Method and apparatus for measurement event prediction in mobile communications

Abstract

Various solutions for measurement event prediction in mobile communications are described. A user equipment (UE) can perform measurements on a first set of radio resources. The UE can obtain predicted measurements of a second set of radio resources based on the measurements by using an artificial intelligence (AI) or machine learning (ML) model in one or a combination of time, space, and frequency domains. The UE can send a measurement report to a network node based on the measurements of the first set of radio resources and / or the predicted measurements.

Description

[0001] Cross-referencing

[0002] This invention is part of a non-provisional application claiming priority to PCT application PCT / CN2024 / 084471 filed on March 28, 2024, the contents of which are incorporated herein by reference in their entirety. Technical Field

[0003] This disclosure generally relates to mobile communications, and more specifically, to the prediction of measurement events based on artificial intelligence (AI) or machine learning (ML) regarding user equipment and network devices in mobile communications. Background Technology

[0004] Unless otherwise stated herein, the methods described in this section are not prior art as claimed in the following claims, and are not considered prior art because they are included in this section.

[0005] In 3GPP 5G New Radio (NR) networks, User Equipment (UE) mobility is a critical consideration. To maintain service continuity and quality, UEs measure and report the environmental quality of radio resource signals. However, traditional handover (HO) designs are controlled by Layer 3 (L3) processes, including Radio Resource Management (RRM) measurements and Radio Resource Control (RRC) reconfiguration, which introduces significant signaling and latency. To address this issue, Low Layer Triggered Mobility (LTM, also known as Layer 1 (L1) / Layer 2 (L2) Triggered Mobility) was introduced to enable serving cell handover through beam management of L1 / L2 signaling, thereby reducing latency, overhead, and interruptions during UE movement.

[0006] On the other hand, with the increasing integration of artificial intelligence (AI) or machine learning (ML) to enhance network performance, leveraging AI / ML to further improve mobility processes, such as enabling proactive measurement reporting (MR) and handover, and / or reducing unnecessary handovers, has become an important issue. Summary of the Invention

[0007] The following abstract is for illustrative purposes only and is not intended to be limiting in any way. That is, the abstract aims to introduce the concepts, key points, benefits, and advantages of the novel and non-obvious techniques described herein. Detailed descriptions will further illustrate these advantages in the following description. Therefore, the following abstract is not intended to define the essential characteristics of the claimed subject matter, nor is it intended to define the scope of the claimed subject matter.

[0008] One objective of this disclosure is to propose a scheme or method for solving the problem of predicting measurement events based on artificial intelligence (AI) or machine learning (ML) in relation to user equipment (UE) and network devices in mobile communications.

[0009] In one aspect, a method may include measuring a first set of radio resources by a device. The method may further include the device, based on the measurement, performing measurement prediction using an AI / ML-based model in one or a combination of the time, spatial, and frequency domains to obtain predicted measurement results for a second set of radio resources. The method may also include the device sending a measurement report to a network node based on the measurement results of the first set of radio resources and / or the predicted measurement results.

[0010] In another aspect, an apparatus may include a transceiver that wirelessly communicates with a network during operation. The apparatus may also include a processor communicatively connected to the transceiver. The processor may perform operations during operation, including measuring a first set of radio resources. The processor may also perform operations during operation, including, based on the measurements, performing measurement predictions using an artificial intelligence (AI) or machine learning (ML) model in one or a combination of the time, spatial, and frequency domains to obtain predicted measurement results for a second set of radio resources. The processor may also perform operations during operation, including, via the transceiver, transmitting a measurement report to a network node based on the measurement results of the first set of radio resources and / or the predicted measurement results.

[0011] In another aspect, a method may include having a device perform an AI / ML-based prediction. The method may further include having the device initiate handover (HO) preparation if the measurement event is predicted to occur based on the AI / ML-based prediction result. The method may also include having the device receive an HO command from a network node when the triggering conditions of the measurement event are met. The method may further include having the device perform an HO operation according to the HO command.

[0012] In another aspect, a method may include receiving a measurement report from the UE by a network node. This measurement report is generated based on AI / ML prediction results. The method may also include making a HO (Housing and Optimization) decision by the network node based on the measurement report. Furthermore, the method may include sending an HO command to the UE by the network node based on the HO decision.

[0013] It is worth noting that although the content described herein may be presented in the context of certain wireless access technologies, networks, and network topologies, such as LTE, LTE-Advanced, LTE-Advanced Pro, 5G, NR, 5G-Advanced, Internet of Things (IoT), Narrowband Internet of Things (NB-IoT), Industrial Internet of Things (IIoT), B5G, and 6G, the proposed concepts, schemes, and any variations / derivatives thereof can be implemented, used, and implemented by other types of wireless access technologies, networks, and network topologies. Therefore, the scope of this disclosure is not limited to the examples described herein. Attached Figure Description

[0014] The accompanying drawings are included in this specification to further understand this disclosure and form part of this disclosure. The drawings illustrate embodiments of this disclosure and, together with the specification, serve to explain the principles of this disclosure. It will be understood that the drawings are not necessarily drawn to scale, as some components may be shown out of proportion to their actual dimensions for clarity in illustrating the concepts of this disclosure.

[0015] Figure 1 The illustration depicts an example scenario of a communication environment that can be implemented according to this disclosure.

[0016] Figure 2 The illustration depicts a predictive example scenario based on an implementation of this disclosure.

[0017] Figure 3 This diagram illustrates the current measurement process.

[0018] Figures 4A-4E The illustration depicts an example of a measurement process that incorporates an artificial intelligence (AI) or machine learning (ML) prediction model, according to an implementation of this disclosure.

[0019] Figure 5 The illustration depicts an example flow of advance measurement reporting according to an embodiment of this disclosure.

[0020] Figure 6 For illustrative purposes, an example flow for predicting measurement events during the trigger time (TTT) and prediction period is described according to an embodiment of this disclosure.

