Managing performance monitoring and reporting for measurement prediction
By implementing performance monitoring and reporting for measurement prediction, the accuracy of UE AI/ML models is ensured, improving handover performance and power efficiency in wireless communication systems.
Patent Information
- Application Number
- PCT/CN2025/093390
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-05-08
- Publication Date
- 2026-11-12
AI Technical Summary
Existing wireless communication systems face issues with inaccurate measurement predictions by user equipment (UE) using AI/ML models, leading to potential degradation in handover performance due to the base station's unawareness of poor prediction performance at the UE.
Implementing performance monitoring and reporting mechanisms for measurement prediction, where both UE and network entities assess the accuracy of AI/ML models, allowing for timely suspension of measurements and improved handover preparation based on predicted results.
Enhances mobility performance and power savings by ensuring accurate measurement predictions, enabling quicker and more reliable handover decisions at the network entity.
Smart Images

Figure CN2025093390_12112026_PF_FP_ABST
Abstract
Description
MANAGING PERFORMANCE MONITORING AND REPORTING FOR MEASUREMENT PREDICTIONFIELD OF THE DISCLOSURE
[0001] This disclosure relates to wireless communications and, more particularly, to managing performance monitoring and reporting for measurement prediction.BACKGROUND
[0002] This background description is provided for the purpose of generally presenting the context of the disclosure. Work of the presently named inventors, to the extent it is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.
[0003] The Third Generation Partnership Project (3GPP) specifies a radio interface referred to as fifth generation (5G) new radio (NR) (5G NR) . An architecture for a 5G NR wireless communication system includes a 5G core (5GC) network, a 5G radio access network (5G-RAN) , a 5G user equipment (5G UE) , etc. The 5G NR architecture seeks to provide increased data rates, decreased latency, and / or increased capacity compared to previous generation cellular communication systems.
[0004] 3GPP is studying measurement prediction based on Artificial Intelligence / Machine Learning (AI / ML) to improve the mobility performance. The study outcome is described in 3GPP technical report (TR) 38.744. In the TR 38.744, 3GPP has identified radio resource management (RRM) measurement event prediction is beneficial to the mobility performance and power saving for the UE. The RRM measurement event prediction can be either indirect measurement prediction or direction measurement prediction. In the indirect measurement event prediction, the UE performs measurements and obtains actual measurement results from the measurements, skips performing measurements and predicts measurement results based on the actual measurement results, and determines whether a configured event occurs (e.g., an entering condition for the configured event is met or an entering condition for the configured event has been satisfied for a duration) based on the predicted measurement results. If the UE determines that the configured event occurs, the UE transmits a prediction report to a base station, indicating that the configured event is predicted to occur. In the direct measurement event prediction, the UE directly predicts whether the configured event occurs, based on the actual measurement results. If the UE predicts that the configured event occurs, the UE transmits a prediction report to a base station, indicating that the configured event is predicted to occur.
[0005] When a UE communicates with a base station, it may report whether measurement prediction is applicable. The UE determines this applicability based on whether its associated AI prediction model was trained using measurement data from a specific cell. Upon receiving an affirmative report, the base station may configure the UE to perform measurement prediction, which the UE then executes. Despite this process, issues can occur. Even ifthe AI model was trained on measurement results that the UE obtained from the specific cell, its prediction performance at the UE may still be poor. Crucially, the base station might not be aware of this performance deficiency. This can lead to the base station receiving inaccurate predicted measurements, thereby potentially degrading handover performance.SUMMARY
[0006] A network entity, such as a base station or a unit of a base station, may communicate with a user equipment (UE) . The UE and / or the network entity may perform radio resource measurement (RRM) event prediction based on artificial intelligence (AI) / machine learning (ML) to increase mobility performance and / or save power at the UE. RRM measurement event prediction can be indirect (i.e., indirect measurement prediction) or direct (i.e., direct measurement prediction) . In indirect measurement prediction, the UE performs measurements and obtains actual measurement results from the measurements. The UE then skips performing the measurements and predicts measurement results (e.g., via an AI / ML model) based on the actual measurement results. The UE determines whether a configured event occurs based on the predicted measurements results. Upon positive determination, the UE transmits a prediction report to the network entity that indicates that the configured event is predicted to occur. In direct measurement prediction, the UE directly predicts whether the configured event occurs based on the actual measurement results. Ifthe UE predicts that the configured event occurs, the UE transmits a prediction report to the network entity that indicates that the configured event is predicted to occur.
[0007] The UE may report to the network entity whether measurement prediction is applicable to the UE based on whether an AI / ML prediction model was trained using measurement data from a specific cell (e.g., a cell that the UE is currently located in, a serving cell of the UE, a neighbor cell, etc. ) . When the network entity receives an indication that the measurement prediction is applicable, the network entity configures the UE to perform the measurement prediction and the UE performs the measurement prediction (e.g., a predicted reference signal received power (RSRP) ) based on the configuration. However, issues may exist with respect to measurement prediction. For example, even in scenarios in which the AI / ML prediction model was trained on measurement results that the UE obtained from the specific cell, prediction performance at the UE might still be poor. Moreover, the network entity might not be aware of the poor prediction performance at the UE.
[0008] Aspects of the present disclosure address the above-noted and other deficiencies by implementing various procedures for managing performance monitoring and reporting for measurement prediction. Some aspects discussed herein pertain to UE-based performance monitoring in which the UE obtains performance metric (s) and / or determines whether performance of measurement prediction (e.g., of an AI / ML model) is good or poor. Some other aspects discussed herein pertain to network-based performance monitoring scenarios in which the network entity obtains performance metric (s) and / or determines whether performance of the measurement prediction is good or poor based on predicted measurement results and actual measurement results received by the network entity from the UE. Additional aspects described herein pertain to determining when to suspend measurements and perform measurement predictions in a prediction window. In an example, the UE transmits a prediction report and / or a performance report to the network entity based on the measurement prediction. In an example, the network entity prepares a handover by selecting a target cell using the measurement prediction more quickly compared to a network entity that prepares a handover without using the measurement prediction.
[0009] According to some aspects, the UE receives, from a network entity, a configuration configuring: measurement prediction of a machine learning (ML) model and performance monitoring for the measurement prediction. The UE transmits, to the network entity, a report associated with at least one of: the measurement prediction or the performance monitoring.
[0010] According to some aspects, the network entity transmits, to a user equipment (UE) , a configuration configuring: measurement prediction of a machine learning (ML) model and performance monitoring for the measurement prediction. The network entity receives, from the UE, a report associated with at least one of: the measurement prediction or the performance monitoring.BRIEF DESCRIPTION OF THE DRAWINGS
[0011] Fig. 1A is a block diagram of an example system in which a radio access network (RAN) and a user equipment (UE) can implement the techniques of this disclosure for managing artificial intelligence / machine learning (AI / ML) techniques according to an embodiment.
[0012] Fig. 1B is a block diagram of an example base station including a central unit (CU) and a distributed unit (DU) that can operate in the system of FIG. lA according to an embodiment.
[0013] Figs. 2A-2B are block diagrams of example protocol stacks according to which the UE communicates with a base station or a unit of a base station (e.g., CU / DU) according to embodiments.
[0014] Figs. 3A-3C illustrate example implementations of time domain prediction by a UE according to an embodiments.
[0015] Fig. 4A illustrates an example signaling diagram for the UE to transmit a performance report to a base station based on a performance monitoring configuration according to an embodiment.
[0016] Fig. 4B illustrates an example signaling diagram for the UE to transmit a prediction report to a base station based on a measurement prediction configuration according to an embodiment.
[0017] Fig. 5A illustrates an example signaling diagram in which the performance report (s) each include respective actual measurement result (s) and predicted measurement result (s) according to an embodiment.
[0018] Fig. 5B illustrates an example signaling diagram in which the prediction report (s) each include respective actual measurement result (s) and predicted measurement result (s) according to an embodiment.
[0019] Fig. 5C illustrates an example signaling diagram in which the prediction report (s) each include respective predicted measurement result (s) and the measurement report (s) each include respective actual measurement result (s) according to an embodiment.
[0020] Figs. 6A-6B are flowcharts of methods of wireless communication at a network entity according to embodiments.
[0021] Figs. 7A-7C are flowcharts of methods of wireless communication at a network entity according to embodiments.
[0022] Figs. 8A-8B are flowcharts of methods of wireless communication at a UE according to embodiments.
[0023] Figs. 9A-9C are flowcharts of methods of wireless communication at a UE according to embodiments.
[0024] Fig. 10 is an example flowchart of a method of wireless communication at a UE according to an embodiment.
[0025] Fig. 11 is an example flowchart of a method of wireless communication at a UE according to an embodiment.
[0026] Fig. 12 is an example flowchart of a method of wireless communication at a UE according to an embodiment.
[0027] Fig. 13 is an example flowchart of a method of wireless communication at a network entity according to an embodiment.
[0028] Fig. 14 illustrates a schematic diagram of modules, components, or circuitry that can provide an example implementation for a UE to perform measurements according to an embodiment. DETAILED DESCRIPTION OF THE DRAWINGS
[0029] Techniques of the disclosure are directed to handover scenarios. A base station can prepare a handover more quickly by selecting a target cell in advance based on measurement predictions. In some example aspects, the base station can use predicted measurement results for layer 3 (L3) mobility (including handovers) and / or the UE can request or initiate handovers based on predicted measurement results. The UE can perform L1 measurements or filtering as described in more detail later herein. Prediction parameters, including observation window size and prediction window size, can be configured by the network to reduce measurement workload for the UE.
[0030] Referring first to Fig. 1A, an example wireless communication system 100 in which communication devices can implement one or more of these techniques. The wireless communication system 100 includes a UE 102, a base station (BS) 104, a base station 106 and a core network (CN) 110. The base stations 104, 106 can operate in a radio access network (RAN) 105. The CN 110 may be or include an evolved packet core (EPC) 111, a fifth generation (5G) core (5GC) 160 and / or a sixth generation (6G) core (6GC) 170, for example. The base station 104 can be an eNB supporting an S1 interface for communicating with the EPC 111, an ng-eNB supporting an NG interface for communicating with the 5GC 160, or a gNB that supports an NR radio interface as well as an NG interface for communicating with the 5GC 160. The base station 104 can also be an 6G base station (BS) supporting the NG interface, a new NG interface, or a 6G BS-to-CN (e.g., N6G) interface for communicating with the 5GC 160 or the 6GC 170. To directly exchange messages with each other during the scenarios discussed below, the base stations 104 and 106 can support an X2, Xn, new Xn, or 6G BS-to-BS (e.g., X6G) interface. Among other components, the EPC 111 can include a Serving Gateway (SGW) 112, a Mobility Management Entity (MME) 114, and a Packet Data Network Gateway (PGW) 116. The SGW 112 is generally configured to transfer user-plane packets related to audio calls, video calls, Internet traffic, etc., and the MME 114 is configured to manage authentication, registration, paging, and other related functions. The PGW 116 provides connectivity from the UE to one or more external packet data networks, e.g., an Internet network and / or an Internet Protocol (IP) Multimedia Subsystem (IMS) network. The 5GC 160 includes a User Plane Function (UPF) 162 and an Access and Mobility Management (AMF) 164, and / or a Session Management Function (SMF) 166. The UPF 162 is generally configured to transfer user-plane packets related to audio calls, video calls, Internet traffic, etc., the AMF 164 is configured to manage authentication, registration, paging, and other related functions, and the SMF 166 is configured to manage PDU sessions. The 6GC 170 includes a 6G UPF 172 and a 6G AMF 174, and / or 6G SMF 176, similar to the UPF 162, the AMF 164 and the SMF 176 with enhanced functions respectively.
[0031] As illustrated in Fig. 1A, the base station 104 supports cell 124, and the base station 106 supports cells 126A, 126B, and 126C. The cells 124, 126A, 126B, and / or 126C can operate on the same carrier frequency or different carrier frequencies. For example, the cells 126A, 126B, and 126C operate on DL carrier frequencies f1, f2, and f3 respectively. The cell 124 may operate on the DL carrier frequency f1. When the cells 124, 126A, 126B, and / or 126C operate in a time division duplex (TDD) mode, the cells 124, 126A, 126B, and / or 126C operate in UL carrier frequencies that are the same as the DL carrier frequencies. When the cells 124, 126A, 126B, and / or 126C operate in a frequency division duplex (FDD) mode, the cells 124, 126A, 126B, and / or 126C operate in UL carrier frequencies different from the DL carrier frequencies. The cells 124, 126A, 126B, and / or 126C can partially overlap to provide seamless service continuity. Thus, while the UE 102 may move among the cells 124, 126A, 126B, and 126C, the UE 102 may still communicate with the CN 110 via these cells. In some other scenarios, the cells 126B and / or 126C may belong to one or more other base stations (e.g., the base station 104 and / or one or more additional base stations not shown in Fig. 1A) . In general, the wireless communication network 100 can include any suitable number of base stations supporting 6G cells, NR cells and / or EUTRA cells. More particularly, the EPC 111 can be connected to any suitable number of base stations supporting EUTRA cells, while the 5GC 160 and / or the 6GC 170 can be connected to any suitable number of base stations supporting 6G cells and / or NR cells. Although the examples below refer specifically to specific CN types (EPC, 5GC, 6GC) and RAT types (6G, 5G NR and EUTRA) , in general the techniques of this disclosure also can apply to other suitable radio access and / or core network technologies such as seventh generation (7G) radio access and / or 7G core network.
