML-capable RAN-based notification area update between the UE 10 and the base station 20 according to the present disclosure. In FIG. 16, an initial data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection.
[0098] The UE capability (e.g., MBS, DRX, AI / ML) and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent RAN-based notification area update configuration to enable online / offline data collection and reporting without increasing signaling overhead and power consumption. The UE in power-saving state (e.g., RRC_IDLE, RRC_INACTIVE) may determine whether to send an actual or a predicted data (e.g., predicted target NG-RAN identifier) based on the intelligent measurement and / or event prediction as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., DRX / DTX configuration) depending on the data required by AI / ML model training / management / inference. The UE in power-saving state may transmit the predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) . Based on data collection and AI / ML prediction (e.g., intelligent measurement prediction, measurement event prediction) , when the UE in power-saving state predicts a RAN-based notification area update occurrence, a predicted RAN-based notification area update for the UE may be performed with the serving NG-RAN, the predicted target NG-RAN, and / or NGC / AIML server. The predicted target NG-RAN may determine whether to paging the UE based on the AI / ML prediction. In some cases, when the UE in power-saving state supports MBS, it may determine whether to send an actual or a predicted data (e.g., predicted target MBS NG-RAN identifier, MBS identifier, frequency preference) based on the intelligent measurement and / or measurement event as those described in the aforesaid and hence are not repeated. Upon data collection and AI / ML prediction, when the UE in power-saving state predicts a RAN-based MBS notification area update occurrence, a predicted RAN-based MBS notification area update for the UE may be operated with the serving MBS NG-RAN, the predicted target MBS NG-RAN, and / or NGC / AIML server. The predicted target MBS NG-RAN may determine whether to transmit MCCH information and MBS for the UE via PTP or PTM based on the MBS transmission threshold, and AI / ML prediction as shown in the dashed line of FIG. 16. The MBS data is transmitted optionally (i.e., the dashed line) when the UE and the network are MBS-capable. Upon AI / ML prediction, the UE in power-saving state can receive MBS data seamlessly from the predicted target MBS NG-RAN without state transition.
[0099] In one embodiment of the present disclosure is as shown in FIG. 17, which depicts implementation scenarios of AI / ML-capable network triggered paging between the UE 10 and the base station 20 according to the present disclosure. In FIG. 17, an initial data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection. The UE AI / ML capability (e.g., DRX, AI / ML) and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent measurement configuration to enable online / offline data collection and measurement reporting without increasing signaling overhead and power consumption. The UE in power-saving state (e.g., RRC_IDLE, RRC_INACTIVE) may determine whether to send an actual or a predicted RAN-based notification area update based on the reporting triggering threshold and / or allowed predicting threshold as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., DRX / DTX configuration) depending on the data required by AI / ML model training / management / inference. The UE in power-saving state may report the collected / predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) . Hence the network can infer and predict UE location / UE trajectory within a pre-determined RAN / CN area to significantly reduce paging overhead and to successfully deliver the incoming data within a delay limitation. According to the intelligent measurement and / or event prediction, the predicted RAN-initiated paging and / or CN-initiated paging can facilitate the network to precisely reach the UE in power-saving state of system information change, emergency service indication, incoming data. The paging DRX cycles are inferred and predicted based on UE-sided, network-sided, two-sided model in addition to network configuration. A new paging cause may be included in the paging message so that the UE can identify the paging predication. When the UE enters RRC_INACTIVE with / without ongoing SDT procedure, it is required to monitor the RAN-initiated paging on the predicted paging occasion within the paging DRX cycle. In some cases, the RAN paging area belongs to multiple NG-RANs (i.e., gNBs) . Each RAN paging area is identified by RAN paging area identity. If the predicted target NG-RAN cannot provide service for the UE (e.g., due to NES) , the serving NG-RAN would try to page the UE and wait for the response from the UE during a valid timer. When the timer is expired, a predicted RAN-initiated paging is transmitted via multicast / broadcast manner through multiple gNBs in the same RAN area. In some cases, based on the paging history and further AI / ML prediction, a predicted CN-initiated paging is transmitted to the UE when the timer is expired.