[0021] Figure 7 The diagram illustrates an example communication system implemented according to this disclosure.

[0022] Figure 8 The flowchart illustrates an example process implemented according to this disclosure.

[0023] Figure 9 The flowchart illustrates another example process implemented according to this disclosure.

[0024] Figure 10 The flowchart illustrates yet another example process implemented according to this disclosure. Detailed Implementation

[0025] Detailed embodiments and implementations of the claimed subject matter are disclosed herein. However, it should be understood that the disclosed embodiments and implementations are merely illustrative of the claimed subject matter, which can be embodied in various forms. This disclosure can be embodied in many different forms and should not be construed as being limited to the exemplary embodiments and implementations set forth herein. Rather, these exemplary embodiments and implementations are provided to make the description of this disclosure exhaustive and complete, and to fully communicate the scope of this disclosure to those skilled in the art. In the following description, details of well-known features and techniques may be omitted to avoid unnecessarily obscuring the presented embodiments and implementations.

[0026] Overview

[0027] According to embodiments of this disclosure, various techniques, methods, schemes, and / or solutions relate to prediction of mobile communication measurement events based on artificial intelligence (AI) or machine learning (ML) (also referred to as "AI or ML" or "AI / ML"). According to this disclosure, multiple possible solutions can be implemented individually or in combination. That is, although these possible solutions are described individually below, two or more solutions can be implemented in some combination.

[0028] Figure 1 Example scenario 100 is illustrated in a communication environment in which various solutions and schemes of this disclosure can be implemented. Scenario 100 involves a user equipment (UE) 110 communicating wirelessly with a wireless network consisting of an access network 125 and a core network (CN) 130. This wireless network can be a 5G NR network, a 5G-Advanced network, or a 6G network, but this disclosure is not limited thereto. UE 110 can be a smartphone, wearable device, Internet of Things (IoT) device, tablet, etc. Alternatively, UE 110 can be a laptop (NB) or personal computer (PC) with a data card inserted or installed, the data card containing a modem and a radio frequency (RF) transceiver to provide wireless communication capabilities. CN 130 can include entities such as User Plane Function (UPF), Mobility Management Function (AMF), Session Management Function (SMF), and Unified Data Management (UDM). Access network 125 can include one or more base stations (BS), such as BS 120. BS 120 can be an evolved Node B (eNB), a next-generation Node B (gNB), or a transceiver point (TRP), but this disclosure is not limited thereto. BS 120 can provide communication coverage for a geographically covered area, supporting communication with UE 110 within that area.

[0029] In scenario 100, UE 110 establishes a connection with BS 120 (e.g., a Radio Resource Control (RRC) connection). BS 120 can provide measurement configurations (e.g., trigger time to trigger (TTT) and offset) to UE 110 via RRC signaling. BS 120 can request UE 110 to perform in-frequency and / or out-of-frequency measurements. UE 110 can perform measurements according to the measurement configuration (e.g., Radio Resource Management (RRM) measurements) and evaluate reporting criteria (also known as reporting conditions) to trigger a measurement report based on the measurement results. In this embodiment, UE 110 utilizes an AI / ML-based prediction module 115, which includes one or more AI / ML-based models, to predict measurement results. For example, UE 110 can use the AI / ML-based prediction module 115 to predict whether a specific measurement event is likely to occur. The output of the AI / ML-based prediction module 115 includes event prediction information indicating the prediction result. The AI / ML-based prediction module 115 may include components implemented in hardware and / or software. AI / ML-based models can be cell-specific, meaning one AI / ML-based model corresponds to one cell. Alternatively, AI / ML-based models can be cluster-specific (i.e., one AI / ML-based model corresponds to a group of cells).

[0030] By utilizing the AI / ML-based prediction module 115, UE 110 performs measurement event prediction and determines when to report model output (e.g., event prediction information) to BS 120. Subsequently, BS 120 can determine how to perform mobility-related procedures (e.g., handover (HO)) based on the event prediction information reported by UE 110. For example, BS 120 can make a handover (HO) decision and send an HO command to UE 110.

[0031] Specifically, UE 110 can use measurement results related to a first set of radio resources to predict measurement results related to a second set of radio resources. The first set of radio resources may correspond to a first set of reference signals for a first set of cells, and the second set of radio resources may correspond to a second set of reference signals for a second set of cells. Reference signal quality may include signal-to-noise ratio (SNR), reference signal received power (RSRP), or received signal strength (RSS). The radio resources of the reference signals may be defined in the time domain or by any combination of the time, spatial, and frequency domains. In one embodiment, UE 110 measures the first set of reference signals for the first set of cells (e.g., the serving cell and co-frequency or inter-frequency neighboring cells) and uses an AI / ML-based prediction module 115 to predict the measurement results for the second set of reference signals for the second set of cells. The second set of cells may be the serving cell and co-frequency or inter-frequency neighboring cells, which may be the same as or different from the first set of cells. UE 110 can derive the cell quality of the first and / or second sets of cells. Furthermore, UE 110 can evaluate reporting criteria and triggering conditions and decide whether to send a measurement report to BS 120. In another embodiment, UE 110 can collect measurement results of a first set of reference signals from a first group of cells and a second set of reference signals from a second group of cells, and train one or more AI / ML-based models based on the collected measurement results (with or without other auxiliary information). More specifically, during the training phase of the AI / ML-based model, UE 110 measures the first set of reference signals from the first group of cells and the second set of reference signals from the second group of cells. UE 110 can train the model by using the measurement results of the first set of reference signals from the first group of cells as model input and the measurement results of the second set of reference signals from the second group of cells as labels.