[0032] With continued reference to Fig. 1A, the base station 104 includes processing hardware 130 that includes one or more general-purpose processors (e.g., CPUs) and a non-transitory computer-readable medium (CRM) storing instructions that the one or more general-purpose processors execute. Additionally or alternatively, the processing hardware 130 may include special-purpose processing units. According to an embodiment illustrated in Fig. 1A, the processing hardware 130 includes a processor 132 to process data that the base station 104 transmits in the downlink direction, or data that the base station 104 receives in the uplink direction. The processing hardware 130 also includes a receiver 134 configured to transmit data in the downlink direction and to receive data in the uplink direction. The processing hardware 130 also includes a measurement controller 136 configured to manage measurement configurations. The processing hardware 130 further includes a measurement prediction controller 138 configured to manage (e.g., configure, release, activate, deactivate, enable, or disable) measurement prediction configurations for UEs as described in more detail later herein. The CRM (not shown) stores executable code that, when executed on the processor 132, enable the processor 132 to perform methods according to embodiments described in this section. The base station 106 includes generally similar components. In particular, components 140, 142, 144, 146, and 148 of the base station 106 may be similar to the components 130, 132, 134, 136, and 138 respectively.
[0033] The UE 102 includes processing hardware 150 that can include one or more general-purpose processors such as CPUs and non-transitory CRM storing machine-readable instructions executable on the one or more general-purpose processors, and / or special-purpose processing units. As schematically illustrated in Fig. 1A, the processing hardware 150 includes a processor 152 to prepare data that the UE 102 transmits in the uplink direction, or to process data that the UE 102 receives in the downlink direction. The processing hardware 150 also includes a transceiver 154 configured to transmit data in the uplink direction and to receive data in the downlink direction. The processing hardware 150 further includes a measurement controller 156 configured to manage measurement configurations, perform measurements and / or transmit measurement results. The process hardware 150 additionally includes a measurement prediction controller 158 configured to manage (e.g., configure, release, activate, deactivate, enable, or disable) measurement predictions.
[0034] Fig. 2A illustrates, in a simplified manner, an example protocol stack 200 according to which the UE 102 can communicate with an eNB / ng-eNB or a gNB 230 or 232 (e.g., one or more of the base stations 104, 106) .
[0035] In the example stack 200, a NR PHY 202B provides transport channels to a NR MAC sublayer 204B, which in turn provides logical channels to a NR RLC sublayer 206B. The NR RLC sublayer 206B in turn provides data transfer services to a NR PDCP sublayer 208B. The NR PDCP sublayer 208B in turn can provide data transfer services to a Service Data Adaptation Protocol (SDAP) 210B sublayer and / or a radio resource control (RRC) sublayer (not shown in Fig. 2A) . Similarly, a physical layer (PHY) 202A of 6G provides transport channels to the 6G MAC sublayer 204A, which in turn provides logical channels to the 6G RLC sublayer 206A. The 6G RLC sublayer 206A in turn provides RLC channels to a 6G PDCP sublayer 208A. The 6G PDCP sublayer 208A in turn can provide data transfer services to a 6G Service Data Adaptation Protocol (SDAP) sublayer 210A or a 6G radio resource control (RRC) sublayer (not shown in Fig. 2A) . In some implementations, the 6G SDAP sublayer 210A can be omitted. In such cases, the PDCP sublayer 208A may support functionalities of the SDAP sublayer 210A. The UE 102, in some implementations, supports both the 6G and the NR stack as shown in Fig. 2A, to support handover between 6G and NR base stations and / or to support DC over 6G and NR interfaces.
[0036] The 6G PDCP sublayer 208A and the NR PDCP sublayer 208B receive packets (e.g., from an Internet Protocol (IP) layer, layered directly or indirectly over the PDCP layer 208A or 208B) that can be referred to as service data units (SDUs) , and output packets (e.g., to the RLC layer 206A or 206B) that can be referred to as protocol data units (PDUs) . Except where the difference between SDUs and PDUs is relevant, this disclosure for simplicity refers to both SDUs and PDUs as “packets. ”
[0037] On a control plane, the 6G PDCP sublayer 208A and the NR PDCP sublayer 208B can provide signaling radio bearers (SRBs) or RRC sublayer (not shown in Fig. 2A) to exchange RRC messages or non-access-stratum (NAS) messages, for example. On a user plane, the 6G PDCP sublayer 208A and the NR PDCP sublayer 208B can provide Data Radio Bearers (DRBs) to support data exchange. Data exchanged on the 6G PDCP sublayer 208A and NR PDCP sublayer 208B can be SDAP PDUs, Internet Protocol (IP) packets or Ethernet packets.
[0038] Fig. 2B illustrates, in a simplified manner, an example protocol stack 250, which the UE 102 can communicate with a DU (e.g., DU 174) and a CU (e.g., CU 172) . The radio protocol stack 200 is functionally split as shown by the radio protocol stack 250 in Fig. 2B. The CU at any of the base stations 104 or 106 can hold all the control and upper layer functionalities (e.g., RRC 212, SDAP 210, PDCP 208) , while the lower layer operations (e.g., RLC 206, MAC 204, and PHY 202) are delegated to the DU. To support connection to a core network (e.g., 5GC or 6GC) , PDCP 208 provides DRBs to SDAP 210 and SRBs to RRC 212. In some implementations, the RRC 212, SDAP 210, and PDCP 208 are the 6G RRC 212A, 6G SDAP 210A, 6G PDCP 208A, respectively. In other implementations, the RRC 212, SDAP 210, and PDCP 208 are the NR RRC 212B, NR SDAP 210B, NR PDCP 208B, respectively.
[0039] Figs. 3A-3C illustrate example implementations in which a UE predicts continuous measurement results of a carrier frequency in a prediction window (PW) based on continuous historical measurement results of the carrier frequency in an observation window (OW) . Note, the historical measurement results in the OW are actual measurement results obtained by a UE based on measurements performed by the UE. In some implementations, the UE obtains the predicted measurement results and the historical measurement results for the same cell which can be a serving cell or a neighbor cell. In other implementations, the UE obtains the predicted measurement results and the historical measurement results for different cells. For example, the UE obtains the historical measurement results for a serving cell and obtains the predicted measurement results for a neighbor cell.
[0040] Fig. 3A illustrates a diagram 300A of a certain mode or operational state of the UE, e.g., the calibration mode during which the UE trains or tunes an ML model (or called AI model) . OW 390A includes a certain number of consecutive observations (measurements) O1, O2, . . . ON, where N >= 1. PW 392A includes a certain number of consecutive predictions P1, P2, . . . PM, where M >= 1. In this example, N = M = 4, but in general the observation and prediction windows need not be of the same size, and the values of N and M can be any suitable positive integers.
[0041] The UE can train the ML model using the O1, O2, . . ., ON to generate the predictions P1, P2, . . . PM. Further, as illustrated in this example, the UE can collect observations within another OW 394A at the same time as generating predictions within PW 392A, to generate predictions in PW 396A. In particular, the UE in this example generates prediction P1 based on OW 390A and, for the same frequency (or set of frequencies) and for the same instance of time (e.g., one or more frames, timeslots, symbols) , generates an observation O'N within OW 394A. The overlap between PW 392A and OW 394A is one. In other words, the offset between OW 390A and OW 394A is one as measured in observation samples O or prediction samples P. More generally, the offset can be any suitable positive integer smaller than the size of the OW.
[0042] The UE can assess the difference between P1 and O'N to generate a feedback signal for training the ML model. A smaller difference P1 and O'N indicates greater accuracy of predictions, and greater difference accordingly indicates smaller accuracy of predictions. When the overlap between PW 392A and OW 394A is L > 1, the UE can assess the cumulative or average difference between groups of samples {P1, . . . PL} and {O'N, ... O'N-L}.
[0043] In some implementations, the UE starts a new OW (not shown) with an offset of one relative to OW 394A, so that there is an overlap of one between the new OW and PW 396A, similar to the overlap between OW 394A and PW 392A. This approach does not provide power saving at the UE, and the UE in some implementations operates in the mode of Fig. 3A only when training or re-calibrating the ML model. Alternatively, the UE can start the new OW after the end of PW 396A, in which the case the overlap between OW 394A and the PW 392A does not provide power saving for P1 (as the UE continues to expend power to obtain the observation O'N) , and results in power saving only for P2 -PN. As another alternative, the UE can start the new OW after the end of PW 392A, in which case the first observation in the new OW overlaps with the last prediction in PW 396A, to result in power saving for P1 -PN-1.
[0044] Although Fig. 3A illustrates observations O2 and O'1 for example as distinct observations, these designations can be only logical, and the UE can perform a single observation that belongs both to OW 390A and OW 394A. In this sense, because each prediction is based on a certain number of prior observations, the UE operates OW 390A and 394A as a sliding OW, and accordingly operates PW394A and 396A as a sliding PW.
[0045] In some implementation, when the UE determines that the ML model is sufficiently accurate, the UE can operate without overlaps between OWs and PWs as illustrated in Fig. 3B. In some implementations, the UE is configured to periodically re-enter the mode of Fig. 3A, so as to re-calibrate or fine-tune the ML model. When predictions are sufficiently accurate, the UE can consider the predictions applicable to the relevant carrier frequency, and provide an indication of applicability to the RAN.
[0046] In Fig. 3B, the diagram 300B illustrates an OW and a PW do not overlap. Thus, the UE generates predictions within PW 392B based on the observations in OW 390B, and then generates predictions within PW 396B based on the observations in OW 394B. The size of each of the OWs and the PWs is two in this example, and thus the UE generates two predictions based on two prior observations and expends approximately half the power required to generate observations at each instance of time.
[0047] In general, the OWs and the PWs in Fig. 3B can be of any suitable size, and the lengths of OWs and PWs need not be the same. In some implementations, the UE switches between the modes of Figs. 3A and 3B depending on the accuracy of predictions or based on timing as discussed above.
[0048] Now referring to Fig. 3C, the diagram 300C illustrates that the UE can implement a sliding OW 390C / 394C but omit certain observations to save power. In this example, the UE does not obtain an observation during gap G1, which is an occasion for a potential observation that occurs between two observations, within an OW 390C / 394C. The PW 392C / 396C in this example spans a single prediction based on three observations within the OW 390C / 394C, but the OW 390C / 394C spans five observation occasions. In this manner, the UE expends less power to generate the set of observations within the OW 390C / 394C.
[0049] Next, several example scenarios in which the UE or the base station operating in the system of Fig. 1A performs measurement prediction and performance monitoring for the measurement prediction. Figs. 4A and 4B are UE-based performance monitoring scenarios, where the UE obtains one or more performance metric (s) or determines performance of the measurement prediction is good or poor. Figs. 5A, 5B, and 5C are network-based performance monitoring scenarios, where the base station obtains one or more performance metric (s) or determines performance of the measurement prediction is good or poor based on predicted measurement results and actual measurement results received from the UE.
[0050] Referring first to Fig. 4A, in a scenario 400A, the UE 102 initially communicates with the base station 106 via a serving cell (e.g., cell 126A shown in Fig. 1A) on a first DL carrier frequency and a first UL carrier frequency. The first DL carrier frequency and the first UL carrier frequency can be the same or different. During the communication, the UE 102 may transmit 402 UE capabilities of the UE 102 to the BS 106. Alternatively, the base station 106 receives 404 the UE capabilities from a RAN node (e.g., the base station 104) or a CN node (e.g., the AMF 164 or 714) . The UE capabilities may include measurement capabilities, measurement prediction capabilities, and / or measurement prediction performance monitoring and / or reporting capabilities. In some implementations, the measurement capabilities include a first capability indicating that the UE 102 supports NR measurements and event A triggered reporting. In some implementations, the measurement capabilities include a second capability indicating that the UE 102 supports EUTRA measurements and event B triggered report. In some implementations, the measurement capabilities include a third capability indicating that the UE 102 supports 6G measurements and event A triggered reporting. In some implementations, the measurement capabilities include a fourth capability indicating that the UE 102 supports 5G measurements and event B triggered report. In some implementations, the measurement capabilities include a fifth capability indicating that the UE 102 supports location-based triggered measurement reporting. In some implementations, the measurement prediction capabilities includes one or more capabilities indicating that the UE 102 supports temporal domain measurement predictions as described for Fig. 3A, 3B and / or 3C. In some implementations, the measurement prediction capabilities includes one or more capabilities indicating a length of a PW and / or a length of an OW. In some implementations, the measurement prediction capabilities includes one or more capabilities indicating support of direct measurement prediction and / or indirect measurement prediction.
[0051] In some implementations, the measurement prediction performance monitoring and / or reporting capabilities includes one or more capabilities indicating supported monitoring type (s) , supported reporting types (s) , and / or supported metric (s) . The supported monitoring type (s) may include periodic monitoring and / or event triggered monitoring. In some implementations, the supported reporting type (s) include periodic report and / or event triggered reporting. In some implementations, the supported metric (s) includes accuracy, true positives, true negatives, false positives, false positives, precision, recall, F1 score, specificity, balanced accuracy, mean absolute error, root mean squared error, mean absolute percentage error, and / or R-squared. In some implementations, the measurement prediction performance monitoring and / or reporting capabilities includes a capability indicating support of a UE-based performance monitoring. In other implementations, the measurement prediction performance monitoring and / or reporting capabilities includes a capability indicating support of a network-based performance monitoring.