[0100] In one embodiment of the present disclosure is as shown in FIG. 18, which depicts implementation scenarios of AI / ML-capable conditional handover between the UE 10 and the base station 20 according to the present disclosure. In FIG. 18, an initial data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection. The UE AI / ML capability (e.g., high / low mobility) and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent measurement configuration to enable online / offline data collection and measurement reporting without increasing signaling overhead and power consumption. The UE may determine whether to send an actual or a predicted measurement / data report based on the intelligent measurement and / or event prediction as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., CSI measurement configuration) depending on the data required by AI / ML model training / management / inference. L1 / L2-triggered mobility (LTM) and L3-based mobility are both considered. The UE may report the collected / predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) . Based on data collection and AI / ML prediction, the serving NG-RAN may establish UE path switching and PDU session with the predicted target NG-RAN and NGC / AIML server in advance and may transmit a predicted conditional handover (CHO) configuration (e.g., RRCReconfiguration) including the predicted CHO candidate cell (s) and execution condition (s) to the UE. Upon the reception of predicted CHO configuration, the UE should measure with the predicted candidate cell (s) to verify which predicted cell is more qualified. If there is a predicted candidate cell can be a qualified target NG-RAN to provide service for the UE, a prediction response is transmitted to the serving NG-RAN so that CHO is triggered and the predicted CHO is completed while performing NES. On the other hand, if the UE cannot measure any predicted candidate cell (e.g., measurement failure) , the UE would feedback the prediction error to the serving NG-RAN for fine-tuning AI / ML prediction. In some cases, the predicted execution condition may consist of more than one triggering condition. The measurement event prediction can be one of the predicted execution conditions. The UE can determine whether to perform conditional handover based on the measurement event prediction. For example, when the event Ai for conditional handover is predicted with a high probability of occurrence, the event Ai is used to trigger the predicted conditional handover procedure. The prediction response is transmitted to the serving NG-RAN and the predicted CHO is triggered and completed upon the establishment of UE connection with the predicted target NG-RAN.
[0101] In one embodiment of the present disclosure is as shown in FIG. 19 which depicts implementation scenarios of AI / ML-capable network energy saving between the UE 10 and the base station 20 according to the present disclosure. In FIG. 19, an initial data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection. The UE capability (e.g., DRX, AI / ML) and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent measurement configuration to enable online / offline data collection and measurement reporting without increasing signaling overhead and power consumption. The UE may determine whether to send an actual or a predicted measurement report and / or incoming UL traffic statistics based on the intelligent measurement and / or event prediction as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., CSI measurement configuration, BSR configuration, SR configuration) depending on the data required by AI / ML model training / management / inference. The UE may report the collected / predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) . Based on data collection and AI / ML prediction, the serving NG-RAN can predict to perform network energy saving (NES) (e.g., intelligent cell on / off switching, SSB-less SCell) for decreasing energy consumption. In some cases, a handover or a conditional handover for NES is predicted. The network can determine whether to perform NES handover whereas the UE can initiate to perform NES conditional handover based on RRM measurement prediction and measurement event prediction. For example, when the event Ai for NES handover or NES conditional handover is predicted with a high probability of occurrence, the event Ai is used to trigger the predicted NES handover or NES conditional handover procedure. Before that, UE path switching and PDU session should be established with the predicted target NG-RAN and NGC / AIML server in advance. The serving NG-RAN may transmit a predicted network energy saving (NES) configuration (e.g., RRCReconfiguration) to contain the predicted NES strategy and target NG-RAN for initiating NES HO, if necessary. In other cases, the predicted network energy saving (NES) configuration may contain the predicted candidate cell (s) and execution condition (s) for performing NES CHO. Upon the reception of predicted NES configuration, the UE should measure with the predicted candidate cell (s) to verify which predicted cell is qualified. If there is a predicted candidate cell can be a qualified target NG-RAN to provide service for the UE, a prediction response is transmitted to the serving NG-RAN so that NES CHO is triggered and the predicted NES CHO is completed upon the establishment of UE connection with the predicted target NG-RAN. On the other hand, if the UE cannot measure any qualified candidate cell (e.g., measurement failure) , the UE would feedback the prediction error to the serving NG-RAN for fine-tuning AI / ML prediction.