[0032] Figure 2This is an illustration depicting an example scenario of prediction using one or more artificial intelligence (AI) or machine learning (ML) based models according to an embodiment of this disclosure. Referring to scenario 210, user equipment (UE) 110 can perform measurement prediction in the time domain using one or more AI / ML based models. That is, the first set of reference signals (i.e., model inputs) of the first group of cells and the second set of reference signals (i.e., model labels) of the second group of cells are identical in both the frequency and spatial domains. In other words, the radio resources of the first and second sets of reference signals are defined by the same frequency and spatial domains. During the training phase, the AI / ML based model is trained using historical measurement results measured within the time period [TN, T] as model input and future measurement results within the time period [T+1, T+M] as labels. By utilizing one or more trained AI / ML based models, UE 110 measures the first set of reference signals of the first group of cells within the time period [TN, T] to predict the second set of reference signals of the second group of cells within the time period [T, T+M].

[0033] Referring to Scenario 220, UE 110 can perform measurement predictions in the time and spatial domains using one or more AI / ML-based models. In other words, the first set of reference signals for the first group of cells and the second set of reference signals for the second group of cells are identical in the frequency domain. That is, the radio resources of the first and second sets of reference signals are defined by the same frequency domain. As shown in Scenario 220, solid lines represent the measured beam sets, and dashed lines represent the predicted beam sets. During the training phase, the AI / ML-based model uses the measurement results of a portion of the beams as model input and is trained with the full beam results or other partial beam results as labels. By utilizing one or more trained AI / ML-based models, UE 110 measures a portion of the beam set of the first set of reference signals for the first group of cells (shown as dashed lines) within the time period [TN, T] to predict the full / other partial beam set of the second set of reference signals for the second group of cells (shown as solid lines) within the time period [T, T+M].

[0034] Referring to Scenario 230, UE 110 can perform measurement predictions in the time and frequency domains using one or more AI / ML-based models. This means that the first set of reference signals for the first group of cells and the second set of reference signals for the second group of cells are identical in the spatial domain. In other words, the radio resources of the first and second sets of reference signals are defined by the same spatial domain. The first and second sets of reference signals can be inter-frequency or intra-frequency. As shown in Scenario 230, solid lines represent the measured bandwidth portion (BWP) / frequency, and dashed lines represent the predicted full or other BWP / frequency. During the training phase, the AI / ML-based model uses the measurement results of a portion of the BWP / frequency within the time period [TN, T] as model input and is trained with the results of the full or other BWP / frequency within the time period [T, T+M] as labels. By utilizing one or more trained AI / ML-based models, UE 110 measures a portion of the BWP / frequency of the first set of reference signals (shown as dashed lines) within the time period [TN, T] to predict the full or other BWP / frequency of the second set of reference signals (shown as solid lines) within the time period [T, T+M].

[0035] In another embodiment, UE 110 can perform measurement predictions using one or more AI / ML-based models in the time, spatial, and frequency domains. For example, during the training phase, historical measurements of a subset of beams (for a subset of BWPs / frequencies) within the time period [TN, T] are used as model input, and future measurements of all / other subsets of beams (for all / other BWPs / frequencies) within the time period [T+1, T+M] are used as labels for training. By utilizing one or more trained AI / ML-based models, UE 110 measures a first set of reference signals for a first group of cells associated with a subset of beams and a set of BWPs / frequencies within the time period [TN, T] to predict a second set of reference signals for a second group of cells associated with all / other subsets of beams and all or other BWPs / frequencies within the time period [T, T+M].

[0036] Figure 3This is a diagram illustrating the flowchart of the current measurement process, including beam-level and cell-level filtering. In connected mode (e.g., RRC connected mode), the UE measures multiple beams (at least one) of a cell and averages the measurement results (power values) to derive cell quality. During this process, the UE is configured to consider a subset of the detected beams. Filtering is performed at two different levels: filtering at the physical layer to obtain beam quality, and then at the RRC level to derive cell quality from multiple beams. The cell quality derived from beam measurements is processed the same for serving and non-serving cells. If the gNB configures the UE, the measurement report can include the measurement results of X best beams. Scenario 300 includes an L1 filtering phase, a beam combining / selection phase, an L3 filtering phase for cell quality, a reporting condition evaluation phase, an L3 beam filtering phase, and a beam selection phase for reporting. In scenario 300, K beams correspond to measurements of Synchronization Signal Blocks (SSBs) or Channel State Information-Reference Signals (CSI-RS) resources configured by the gNB for Layer 3 (L3) mobility and detected by the UE in Layer 1 (L1). Point A is a measurement within the physical layer (beam-specific sample). The L1 filtering phase includes internal L1 filtering of the input measured at point A, the specific implementation of which is determined by the UE. Point A1 represents the measurement reported from L1 to L3 after L1 filtering. The beam combining / selection phase combines beam-specific measurements to derive cell quality. The L3 filtering phase for cell quality filters the measurements provided at point B. Point C is the measurement after L3 filtering for cell quality. The reporting condition assessment phase checks at point D whether an actual measurement report is required. That is, measurement report information (messages) can be sent at point D. The L3 beam filtering phase filters the measurements provided at point A1. The beam selection phase for reporting selects X measurements from the measurements provided at point E. The behavior of the beam combining / selection phase, the L3 filtering phase for cell quality, the reporting condition assessment phase, the L3 beam filtering phase, and the beam selection phase for reporting is standardized. The configuration of these phases is provided by RRC signaling.

[0037] In this disclosure, one or more artificial intelligence (AI) or machine learning (ML) based models can be incorporated into the measurement process to perform measurement event prediction, and the prediction can be cell-level radio resource management (RRM) measurement prediction or beam-level RRM measurement prediction. Figures 4A to 4E This illustration depicts an example of incorporating an AI or ML-based predictive model into a measurement process according to an embodiment of this disclosure. However, this disclosure is not limited to... Figures 4A to 4EIn other words, the output of an AI or ML-based model can serve as input to one or a combination of the following stages: Layer 1 (L1) filtering, beam combining / selection, Layer 3 (L3) filtering for cell quality, reporting condition assessment, Layer 3 (L3) beam filtering, and beam selection for reporting. In one example, by utilizing one or more AI or ML-based models, L1 beam-level measurement results can be predicted based on actual L1 beam-level measurements, and then L3 cell-level measurement results can be generated. In another example, L3 cell-level measurement results can be predicted based on actual L1 beam-level measurements. In another example, L1-filtered beam-level measurement results can be predicted based on actual L1 beam-level measurements, and then L3 beam-level measurement results can be generated. In another example, L3 beam-level measurement results can be predicted based on actual L3 beam-level measurements. In another example, L3 beam-level measurement results can be predicted based on actual L1 beam-level measurements.

