Method executed by means of user equipment, and user equipment

Abstract

Provided in the present disclosure are a method executed by means of a user equipment, and a user equipment. The method comprises: acquiring an RLF prediction result from an RLF prediction model; and when the RLF prediction result is acquired and at least one of the following is satisfied, declaring that an RLF has occurred: a first condition, in which RLF prediction configuration information comprises an information element representing enabling of an RLF prediction-based RLF declaration process, an RLF prediction-based RRC re-establishment process or an RLF prediction-based MCG failure information process; a second condition, in which a time difference between an RLF occurrence moment indicated by the RLF prediction result and the current moment is not greater than a first threshold value; a third condition, in which the current moment is within a prediction time window; a fourth condition, in which a probability value is not less than a second threshold value; a fifth condition, in which there is currently a running timer T310 or timer T312; a sixth condition, in which there are currently one or more triggered measurement events; and a seventh condition, in which a UE, after sending the RLF prediction result to a network side, has not received a handover command or an RRC release message from the network side within a given period of time.

Description

Methods executed by user equipment and user equipment Technical Field

[0001] This disclosure relates to the field of wireless communication technology, and more specifically, to a method performed by a user equipment and a user equipment. Background Technology

[0002] Artificial Intelligence / Machine Learning (AI / ML) represents a significant revolution in computer science and data processing. AI / ML typically refers to processes and algorithms that simulate human intelligence, using the collection, analysis, learning, and deduction of existing data to solve problems in various fields. At the 3GPP RAN plenary meeting in September 2024, a research project on the application of AI / ML in NR mobility was approved (see 3GPP non-patent document RP-242393). This research project focuses on enhancing air interface mobility in Radio Resource Connection (RRC) states, referring to changes in the Primary Cell (PCell) in the NR system. This project studies and evaluates the benefits and gains of AI / ML-assisted, network-triggered Layer 3 handovers, primarily considering the following aspects:

[0003] • AI / ML-based Radio Resource Management (RRM) measurement and event prediction. This includes cell-level measurement prediction, encompassing intra-frequency and inter-frequency measurements; Radio Link Failure (RLF) prediction; and handover failure prediction, among others.

[0004] • Research the necessity and benefits of UE auxiliary information in network-side models.

[0005] The impact of AI / ML-assisted mobility on 3GPP specifications.

[0006] In existing NR systems, the UE performs measurements for RRM and reports the obtained measurement results to the network based on the reporting configuration (such as measurement events) configured on the network side. The network side then performs RRC connected-state mobility management and decisions based on the received measurement results actually performed by the UE. If the UE detects a radio link failure before receiving a handover command from the network side, the UE initiates an RRC re-establishment procedure to restore its connection with the network.

[0007] This disclosure aims to address the issue of enhancing the RRC re-establishment process based on AI / ML-assisted RLF prediction in NR networks, and further, to address the issue of how a UE can declare an RLF or perform an RRC re-establishment process in advance based on the obtained RLF prediction results in a system with AI / ML-assisted RLF prediction enabled. Summary of the Invention

[0008] The main objective of this disclosure is to provide an RLF detection method and user equipment, so that in a system with AI / ML-assisted RLF prediction enabled, the UE can promptly declare an RLF or initiate an RRC re-establishment process when there is a predicted RLF result, thereby reducing the service interruption time caused by the waiting time before the actual RLF is detected.

[0009] According to a first aspect of this disclosure, a method performed by a user equipment is provided, comprising: obtaining an RLF prediction result from a Radio Link Failure (RLF) prediction model, the RLF prediction model being trained to output the RLF prediction result relating to whether an RLF occurs at a future time or within a future time window when inputting channel measurement results obtained from actual measurements, the RLF prediction result being a probability value representing the likelihood of an RLF occurring, or a predicted value from which the probability value can be obtained; and declaring an RLF occurrence when the RLF prediction result is obtained and at least one of the following conditions is met: a first condition, RLF prediction configuration information received from the network side includes an indication of enabling RLF declaration based on RLF prediction, and an RLF prediction based on RLF prediction. The information elements of the RRC re-establishment process based on LF prediction or MCG failure information process based on RLF prediction; the second condition is that the time difference between the RLF occurrence time represented by the RLF prediction result and the current time is not greater than the first threshold value; the third condition is that the input corresponding to the RLF prediction result includes a prediction time window representing the future time window, and the current time is within the prediction time window; the fourth condition is that the probability value is not lower than the second threshold value; the fifth condition is that there is currently a running timer T310 or timer T312; the sixth condition is that there are currently one or more triggered measurement events; the seventh condition is that after the user equipment sends the RLF prediction result to the network side, it does not receive a handover command or RRC release message from the network side within a given time period.