[0102] In one embodiment of the present disclosure is as shown in FIG. 20, which depicts implementation scenarios of AI / ML-capable network load balancing between the UE 10 and the base station 20 according to the present disclosure. In FIG. 20, an initial data collection configuration may be configured to the UE based on the pre-configured AI / ML parameters or history-data collection. The UE AI / ML capability and UE-sided / network-sided additional condition (s) exchange would be performed for UE-sided / network-sided model management (e.g., functionality / model selection, (de) activation, switching, fallback, etc. ) . The enhanced data collection configuration would be configured based on the AI / ML capability and UE-sided / network-sided additional condition (s) . For example, the enhanced data collection configuration may include an intelligent measurement configuration to enable online / offline data collection and measurement reporting without increasing signaling overhead and power consumption. The UE may determine whether to send an actual or a predicted measurement / data report and / or incoming UL traffic statistics based on the intelligent measurement and / or event prediction as those described in the aforesaid and hence are not repeated. In some cases, the enhanced data collection configuration may include the other configuration (e.g., CSI measurement configuration, BSR configuration, SR configuration) depending on the data required by AI / ML model training / management / inference. L1 / L2-triggered mobility (LTM) and L3-based mobility are both considered. The UE may report the collected / predicted data via 3GPP signaling (e.g., NAS, RRC, MAC-CE, UCI, DCI, L1, LPP, UAI) . Based on data collection and AI / ML prediction, the serving NG-RAN transmits the predicted load balancing request and negotiates with the predicted target NG-RAN (s) and NGC / AIML server. In some cases, UE path switching and PDU session with the predicted target NG-RAN and NGC / AIML server may be established after the negotiation. Then the serving NG-RAN determines at least one UE (e.g., the UE with lower or higher traffic attempts, the UE with low or high mobility) to perform cell switching and transmits the predicted load balancing configuration (e.g., RRCReconfiguration with mobility control information) to the UE for load balancing purposes. Upon the reception of predicted load balancing configuration, the UE should switch to the predicted target NG-RAN for load balancing. However, if the UE cannot connect with the predicted target NG-RAN (e.g., cell switching failure) , the UE would feedback the prediction error to the serving NG-RAN for fine-tuning AI / ML prediction. In some cases, the predicted cell switching procedure can be applied to self-organizing network (SON) . The serving NG-RAN can initiate the predicted SON with the neighboring NG-RAN (s) and NGC / AIML server based on data collection and AI / ML prediction. In AI / ML model training / management / inference, the output of the prediction model can be used to enable self-configuration and self-optimization. For example, the predicted inter-cell interference coordination (ICIC) can be used to minimize the interference between cells using the same spectrum. The serving NG-RAN can adjust the transmitted power / beam for self-optimization based on AI / ML prediction.
[0103] In one embodiment for the mobility scenario, the serving NG-RAN and / or UE can predict whether to perform MBS handover / MBS tracking area update based on data collection from UEs, seamless MBS thresholds, and AI / ML prediction. Upon the reception of predicted MCCH change notification and / or reconfiguration, the network-initiated MBS handover / MBS tracking area update would be performed so that the seamless MBS transmission within multiple MBS service area is achieved. In some cases, the UE may trigger the MBS handover / MBS tracking area update based on the AI / ML prediction. Upon the reception of UE prediction, the serving NG-RAN or target NG-RAN may provide the MCCH change notification and MBS transmission for the UE.
[0104] In one embodiment for the other applications (e.g., positioning, beam management, load balancing, network energy saving, self-organization network) , the UE may determine whether to send an actual or a predicted data (e.g., UE location, UE trajectory, beam selection, buffer status report, scheduling request) based on the reporting triggering threshold and / or allowed predicting threshold as those described in the aforesaid and hence are not repeated. Based on data collection and AI / ML prediction for AI / ML model training / management / inference, the predicted backhaul establishment for the UE may be established with the NGC / AIML server. The serving NG-RAN may determine whether to serve the UE based on the data collection and AI / ML prediction. The associated actions between the network and UE for the application and scenario.