[0038] See Figure 4A As shown in process 400a, the measurement results at point A without L1 filtering are used for AI- or ML-based RRM prediction (i.e., as input to one or more AI- or ML-based models). Specifically, using the measurement results of the first set of reference signals without L1 filtering as model input, the user equipment (UE) 110 can predict the measurement results of the second set of reference signals at point A', and then perform L1 and L3 filtering on the measurement results of the first and second sets of reference signals. Afterwards, the UE 110 performs beam combining / selection to ultimately determine the cell quality. In one embodiment, the first and second sets of reference signals are defined by one or a combination of spatial, temporal, and frequency domains.

[0039] See Figure 4B As shown in process 400b, the L1 measurement results at point A1 (after L1 filtering) are used for AI- or ML-based RRM prediction (i.e., as input to one or more AI- or ML-based models). Specifically, using the L1-filtered measurement results of the first set of reference signals as model input, UE 110 can first predict the measurement results of the second set of reference signals at point A1', and then perform L3 filtering on the measurement results of the first and second sets of reference signals. UE 110 can then perform beam combining / selection to ultimately determine the cell quality. The first and second sets of reference signals can be defined by one or a combination of spatial, temporal, and frequency domains.

[0040] See Figure 4CIn the illustrated process 400c, the L3 measurement results at point E are used for AI- or ML-based RRM prediction (i.e., as input to one or more AI- or ML-based models). For example, by using the measurement results of the first set of reference signals after L3 beam filtering as model input, UE 110 can predict the measurement results of the second set of reference signals at point E'. In one embodiment, UE 110 can first perform L3 filtering on the measurement results of the first set of reference signals, perform measurement prediction on the second set of reference signals, and then perform beam combining / selection to finally determine the cell quality. The first and second sets of reference signals can be defined by one or a combination of the spatial, temporal, and frequency domains.

[0041] Cell quality prediction based on RRM measurements can be achieved through... Figure 4D The process shown is 400d or Figure 4E The illustrated process 400e is implemented. In process 400d, the measurement results at point B are used for AI or ML-based RRM prediction; these measurements are L1 measurements after L1 filtering and beam combining / selection. Figure 4E In the illustrated process 400e, the measurement result at point C is used for AI or ML-based RRM prediction. This measurement result is an L3 measurement result after L3 cell filtering. UE 110 can measure the first group of cells and predict the measurement results for the second group of cells. For example, the measured first group of cells can be the same as the predicted second group of cells. Alternatively, the measured first group of cells can be different from the predicted second group of cells. The first and second sets of reference signals can be defined by one or a combination of spatial, time, and frequency domains.

[0042] Figure 5The following is an example flow of advance measurement reporting according to an embodiment of the present disclosure, illustrated for illustrative purposes. In flow 500, UE 110 can evaluate configured reporting conditions using predicted RRM results and report predicted measurement events during the measurement period. In one example, for temporal beam quality prediction, UE 110 can perform a full beam scan within the observation time window and stop beam scanning within the prediction time window. In another example, for temporal-spatial beam quality prediction, UE 110 can perform a partial beam scan within the observation time window. The evaluation of reporting conditions, considering measurement results, prediction results, or a combination of both, is configured by base station (BS) 120. Furthermore, within the measurement period, UE 110 can report measurement events based on prediction results, measurement results, or a combination of both, configured by BS 120. By reporting measurement events in advance, BS 120 can make a handover (HO) decision in advance and send an HO command to UE 110. Accordingly, UE 110 can perform an HO operation based on the HO command. For example, UE 110 initiates HO preparation when the predicted event (A3 event) is about to occur, and sends an HO command when the A3 entry condition is met based on actual / real measurements and the predicted event is predicted to be met during TTT.

[0043] Figure 6 This is a schematic diagram illustrating an example flow for predicting measurement events during the TTT and prediction period, according to an embodiment of this disclosure. In flow 600, UE 110 can evaluate reporting conditions for a longer period including the TTT and prediction period, and report measurement events at the prediction time when reporting conditions with a longer TTT are met. In one example, for temporal beam quality prediction, UE 110 can perform a full beam scan within the observation time window and stop beam scanning within the prediction time window. In another example, for temporal-spatial beam quality prediction, UE 110 can perform a partial beam scan within the observation time window. UE 110 can evaluate reporting conditions within the TTT, the usage of which is configured by BS 120, and use predicted measurement results during the prediction period, or use both actual and predicted measurement results simultaneously. Furthermore, UE 110 can report measurement events that meet the reporting conditions for both the TTT and prediction period at a certain prediction time. That is, UE 110 evaluates reporting conditions based on measurement results from the TTT and prediction periods. If the conditions are met, UE 110 reports the measurement event at the specific prediction time instead of waiting for the TTT to end. In other words, UE 110 can report not only at the end of TTT, but also at a predicted time during TTT, based on a joint evaluation of measurements during TTT and the prediction period. A prediction time longer than TTT can avoid unnecessary measurement reports and HOs.

[0044] Exemplary Implementation

[0045] Figure 7 An example communication system 700 according to an embodiment of this disclosure is shown, comprising at least one example communication device 710 and one example network device 720. Both the communication device 710 and the network device 720 can perform various functions to implement the schemes, techniques, processes, and methods described herein related to AI / ML-based mobile communication measurement event prediction, including the aforementioned scenarios / schemes and processes 800, 900, and 1000 described below.