[0010] Optionally, if multiple RLF prediction results are obtained within the prediction time window, in the second condition, the RLF occurrence time can be the RLF occurrence time of the RLF measurement result corresponding to the maximum value among the multiple probability values ​​of the multiple RLF prediction results.

[0011] Optionally, if multiple RLF prediction results are obtained within the prediction time window, in the fourth condition, the maximum value among the multiple probability values ​​based on the multiple RLF prediction results can be compared with the second threshold value.

[0012] Optionally, in the sixth condition, one or more of the triggered measurement events may be triggered based on actual measurement results.

[0013] Optionally, the method may further include: performing a judgment on the sixth condition when the measurement event involved in the sixth condition, or the measurement reporting configuration, measurement identifier, or measurement object corresponding to the measurement event is enabled by an RLF claim based on RLF prediction, an RRC re-establishment based on RLF prediction, or an MCG failure information process based on RLF prediction.

[0014] Optionally, the information elements can be used to enable RLF claim based on RLF prediction, RRC re-establishment based on RLF prediction, or MCG failure information process based on RLF prediction for the measurement event, or the measurement reporting configuration, measurement identifier, or measurement object corresponding to the measurement event.

[0015] Optionally, the method may further include: if the first condition is not met, even if other conditions are met, an RLF will not be declared to have occurred.

[0016] Optionally, the second threshold value is configured by the network side and represented by the information element; or the first threshold value is configured by the network side and represented by the information element.

[0017] Optionally, the prediction time window can be determined by the user equipment and reported to the network side.

[0018] According to another aspect of this disclosure, a user equipment is also provided, comprising: a processor; and a memory storing instructions, wherein the instructions, when executed by the processor, are the method described above.

[0019] According to this disclosure, enhanced RLF monitoring can be achieved, and the problem of inappropriately premature RLF declaration or premature initiation of RRC re-establishment or MCG failure information process can be avoided in the mechanism for implementing early RLF declaration or early initiation of RRC re-establishment or MCG failure information process. Attached Figure Description

[0020] The above and other features of this disclosure will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, wherein:

[0021] Figure 1 is a block diagram showing an outline of the measurement model for NR.

[0022] Figure 2 is a schematic flowchart illustrating a method executed by a user equipment in one embodiment.

[0023] Figure 3 is a schematic flowchart illustrating a method performed by a user equipment according to a specific embodiment.

[0024] Figure 4 shows a block diagram of a user equipment according to an embodiment of the present disclosure. Detailed Implementation

[0025] Other aspects, advantages, and key features of this disclosure will become apparent to those skilled in the art from the following detailed description of exemplary embodiments of the disclosure taken in conjunction with the accompanying drawings.

[0026] In this disclosure, the terms “comprising” and “containing” and their derivatives are meant to include rather than limit; the term “or” is inclusive and may be equivalent to “and” or “and / or”.

[0027] In this specification, the various embodiments described below to illustrate the principles of this disclosure are merely illustrative and should not be construed as limiting the scope of the disclosure in any way. The following description, with reference to the accompanying drawings, is intended to aid in a comprehensive understanding of exemplary embodiments of this disclosure as defined by the claims and their equivalents. The following description includes various specific details to aid understanding, but these details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Furthermore, for clarity and brevity, descriptions of well-known functions and structures have been omitted. Additionally, throughout the drawings, the same reference numerals are used for similar functions and operations.

[0028] The following description uses an NR mobile communication system as an example application environment to illustrate several implementations according to this disclosure. However, it should be noted that this disclosure is not limited to the following implementations, but is applicable to many other wireless communication systems.

[0029] The base station in this disclosure can be any type of base station, including NodeB, enhanced base station eNB, 5G communication system base station gNB; or micro base station, pico base station, macro base station, home base station, etc.; the network side generally refers to the base station. The cell can also be a cell under any of the above-mentioned types of base stations. Unless otherwise specified, cell, beam, and transmission point (TRP) can be interchanged, and the base station can also be the central unit (gNB-Central Unit, gNB-CU) or distributed unit (gNB-Distributed Unit, gNB-DU) that makes up the base station. Different embodiments can also be combined, for example, the same variables / parameters / terms in different embodiments can be interpreted in the same way. Cancel, release, delete, clear, and clear can be replaced. Execute, use, and apply can be replaced. Configure and reconfigure can be replaced. Monitor and detect can be replaced. Initiate and trigger can be replaced. If..., if..., and under... circumstances can be replaced, or can be replaced with When the UE determines... The handover command refers to an RRC reconfiguration message containing synchronization reconfiguration information elements for the Master Cell Group (MCG). RLF and RLF events are interchangeable.