[0052] After receiving the UE capabilities, the base station 106 transmits 406 one or more measurement configuration (s) to the UE 102. In some implementations, the base station 106 transmits 406 the measurement configuration (s) the UE 102 based on the measurement capabilities. The UE 102 performs 408 measurements based on the measurement configuration (s) . In some implementations, each of the measurement configuration (s) indicates or includes a measurement object configuration, and / or a measurement reporting configuration. Each of the measurement object configuration (s) configures a respective measurement object. The UE 102 may perform 408 measurements on the measurement object (s) . Each of the measurement configuration (s) is identified by a respective measurement configuration identity / identifier (ID) . Each of the measurement object configuration (s) is identified by a respective measurement object ID. Each of the measurement reporting configuration (s) is identified by a respective measurement reporting configuration ID. In some implementations, each of the measurement configuration (s) includes a respective measurement ID, a respective measurement object ID, and / or a respective measurement reporting configuration ID. The base station 106 transmits the measurement object configuration (s) , the measurement object ID (s) , the measurement reporting configuration (s) , and the measurement reporting configuration ID (s) to the UE 102.
[0053] In some implementations, the measurement configuration (s) , the measurement object configuration (s) , and the measurement reporting configuration (s) includes a first measurement configuration, a first measurement object configuration, and a first measurement reporting configuration, respectively. The first measurement configuration, the first measurement object configuration, and the first measurement reporting configuration are identified by a first measurement ID, a first measurement object ID and a first measurement reporting configuration ID respectively. The first measurement configuration includes the first measurement object ID and the first measurement reporting configuration ID to indicate the first measurement object configuration and the first measurement reporting configuration, respectively. The first measurement configuration may include the first measurement ID.
[0054] In some implementations, the measurement configuration (s) , the measurement object configuration (s) , and the measurement reporting configuration (s) includes a second measurement configuration, a second measurement object configuration, and a second measurement reporting configuration, respectively. The second measurement configuration, the second measurement object configuration, and the second measurement reporting configuration are identified by a second measurement ID, a second measurement object ID and a second measurement reporting configuration ID respectively. The second measurement configuration includes the second measurement object ID and the second measurement reporting configuration ID to indicate the second measurement object configuration and the second measurement reporting configuration, respectively. The second measurement configuration may include the second measurement ID.
[0055] In some implementations, after transmitting the UE capabilities or receiving 406 the measurement configuration (s) , the UE 102 determines whether measurement prediction is applicable. If the UE 102 determines that measurement prediction is applicable, the UE 102 transmits 410 an applicability report to the base station 106, indicating that the measurement prediction is applicable. In some implementations, the UE 102 determines whether measurement prediction is applicable to a particular measurement configuration, a particular measurement object (configuration) , and / or a particular measurement reporting configuration. In some implementations, the UE 102 determines that measurement prediction is applicable to the first measurement configuration. In response to the determination, the UE 102 transmits 410 an applicability report to the base station 106, indicating that the measurement prediction is applicable to the first measurement configuration. The UE 102 may include the first measurement ID in the applicability report, indicating the measurement prediction is applicable to the first measurement configuration. In some implementations, the UE 102 determines that measurement prediction is applicable to the first measurement object. In response to the determination, the UE 102 indicates, in the applicability report, that the measurement prediction is applicable to the first measurement object (configuration) . In some implementations, the UE 102 includes the first measurement object ID in the applicability report, indicating the measurement prediction is applicable to the first measurement object (configuration) . In some implementations, the UE 102 determines that measurement prediction is applicable to the first measurement reporting configuration. In response to the determination, the UE 102 indicates, in the applicability report, that the measurement predication is applicable to the first measurement reporting configuration. In some implementations, the UE 102 includes the first measurement reporting configuration ID in the applicability report, indicating that the measurement prediction is applicable to the first measurement reporting configuration.
[0056] In some implementations, the UE 102 determines that the measurement prediction is applicable to the second measurement configuration, the second measurement object, or the second measurement reporting configuration as described above for the first measurement configuration, the first measurement object, or the first measurement reporting configuration. In other implementations, the UE 102 determines that the measurement prediction is not applicable to the second measurement configuration, the second measurement object, or the second measurement reporting configuration.
[0057] In some implementations, the UE 102 includes one or more cell ID (s) in the applicability report and the one or more cell ID (s) indicates one or more respective cell (s) . Thus, the cell ID (s) indicate measurement prediction (s) is / are applicable to the respective cell (s) . In some implementations, the UE 102 includes or indicates one or more preferred or applicable {OW and / or PW} s in the applicability report. In some implementations, the UE 102 includes or indicates, in the applicability report, one or more tuple (s) of {a length of OW, the number of actual (i.e., non-predicted) measurement results in the OW, and / or a TTT and / or a length of PW for predicting occurrence of the event} for direct measurement prediction (s) . Examples of the OWs are as described for Figs. 3A-3C.
[0058] In other implementations, the UE 102 includes or indicates, in the applicability report, one or more tuples of {a length OW, the number of actual (i.e., non-predicted) measurement results in the OW, PW, and / or the number of predicted measurement results in the PW} for indirect measurement prediction (s) . For example, the UE 102 includes a length of a first OW, the number of actual (i.e., non-predicted) measurement results in the first OW, a length of a first PW, and / or the number of predicted measurement results in the first PW for a first (indirect) measurement prediction. In another example, the UE 102 includes a length of a second OW, the number of actual (i.e., non-predicted) measurement results in the second OW, a length of a second PW, and / or the number of predicted measurement results in the second PW for a first measurement prediction. Examples of the OWs and / or the PWs are as described for Figs. 3A-3C.
[0059] After event 402, 404, 406 or 410 or before event 410, the base station 106 transmits 412A a measurement prediction configuration to the UE 102. The UE 102 performs 416 one or more measurement prediction (s) based on (e.g., in response to or in accordance with) the measurement prediction configuration. In some implementations, the UE 102 suspends 414A performing measurements in a periodic PW (e.g., recurred periodically) and performs measurement prediction (s) in the periodic PW, as described for Fig. 3A, 3B, or 3C. The UE 102 may obtain one or more predicted measurement result (s) in the measurement prediction (s) . Based on the measurement prediction (s) , the UE 102 may predict 416 the event occurs and transmit 418A a prediction report to the base station 106. The events 412A, 414A, 416, and 418A may be collectively referred to as a measurement prediction configuration and reporting procedure 490A. In the description, “OW” can be used to represent “periodic OW” , “OW instances” , or “a specific OW” . In the description, “OW” can be used to represent “periodic PW” , “PW instances” , or “a specific PW” . In some implementations, the performance report (s) are measurement report message (s) (e.g., MeasurementReport message (s) ) . In other implementations, the prediction report (s) are UE assistance information message (s) (e.g., UEAssistanceInformation message (s) ) . In yet other implementations, the prediction report (s) are RRC message (s) .
[0060] In some implementations, the measurement prediction configuration includes one or more configuration parameter (s) for a measurement prediction (e.g., indirect measurement prediction or direct measurement prediction) . In some implementations, the measurement prediction configuration includes an indication indicating indirect measurement prediction or direct measurement prediction. In some implementations, the measurement prediction configuration includes an indication indicating prediction domain information. The prediction domain can be at least one of the following: temporal domain, frequency domain, or spatial domain. In some implementations, the base station 106 includes the first measurement ID in the measurement prediction configuration, indicating that the measurement prediction configuration is associated with the first measurement configuration. In some implementations, the base 106 includes the first measurement object ID in the measurement prediction configuration, indicating that the measurement prediction configuration is associated with the first measurement object. In some implementations, the base 106 includes the first report configuration ID in the measurement prediction configuration, indicating that the measurement prediction configuration is associated with the first report configuration.
[0061] In some implementations, the measurement prediction configuration includes an indication indicating prediction type information. The prediction type information can include at least one of the following: intra-frequency intra-cell prediction, intra-frequency inter-cell prediction, inter-frequency inter-cell prediction, inter-RAT prediction or others. In some implementations, the measurement prediction configuration includes an indication indicating cluster-based prediction, e.g., multiple inputs cells to predict one or multiple output cells. In some implementations, the measurement prediction configuration includes an indication indicating non-sliding prediction or sliding prediction. In some implementations, the base station 106 assigns a measurement prediction configuration ID (e.g., a first measurement prediction configuration ID) indicating (e.g., identifying) the measurement prediction configuration and transmits the measurement prediction configuration ID to the UE 102 in event 412A. In one implementation, the measurement prediction configuration includes the measurement prediction configuration ID. In another implementation, in event 412A, the base station 106 includes the measurement prediction configuration ID and the measurement prediction configuration in a container (e.g., information element (IE) ) and transmits the container to the UE 102. To simplify the following description, the measurement prediction configuration can represent the measurement prediction configuration alone, the measurement prediction configuration and the measurement prediction configuration ID, or the container.
[0062] In some implementations, the UE 102 may transmit 410 the applicability report after receiving the measurement prediction configuration. In such cases, the UE 102 may include, in the applicability report, the measurement prediction configuration ID indicating the measurement prediction configured by the measurement prediction configuration is applicable.
[0063] In some implementations, the measurement prediction configuration includes one or more configurations for measurement predictions. For example, the measurement prediction configuration includes an OW configuration configuring an OW. In another example, the measurement prediction configuration includes a PW configuration configuring a PW. In some implementations, the OW and the PW recur periodically as described for Figs. 3A-3C. In some implementations, the OW and PW are indicated in the UE capabilities. In other implementations, the OW and PW are indicated in the applicability report.
[0064] In some implementations, the measurement prediction configuration includes one or more configurations for prediction reporting. For example, the measurement prediction configuration may include a prediction report configuration configuring an event (e.g., predicted event) for the UE 102 to predict. The prediction report configuration may include parameters for evaluating (e.g., determine) whether an event (e.g., predicted event) occurs. The prediction report configuration may configure first criterion / criteria for the UE 102 to determine whether the event occurs or to send a prediction report. The prediction report configuration may configure a reference signal (RS) type indicating a RS that the UE uses to obtain (beam or cell) predicted measurement results. The RS may be a synchronization signal (SS) / physical broadcast channel (PBCH) block or a channel state information RS (CSI-RS) . The prediction report configuration may configure a reporting format that may configure one or more quantities per cell and / or per beam that the UE 102 includes in a prediction report. The one or more quantities may include or be a RS received power (RSRP) , Reference Signal Received Quality (RSRQ) , Received Signal Strength Indicator (RSSI) , Signal to Interference plus Noise Ratio (SINR) , or Precoding matrix indicator (PMI) . Optionally, the one or more quantities can be L3 filtered result (s) . The reporting format may include other associated information such as the maximum number of cells and / or the maximum number of beams per cell to report.
[0065] In some implementations, the first criterion / criteria includes a first inequality for evaluating whether a first entering condition for the event is fulfilled. The prediction report configuration may include first parameters for evaluating (e.g., determining) whether the first entering condition (i.e., the first inequality) is fulfilled. In some implementations, the first criterion / criteria includes a second inequality for evaluating whether a first leaving condition (i.e., the second inequality) for the event is fulfilled. The prediction report configuration may include second parameters for evaluating (e.g., determine) whether the first leaving condition (i.e., the second inequality) is fulfilled. The second parameters may be the same as or different from the first parameters. In some implementations, the measurement prediction configuration or the prediction report configuration may configure a first time to trigger (TTT) (i.e., a value) indicating a time duration for which the first entering condition for the event has to be met in order to trigger a prediction report. If the first entering condition has been fulfilled for the first TTT, the UE 102 determines that the event occurs.
[0066] In other implementations, the UE 102 is not configured with a TTT (i.e., a value) indicating a time duration for which the first entering condition for the event has to be met in order to trigger a prediction report. That is, a TTT for evaluating whether the event occurs is zero. For example, the measurement prediction configuration or the prediction report configuration does not configure a TTT (e.g., the first TTT) . In one implementation, if the first entering condition is fulfilled, the UE 102 determines that the event occurs. In another implementation, if the UE 102 is configured with the second TTT as describe below, the UE 102 may use the second TTT. In this case, if the first entering condition has been fulfilled for the second TTT, the UE 102 determines that the event occurs. In some implementations, if the UE 102 is not configured with the second TTT (and the first TTT) and the first entering condition is fulfilled, the UE 102 determines that the event occurs.
[0067] In some alternative implementations, the first measurement configuration or the first measurement reporting configuration configures the first criterion / criteria instead of the measurement prediction configuration. In such cases, the first measurement configuration or the first measurement reporting configuration includes the first parameters, the second parameters, and / or the first TTT instead of the measurement prediction configuration.
[0068] In other implementations, the measurement prediction configuration does not include the prediction report configuration. In such cases, the UE 102 may use the first measurement reporting configuration to determine whether to send a prediction report.
[0069] In some implementations, the base station 106 transmits the first measurement configuration and the measurement prediction configuration in a single DL message (e.g., RRC reconfiguration message) to the UE 102. In response, the UE 102 may transmit a UL response message (e.g., a RRC reconfiguration complete message) . In other implementations, the base station 106 transmits the first measurement configuration and the measurement prediction configuration in a first DL message (e.g., RRC reconfiguration message) and a second DL message (e.g., a RRC reconfiguration message) to the UE 102, respectively. In response to the first DL message and the second DL message, the UE 102 may transmit a first UL response message and a second UL response message, respectively. The base station 106 may transmit the other measurement configuration (s) in the first DL message, the second DL message and / or additional DL message (s) to the UE 102. Each of the additional DL message (s) may be a RRC message such as a RRC reconfiguration message. The UE 102 may transmit a UL message to the base station 106 in response to each of the additional RRC message (s) . The UL message may be a RRC response message such as a RRC reconfiguration complete message.