[0105] In one embodiment for the NG-RAN which is a RAN functional split node according to the embodiment of the present disclosure. The associated gNB-CU / gNB-DU F1AP identifier and parameters during the enhanced data collection configuration and report would be used. The data collection / prediction from UE is performed as the aforesaid embodiments, wherein the RRC message proposed in this embodiment is encapsulated in the F1AP message (s) (e.g., UE Context Modification, DL / UL RRC Message Transfer, etc. ) and will not be described again.
[0106] Commercial interests for some embodiments are as follows. 1. solving issues in the prior art. 2. enhancing data collection procedure. 3. realizing measurement result / event prediction. 4. providing a good communication performance. Some embodiments of the present disclosure are used by 5G-NR chipset vendors, V2X communication system development vendors, automakers including cars, trains, trucks, buses, bicycles, moto-bikes, helmets, and etc., drones (unmanned aerial vehicles) , smartphone makers, communication devices for public safety use, AR / VR device maker for example gaming, conference / seminar, education purposes. Some embodiments of the present disclosure are a combination of “techniques / processes” that can be adopted in 3GPP specification to create an end product. Some embodiments of the present disclosure could be adopted in the 5G NR unlicensed band communications. Some embodiments of the present disclosure propose technical mechanisms.
[0107] The embodiment of the present disclosure further provides a user equipment including a memory and a processor coupled to the memory, the processor configured to call and run program instructions stored in a memory, to execute each of the methods of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0108] The embodiment of the present disclosure further provides a base station including a memory and a processor coupled to the memory, the processor configured to call and run program instructions stored in a memory, to execute each of the methods of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0109] The embodiment of the present disclosure further provides a non-transitory computer readable storage medium for storing a computer program. The non-transitory computer readable storage medium enables a computer to execute corresponding processes implemented by the UE / BS in each of the methods of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0110] The embodiment of the present disclosure further provides a computer program product including computer program instructions. The computer program product enables a computer to execute corresponding processes implemented by the UE / BS in each of the methods of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0111] The embodiment of the present disclosure further provides a computer program. The computer program enables a computer to execute corresponding processes implemented by the UE / BS in each of the methods of the embodiment of the present disclosure. For brevity, details will not be described herein again.
[0112] Although not shown in detail any of the devices or apparatus that form part of the network may include at least a processor, a storage unit and a communications interface, wherein the processor unit, storage unit, and communications interface are configured to perform the method of any aspect of the present invention. Further options and choices are described below.
[0113] The signal processing functionality of the embodiments of the invention especially the gNB and the UE may be achieved using computing systems or architectures known to those who are skilled in the relevant art. Computing systems such as, a desktop, laptop or notebook computer, hand-held computing device (PDA, cell phone, palmtop, etc. ) , mainframe, server, client, or any other type of special or general purpose computing device as may be desirable or appropriate for a given application or environment can be used. The computing system can include one or more processors which can be implemented using a general or special-purpose processing engine such as, for example, a microprocessor, microcontroller or other control module.
[0114] The computing system can also include a main memory, such as random access memory (RAM) or other dynamic memory, for storing information and instructions to be executed by a processor. Such a main memory also may be used for storing temporary variables or other intermediate information during execution of instructions to be executed by the processor. The computing system may likewise include a read only memory (ROM) or other static storage device for storing static information and instructions for a processor.
[0115] The computing system may also include an information storage system which may include, for example, a media drive and a removable storage interface. The media drive may include a drive or other mechanism to support fixed or removable storage media, such as a hard disk drive, a floppy disk drive, a magnetic tape drive, an optical disk drive, a compact disc (CD) or digital video drive (DVD) read or write drive (R or RW) , or other removable or fixed media drive. Storage media may include, for example, a hard disk, floppy disk, magnetic tape, optical disk, CD or DVD, or other fixed or removable medium that is read by and written to by media drive. The storage media may include a computer-readable storage medium having particular computer software or data stored therein.