[0046] The communication device 710 may be part of an electronic device, which may be a UE (User Equipment), such as a portable or mobile device, a wearable device, a wireless communication device, or a computing device. For example, the communication device 710 may be implemented in a smartphone, smartwatch, personal digital assistant, digital camera, tablet, laptop, or other computing device. The communication device 710 may also be part of a machine-type device, which may be an Internet of Things (IoT), Narrowband Internet of Things (NB-IoT), or Industrial Internet of Things (IIoT) device, such as a non-movable or fixed device, a home appliance, a wired communication device, or a computing device. For example, the communication device 710 may be implemented in a smart thermostat, a smart refrigerator, a smart door lock, a wireless speaker, or a home control center. Alternatively, the communication device 710 may be implemented in the form of one or more integrated circuit (IC) chips, such as, but not limited to, one or more single-core processors, one or more multi-core processors, one or more Reduced Instruction Set Computing (RISC) processors, or one or more Complex Instruction Set Computing (CISC) processors. The communication device 710 may contain... Figure 7 The at least some components shown are, for example, the processor 712. The communication device 710 may also include one or more other components unrelated to the present disclosure (e.g., internal power supply, display device, and / or user interface device), and therefore these components are... Figure 7 It is not shown in the text, and will not be described in the following text for the sake of brevity.

[0047] Network device 720 may be part of a network device, which may be a network node such as a satellite, base station, cell, router, or gateway. For example, network device 720 may be implemented in an eNB in ​​an LTE network, a gNB in ​​a 5G / NR, IoT, NB-IoT, or IIoT network, or a satellite or base station in a 6G network. Network device 720 may include... Figure 7 The diagram shows at least some components, such as processor 722. Processor 722 may also include a protocol stack and a set of control function modules and circuitry. Network device 720 may also include one or more other components unrelated to the present disclosure (e.g., internal power supply, display device, and / or user interface device), therefore these components... Figure 7 It is not shown in the text, and will not be described in the following text for the sake of brevity.

[0048] In one aspect, both processor 712 and processor 722 may be implemented as one or more single-core processors, one or more multi-core processors, or one or more CISC processors. That is, although the singular term "processor" is used herein to refer to processor 712 and processor 722, in some embodiments of this disclosure, processor 712 and processor 722 may comprise multiple processors, or in other embodiments, a single processor. In another aspect, both processor 712 and processor 722 may be implemented in hardware (and optionally firmware) and include, for example, but not limited to, one or more transistors, one or more diodes, one or more capacitors, one or more resistors, one or more inductors, one or more memristors, and / or one or more transformers, these electronic components being configured and arranged to achieve the specific purposes of this disclosure. In other words, in at least some embodiments, processor 712 and processor 722 are dedicated machines specifically designed, arranged, and configured to perform specific tasks of various embodiments of this disclosure in devices (e.g., communication device 710) and networks (e.g., network device 720).

[0049] In some embodiments, the communication device 710 may further include a memory 714 coupled to a processor 712, which can access and store data therein. In some embodiments, the communication device 710 may further include a transceiver 716 coupled to the processor 712, capable of wirelessly transmitting and receiving data.

[0050] In some embodiments, network device 720 may further include a memory 724 coupled to processor 722, in which processor 722 can access and store data, and a transceiver 726 coupled to processor 722, capable of wirelessly transmitting and receiving data. Therefore, communication device 710 and network device 720 can communicate wirelessly via transceiver 716 and transceiver 726, respectively.

[0051] For ease of explanation and without limitation, the capabilities of communication device 710 and network device 720 are described below, in conjunction with procedures 800, 900, and 1000. In these procedures, communication device 710 is implemented as a communication device or UE, and network device 720 is implemented as a network node of a communication network (e.g., a base station).

[0052] Exemplary process

[0053] Figure 8An example flow 800 of an embodiment of this disclosure is illustrated. Flow 800 can be an example implementation of the above-described scenario / solution, whether partially or entirely, for AI or ML-based measurement event prediction in mobile communications. Flow 800 can represent one aspect of the functional implementation of communication device 710. Flow 800 may include one or more operations, actions, or functions as shown in modules 810, 820, and 830. Although shown as independent modules, the various modules of flow 800 may be divided into more modules, merged into fewer modules, or omitted depending on the desired implementation. Furthermore, the modules of flow 800 may be arranged according to... Figure 8 The process can be executed in the order shown, or in a different order. Process 800 can be implemented by communication device 710 or any suitable UE (e.g., UE 110) or machine-type device. For illustrative purposes only and without limitation, process 800 is described below with communication device 710 as the UE. Process 800 may begin at block 810.

[0054] In block 810, process 800 may involve the processor 712 of communication device 710 measuring a first set of radio resources. Process 800 may proceed from block 810 to block 820.

[0055] In block 820, process 800 may involve processor 712 obtaining predicted measurements of a second set of radio resources based on the measurement, by using an AI / ML-based model to predict the measurement in one or a combination of the time, spatial, and frequency domains. Process 800 may proceed from block 820 to block 830.

[0056] In block 830, process 800 may involve processor 712 sending a measurement report to network node (e.g., network device 720) via transceiver 716 based on the measurement results and / or predicted measurement results of the first set of radio resource measurements.

[0057] In some implementations, the measurement prediction may include cell-level RRM measurement prediction or beam-level RRM measurement prediction.

[0058] In some implementations, the measurement result may correspond to an L1 measurement result, an L1 filter output, an L3 beam filter output, a beam combining or selection output, or an L3 filter output of cell quality.

[0059] In some embodiments, the measurement results used for the measurement prediction include L1 measurement results. Process 800 may also involve processor 712 predicting the predicted measurement results for a second set of radio resources. Process 800 may further involve processor 712 performing L3 filtering on the measurement results associated with the first set of radio resources and the predicted measurement results associated with the second set of radio resources.