[0070] In some implementations, each of the measurement object (s) configures a respective DL carrier frequency for measurement. For example, one of the measurement object (s) configures a second DL carrier frequency. In some implementations, the second DL carrier frequency is or is not within the first DL carrier frequency. In other implementations, the second DL carrier frequency does not overlap with the first DL carrier frequency. In some implementations, each of the measurement object (s) indicates a frequency / time location and / or a subcarrier spacing of reference signal (s) to be measured.
[0071] Each of the measurement reporting configuration (s) configures measurement result reporting. The measurement reporting configuration may configure (e.g., include) a reporting type. The measurement reporting configuration may configure second criterion / criteria for the UE 102 to determine whether to send a measurement report. The measurement reporting configuration may configure a reference signal (RS) type indicating a RS that the UE 102 uses to obtain (beam or cell) measurement results. The RS may be a SS / PBCH block or a CSI-RS. The measurement reporting configuration may configure a reporting format that may configure one or more quantities per cell and / or per beam that the UE includes in a measurement report. The descriptions of one or more quantities can refer to the above ones and will not be repeated herein. The reporting format may include other associated information such as the maximum number of cells and / or the maximum number of beams per cell to report. In some implementations, the reporting type is periodical or event triggered. If the reporting type is periodical, the UE 102 periodically transmits a measurement report (e.g., MeasurementReport message) to the base station 106. The measurement reporting configuration may include a periodicity for the periodical measurement reports. If the reporting type is event triggered, the measurement reporting configuration may include an event trigger configuration configuring an event (e.g., trigger event) for the UE 102 to detect. The event in the event trigger configuration may be the same as the event (if configured or included) in the measurement prediction configuration. If the UE 102 detects the event (i.e., actual event) occurs as described below, the UE 102 transmits a measurement report to the base station 106. The event trigger configuration may include parameters for evaluating (e.g., determine) whether the event occurs. In some implementations, the second criterion / criteria includes a third inequality for evaluating whether a second entering condition for the event is fulfilled. The event trigger configuration may include third parameters for evaluating (e.g., determine) whether the second entering condition (i.e., the third inequality) is fulfilled. The second entering condition may be the same as or different from the first entering condition. The third inequality may be the same as or different form the first inequality. The third parameters may be the same as or different from the first parameters. In some implementations, the second criterion / criteria includes a fourth inequality for evaluating whether a second leaving condition (i.e., the fourth inequality) for the event is fulfilled. The event trigger configuration may include fourth parameters for evaluating (e.g., determine) whether the second leaving condition (i.e., the fourth inequality) is fulfilled. The second leaving condition may be the same as or different from the first leaving condition. The fourth inequality may be the same as or different form the second inequality. The fourth parameters may be the same as or different from the second parameters or the third parameters. The event trigger configuration may include a second TTT (i.e., a value) indicating a time duration during an entering condition (e.g., the second entering condition) for the event needs to be met in order to trigger a measurement report.
[0072] In some implementations, the third / fourth parameters in the event trigger configuration may include one or more offsets, one or more thresholds, a maximum number of RS indexes to report, a maximum number of cells to include in a measurement report, an indication indicating to include the best neighbor cells per serving frequency, an indication indicating to send a measurement report when the second leaving condition is met for a cell in a cell triggered list, and / or a report quantity for measurement result (s) included in a measurement report. The second TTT may be shorter than, longer than, or equal to the first TTT.
[0073] In some implementations, the second criterion / criteria are configured for determining whether the event occurs regardless the event is a predicted event or an actual event. The descriptions with respect to the first criterion / criteria can apply with the second criterion / criteria. In such cases, there is no need to configure the first criterion / criteria in the measurement prediction configuration, which saves signaling overhead.
[0074] In some implementations, the even trigger configuration includes an event ID indicating the event. In one implementation, the event ID is event A3 and the event is that the UE transmits a measurement report when the UE detects that a neighbor cell becomes offset better than a serving cell, e.g., in terms of the reporting quantity. In some implementations, the entering condition and the leaving condition for event A3 are as described in 3GPP specification 38.331. In other implementations, the event ID is event A5 and the event is that the UE detects that a service cell becomes worse than threshold 1 and a neighbor cell becomes better than threshold 2. The serving cell may be the cell 124 which may be primary cell or a primary secondary cell group (SCG) cell. In some implementations, the entering condition and the leaving condition for event A5 are as described in 3GPP specification 38.331.
[0075] In some implementations, the first measurement configuration includes the measurement prediction configuration. In other implementations, the first measurement configuration includes the first measurement prediction configuration ID to indicate that the measurement prediction configuration is associated with the first measurement configuration. In some implementations, the first measurement object configuration includes the measurement prediction configuration. In other implementations, the first measurement object configuration includes the first measurement prediction configuration ID to indicate that the measurement prediction configuration is associated with the first measurement object configuration. In some implementations, the first measurement reporting configuration includes the measurement prediction configuration. In other implementations, the first measurement reporting configuration includes the first measurement prediction configuration ID to indicate that the measurement prediction configuration is associated with the first measurement reporting configuration.
[0076] In some implementations, the measurement prediction configuration configures an OW and / or a PW for a measurement prediction. Examples of the OW and / or the PW are as described for Figs. 3A-3C. The base station 106 may configure the OW and / or the PW based on the UE capabilities and / or the applicability report. In some implementations, the measurement prediction configuration configures a length of the OW and / or the number of actual (i.e., non-predicted) measurement results in the OW. In some implementations, the base station 106 configures the length of the OW and the number of actual measurement results in the OW based on the UE capabilities. For example, the UE capabilities include or indicate the length of the OW and the number of actual measurement results in the OW. In other implementations, the base station 106 configure the length of the OW and the number of actual measurement results in the OW based on the applicability report. For example, the applicability report includes or indicates the length of the OW and the number of actual measurement results in the OW. In the case of indirect measurement prediction, the configuration parameter (s) may configure a length of a PW and / or the number of measurement results to predict in the PW as described for Figs. 3A-3C. In some implementations, the base station 106 configures the length of the PW and the number of predicted measurement results in the PW based on the UE capabilities. For example, the UE capabilities include or indicate the length of the PW and the number of predicted measurement results in the PW. In other implementations, the base station 106 configures the length of the PW and the number of predicted measurement results in the PW based on the applicability report. For example, the applicability report includes or indicates the length of the PW and the number of predicted measurement results in the PW. In the case of direct measurement prediction, the configuration parameter (s) may configure a PW (e.g., a length of the PW) for predicting occurrence of the event. In some implementations, the base station 106 may configure the PW based on the UE capabilities and / or the applicability report. For example, the UE capabilities include or indicate the PW and / or a TTT. In another example, the applicability report includes or indicates the PW and / or a TTT. The base station 106 may configure the first TTT or the second TTT in accordance with the TTT indicated in the UE capabilities and / or the applicability report.
[0077] In some implementations, the UE 102 may perform measurements for the serving cell and a neighbor cell based on the measurement configuration (s) or the measurement object configuration (s) . In some implementations, the UE 102 performs measurements in the OW recurring periodically. The UE 102 may perform measurements for the serving cell and a neighbor cell in the OW. Based on the measurement prediction configuration, the UE 102 suspends (i.e., skips) measurements and performs measurement prediction (s) for the skipped measurements using an AI model. In some implementations, the UE 102 skips measurements in the PW and performs measurement prediction (s) for the skipped measurements in the PW using the AI model based on the measurements in the OW, as described for Figs. 3B and 3C. The UE 102 may perform measurement prediction (s) in the PW to obtain predicted measurement result (s) for the serving cell. The UE 102 may perform measurement prediction (s) in the PW to obtain predicted measurement result (s) for the neighbor cell. The AI model may be a classification model or a regression model.
[0078] In some implementations, the UE 102 obtains intermediate measurement results (i.e., actual or non-predicted measurement results) from measurements performed by the UE 102 and generates at least one measurement result (i.e., actual or non-predicted measurement result) using the intermediate measurement results as described for Fig. 12. In some implementations, the UE 102 performs 416 the measurement prediction (s) based on the measurements, the intermediate measurement results, and / or the at least one measurement result. In some implementations, in the measurement prediction (s) , the UE 102 may predict one or more measurement result (s) based on the measurements, the intermediate measurement results, and / or the at least one measurement result. In some implementations, the predicted measurement result (s) include predicted RSRP value (s) , predicted RSRQ value (s) , and / or predicted SINR value (s) . In some implementations, the UE 102 may predict one or more measurement result (s) for the serving cell based on the measurements for the serving cell and / or one or more other cell (s) , the intermediate measurement results for the serving cell and / or one or more other cell (s) , and / or the at least one actual measurement result for the serving cell and / or one or more the other cell (s) . In some implementations, the UE 102 may predict one or more measurement result (s) for the neighbor cell based on the measurements for the serving cell and / or one or more other cell (s) , the intermediate measurement results for the serving cell and / or one or more other cell (s) , and / or the at least one actual measurement result for the serving cell and / or one or more other cell (s) . The one or other cells described above may include the serving cell, the neighbor cell, and / or one or more additional cell (s) other than the serving cell and the neighbor cell.
[0079] In some implementations, the UE 102 determines (i.e., indirectly predict) whether the event occurs based on the predicted measurement result (s) for the serving cell and / or the predicted measurement result (s) for the neighbor cell. In some implementations, the UE 102 determines whether an entering condition is met based on the predicted measurement result (s) for the serving cell and / or the predicted measurement result (s) for the neighbor cell. The entering condition may be the first entering condition or the second enter condition described above. If the entering condition has been met for a TTT (e.g., the first TTT, the second TTT, or the TTT = 0) , the UE 102 determines the event occurs (e.g., the event is triggered) . In other implementations, in the measurement prediction (s) , the UE 102 directly predicts whether the event occurs based on the measurements, the intermediate measurement results, and / or the at least one non-predicted measurement result.
[0080] In some implementations, the UE 102 predicts the event occurs as described above. In response to the prediction, the UE 102 transmits 418A the prediction report to the base station 106, including a physical cell ID (PCI) of a neighbor cell, the first measurement ID, the measurement prediction configuration ID, at least one measurement result (s) . In some implementations, the measurement result (s) includes a first predicted measurement result and / or a second predicted measurement result for the serving cell and / or the neighbor cell respectively. In another implementation, the measurement result (s) includes at least one first actual measurement result (s) for the serving cell, which the UE 102 obtains as described above. In yet another implementation, the measurement result (s) includes at least one second actual measurement result (s) for the neighbor cell, which the UE 102 obtains as described above. The UE 102 may use the first actual measurement result (s) and / or the second actual measurement result (s) to predict the event occurs as described above. The UE 102 may use the first actual measurement result (s) to predict the first predicted measurement result as described above. The UE 102 may use the second actual measurement result (s) to predict the second predicted measurement result as described above.
[0081] In some implementations, the first measurement ID (if included) in the prediction report indicates that the prediction report is associated with the first measurement configuration and / or the measurement prediction configuration. In some implementations, the measurement prediction configuration ID (if included) in the prediction report indicates that the prediction report is associated with the measurement prediction configuration.
[0082] In some implementations, the prediction report is a measurement report message (e.g., MeasurementReport message) . In some implementations, the UE 102 includes a prediction indication in the measurement report message to indicate that the measurement report message includes predicted information. In other implementations, the UE 102 includes a prediction indication in the measurement report message to indicate the measurement result is a predicted measurement result.
[0083] After event 402, 404, 406, 410, or 412A, the base station 106 determines to transmit 420 a performance monitoring configuration to the UE 102, configuring the UE 102 to perform performance monitoring for measurement prediction (s) configured by the measurement prediction configuration. The UE 102 performs performance monitoring for measurement prediction (s) configured by the measurement prediction configuration based on (e.g., in response to or in accordance with) the performance monitoring configuration. In some implementations, to perform the performance monitoring, the UE 102 resumes 422 performing measurements in PW instances 1, ..., Z of the PW while performing measurement prediction (s) for the PW instances 1, ..., Z. Z is an integer and larger than zero. That is, the UE 102 performs measurements in the PW instances 1, ..., Z. In some implementations, the UE 102 obtains 424 one or more performance metric (s) for the measurement prediction (s) . In some implementations, the performance monitoring configuration may configure the UE 102 to obtain the performance metric (s) of the measurement prediction and / or determine performance of the measurement prediction (i.e., UE-based performance monitoring) . The base station 106 configures the UE-based performance monitoring in the performance monitoring configuration based on the UE capabilities or the applicability report. In some implementations, the UE 102 indicates support of the UE-based performance monitoring or supported performance metric type (s) in the UE capabilities (e.g., measurement prediction performance monitoring and / or reporting capabilities) . The base station 106 configures the UE-based performance monitoring in the performance monitoring configuration based on the support indication or supported performance metric type (s) . In other implementations, the UE 102 indicates support of the UE-based performance monitoring or supported performance metric type (s) in the applicability report. The base station 106 configures the UE-based performance monitoring in the performance monitoring configuration based on the support indication or supported performance metric type (s) .