[0030] The following section will first explain some existing mechanisms involved in this disclosure. It is worth noting that some names in the following description are merely illustrative and not restrictive, and may be used in other ways.

[0031] Artificial Intelligence / Machine Learning (AI / ML)

[0032] In this disclosure, an AI / ML model is used to represent the application of AI / ML technology in the NR air interface. An AI / ML model can also be referred to as an AI / ML function. AI / ML technology can be divided into the following five aspects:

[0033] - AI / ML model training

[0034] The training of AI / ML models involves obtaining an inference relation (e.g., a function) based on the combination of input and output parameters, which is then used for subsequent inference. Taking the RLF prediction model as an example, this model can be trained on the network side or by the UE. The input parameters of this model are the raw channel data (e.g., actual measured cell or beam measurements), and the output parameters are the predicted measurement results or the RLF prediction probability.

[0035] - AI / ML model transfer

[0036] If an AI / ML model is trained by a network, the trained model can be sent from the network to the user (UE) for inference based on that model. This sending of the model is called AI / ML model transfer.

[0037] - AI / ML model inference

[0038] Taking the RRM measurement model as an example, the process of the UE inputting relevant input information (such as measurement events, measurement results of one or more cells actually measured) into an RRM measurement model and generating the predicted RRM measurement results by using the model is the inference process of the AI / ML model.

[0039] AI / ML model monitoring

[0040] The network or UE needs to monitor the AI / ML model used to determine whether the model is suitable for the current link state or network environment.

[0041] - AI / ML model update

[0042] When the network or UE deems the model no longer applicable, the AI / ML model will be updated.

[0043] RRM measurement

[0044] RRM measurements in connected state are primarily used for mobility management, such as PCell handover. RRM measurements include intra-cell measurements, intra-frequency measurements, inter-cell measurements, and inter-frequency measurements. The NR measurement model is shown in Figure 1.

[0045] -A: Measurement sample of a single beam inside the physical layer.

[0046] - Layer 1 filtering: The Layer 1 (L1) filtering process performed within the physical layer based on a single beam measurement sample.

[0047] -A1: Measurement results of a single beam obtained after layer 1 filtering; this result is reported to RRC from the physical layer.

[0048] Layer (i.e., reporting from L1 to L3).

[0049] - Beam selection / merging: Select / merge several measurements from the beam measurement results reported by the physical layer to obtain the measurement results of the cell.

[0050] -B: Cell measurement results are reported to the RRC layer (i.e., L3).

[0051] - Layer 3 Cell Quality Filtering: The cell measurement results are filtered based on filtering parameters, etc.

[0052] -C: Measurement results, used as input for reporting criteria evaluation.

[0053] - Evaluate the reporting criteria: Based on the configuration parameters of the measurement report, evaluate whether it is necessary to trigger the measurement report reporting.

[0054] -D: Report a measurement report containing cell measurement results to the base station over the air interface.

[0055] - Layer 3 beam filtering: Layer 3 (L3) filtering is applied to the measurement results of a single beam.

[0056] -E: Measurement results of a single beam obtained after filtering.

[0057] - Beam selection reporting: Select the measurement results of X beams from the measurement results of K beams obtained from point E.

[0058] -F: Report a measurement report containing beam measurement results to the base station over the air interface.

[0059] In RRC connected mode, the network sends measurement configuration (such as that contained in the MeasConfig information element) to the UE via RRC messages (e.g., RRC reconfiguration messages). This measurement configuration can include: measurement object configuration, measurement report configuration, measurement identifier configuration, measurement quantity configuration, measurement interval configuration, etc. The UE stores the measurement configuration in its measurement configuration variable (VarMeasConfig).

[0060] Measurement objects: These are the objects that the UE needs to measure. The network can configure a list of measurement objects containing multiple objects. The measurement object configuration mainly includes: the time-frequency resource location and subcarrier spacing of the reference signal used for measurement, the frequency information of the measurement, and the cell information of the measurement. Measurement objects are identified by measurement object identifiers; each measurement object corresponds to a unique measurement object identifier.

[0061] Reporting configurations: Each measurement object can correspond to one or more reporting configurations, and the network can configure a list of reporting configurations containing multiple measurement report configurations. Measurement report configurations mainly include: reporting criteria, measurement events, reference signal types, and report formats. Measurement reports are identified by a measurement report identifier, and each measurement report corresponds to a unique identifier.

[0062] Measurement identities (Measurement IDs): Used to associate measurement object identifiers (IDs) and measurement report identifiers (IDs). Each measurement identity is associated with one measurement object identifier and one measurement report identifier. The network can be configured to contain a list of multiple measurement identities.