[0060] In some embodiments, the measurement results used for the measurement prediction include L3 measurement results. Process 800 may also involve processor 712 performing L3 filtering on the measurement results associated with the first set of radio resources. Process 800 may further involve processor 712 predicting the predicted measurement results for the second set of radio resources.

[0061] In some implementations, process 800 may also involve processor 712 performing L3 filtering on measurement results related to the first group of radio resources, obtaining predicted measurement results for the second group of radio resources by performing measurement prediction, performing beam combining or selection, and deriving cell quality.

[0062] In some implementations, the first set of radio resources corresponds to a first set of reference signals for a first set of cells, and the second set of radio resources corresponds to a second set of reference signals for a second set of cells. The first set of reference signals and the second set of reference signals are in the same frequency and spatial domains. Process 800 may also involve a processor 712 performing measurement prediction in the time domain by using measurement results measured in a first time period as input to an AI or ML-based model, and using predicted measurement results related to a second time period as labels for the AI ​​or ML-based model, wherein the first time period is earlier than the second time period.

[0063] In some implementations, the first set of radio resources corresponds to a first set of reference signals for a first group of cells, and the second set of radio resources corresponds to a second set of reference signals for a second group of cells. The first set of reference signals and the second set of reference signals are in the same frequency domain. Process 800 may also involve a processor 712 performing measurement prediction in a combination of the temporal and spatial domains by using measurement results of a first portion of the beam measured in a first time period as input to an AI or ML-based model, and using predicted measurement results of a second portion or all of the beam related to a second time period as labels for the AI ​​or ML-based model, wherein the first time period is earlier than the second time period.

[0064] In some implementations, the first set of radio resources corresponds to a first set of reference signals for a first set of cells, and the second set of radio resources corresponds to a second set of reference signals for a second set of cells. The first set of reference signals and the second set of reference signals are in the same spatial domain. Process 800 may also involve processor 712 performing measurement prediction in a combination of the time and frequency domains by using measurement results of a first BWP or frequency set measured in a first time period as input to an AI or ML-based model, and using predicted measurement results of a second or all BWPs or frequency sets related to a second time period as labels for the AI ​​or ML-based model, wherein the first time period is earlier than the second time period.

[0065] In some implementations, process 800 may also involve processor 712 using measurement results of a first portion of the beams of a first BWP or frequency set measured in a first time period as input to an AI or ML-based model, and using predicted measurement results of a second portion or all of the beams of a second or all of the BWP or frequency set associated with a second time period as labels for the AI ​​or ML-based model, to perform measurement predictions in a combination of time, spatial, and frequency domains, wherein the first time period is earlier than the second time period.

[0066] In some implementations, the first set of radio resources corresponds to a first set of cells, and the second set of radio resources corresponds to a second set of cells. Process 800 may involve processor 712 deriving the cell quality of the first set of cells, the second set of cells, or both.

[0067] In some implementations, the prediction information indicates that a predicted measurement event will occur. Process 800 may involve processor 712 evaluating reporting conditions for triggering the transmission of prediction information, wherein the measurement results and / or predicted measurement results are used to evaluate the reporting conditions based on network configuration.

[0068] In some implementations, process 800 may involve processor 712 evaluating reporting conditions at the prediction time. Further, process 800 may involve processor 712 reporting a measurement event at the measurement time, provided that the reporting conditions are met.

[0069] In some implementations, process 800 may involve processor 712 evaluating reporting conditions during a period including TTT and the prediction time. Further, process 800 may involve processor 712 reporting measurement events at the prediction time, provided that reporting conditions with a longer TTT are met.

[0070] In some implementations, the AI / ML-based model corresponds to one or a group of cells.

[0071] In some implementations, the quality of the reference signal associated with the measurement result includes SNR, RSRP, or RSS.

[0072] In some implementations, process 800 may involve processor 712 training an AI / ML-based model based on measurement results and predicted measurement results.

[0073] Figure 9An example flow 900 according to an embodiment of this disclosure is illustrated. Flow 900 can serve as an example implementation of the above-described scenario / solution, whether partially or entirely, for AI / ML-based measurement event prediction in mobile communications. Flow 900 may represent one aspect of the functional implementation of communication device 710. Flow 900 may include one or more operations, actions, or functions, as shown by one or more modules 910, 920, 930, and 940. Although shown as independent modules, the various modules of flow 900 may be divided into more modules, merged into fewer modules, or omitted depending on the desired implementation. Furthermore, the modules of flow 900 may be arranged according to... Figure 9 The process 900 may be executed in the order shown, or in a different order. Process 900 may be implemented by communication device 710 or any applicable user equipment (e.g., UE 110) or machine-type device. For illustrative purposes only and without limitation, process 900 is described below with communication device 710 as the UE. Process 900 may begin with module 910.

[0074] In module 910, process 900 may involve the processor 712 of communication device 710 performing AI / ML-based predictions. Process 900 may proceed from module 910 to module 920.

[0075] In module 920, process 900 may involve processor 712 initiating HO preparation if a measurement event is predicted to occur based on AI / ML prediction results. Process 900 may proceed from module 920 to module 930.

[0076] In module 930, process 900 may involve processor 712 receiving a HO command from a network node (e.g., network device 720) via transceiver 716 when the triggering condition of a measurement event is met. Process 900 may proceed from module 930 to module 940.

[0077] In module 940, process 900 may involve processor 712 performing HO operations according to HO commands.

[0078] In some implementations, the measurement event is predicted to occur based on AI / ML predictions and measurement results.

[0079] In some implementations, process 900 may involve processor 712 sending a measurement report to network nodes based on AI / ML prediction results via transceiver 716.

[0080] In some implementations, process 900 may involve processor 712 evaluating report conditions based on AI / ML prediction results.