[0084] In each of the PW instances 1, ..., Z, the UE 102 obtains L actual measurement result (s) for the respective PW instance, based on the measurements performed in the respective PW instance, as described above or for Fig. 12. L is an integer and larger than zero. The UE 102 performs measurements in OW instances 1, ..., Z of the OW. The OW instances 1, ..., Z is corresponding to (e.g., associated to or paired with) the PW instances 1, ..., Z, respectively. For example, the OW instance K and the PW instance K are similar to the OW 390B and the PW 392B, respectively. 1 ≤ K ≤ Z. In another example, the OW instance K and the PW instance K are similar to the OW 390C and the PW 392C, respectively. 1 ≤ K ≤ Z. In each of the OW instances 1, ..., Z, the UE 102 may obtain N actual measurement result (s) based on the measurements in the respective OW instance. N is an integer and larger than zero. N may be the same as or different from L. For the PW instances 1, ..., Z, the UE 102 performs measurement prediction (s) 1, ..., Z based on the N actual measurement result (s) in OW instances 1, ..., Z, respectively. In each of the measurement prediction (s) 1, ..., Z, the UE 102 obtains L predicted measurement result (s) for the respective PW instance. For the respective PW instance, the L predicted measurement result (s) correspond to or are associated with the L actual measurement result (s) respectively. In some implementations, the UE 102 calculates the performance metric (s) based on the L predicted measurement result (s) and the L actual measurement result (s) for the PW instances 1, ..., Z. In other implementations, the UE 102 predicted whether the event occurs (i.e., occur or not occur) for each of the PW instances 1, ..., Z based on the measurements or L actual measurement results for the corresponding OW instance. The UE 102 also determines whether the event occurs (i.e., occur or not occur) for each of the PW instances 1, ..., Z based on the measurements or L actual measurement results for the respective PW instance. The UE 102 calculates the performance metric (s) based on the (occur or not occur) information for the PW instances 1, ..., Z.
[0085] The UE 102 transmits 426 one or more performance report (s) to the base station 106. In some implementations, the performance report (s) may include the performance metric (s) . The base station 106 may determine performance of the measurement prediction (s) is good or poor based on the performance metric (s) and / or performance evaluation criterion / criteria. In some implementations, the performance evaluation criterion / criteria includes one or inequalities and one or more threshold (s) . In some implementations, the threshold (s) are predetermined by the base station 106 or defined in a 3GPP specification. In some implementations, if the performance metric (s) is above or equal to the threshold (s) , the base station 106 determines performance of the measurement prediction (s) is good. Otherwise, ifthe performance metric (s) is below or equal to the threshold (s) , the base station 106 determines performance of the measurement prediction (s) is poor. In other implementations, if the performance metric (s) is below or equal to the threshold (s) , the base station 106 determines performance of the measurement prediction (s) is good. Otherwise, if the performance metric (s) is above or equal to the threshold (s) , the base station 106 determines performance of the measurement prediction (s) is poor.
[0086] In other implementations, the UE 102 may indicate performance of the measurement prediction (s) is good or poor in the performance report (s) . In some implementations, the UE 102 determines performance of the measurement prediction (s) is good or poor based on the performance metric (s) and / or performance evaluation criterion / criteria. In some implementations, the performance evaluation criterion / criteria includes one or inequalities and one or more threshold (s) . In some implementations, the performance monitoring configuration includes the threshold (s) . In other implementations, the threshold (s) are predetermined by the UE 102 or defined in a 3GPP specification. In some implementations, if the performance metric (s) is above or equal to the threshold (s) , the UE 102 determines performance of the measurement prediction (s) is good. Otherwise, if the performance metric (s) is below or equal to the threshold (s) , the UE 102 determines performance of the measurement prediction (s) is poor. In other implementations, if the performance metric (s) is below or equal to the threshold (s) , the UE 102 determines performance of the measurement prediction (s) is good. Otherwise, if the performance metric (s) is above or equal to the threshold (s) , the UE 102 determines performance of the measurement prediction (s) is poor.
[0087] In other implementations, the performance evaluation criterion / criteria includes one or inequalities, a first threshold, and a second threshold. In some implementations, if a first one of the performance metric (s) is above or equal to the first threshold and a second one of the performance metric (s) is below or equal to the second threshold, the UE 102 determines performance of the measurement prediction (s) is good. Otherwise, if the first performance metric is below or equal to the first threshold and the second performance metric is above or equal to the second threshold, the UE 102 determines performance of the measurement prediction (s) is poor.
[0088] In some implementations, the performance monitoring configuration includes one or more configuration parameters that configure a monitoring type. For example, the monitoring type may be an event-triggered monitoring or a periodic monitoring. In the case of the event-triggered monitoring, the performance monitoring configuration may include a first configuration (e.g. a monitoring event trigger configuration) configuring a first event for performing performance monitoring. Upon detecting occurrence of the first event, the UE 102 starts performing performance monitoring for the measurement prediction (s) . In one implementation, the first event is that a predicted measurement result for a cell (e.g., the serving cell or a neighbor cell) is an offset below or above an actual measurement result for the cell. The UE 102 obtains the actual measurement result associated or corresponding to the predicted measurement result. The predicted measurement result and the actual measurement result may be RSRP values, RSRQ values, or SINR values. The UE 102 may obtain the predicted measurement result for a first time instance as described above. The UE 102 may obtain the actual measurement result at the second time instance as described above or for Fig. 12. The first time instance may be earlier or later than the second time instance. In some implementations, the first time instance and the second time instance are within a time duration predetermined by the UE 102 or defined in a 3GPP specification.
[0089] In the case of the periodic monitoring, the performance monitoring configuration may configure the UE 102 to periodically perform the performance monitoring. Accordingly, the UE 102 periodically perform the performance monitoring. In one implementation, the performance monitoring configuration includes a second configuration configuring a periodicity for performing the performance monitoring. In some implementations, the periodicity is be N * (OW + PW) . In such implementations, the UE 102 may perform the performance monitoring for at least one PW (e.g., Mth, ..., (M+X) th PW (s) ) in every N consecutive (OW+PW) s and not perform the performance monitoring for the other PWs. M, X, and N are integers. 1 ≤ M ≤ M+X ≤ N. X ≥ 0 or may be omitted. In other implementations, the periodicity is units of time (e.g., milliseconds, seconds, slots, subframes, or frames) . In some alternative implementations, the performance monitoring configuration may not configure a monitoring type (e.g., the monitoring type described above) . In such cases, the UE 102 may perform performance monitoring for each instance of the PW.
[0090] In some implementations, the performance monitoring configuration includes a third configuration indicating a monitoring duration for the UE 102 to perform the performance monitoring. The UE 102 performs the performance monitoring for the monitoring duration. If the performance monitoring configuration does not configure a monitoring duration, the UE 102 may perform the performance monitoring for a monitoring duration determined by the UE 102 or defined in a 3GPP specification. In some implementations, the monitoring duration spans across multiple OWs and / or PWs. In one implementation, the monitoring duration is units of multiple {OW+PW} s. In another implementation, the monitoring duration is units of time (e.g., milliseconds, seconds, slots, subframes, or frames) . After the monitoring duration passes, the UE 102 stops performing the performance monitoring and / or suspends 430 performing measurements in the PW as described for event 414A. In some alternative implementations, if the performance monitoring configuration does not configure a / the monitoring duration, the UE 102 may stop performing the performance monitoring upon receiving an indication of stopping performance monitoring as described for event 428. In some implementation, upon stopping performing the performance monitoring, the UE 102 obtains one or more performance metric (s) and / or transmits a performance report including the performance metric (s) to the base station 106. The performance report may be the performance report of event 426 or another performance report other than the performance report of event 426.
[0091] In some implementations, the performance monitoring configuration includes one or more configuration parameters configuring a performance reporting type. For example, the performance reporting type may be an event-triggered reporting or a periodic reporting. In the case of the event-triggered reporting, the performance monitoring configuration may include a fourth configuration (e.g. a performance event trigger configuration) configuring a reporting event for reporting the performance metric (s) . In such cases, the UE 102 transmits 426 the performance report upon detecting occurrence of the reporting event. In one implementation, the reporting event is that the performance metric (s) is / are worse than one or more performance threshold (s) . The fourth configuration may configure the performance threshold (s) . In the case of the periodic reporting, the performance monitoring configuration may configure the UE 102 to periodically transmit a performance report. Accordingly, the UE 102 periodically transmits 426 the performance reports including respective performance metric (s) that the UE 102 obtains in each periodicity. In one implementation, the performance monitoring configuration includes a fifth configuration configuring a reporting periodicity for the UE 102 to periodically report performance metric (s) for the measurement prediction (s) . In some implementations, the reporting periodicity is be N * (OW + PW) . N is an integer and 1 ≤ N. In other implementations, the periodicity is a unit of time (e.g., milliseconds, seconds, slots, subframes, or frames) . In some implementations, the fifth configuration may indicate a number of periods. After the number of periods pass, the UE 102 stops performing the performance monitoring. If the fifth configuration does not indicate the number of periods, the UE 102 may not stop performing the performance monitoring until receiving an indication of stopping the performance monitoring as event 428.
[0092] After receiving the performance report (s) transmitted by the UE 102 as described above, the base station 106 may transmit 428 an indication of stopping the performance monitoring to the UE 102. In one implementation, the UE 102 transmits 428 the indication in a DL message (e.g., a RRC reconfiguration message) to the UE 102. In another implementation, the UE 102 transmits 428 the indication in a DL MAC control element (CE) to the UE 102. The UE 102 stops performing the performance monitoring in response to receiving 428 the indication. Upon stopping performing the performance monitoring or receiving 428 the indication, the UE 102 may suspend 430 performing measurements and / or perform (new) measurement prediction (s) in the PW (i.e., PW instances after event 430) as described for event 414A. Based on the new measurement prediction (s) , the UE 102 may transmit a prediction report the base station 106 as described above. In some implementations, the indication 428 may configure the UE 102 to release the performance monitoring configuration. In response, the UE 102 releases the performance monitoring configuration. In other implementations, the indication 428 does not configure the UE 102 to release the performance monitoring configuration.
[0093] Based on the performance report (s) , the base station 106 may determine whether to configure the UE 102 to stop performing the measurement prediction (s) or release the measurement prediction configuration. If performance of the measurement prediction (s) is poor (as indicated in or determined based on the performance report (s) as described above) , the base station 106 transmits an indication of stopping performing the measurement prediction (s) or releasing the measurement prediction configuration to the UE 102. In one implementation, the UE 102 transmits the indication in a DL message (e.g., a RRC reconfiguration message) to the UE 102. In another implementation, the UE 102 transmits the indication in a DL MAC control element (CE) to the UE 102. The UE 102 stops performing the measurement prediction (s) in response to receiving the indication. Otherwise, if performance of the measurement prediction (s) is good (as indicated in or determined based on the performance report (s) as described above) , the base station 106 may transmit 428 the indication.
[0094] In some implementations, the base station 106 determines to release the measurement prediction configuration as described above. In such cases, the base station 106 transmits 432 an indication to the UE 102, configuring the UE 102 to release the measurement prediction configuration. In response to the indication 432, the UE 102 releases the measurement prediction configuration. Later in time, the base station 106 may perform 490B a measurement prediction configuration and reporting procedure with the UE 102, similar to the procedure 490A.
[0095] In some implementations, if the UE 102 receives 432 the indication and does not receive 430 the indication, the UE 102 stops performing the performance monitoring. Alternatively, if the UE 102 receives 432 the indication and does not receive 430 the indication, the UE 102 continues performing the performance monitoring as described above.
[0096] In some implementations, there are different types of performance metrics to evaluate performance of the measurement prediction (s) configured by the measurement prediction configuration. The types (i.e., performance metric types) may include confusion matrix, accuracy, precision, recall (sensitive or true positive rate) , F1 score, logarithmic loss, Mean Absolute Error (MAE) , Mean Squared Error (MSE) , Root MSE (RMSE) , R-squared (Coefficient of Determination) , Adjusted R-squared, Mean Absolute Percentage Error (MAPE) , and / or Root Mean Squared Logarithmic Error (RMSLE) . If the AI model for the measurement prediction (s) is a classification model, the types may include confusion matrix, accuracy, precision, recall (sensitive or true positive rate) , F1 score, and / or logarithmic loss. If the AI model for the measurement prediction (s) is a regression model, the types may include MAE, MSE, RMSE, R-squared (Coefficient of Determination) , MAPE, and / or Root Mean Squared Logarithmic Error (RMSLE) . In some implementations, the UE 102 obtains 424 the performance metric (s) in accordance with at least one of the performance metric types. In some implementations, the performance monitoring configuration includes one or more configurations configuring the at least one performance metric type. In other implementations, the UE 102 determines the at least one performance metric type.