[0116] In alternative embodiments, an information storage system may include other similar components for allowing computer programs or other instructions or data to be loaded into the computing system. Such components may include, for example, a removable storage unit and an interface, such as a program cartridge and cartridge interface, a removable memory (for example, a flash memory or other removable memory module) and memory slot, and other removable storage units and interfaces that allow software and data to be transferred from the removable storage unit to computing system.
[0117] The computing system can also include a communications interface. Such a communications interface can be used to allow software and data to be transferred between a computing system and external devices. Examples of communications interfaces can include a modem, a network interface (such as an Ethernet or other NIC card) , a communications port (such as for example, a universal serial bus (USB) port) , a PCMCIA slot and card, etc. Software and data transferred via a communications interface are in the form of signals which can be electronic, electromagnetic, and optical or other signals capable of being received by a communications interface medium.
[0118] In this document, the terms ‘computer program product’ , ‘computer-readable medium’a nd the like may be used generally to refer to tangible media such as, for example, a memory, storage device, or storage unit. These and other forms of computer-readable media may store one or more instructions for use by the processor including the computer system to cause the processor to perform specified operations. Such instructions, generally referred to as ‘computer program code’ (which may be grouped in the form of computer programs or other groupings) , when executed, enable the computing system to perform functions of embodiments of the present invention. Note that the code may directly cause a processor to perform specified operations, be compiled to do so, and / or be combined with other software, hardware, and / or firmware elements (e.g., libraries for performing standard functions) to do so.
[0119] The non-transitory computer readable medium may include at least one from a group consisting of: a hard disk, a CD-ROM, an optical storage device, a magnetic storage device, a Read Only Memory, a Programmable Read Only Memory, an Erasable Programmable Read Only Memory, EPROM, an Electrically Erasable Programmable Read Only Memory and a Flash memory. In an embodiment where the elements are implemented using software, the software may be stored in a computer-readable medium and loaded into computing system using, for example, removable storage drive. A control module (in this example, software instructions or executable computer program code) , when executed by the processor in the computer system, causes a processor to perform the functions of the invention as described herein.
[0120] Furthermore, the inventive concept can be applied to any circuit for performing signal processing functionality within a network element. It is further envisaged that, for example, a semiconductor manufacturer may employ the inventive concept in a design of a stand-alone device, such as a microcontroller of a digital signal processor (DSP) , or application-specific integrated circuit (ASIC) and / or any other sub-system element.
[0121] It will be appreciated that, for clarity purposes, the above description has described embodiments of the invention with reference to a single processing logic. However, the inventive concept may equally be implemented by way of a plurality of different functional units and processors to provide the signal processing functionality. Thus, references to specific functional units are only to be seen as references to suitable means for providing the described functionality, rather than indicative of a strict logical or physical structure or organization.
[0122] Aspects of the invention may be implemented in any suitable form including hardware, software, firmware or any combination of these. The invention may optionally be implemented, at least partly, as computer software running on one or more data processors and / or digital signal processors or configurable module components such as FPGA devices.
[0123] Thus, the elements and components of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way. Indeed, the functionality may be implemented in a single unit, in a plurality of units or as part of other functional units. Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in accordance with the invention. In the claims, the term ‘comprising’ does not exclude the presence of other elements or steps.
[0124] Furthermore, although individually listed, a plurality of means, elements or method steps may be implemented by, for example, a single unit or processor. Additionally, although individual features may be included in different claims, these may possibly be advantageously combined, and the inclusion in different claims does not imply that a combination of features is not feasible and / or advantageous. Also, the inclusion of a feature in one category of claims does not imply a limitation to this category, but rather indicates that the feature is equally applicable to other claim categories, as appropriate.