[0081] In some implementations, process 900 may involve processor 712 evaluating reporting conditions, including TTT and prediction time, based on measurement results and AI or ML prediction results.

[0082] Figure 10 An example flow 1000 according to an embodiment of this disclosure is illustrated. Flow 1000 can serve as an example implementation of the aforementioned scenario / solution, whether partially or entirely, for AI / ML-based measurement event prediction in mobile communications. Flow 1000 can represent one aspect of the functional implementation of network device 720. Flow 1000 may include one or more operations, actions, or functions, as shown by one or more modules 1010, 1020, and 1030. Although shown as independent modules, the various modules of flow 1000 may be divided into more modules, merged into fewer modules, or omitted depending on the desired implementation. Furthermore, the modules of flow 1000 may be arranged according to... Figure 10 The process can be executed in the order shown, or in a different order. Process 1000 can be implemented by network device 720 or any base station (e.g., BS 120) or network node. For illustrative purposes only and without limitation, process 1000 is described below in the context of network device 720. Process 1000 may begin with module 1010.

[0083] In module 1010, process 1000 may involve the processor 722 of network device 720 receiving a measurement report from the UE (e.g., communication device 710) via transceiver 726. This measurement report is generated based on AI / ML prediction results. Process 1000 may proceed from module 1010 to module 1020.

[0084] In module 1020, process 1000 may involve processor 722 making HO decisions based on measurement reports. Process 1000 may proceed from module 1020 to module 1030.

[0085] In module 1030, process 1000 may involve processor 712 sending HO commands to UE via transceiver 726 based on HO decision.

[0086] Additional Notes

[0087] The topics described herein sometimes demonstrate that different components are contained within or connected to other different components. It should be understood that such architectures are merely examples, and many other architectures can actually be implemented to achieve the same functionality. Conceptually, any arrangement of components to achieve the same functionality is effectively “associated” to achieve the desired function. Therefore, any two components combined in this document to achieve a specific function can be considered “associated with each other” to achieve the desired function, regardless of the architecture or intermediate components. Similarly, any two components so associated can be considered “operationally connected” or “operationally coupled” to achieve the desired function, and any two components that can be so associated can be considered “operationally coupled” to achieve the desired function. Specific examples of operational coupling include, but are not limited to, physically matable and / or physically interactive components, wirelessly interactive and / or wirelessly interactive components, and logically interactive and / or logically interactive components.

[0088] Furthermore, regarding the use of almost all plural and / or singular terms in this document, those skilled in the art can appropriately convert from plural to singular and / or from singular to plural depending on the context and / or application. For clarity, various singular / plural arrangements are explicitly listed herein.

[0089] Furthermore, those skilled in the art will understand that terms commonly used herein, particularly in appended claims, such as the body portion of appended claims, are generally considered "open-ended" terms. For example, the word "comprising" should be interpreted as "including but not limited to," the word "having" should be interpreted as "having at least," and the word "includes" should be interpreted as "including but not limited to," etc. Those skilled in the art will also further understand that if a particular quantity introduced in a claim is intentional, that intention will be explicitly stated in the claim; if no such statement is made, then that intention does not exist. For example, for ease of understanding, the appended claims described below may contain the use of the introductory phrases "at least one" and "one or more" to introduce the content of the claim. However, the use of such phrases should not be interpreted as meaning that when the content of a claim is introduced by the indefinite article "a" or "an," any specific claim containing that content is limited to containing only one instance of that content, even if the same claim contains the introductory phrase "one or more" or "at least one" and the indefinite article such as "a" or "an," for example, "a" and / or "an" should be interpreted as "at least one" or "one or more"; the same applies to the use of definite articles used to introduce the content of a claim. Furthermore, even if a specific quantity is explicitly stated in the claims, those skilled in the art will recognize that such a statement should be interpreted as at least the stated quantity. For example, stating "two items" alone, without further modification, means at least two items, or two or more items. Additionally, when using conventions such as "at least one A, B, and C, etc.", such structures are generally intended for those skilled in the art to understand the meaning of the convention. For example, "a system having at least one A, B, and C" includes, but is not limited to, systems having only A, only B, only C, both A and B, both A and C, both B and C, and systems having both A, B, and C, etc. Similarly, when using conventions such as "at least one A, B, or C, etc.", such structures are generally intended for those skilled in the art to understand the meaning of the convention. For example, "a system having at least one A, B, or C" includes, but is not limited to, systems having only A, only B, only C, both A and B, both A and C, both B and C, and systems having both A, B, and C, etc. Those skilled in the art will further understand that virtually any extractive word and / or phrase presenting two or more alternative terms, whether in the specification, claims, or drawings, should be understood to include the possibility of including only one term, any two terms, or all of the terms. For example, the phrase "A or B" would be understood to include the possibility of including "A" or "B" or "A and B".

[0090] As can be seen from the foregoing, various embodiments of this disclosure have been described for illustrative purposes, and various modifications can be made without departing from the scope and spirit of this disclosure. Therefore, the various embodiments disclosed herein are not intended to be limiting, and the true scope and spirit are defined by the following claims.

Claims

1. A method comprising: The device's processor measures the first set of radio resources; Based on this measurement, the processor uses artificial intelligence (AI) or machine learning (ML) models in one or a combination of the time, spatial, and frequency domains to make measurement predictions and obtain the predicted measurement results of the second set of radio resources. as well as The processor sends a measurement report to the network node based on the measurement results of the first set of radio resources and / or the predicted measurement results.

2. The method of claim 1, wherein: This measurement prediction includes cell-level radio resource management (RRM) measurement prediction or beam-level RRM measurement prediction; or The measurement results correspond to Layer 1 measurement results, Layer 1 filter output, Layer 3 beam filter output, beam combining or selection output, or Layer 3 filter output for cell quality.