[0097] In some implementations, the base station 106 transmits a DL message (e.g., RRC reconfiguration message) including an applicability reporting configuration (not shown in Fig. 4A) to the UE 102. In some implementations, the applicability reporting configuration may include the measurement prediction configuration. Alternatively, the application reporting configuration includes a portion of the measurement prediction configuration. In this case, the base station 106 may transmit the remaining portion of the measurement prediction configuration to the UE 102 at event 412A. Yet alternatively, the base station 106 transmits the applicability reporting configuration to the UE 102 after event 412A. Based on the applicability reporting configuration, the UE 102 determines whether the measurement prediction is applicable and transmits an applicability report (e.g., event 410) as described above. In some implementations, the base station 106 may include the performance monitoring configuration in the applicability reporting configuration. In such cases, the UE 102 may include, in the applicability report, the performance metric (s) for the measurement prediction. Based on the performance metric (s) , the base station 106 determines performance of the measurement prediction, as described above. Alternatively or additionally, the UE 102 may determine performance of the measurement prediction is good or poor based on the performance metric (s) as described above, and include, in the applicability report, an indication indicating the determined result (i.e., performance of the measurement prediction is good or poor) . Based on the performance metric (s) or the indication that performance of the measurement prediction is good or poor, the base station 106 may determine whether to activate or configure the measurement prediction for the UE 102. In some implementations, if the base station 106 determine to activate or configure the measurement prediction for the UE 102, the base station 106 may transmit the measurement prediction configuration or the remaining portion of the measurement prediction configuration to the UE 102 at event 412A. In other implementations, if the base station 106 determine to activate the measurement prediction for the UE 102, the base station 106 may transmit an AI activation command (not shown in Fig. 4A) to the UE 102, configure the UE 102 to activate the measurement prediction configured by the measurement prediction configuration. The AI activation command may include the first measurement prediction configuration ID. Based on the first measurement prediction configuration ID, the UE 102 identifies the measurement prediction configuration. In response to the AI activation command, the UE 102 performs the measurement prediction (s) based on the measurement prediction configuration as described above. The Al activation command may be a MAC CE or a RRC message. In some implementations, the applicability report is a MAC CE, a RRC message (e.g., a RRC reconfiguration complete message) or a UE assistance information message (e.g., UEAssisteanceInformation message) .
[0098] The events 412A, 414, 416, and 418A are collectively referred to in Fig. 4A as a measurement prediction configuration and reporting procedure 490A.
[0099] Fig. 4B depicts a scenario 400B similar to the scenario 400A. The differences between the scenarios 400A and 400B are described below. In the scenario 400B, the base station 106 includes the performance monitoring configuration in the measurement prediction configuration and transmits 412B the measurement prediction configuration to the UE 102. As described for event 418A of Fig. 4A, the UE 102 transmits 418B a prediction report to the base station 106. The UE 102 may obtain 424 the performance metric (s) as described for Fig. 4A and includes the performance metric (s) in the prediction report. The prediction report may be a RRC message defined in 3GPP specification v20.0.0 or later version. The prediction report may be a measurement report message (e.g., MeasurementReport message) .
[0100] In some alternative implementations, the base station 106 includes the measurement prediction configuration in the performance monitoring configuration. In other alternative implementations, the base station 106 directly configures the UE 102 to perform the measurement prediction in the performance monitoring configuration instead of using a measurement prediction configuration.
[0101] Fig. 5A depicts a scenario 500A similar to the scenario 400A. The differences between the scenarios 500A and 400A are described below. Event 520 is similar to event 420. In some implementations, the performance monitoring configuration of event 520 configures reporting of paired actual measurement result (s) and predicted measurement result (s) (i.e., network-based performance monitoring) . The base station 106 configures the network-based performance monitoring in the performance monitoring configuration based on the UE capabilities or the applicability report. In some implementations, the UE 102 indicates support of the network-based performance monitoring in the UE capabilities (e.g., measurement prediction performance monitoring and / or reporting capabilities) . The base station 106 configures the network-based performance monitoring in the performance monitoring configuration based on the support indication. In other implementations, the UE 102 indicates support of the network-based performance monitoring in the applicability report. In one implementation, the base station 106 configures the network-based performance monitoring in the performance monitoring configuration based on the support indication.
[0102] Based on (e.g., in response to receiving) the performance monitoring configuration, the UE 102 resumes 422 performing measurements in PW instances 1, ..., Z of the PW while performing measurement prediction (s) for the PW instances 1, ..., Z. As described for Fig. 4A, the UE 102 obtains L actual measurement result (s) and L predicted measurement result (s) as a pair for each of the PW instances 1, ..., Z. The UE 102 transmits 526 performance report (s) 1, ..., Z to the base station 106, each including the respective L actual measurement result (s) and L predicted measurement result (s) . Z and L are integers and larger than zero. The base station 106 obtains one or more performance metric (s) based on the L actual measurement result (s) and the L predicted measurement result (s) , as described the UE 102 for Fig. 4A. Based on the performance metric (s) , the base station 106 may determine whether to send 428 the indication and / or 430 the indication as described for Fig. 4A.
[0103] In some implementations, the UE 102 transmits 526 the performance reports in a periodic manner. The performance monitoring configuration may configure a periodicity and the UE 101 transmits 526 the performance reports with the periodicity. In other implementations, the UE 102 transmits each of the performance reports after obtaining the respective L actual measurement result (s) and L predicted measurement result (s) . In some other implementations, a performance report may include one or multiple pairs of actual measurement (s) and predicted measurement result (s) . In such cases, the UE 102 may transmit, to the base station 106, the Z pairs of L actual measurement result (s) and L predicted measurement result (s) in one or more performance reports where the number is less than Z. In some implementations, the UE 102 transmits the performance report (s) to the base station 106 after obtaining the Z pairs. The base station 106 obtains 525 metric (s) based on the actual measurement result (s) and the predicted measurement result (s) .
[0104] Fig. 5B depicts a scenario 500B similar to the scenarios 400A, 400B, and 500A. The differences among the scenarios 500B, 400A, 400B, and 500B are described below. In the scenario 500B, the base station 106 transmits 512 a measurement prediction configuration to the UE 102, including a performance monitoring configuration, similar to event 412B. In some implementations, the performance monitoring configuration configures a network-based performance monitoring as described for Fig. 5A. Based on the measurement prediction configuration and / or the performance monitoring configuration, the UE 102 transmits 518B prediction report (s) 1 , .., Z to the base station 106, each including the respective L actual measurement result (s) and L predicted measurement result (s) . Descriptions for the performance report (s) described for Fig. 5A can apply to the prediction report (s) .
[0105] Fig. 5C depicts a scenario 500C similar to the scenarios 400A, 400B, 500A, and 500C. The differences among the scenarios 500C, 400A, 400B, 500A, and 500B are described below. Based on the measurement prediction configuration and / or the performance monitoring configuration, the UE 102 transmits 518C prediction report (s) 1, ..., Z to the base station 106, each including the respective L predicted measurement result (s) . Based on the measurement prediction configuration and / or the performance monitoring configuration, the UE 102 transmits 534 measurement report (s) 1 , .., Z to the base station 106, each including the respective L actual measurement result (s) . In some implementations, the prediction report (s) 1 ..., Z and the measurement report (s) 1, ..., Z include association information 1, ..., Z, respectively. Thus, the base station 106 associates prediction report (s) 1, ..., Z with the measurement report (s) 1, ..., Z based on the association information 1, ..., Z, respectively. In some implementations, the association information 1, ..., Z are timestamps. In some implementations, the association information K include a timestamp where the UE 102 obtains the L actual measurement result (s) measurement result (s) or the L predicted measurement result (s) for the PW instance 1. K is an integer and 1 ≤ K ≤ Z. In one implementation, the timestamp can be an absolute time. In another implementation, the timestamp includes a SFN number and / or a slot number. In other implementations, the association information K is an ID or number identifying the PW instance K.
[0106] Next, several example methods, which can be implemented in a UE (e.g., the UE 102 in Figs. 1, 2, or 4A-5C) or a base station (e.g., the base station 104 or the base station 106 in Figs. 1, 2, or 4A-5C) , are discussed with reference to Figs. 6-11. Descriptions described for Figs. 1, 2, or 4A-5C can apply to Figs. 6-11. Generally speaking, similar events in Figs. 4A-11 are labeled with similar reference numbers that share two least significant digits, with differences discussed below where appropriate. For example, event 412A of Figs. 4A and 5A is similar to event 412B of Fig 4B, event 512 of Figs. 5B and 5C, event 612A of Figs. 6A and 7A, event 612B of Fig. 6B, event 712 of Figs. 7B and 7C, event 812A of Figs. 8A, 9A and, 10, event 812B of Fig. 8B, and event 912 of Figs. 9B and 9C. With the exception of the differences shown in the figures and discussed below, any of the other implementations discussed with respect to a particular event (e.g., for messaging and processing) may apply to events labeled with similar reference numbers in other figures.
[0107] Fig. 6A illustrates an example method 600A, which can be implemented by a BS. The method 600A begins at block 603, where the BS receives UE capabilities of a UE, including measurement prediction capabilities, and / or measurement prediction performance monitoring and / or reporting capabilities (e.g., event 402 or 404) . At block 612A, the BS transmits a measurement prediction configuration to the UE, configuring measurement prediction. At block 620, the BS transmits a performance monitoring configuration to the UE, configuring performance monitoring and / or reporting for the measurement prediction. At block 626, the BS receives a performance report from the UE, including at least one performance metric for the measurement prediction. The flow proceeds to block 628 and / or block 632. At block 628, the BS transmits a message to the UE, configuring to stop the performance monitoring and / or reporting. At block 632, the BS transmits a message to the UE, configuring to release the measurement prediction configuration.
[0108] In some implementations, the performance monitoring configuration includes a measurement prediction ID indicating that the performance monitoring configuration is associated with or applies to the measurement prediction configuration. The measurement prediction ID may identify the measurement prediction configuration. In some implementations, the performance report includes the measurement prediction ID so that the BS determines that the performance report is associated with the performance monitoring configuration. In some implementations, the BS transmits a measurement configuration to the UE, configuring the UE to perform measurements. In some implementations, the BS may include a measurement ID in the measurement prediction configuration, indicating the measurement prediction configuration is associated with or applies to the measurement configuration. The measurement ID identifies the measurement configuration. Additionally or alternatively, the prediction report may include a measurement reporting configuration ID and / or a measurement object ID. The measurement configuration may include the measurement reporting configuration ID and / or the measurement object ID.
[0109] Fig. 6B is a flow diagram of an example method 600B similar to the method 600A, except that the method 600B includes blocks 612B, and 618 instead of blocks 612A, 620 and 626. At block 612B, the BS transmits a measurement prediction configuration to the UE, configuring measurement prediction and performance monitoring and / or reporting for the measurement prediction. At block 618, the BS receives a prediction report from the UE, including at least one performance metric for the measurement prediction.
[0110] In some implementations, the prediction report includes the measurement prediction ID so that the BS determines that the prediction report is associated with the measurement prediction configuration. In some implementations, the prediction report includes the measurement ID so that the BS determines that the prediction report is associated with the measurement configuration.
[0111] Fig. 7A illustrates an example method 700A, which can be implemented by a BS. The method 700A begins at block 703, where the BS receives UE capabilities of a UE, including measurement prediction capabilities and / or measurement prediction performance monitoring and / or reporting capabilities (e.g. event 402 or 404) . The flow proceeds to block 612A described for Fig. 6A. At block 720, the BS transmits a performance monitoring configuration to the UE, configuring performance monitoring and / or reporting for the measurement prediction. At block 726, the BS receives at least one performance report from the UE, each including one or more tuples of {actual measurement result, predicted measurement result} . The flow proceeds to block 628 and / or 632.
[0112] Fig. 7B is a flow diagram of an example method 700B similar to the method 700A, except that the method 700B includes blocks 712 and 718B instead of blocks 612A, 720 and 726. At block 712, the BS transmits a measurement prediction configuration to the UE, configuring measurement prediction and performance monitoring and / or reporting for the measurement prediction. At block 718B, the BS receives at least one prediction report from the UE, each including one or more tuples of {actual measurement result, predicted measurement result} . The flow proceeds to block 628 and / or block 632.
[0113] Fig. 7C is a flow diagram of an example method 700C similar to the methods 700A and 700B, except that the method 700C includes blocks 718C and 734 instead of block 718B. At block 718C, the BS receives at least one prediction report from the UE, each including one or more predicted measurement results. At block 734, the BS receives at least one measurement report from the UE, each including one or more actual measurement results. In some implementations, the one or more actual measurement results correspond to the one or more predicted measurement results, respectively.
[0114] Fig. 8A illustrates an example method 800A, which can be implemented by a UE. The method 800A begins at block 802, where the UE transmits UE capabilities of the UE to a RAN, including measurement prediction capabilities and / or measurement prediction performance monitoring and / or reporting capabilities. At block 812A, the UE receives a measurement prediction configuration from the RAN, configuring measurement prediction. At block 820, the UE receives a performance monitoring configuration from the RAN, configuring performance monitoring and / or reporting for the measurement prediction. At block 826, the UE transmits a performance report to the RAN, including at least one performance metric for the measurement prediction. The flow proceeds to block 828 and / or block 832. At block 828, the UE stops the performance monitoring and / or reporting. At block 832, the UE releases the measurement prediction configuration.
[0115] Fig. 8B is a flow diagram of an example method 800B similar to the method 800A, except that the method 800B includes blocks 812B and 818 instead of blocks 812A, 820, and 826. At block 812B, the UE receives a measurement prediction configuration from the RAN, configuring measurement prediction and performance monitoring and / or reporting for the measurement prediction. At block 818, the UE transmits a prediction report to the RAN, including at least one performance metric for the measurement prediction.
[0116] Fig. 9A illustrates an example method 900A, which can be implemented by a UE. The method 900A begins at block 902, where the UE transmits UE capabilities of the UE to a RAN, including measurement prediction capabilities and / or measurement prediction performance monitoring and / or reporting capabilities. The flow proceeds to block 812A. At block 920, the UE receives a performance monitoring configuration from the RAN, configuring performance monitoring and / or reporting for the measurement prediction. At block 926, the UE transmits at least one performance report to the RAN, each including one or more tuples of {actual measurement result, predicted measurement result} . The flow proceeds to block 828 and / or block 832 described for Fig. 8A.