[0125] Furthermore, the order of features in the claims does not imply any specific order in which the features must be performed and in particular the order of individual steps in a method claim does not imply that the steps must be performed in this order. Rather, the steps may be performed in any suitable order. In addition, singular references do not exclude a plurality. Thus, references to ‘a’ , ‘an’ , ‘first’ , ‘second’ , etc. do not preclude a plurality.
[0126] While the present disclosure has been described in connection with what is considered the most practical and preferred embodiments, it is understood that the present disclosure is not limited to the disclosed embodiments but is intended to cover various arrangements made without departing from the scope of the broadest interpretation of the appended claims.
Claims
1.A wireless communication method for applying AI / ML functionality, executable by a user equipment (UE) , the method comprising:receiving from the network a measurement configuration including configurations of actual measurement and predicted measurement to start data collection;performing predicted measurement by using an artificial intelligence (AI) / machine learning (ML) based on the measurement configuration; andsending a predicted measurement report to the network for AI / ML model training, management, and inference.2.The method of claim 1, wherein the predicted measurement report comprises a predicted measurement result.3.The method of claim 2, wherein the predicted measurement result sent in the predicted measurement report is within an allowed predicting threshold.4.The method of claim 2, wherein the predicted measurement result is cell-level or beam-level and comprises predicted reference signal received power (RSRP) , received signal strength indicator (RSSI) , RSRP change and / or RSSI change.5.The method of claim 3, wherein the allowed predicting threshold comprises an adaptive RSRP or RSSI threshold or an adaptive RSRP or RSSI change threshold.6.The method of claim 5, wherein the adaptive RSRP or RSSI change threshold is at least one criterion of a difference between two measurements, and the two measurements are two predicted measurements, or an actual measurement and a predicted measurement.7.The method of claim 1, wherein the predicted measurement report is sent to the network via 3GPP signaling including any one of non-access stratum (NAS) , radio resource control (RRC) , medium access control control element (MAC-CE) , uplink control information (UCI) , downlink control information (DCI) , layer 1 (L1) signaling, location positioning protocol (LPP) , and UE assistance information (UAI) .8.The method of claim 1, wherein the UE reports an actual measurement result to the network within a first time duration, predicts the measurement to obtain a predicted measurement result within a second time duration that is after the first time duration, and reports at least one of the actual measurement result and the predicted measurement result at an end of the second time duration when at least one reporting criterion is met.9.The method of claim 1, wherein the measurement configuration comprises a measurement gap that is a time duration the UE uses to perform the predicted measurement, and a predicted measurement result is checked with an allowed predicting threshold within the measurement gap.10.The method of claim 9, wherein the predicted measurement result is reported when the allowed predicting threshold is met.11.The method of claim 9, wherein a data report is sent to the network at an end of the measurement gap if no other predicted data meets at least one reporting criterion during the measurement gap.12.The method of claim 9, wherein the measurement gap is enlarged to reduce a number of actual measurement time instances for actual measurements.13.The method of claim 12, wherein a measurement reduction rate is associated with skipped actual measurement time instances within total measurement time instances.14.The method of claim 1, wherein the predicted measurement report comprises a predicted measurement event.15.The method of claim 14, wherein the predicted measurement event is obtained by direct measurement event prediction, wherein an actual measurement result is a part of model inputs.16.The method of claim 15, wherein in the direct measurement event prediction, UE mobility information is also a part of the model inputs.17.The method of claim 14, wherein the predicted measurement event is obtained by indirect measurement event prediction, where a predicted measurement result is used for the indirect measurement event prediction.18.The method of claim 17, wherein in the indirect measurement event prediction, estimated information is used as an input of a model.19.The method of claim 14, wherein an output of the AI / ML model is a probability of measurement event occurrence within a time duration.20.The method of claim 14, wherein the predicted measurement event is used to provide predicted configuration and / or reporting.21.The method of claim 14, wherein the predicted measurement event is used for UE mobility enhancement and / or network resource management.22.The method of claim 1, wherein the predicted measurement report is sent to the network when the at least one reporting criterion is met.23.The method of claim 1, wherein a predicted