3. The method of claim 1, wherein the measurement result used for the measurement prediction includes the layer 1 measurement result, and the method further includes: The processor predicts the predicted measurement results for the second set of radio resources; as well as The processor performs layer 3 filtering on the measurement results associated with the first group of radio resources and the predicted measurement results associated with the second group of radio resources.

4. The method of claim 1, wherein the measurement result used for the measurement prediction includes layer 3 measurement results, and the method further includes: The processor performs layer 3 filtering on the measurement results related to the first set of radio resources; as well as The processor predicts the results of the predicted measurements for the second set of radio resources.

5. The method of claim 4, further comprising: The processor performs layer 3 filtering on the measurement results related to the first set of radio resources; The processor obtains the predicted measurement results for the second set of radio resources by performing the measurement prediction. The processor performs beam combining or selection; as well as The processor determines the cell quality.

6. The method of claim 1, wherein the first set of radio resources corresponds to a first set of reference signals for a first set of cells, and the second set of radio resources corresponds to a second set of reference signals for a second set of cells, wherein the first set of reference signals and the second set of reference signals are in the same frequency domain and spatial domain, and obtaining the prediction measurement result further comprises: The measurement prediction is performed in the time domain by using the measurement result obtained in the first time period as input to the AI ​​or ML model and using the predicted measurement result related to the second time period as the label of the AI ​​or ML model, wherein the first time period is earlier than the second time period.

7. The method of claim 1, wherein the first set of radio resources corresponds to a first set of reference signals for a first set of cells, and the second set of radio resources corresponds to a second set of reference signals for a second set of cells, wherein the first set of reference signals and the second set of reference signals are in the same frequency domain, and obtaining the predictive measurement result further comprises: The measurement prediction is performed in the combination of the time domain and the spatial domain by using the measurement result of the first portion of the beam measured in the first time period as input to the AI ​​or ML model, and using the predicted measurement result of the second portion or all of the beam related to the second time period as the label of the AI ​​or ML model, wherein the first time period is earlier than the second time period.

8. The method of claim 1, wherein the first set of radio resources corresponds to a first set of reference signals for a first set of cells, and the second set of radio resources corresponds to a second set of reference signals for a second set of cells, wherein the first set of reference signals and the second set of reference signals are in the same spatial domain, and obtaining the prediction measurement result further comprises: The measurement prediction is performed in a combination of the time domain and the frequency domain by using the measurement result of the first bandwidth portion (BWP) or frequency set measured in the first time period as input to the AI ​​or ML model, and using the predicted measurement result of the second or all BWP or frequency set related to the second time period as the label of the AI ​​or ML model, wherein the first time period is earlier than the second time period.

9. The method of claim 1, wherein obtaining the predicted measurement result further comprises: The measurement prediction is performed in the combination of the time domain, the spatial domain, and the frequency domain by using the measurement result of the first portion of the beams of the first BWP or frequency set measured in the first time period as input to the AI ​​or ML model, and using the predicted measurement result of the second portion or all of the second BWP or frequency set related to the second time period as the label of the AI ​​or ML model, wherein the first time period is earlier than the second time period.

10. The method of claim 1, wherein the first group of radio resources corresponds to a first group of cells, and the second group of radio resources corresponds to a second group of cells, and the method further comprises: The processor determines the cell quality of the first group of cells, the second group of cells, or both.

11. The method of claim 1, wherein the prediction information indicates that a predicted measurement event will occur, and the method further comprises: The processor evaluates the reporting conditions used to trigger the transmission of the predictive information, wherein the measurement results and / or the predictive measurement results are used to evaluate the reporting conditions based on the network configuration.

12. The method of claim 1, further comprising: The processor evaluates the reporting conditions for the predicted time. as well as The processor reports the measurement event at the measurement time if the reporting conditions are met.

13. The method of claim 1, further comprising: The processor evaluates the reporting conditions during the period including the trigger time (TTT) and the prediction time. as well as The processor reports the measurement event at the predicted time if the reporting condition with a longer TTT is met.

14. The method of claim 1, wherein the AI ​​or ML-based model corresponds to one or a group of cells.

15. The method of claim 1, wherein the quality of the reference signal associated with the measurement result includes the signal-to-noise ratio (SNR), the reference signal received power (RSRP), or the received signal strength (RSS).

16. The method of claim 1, further comprising: The processor trains the AI ​​or ML-based model based on the measurement results and the predicted measurement results.

17. A method comprising: The device's processor performs AI or ML-based predictions; The processor initiates HO switching preparation when it predicts that a measurement event will occur based on the results of the AI ​​or ML-based prediction. The processor receives the HO command from the network node when the triggering condition of the measurement event is met; as well as The processor executes the HO operation according to the HO command.

18. The method of claim 17, wherein the measurement event is predicted to occur based on the results of the AI ​​or ML-based prediction and the measurement results.

19. The method of claim 17, further comprising: The processor sends a measurement report to the network node based on the AI ​​or ML-based prediction.

20. The method of claim 19, further comprising: The processor evaluates the report conditions based on the results of the AI ​​or ML-based prediction.

21. The method of claim 19, further comprising: The processor evaluates the reporting conditions with trigger time (TTT) and prediction time based on the measurement results and the AI ​​or ML-based prediction results.

22. A method comprising: The network node's processor receives a measurement report from the user equipment (UE), which is generated based on AI or ML prediction results; The processor makes the HO switching decision based on the measurement report; as well as The processor sends an HO command to the UE based on the HO decision.

23. An apparatus comprising: A transceiver that enables wireless communication during operation; as well as A processor, communicatively coupled to the transceiver, performs the following operations during operation: The first group of radio resources was measured; Based on this measurement, a second set of predicted measurements of radio resources is obtained by using AI- or ML-based models to predict measurements in one or a combination of the time, spatial, and frequency domains; and Based on the measurement results of the first set of radio resources and / or the predicted measurement results, a measurement report is sent to the network node via the transceiver.

Images