[0117] Fig. 9B is a flow diagram of an example method 900B similar to the method 900A, except that the method 900B includes blocks 912 and 918B instead of blocks 812A, 920, 926. At block 912, the UE receives a measurement prediction configuration from the RAN, configuring measurement prediction and performance monitoring and / or reporting for the measurement prediction. At block 918B, the UE transmits at least one prediction report to the RAN, each including one or more tuples of {actual measurement result, predicted measurement result} . The flow proceeds to block 828 and / or block 832.
[0118] Fig. 9C is a flow diagram of an example method 900C similar to the methods 900A and 900B, except that the method 900C includes blocks 918C and 934 instead of block 918B. At block 918C, the UE transmits at least one prediction report to the RAN, each including one or more predicted measurement results. At block 934, the UE receives at least one measurement report from the RAN, each including one or more actual measurement results.
[0119] Fig. 10 illustrates an example method 1000, which can be implemented by a UE. The method 1000 begins at block 812A described for Fig. 8A. At block 1008, the UE performs measurements in an observation window. At block 1016, the UE configures a lower layer to skip measurements in a prediction window and perform measurement prediction (s) in the prediction window. At block 1020, the UE receives, from the RAN, a performance monitoring configuration for the measurement prediction. Alternatively, the UE is preconfigured with the performance monitoring configuration. Yet alternatively, the UE receives the performance monitoring configuration from a server. The flow proceeds to block 910. At block 1022, the UE configures a lower layer to perform measurements in the prediction window.
[0120] Fig. 11 illustrates an example method 1100, which can be implemented by a UE. The method 1100 begins at block 1108, where the UE performs measurements in an observation window. At block 1116, the UE performs measurement predictions in a prediction window. At block 1134, the UE determines whether to perform performance monitoring. If the UE determines to perform performance monitoring (i.e., “Yes” branch of block 1134) , the flow proceeds to block 1136. At block 1136, the UE performs measurements in the prediction window. Otherwise, if the UE determines not to perform performance monitoring (i.e., “No” branch of block 1134) , the flow proceeds to block 1138. At block 1138, the UE skips performing measurements in the prediction window.
[0121] Fig. 12 illustrates a flowchart 1200 of a method of wireless communication at a UE. With reference to Figs. 4A-5C, the method may be performed by the UE 102. In embodiments, the UE 102 transmits 1202, to a network entity, an indication of a capability of the UE for at least one of: measurement, measurement prediction, measurement prediction performance monitoring, or reporting. For example, referring to Figs. 4A-5C, the UE 102 transmits 402, to the network entity 106, UE capabilities including a measurement prediction capability, a measurement prediction performance monitoring capability, and / or a reporting capability, etc.
[0122] The UE 102 receives 1220, from the network entity, a configuration configuring at least one of: measurement prediction of an ML model, or performance monitoring for the measurement prediction. For example, referring to Figs. 4A and 5A, the UE 102 receives 412A, 420, 520, from the network entity 106, separate measurement prediction and performance monitoring configurations. Referring to Figs. 4B and 5B, the UE 102 receives 412B, 512, from the network entity 106, a combined measurement prediction configuration and performance monitoring configuration. Referring to Fig. 5C, the UE 102 receives 512, from the network entity 106, a measurement prediction configuration without a performance monitoring configuration.
[0123] The UE 102 transmits 1226, to the network entity, a report associated with at least one of: the measurement prediction or the performance monitoring. For example, referring to Figs. 4A-5C, the UE 102 transmits 418A, 418B, 518B, 518C, to the network entity 106, a prediction report including, e.g., prediction report metric (s) , actual measurement result (s) , and / or predicted measurement result (s) . For example, referring to Figs. 4A and 5A, the UE 102 transmits 426, 526, to the network entity 106, a performance report including, e.g., performance report metric (s) , actual measurement result (s) , and / or predicted measurement result (s) .
[0124] The UE 102 receives 1228, from the network entity, a first indication to stop the performance monitoring. For example, referring to Figs. 4A-5C, the UE 102 receives 428, from the network entity 106, an indication of stopping the performance monitoring.
[0125] The UE 102 receives 1232, from the network entity, a second indication to release a measurement prediction configuration in the configuration. For example, referring to Figs. 4A-5C, the UE 102 receives 432, from the network entity 106, an indication of releasing the measurement prediction configuration. Fig. 12 describes a method from a UE-side of a wireless communication link, whereas Fig. 13 describes a method from a network-side of the wireless communication link.
[0126] Fig. 13 is a flowchart 1300 of a method of wireless communication at a network entity. With reference to Figs. 4A-5C, the method may be performed by one or more network entities, which may correspond to a base station or a unit of the base station, such as an RU, a DU, and / or a CU. In embodiments, the network entity 106 receives 1202, from the UE, an indication of a capability of the UE for at least one of: measurement, measurement prediction, measurement prediction performance monitoring, or reporting. For example, referring to Figs. 4A-5C, the network entity 106 receives 402, from the UE 102, UE capabilities including a measurement prediction capability, a measurement prediction performance monitoring capability, and / or a reporting capability, etc.
[0127] The network entity 106 transmits 1220, to the UE, a configuration configuring at least one of: measurement prediction of an ML model, or performance monitoring for the measurement prediction. For example, referring to Figs. 4A and 5A, the network entity 106 transmits 412A, 420, 520, to the UE 102, separate measurement prediction and performance monitoring configurations. Referring to Figs. 4B and 5B, the network entity 106 transmits 412B, 512, to the UE 102, a combined measurement prediction configuration and performance monitoring configuration. Referring to Fig. 5C, the network entity 106 transmits 512, to the UE 102, a measurement prediction configuration without a performance monitoring configuration.
[0128] The network entity 106 receives 1226, from the UE, a report associated with at least one of: the measurement prediction or the performance monitoring. For example, referring to Figs. 4A-5C, the network entity 106 receives 418A, 418B, 518B, 518C, from the UE 102, a prediction report including, e.g., prediction report metric (s) , actual measurement result (s) , and / or predicted measurement result (s) . For example, referring to Figs. 4A and 5A, the network entity 106 receives 426, 526, from the UE 102, a performance report including, e.g., performance report metric (s) , actual measurement result (s) , and / or predicted measurement result (s) .
[0129] The network entity 106 transmits 1228, to the UE, a first indication to stop the performance monitoring. For example, referring to Figs. 4A-5C, the network entity 106 transmits 428, to the UE 102, an indication of stopping the performance monitoring.
[0130] The network entity 106 transmits 1232, to the UE, a second indication to release a measurement prediction configuration in the configuration. For example, referring to Figs. 4A-5C, the network entity 106 transmits 432, to the UE 102, an indication of releasing the measurement prediction configuration.
[0131] Fig. 14 illustrates a schematic diagram 1400 of modules, components or circuitry that can provide an example implementation for a UE to perform reference signal measurements. When the UE is in RRC_CONNECTED mode, the UE measures multiple beams 1462 (or at least one beam) of a cell. The UE then averages the measurements results or power values to derive the cell quality. In doing so, the UE is configured to consider a subset of the detected beams. Filtering 1464a-146c takes place at two different levels: at the physical layer filtering 1464a is done to derive beam quality and then filtering 1464b-1464c is done at RRC level to derive cell quality from multiple beams. The UE derives cell quality from beam measurements in the same way for the serving cell (s) and for the non-serving cell (s) . The gNB may configure the UE to provide measurement reports 1466 that contain the measurement results of the X best beams. K beams correspond to the measurements on SSB or CSI-RS resources configured for L3 mobility by gNB and detected by UE at L1. The control points / blocks / modules shown in Figure 14 are further described below.
[0132] Point A shows measurements (beam specific samples) internal to the physical layer for internal layer 1 filtering of the inputs measured at point A. Exact filtering is implementation dependent. How the measurements are actually executed in the physical layer by an implementation (inputs A and Layer 1 filtering) is not constrained by the standard. Point A1 shows measurements (i.e. beam specific measurements) reported by layer 1 to layer 3 after layer 1 filtering.
[0133] For beam consolidation / selection, beam specific measurements are consolidated to derive cell quality. The behavior of the beam consolidation / selection is standardized and the configuration of this module is provided by RRC signaling. The reporting period at point B equals one measurement period at A1. At point B, a measurement (i.e. cell quality) is derived from beam-specific measurements reported to layer 3 after beam consolidation / selection, followed by Layer 3 filtering for cell quality, where filtering is performed on the measurements provided at point B. The behavior of the Layer 3 filters is standardized and the configuration of the layer 3 filters is provided by RRC signaling. The filtering reporting period at point C equals one measurement period at B, where C corresponds to a measurement after processing in the layer 3 filter. The reporting rate is the same as the reporting rate at point B. This measurement is used as input for one or more evaluation of reporting criteria. The evaluation of reporting criteria checks whether actual measurement reporting is necessary at point D. The evaluation can be based on more than one flow of measurements at reference point C, e.g., to compare between different measurements. This is illustrated by input C and C1. The UE evaluates the reporting criteria at least every time a new measurement result is reported at point C, C1. The reporting criteria are standardized and the configuration is provided by RRC signaling (UE measurements) . Point D corresponds to measurement report information (message) sent on the radio interface.
[0134] For L3 Beam filtering, filtering is performed on the measurements (i.e. beam specific measurements) provided at point A1. The behavior of the beam filters is standardized and the configuration of the beam filters is provided by RRC signaling. The filtering reporting period at E equals one measurement period at A1, where E corresponds to a measurement (i.e. beam-specific measurement) after processing in the beam filter. The reporting rate is the same as the reporting rate at point A1. This measurement is used as input for selecting the X measurements to be reported. Beam selection for beam reporting is based on selecting the X measurements from the measurements provided at point E. The behavior of the beam selection is standardized and the configuration of this module is provided by RRC signaling. At point F, the beam measurement information is included in a measurement report (sent) on the radio interface.
[0135] Layer 1 filtering introduces a certain level of measurement averaging. How and when the UE exactly performs the required measurements is implementation specific to the point that the output at B fulfils the performance requirements set in 3GPP TS 38.133. Layer 3 filtering for cell quality and related parameters used are specified in 3GPP TS 38.331 and do not introduce any delay in the sample availability between B and C. Measurement at point C, C1 is the input used in the event evaluation. L3 Beam filtering and related parameters used are specified in TS 38.331 and do not introduce any delay in the sample availability between E and F.
[0136] The following description may be applied to the description above.
[0137] Generally speaking, description for one of the above figures can apply to another of the above figures. Examples, implementations and methods described above can be combined, if there is no conflict. An event or block described above can be optional or omitted. For example, an event or block with dashed lines in the figures can be optional. The description described from the perspective of the receiving node also applies to the sending node. For example, a description that a receiving node (e.g., DU) receives a message from a sending node (e.g., CU) may be replaced by the sending node sending a message to the receiving node. Similarly, a description that a receiving node (e.g., CU) receives a message from a sending node (e.g., DU) may be replaced by the sending node sending a message to the receiving node.
[0138] In some implementations, “message” is used and can be replaced by “information element (IE) ” , and vice versa. In some implementations, “IE” is used and can be replaced by “field” , and vice versa. In some implementations, “configuration” can be replaced by “configurations” or “configuration parameters” , and vice versa. In some implementations, the “indication” can be replaced by “indicator” , and vice versa. In some implementations, the “measurement prediction” can be replaced by “measurement prediction function” , “measurement AI / ML inference” . In some implementations, the “non-predicted” can be replaced by “actual” , or vice versa. In some implementations, the “performance monitoring” can be replaced by “performance monitoring and reporting” .
[0139] A user device in which the techniques of this disclosure can be implemented (e.g., the UE 102) can be any suitable device capable of wireless communications such as a smartphone, a tablet computer, a laptop computer, a mobile gaming console, a point-of-sale (POS) terminal, a health monitoring device, a drone, a camera, a media-streaming dongle or another personal media device, a wearable device such as a smartwatch, a wireless hotspot, a femtocell, or a broadband router. Further, the user device in some cases may be embedded in an electronic system such as the head unit of a vehicle or an advanced driver assistance system (ADAS) . Still further, the user device can operate as an internet-of-things (IoT) device or a mobile-internet device (MID) . Depending on the type, the user device can include one or more general-purpose processors, a computer-readable memory, a user interface, one or more network interfaces, one or more sensors, etc.
[0140] Certain embodiments are described in this disclosure as including logic or a number of components or modules. Modules may can be software modules (e.g., code, or machine-readable instructions stored on non-transitory machine-readable medium) or hardware modules. A hardware module is a tangible unit capable of performing certain operations and may be configured or arranged in a certain manner. A hardware module can comprise dedicated circuitry or logic that is permanently configured (e.g., as a special-purpose processor, such as a field programmable gate array (FPGA) or an application-specific integrated circuit (ASIC) , a digital signal processor (DSP) , etc. ) to perform certain operations. A hardware module may also comprise programmable logic or circuitry (e.g., as encompassed within a general-purpose processor or other programmable processor) that is temporarily configured by software to perform certain operations. The decision to implement a hardware module in dedicated and permanently configured circuitry, or in temporarily configured circuitry (e.g., configured by software) may be driven by cost and time considerations.