measurement result is reported to the network only when the predicted measurement result meets an allowed predicting threshold.24.The method of claim 1, wherein a predicted RSRP or RSSI change that meets an adaptive RSRP or RSSI change threshold is reported to the network in the predicted measurement report.25.The method of claim 1, wherein in a case that an allowed predicting threshold is not set or configured, an actual or predicted measurement result is reported to the network when a triggering threshold for the reporting is met.26.The method of claim 1, wherein at least one criterion that triggers the UE to perform measurement or send a measurement report is provided by the network, and whether the UE performs the actual or predicted measurement or report an actual or predicted measurement result is determined by the UE based on the at least one criterion.27.The method of claim 1, wherein at least one of a triggering threshold and an allowed predicting threshold is configured for the UE for intelligent measurement and / or reporting.28.The method of claim 1, wherein a measurement triggering threshold is configured for the UE to determine whether to perform the actual or predicted measurement.29.The method of claim 1, wherein a report triggering threshold is configured for the UE to determine whether to report an actual or predicted measurement result.30.The method of claim 1, wherein a triggering threshold is configured for the UE to determine whether to perform the actual or predicted measurement and report an actual or predicted measurement result.31.The method of claim 1, wherein the predicted measurement is not transmitted by the UE even if at least one reporting criterion and / or an adaptive RSRP change threshold is met.32.The method of claim 1, wherein according to an output of the AI / ML model and / or the predicted measurement report sent to the network, whether to carry out at least one of the following applications is determined: enhanced mobility, multicast and broadcast service (MBS) , load balancing, self-organizing network (SON) , and network energy saving (NES) .33.A wireless communication method for applying AI / ML functionalities, executable by a base station (BS) , the method comprising:transmitting to a user equipment (UE) in a network a measurement configuration including configurations of actual measurement and predicted measurement for the UE to start data collection; andreceiving from the UE a predicted measurement report for AI / ML model training, management, and inference, wherein the predicted measurement report is obtained by performing predicted measurement by using an artificial intelligence (AI) / machine learning (ML) based on the measurement configuration by the UE.34.The method of claim 33, wherein the predicted measurement report comprises a predicted measurement result.35.The method of claim 34, wherein the predicted measurement result received in the predicted measurement report is within an allowed predicting threshold.36.The method of claim 34, wherein the predicted measurement result is cell-level or beam-level and comprises predicted reference signal received power (RSRP) , received signal strength indicator (RSSI) , RSRP change and / or RSSI change.37.The method of claim 35, wherein the allowed predicting threshold comprises an adaptive RSRP or RSSI threshold or an adaptive RSRP or RSSI change threshold.38.The method of claim 37, wherein the adaptive RSRP or RSSI change threshold is a criterion of a difference between two measurements, and the two measurements are two predicted measurements, or an actual measurement and a predicted measurement.39.The method of claim 33, wherein the predicted measurement report is received from the UE via 3GPP signaling including any one of non-access stratum (NAS) , radio resource control (RRC) , medium access control control element (MAC-CE) , uplink control information (UCI) , downlink control information (DCI) , layer 1 (L1) signaling, location positioning protocol (LPP) , and UE assistance information (UAI) .40.The method of claim 33, wherein an actual measurement result is received from the UE within a first time duration, measurement prediction is performed by the UE to obtain a predicted measurement result within a second time duration that is after the first time duration, and at least one of the actual measurement result and the predicted measurement result is received from the UE at an end of the second time duration when at least one reporting criterion is met.41.The method of claim 33, wherein the measurement configuration comprises a measurement gap that is a time duration the UE uses to perform the predicted measurement, and a predicted measurement result is checked with an allowed predicting threshold within the measurement gap.42.The method of claim 41, wherein the predicted measurement result is received from the UE when the allowed predicting threshold is met.43.The method of claim 41, wherein a data report is received from the UE at an end of the measurement gap if no other predicted data meets at least one reporting criterion during the measurement gap.44.The method of claim 41, wherein the measurement gap is enlarged to reduce a number of actual measurement time instances for actual measurements.45.The method of claim 44, wherein a measurement reduction rate is associated with