[0141] When implemented in software, the techniques can be provided as part of the operating system, a library used by multiple applications, a particular software application, etc. The software can be executed by one or more general-purpose processors or one or more special-purpose processors.
[0142] Upon reading this disclosure, those of skill in the art will appreciate still additional and alternative structural and functional designs for handling mobility between base stations through the principles disclosed herein. Thus, while particular embodiments and applications have been illustrated and described, it is to be understood that the disclosed embodiments are not limited to the precise construction and components disclosed herein. Various modifications, changes and variations, which will be apparent to those of ordinary skill in the art, may be made in the arrangement, operation and details of the method and apparatus disclosed herein without departing from the spirit and scope defined in the appended claims.
[0143] The following examples are illustrative only and may be combined with other examples or teachings described herein, without limitation.
[0144] Example 1 is a method of wireless communication at a UE, including: receiving, from a network entity, a configuration configuring: measurement prediction of an ML model; and performance monitoring for the measurement prediction; and transmitting, to the network entity, a report associated with at least one of: the measurement prediction or the performance monitoring.
[0145] Example 2 may be combined with Example 1 and includes that the receiving the configuration comprises: receiving, from the network entity, a measurement prediction configuration configuring the measurement prediction of the ML model; and receiving, from the network entity, a performance monitoring configuration configuring the performance monitoring, or wherein the receiving the configuration comprises: receiving, from the network entity, the measurement prediction configuration configuring: the measurement prediction of the ML model; and the performance monitoring for the measurement prediction.
[0146] Example 3 may be combined with any of Examples 1-2 and further includes transmitting, to the network entity, an indication of a capability of the UE for at least one of: measurement; measurement prediction; measurement prediction performance monitoring; or reporting.
[0147] Example 4 may be combined with any of Examples 1-3 and further includes receiving, from the network entity, a measurement configuration comprising at least one of: a measurement object configuration; a measurement reporting configuration; a measurement ID; a measurement object ID; or a measurement reporting configuration ID; and performing a measurement in an observation window based on the measurement configuration.
[0148] Example 5 may be combined with Example 4 and further includes performing the measurement prediction in a PW; determining to perform the performance monitoring; and upon positive determination, performing a second measurement in the PW; or upon negative determination, skipping the second measurement in the PW.
[0149] Example 6 may be combined with any of Examples 4-5 and further includes transmitting, to the network entity, an applicability report indicating that the measurement prediction is applicable, wherein the receiving the configuration is based on the transmitted applicability report.
[0150] Example 7 may be combined with any of Examples 1-6 and further includes suspending performance of measurements in the PW based on the configuration; and performing the measurement prediction in the PW.
[0151] Example 8 may be combined with Example 7 and further includes predicting, based on the measurement prediction, an event, wherein the transmitting the report comprises transmitting a prediction report based on the event.
[0152] Example 9 may be combined with Example 8 and further includes resuming the performance of the measurements in the PW; and obtaining a performance metric for the measurement prediction, wherein the report includes the performance metric.
[0153] Example 10 may be combined with any of Examples 1-9 and further includes receiving, from the network entity, a first indication to stop the performance monitoring; and suspending the performance of the measurements in the period PW (392, 396) based on the first indication.
[0154] Example 11 may be combined with any of Examples 1-10 and further includes receiving, from the network entity, a second indication to release a measurement prediction configuration in the configuration.
[0155] Example 12 may be combined with any of Examples 1-11 and includes that the transmitting the report comprises at least one of: transmitting, to the network entity, a performance report including an actual measurement result and a predicted measurement result; transmitting, to the network entity, a prediction report including the actual measurement result and the predicted measurement result; or transmitting, to the network entity, a first prediction report including the predicted measurement result; and transmitting, to the network entity (106) , a measurement report including the actual measurement result.
[0156] Example 13 may be combined with any of Examples 1-12 and includes that the configuration includes at least one of: an indication of indirect measurement prediction; an indication of direct measurement prediction; prediction domain information; a measurement ID indicating that a measurement prediction configuration is associated with at least one of: a measurement configuration or a measurement object; a report configuration ID; prediction type information; an indication of cluster based prediction; a measurement prediction configuration ID; an indication of non-sliding prediction; an indication of sliding prediction; an indication of an observation window; an indication of a PW; a prediction report configuration configuring an event for prediction; a TTT; an indication of a supported performance metric; performance evaluation criteria including an inequality and a threshold; an indication of an event for triggering the performance monitoring; a monitoring type for the performance monitoring; a periodicity for the performance monitoring; a monitoring duration for the performance monitoring; a reporting event for reporting a performance metric; a reporting periodicity.
[0157] Example 14 is a method of wireless communication at a network entity, comprising: transmitting, to a UE, a configuration configuring: measurement prediction of an ML model; and performance monitoring for the measurement prediction; and receiving, from the UE, a report associated with at least one of: the measurement prediction or the performance monitoring.
[0158] Example 15 may be combined Example 14 and includes that the transmitting the configuration comprises: transmitting, to the UE, a measurement prediction configuration configuring the measurement prediction of the ML model; and transmitting, to the UE, a performance monitoring configuration configuring the performance monitoring, or wherein the transmitting the configuration comprises: transmitting, to the UE, the measurement prediction configuration configuring: the measurement prediction of the ML model; and the performance monitoring for the measurement prediction.
[0159] Example 16 may be combined with any of Examples 14-15 and further includes receiving, from at least one of: the UE, a RAN node, or a CN node, an indication of a capability of the UE for at least one of: measurement; measurement prediction; measurement prediction performance monitoring; or reporting.
[0160] Example 17 may be combined with any of Examples 14-16 and further includes transmitting, to the UE, a measurement configuration comprising at least one of: a measurement object configuration; a measurement reporting configuration; a measurement ID; a measurement object ID; or a measurement reporting configuration ID.
[0161] Example 18 may be combined with any of Examples 14-17 and further includes receiving, from the UE, an applicability report indicating that the measurement prediction is applicable, wherein the transmitting the configuration is based on the transmitted applicability report.
[0162] Example 19 may be combined with any of Examples 14-18 and includes that the receiving the report comprises receiving a prediction report based on an event.
[0163] Example 20 may be combined with any of Examples 14-19 and further includes transmitting, to the UE, a first indication to stop the performance monitoring.
[0164] Example 21 may be combined with any of Examples 14-20 and further includes transmitting, to the UE, a second indication to release a measurement prediction configuration in the configuration.
[0165] Example 22 may be combined with any of Examples 14-21 and includes that the receiving the report comprises at least one of: receiving, from the UE, a performance report including an actual measurement result and a predicted measurement result; receiving, from the UE, a prediction report including the actual measurement result and the predicted measurement result; or receiving, from the UE, a first prediction report including the predicted measurement result; and receiving, from the UE, a measurement report including the actual measurement result.
[0166] Example 23 may be combined with Example 22 and further includes obtaining a performance metric based on the actual measurement result and the predicted measurement result.
[0167] Example 24 may be combined with any of Examples 14-23 and includes that the configuration includes at least one off an indication of indirect measurement prediction; an indication of direct measurement prediction; prediction domain information; a measurement ID indicating that a measurement prediction configuration is associated with at least one of: a measurement configuration or a measurement object; a report configuration ID; prediction type information; an indication of cluster based prediction; a measurement prediction configuration ID; an indication of non-sliding prediction; an indication of sliding prediction; an indication of an observation window; an indication of a PW; a prediction report configuration configuring an event for prediction; a TTT; an indication of a supported performance metric; performance evaluation criteria including an inequality and a threshold; an indication of an event for triggering the performance monitoring; a monitoring type for the performance monitoring; a periodicity for the performance monitoring; a monitoring duration for the performance monitoring; a reporting event for reporting a performance metric; a reporting periodicity.
[0168] Example 25 is an apparatus for wireless communication for implementing a method as in any of Examples 1-24.
[0169] Example 26 is an apparatus for wireless communication including means for implementing a method as in any of Examples 1-24.
[0170] Example 27 is a non-transitory computer-readable medium storing computer executable code, the code when executed by a processor causes the processor to implement a method as in any of Examples 1-24.
[0171] Example 28 is a computer program product for implementing a method as in any of Examples 1-24.
Claims
1.A method of wireless communication performed at a user equipment, UE, (102) , comprising:receiving (412A, 412B, 420, 512, 520) , from a network entity (106) , a configuration configuring:measurement prediction of a machine learning, ML, model; andperformance monitoring for the measurement prediction; andtransmitting (418A, 418B, 426, 518B, 518C, 526, 534) , to the network entity (106) , a report associated with at least one of: the measurement prediction or the performance monitoring.2.The method of claim 1, wherein the receiving (412A, 412BB, 420, 512, 520) the configuration comprises:receiving (412A) , from the network entity (106) , a measurement prediction configuration configuring the measurement prediction of the ML model; andreceiving (420, 520) , from the network entity (106) , a performance monitoring configuration configuring the performance monitoring, orwherein the receiving (412A, 412B, 420, 512, 520) the configuration comprises:receiving (412B, 512) , from the network entity (106) , the measurement prediction configuration configuring:the measurement prediction of the ML model; andthe performance monitoring for the measurement prediction.3.The method of any of claims 1-2, further comprising:transmitting (402) , to the network entity (106) , an indication of a capability of the UE (102) for at least one of:measurement;measurement prediction;measurement prediction performance monitoring; orreporting.4.The method of any of claims 1-3, further comprising:receiving (406) , from the network entity (106) , a measurement configuration comprising at least one of:a measurement object configuration;a measurement reporting configuration;a measurement identity, ID;a measurement object ID; ora measurement reporting configuration ID; andperforming (408) a measurement in an observation window (390) based on the measurement configuration.5.The method of claim 4, further comprising:performing (1116) the measurement prediction in a prediction window, PW, (392, 396) ;determining (1134) to perform the performance monitoring; andupon positive determination, performing (1136) a second measurement in the PW; orupon negative determination, skipping (1138) the second measurement in the PW.6.The method of any of claims 4-5, further comprising:transmitting (410) , to the network entity (106) , an applicability report indicating that the measurement prediction is applicable, wherein the receiving (412A, 412B, 420, 512, 520) the configuration is based on the transmitted applicability report.7.The method of any of claims 1-6, further comprising:suspending (414A, 414B) performance of measurements in the PW (392, 396) based on the configuration; andperforming (414A, 414B) the measurement prediction in the PW (392, 396) .8.The method of claim 7, further comprising:predicting (416) , based on the measurement prediction, an event, wherein the transmitting (418A, 418B, 426, 518B, 518C, 526, 534) the report comprises transmitting (418A) a prediction report based on the event.9.The method of claim 8, further comprising:resuming (422) the performance of the measurements in the PW (392, 396) ; andobtaining (424) a performance metric for the measurement prediction, wherein the report includes the performance metric.10.The method of any of claims 1-9, further comprising:receiving (428) , from the network entity (106) , a first indication to stop the performance monitoring; andsuspending (430) the performance of the measurements in the period PW (392, 396) based on the first indication.11.The method of any of claims 1-10, further comprising:receiving (432) , from the network entity (106) , a second indication to release a measurement prediction configuration in the configuration.12.The method of any of claims 1-11, wherein the transmitting (418A, 418B, 426, 518B, 518C, 526, 534) the report comprises at least one of:transmitting (526) , to the network entity (106) , a performance report including an actual measurement result and a predicted measurement result;transmitting (518B) , to the network entity (106) , a prediction report including the actual measurement result and the predicted measurement result; ortransmitting (518C) , to the network entity (106) , a first prediction report including the predicted measurement result; andtransmitting (534) , to the network entity (106) , a measurement report including the actual measurement result.13.The method of any of claims 1-12, wherein the configuration includes at least one of:an indication of indirect measurement prediction;an indication of direct measurement prediction;prediction domain information;a measurement ID indicating that a measurement prediction configuration is associated with at least one of: a measurement configuration or a measurement object;a report configuration ID;prediction type information;an indication of cluster based prediction;a measurement prediction configuration ID;an indication of non-sliding prediction;an indication of sliding prediction;an indication of an observation window;an indication of a PW;a prediction report configuration configuring an event for prediction;a time to trigger, TTT;an indication of a supported performance metric;performance evaluation criteria including an inequality and a threshold;an indication of an event for triggering the performance monitoring;a monitoring type for the performance monitoring;a periodicity for the performance monitoring;a monitoring duration for the performance monitoring;a reporting event for reporting a performance metric;a reporting periodicity.14.A method of wireless communication at a network entity (106) , comprising:transmitting (412A, 412B, 420, 512, 520) , to a user equipment, UE, (102) , a configuration configuring:measurement prediction of a machine learning, ML, model; andperformance monitoring for the measurement prediction; andreceiving (418A, 418B, 426, 518B, 518C, 526, 534) , from the UE (102) , a report associated with at least one of: the measurement prediction or the performance monitoring.15.The method of claim 14, wherein the transmitting the configuration comprises:transmitting (412A) , to the UE (102) , a measurement prediction configuration configuring the measurement prediction of the ML model; andtransmitting (420, 520) , to the UE (102) , a performance monitoring configuration configuring the performance monitoring.16.The method of claim 14, wherein the transmitting the configuration comprises:transmitting (412B, 512) , to the UE (102) , the measurement prediction configuration configuring:the measurement prediction of the ML model; andthe performance monitoring for the measurement prediction.17.An apparatus for wireless communication comprising a memory, a transceiver, and a processor coupled to the memory and the transceiver, the apparatus being configured to implement a method as in any of claims 1-16.