skipped actual measurement time instances within total measurement time instances.46.The method of claim 33, wherein the predicted measurement report comprises a predicted measurement event.47.The method of claim 46, wherein the predicted measurement event is obtained by direct measurement event prediction, wherein an actual measurement result is a part of model inputs.48.The method of claim 47, wherein in the direct measurement event prediction, UE mobility information is also a part of the model inputs.49.The method of claim 46, wherein the predicted measurement event is obtained by indirect measurement event prediction, where a predicted measurement result is used for the indirect measurement event prediction.50.The method of claim 49, wherein in the indirect measurement event prediction, estimated information is used as an input of a model.51.The method of claim 46, wherein an output of the AI / ML model is a probability of measurement event occurrence within a time duration.52.The method of claim 46, wherein the predicted measurement event is used to provide predicted configuration and / or reporting.53.The method of claim 46, wherein the predicted measurement event is used for UE mobility enhancement and / or network resource management.54.The method of claim 33, wherein the predicted measurement report is received from the UE when the at least one reporting criterion is met.55.The method of claim 33, wherein a predicted measurement result is received from the UE only when the predicted measurement result meets an allowed predicting threshold.56.The method of claim 33, wherein a predicted RSRP or RSSI change that meets an adaptive RSRP or RSSI change threshold is received from the UE in the predicted measurement report.57.The method of claim 33, wherein in a case that an allowed predicting threshold is not set or configured, an actual or predicted measurement result is received from the UE when a triggering threshold for the reporting is met.58.The method of claim 33, wherein at least one criterion that triggers the UE to perform measurement or send a measurement report is provided to the UE, and whether the UE performs the actual or predicted measurement or report an actual or predicted measurement result is determined by the UE based on the at least one criterion.59.The method of claim 33, wherein at least one of a triggering threshold and an allowed predicting threshold is configured for the UE for intelligent measurement and / or reporting.60.The method of claim 33, wherein a measurement triggering threshold is configured for the UE to determine whether to perform the actual or predicted measurement.61.The method of claim 33, wherein a report triggering threshold is configured for the UE to determine whether to report an actual or predicted measurement result.62.The method of claim 33, wherein a triggering threshold is configured for the UE to determine whether to perform the actual or predicted measurement and report an actual or predicted measurement result.63.The method of claim 33, wherein the predicted measurement is not received from the UE even if at least one reporting criterion and / or an adaptive RSRP change threshold is met.64.The method of claim 33, wherein according to an output of the AI / ML model and / or the predicted measurement report received from the UE, whether to carry out at least one of the following applications is determined: enhanced mobility, multicast and broadcast service (MBS) , load balancing, self-organizing network (SON) , and network energy saving (NES) .65.The method of claim 33, further comprising:checking performance metrics based on AI / ML model output and actual measurement results to determine functionality management including at least one of AI / ML mode selection, activation, deactivation, switching, and fallback.66.A user equipment (UE) , comprising a memory and a processor coupled to the memory, the processor configured to call and run program instructions stored in a memory to execute the method of any of claims 1 to 32.67.A base station (BS) , comprising a memory and a processor coupled to the memory, the processor configured to call and run program instructions stored in a memory to execute the method of any of claims 33 to 65.68.A non-transitory machine-readable storage medium having stored thereon instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 65.69.A chip, comprising:a processor, configured to call and run a computer program stored in a memory, to cause a device in which the chip is installed to execute the method of any one of claims 1 to 65.70.A non-transitory computer readable storage medium, in which a computer program is stored, wherein the computer program causes a computer to execute the method of any one of claims 1 to 65.71.A computer program product, comprising a computer program, wherein the computer program causes a computer to execute the method of any one of claims 1 to 65.72.A computer program, wherein the computer program causes a computer to execute the method of any one of claims 1 to 65.
Citation Information
Patent Citations
Inference error information feedback for machine learning-based inferences
WO2023206114A1
Method and apparatus for measurement prediction in a wireless communication system
WO2023234666A1
Measurement reporting of mixed types in wireless communication system
WO2024072037A1