[0063] The UE reports the measured results to the network via RRC messages (Measurement Report Messages). The UE includes the measurement results from the UE Variable Measurement Report List in the RRC message and indicates them to the network. The measurement results generally include at least one or more cell-level measurement results / signal quality or beam-level measurement results / signal quality for one or more cells or frequencies. The measurement results / signal quality are usually characterized by Reference Signal Received Power (RSRP), Reference Signal Received Quality (RSRQ), or Received Signal Strength Indicator (RSSI).

[0064] AI / ML-assisted RLF prediction

[0065] In the AI / ML-assisted mobility research project currently being studied by 3GPP, one aspect involves AI / ML-assisted prediction of Recurrent LRF (Recurrent Starting Fault) events. Its main objective is to achieve time-domain measurement prediction and RLF event prediction through inference using a pre-trained AI / ML model. This predictive approach aims to improve mobility performance, such as handover robustness, or reduce latency caused by data terminals.

[0066] For ease of description, AI / ML models are simply referred to as models, and RLF event prediction and RLF prediction can be used interchangeably.

[0067] In current 3GPP discussions, there are two methods for predicting RLF events: direct and indirect. In the direct method, the model input includes actual channel measurement results, and the model output is the probability of an RLF event occurring; more specifically, the probability of an RLF event occurring within a time instance or a time period (such as a time window). In the indirect method, the model input includes actual channel measurement results, and the model output is the channel measurement results of the serving cell or primary cell at one or more time points, such as the Signal-to-Interference-plus-Noise Ratio (SINR). The output of the indirect method can be considered an intermediate output, and the final output is based on this intermediate output to predict the expected occurrence time and probability of an RLF event. Clearly, the model's involvement is not always necessary to arrive at the final output from the intermediate output. In other words, the direct output of the indirect prediction model is the prediction of the signal quality of the serving cell or primary cell at one or more future time points, and then, based on this signal quality and other RLF prediction configurations, the probability of an RLF event occurring at those time points is indirectly obtained. In the direct prediction model, the model output is the probability of a measured event occurring within a future time instance or time period.

[0068] Typically, the model's input parameters may also include an observation window. In the time domain, the observation window is a time window or a period of time, such as 100ms. Additionally, the model's input parameters may include a prediction window. The user interface (UE) uses the (actually measured) measurements within the observation window to predict the output within the prediction window.

[0069] RLF detection

[0070] In the NR system, UEs in RRC connected state perform Radio Link Monitoring (RLM) on special cells (SpCells). During RLM, lower layers of the UE, such as the physical layer, monitor link quality. When the UE receives "out-of-sync" indications for N310 links from lower layers, the UE starts timer T310 for the corresponding special cell. When timer T310 associated with RLM / RLF monitoring times out, the UE considers an RLF (Recurrent Link Failure) detected for the corresponding cell group. RLF monitoring is performed on a cell group-by-cell basis. The UE also considers an RLF detected for the corresponding cell group when one or more of the following conditions occur: timer T312 times out; a random access problem indication or a consistent uplink listen-before-talk failure indication is received from the Medium Access Control (MAC) layer of the corresponding cell group; or an indication that the maximum number of retransmissions has been reached is received from the Radio Link Control (RLC) layer of the corresponding cell group. When one or more of the above situations occur for the primary cell, if the UE believes that an RLF of the primary cell group has been detected, it can rebuild the connection with the network side by initiating an RRC re-establishment procedure or an MCG failure information procedure.

[0071] RRC Re-establishment Process

[0072] The RRC re-establishment process is initiated by the UE and is used to re-establish or continue the RRC connection in RRC connected state. In NR systems, when the UE experiences a connection failure with the network, such as detecting an RLF in the primary cell group, or when both the primary cell group link and the secondary cell link fail or are suspended in a dual-connectivity configuration, the UE initiates an RRC connection re-establishment process to rebuild / restore the connection with the network. When initiating the RRC connection re-establishment process, the UE starts a start timer T311 and executes a cell selection procedure. If the UE selects a suitable NR cell before timeout T311, the UE further determines whether this cell is one of its saved conditional reconfiguration candidate cells or Layer 1 / Layer 2 Triggered Mobility (LTM) candidate cells. If so, the UE performs the corresponding conditional reconfiguration or LTM cell change to restore the connection with the network. If not, the UE initiates an RRC re-establishment request message and starts timer T301 to restore the connection with the network through RRC re-establishment. When the UE receives an RRC re-establishment message or RRC establishment message in response to the RRC re-establishment request message, it executes the RRC configuration contained in the message and returns the corresponding RRC re-establishment completion message or RRC establishment completion message to the network, successfully ending the RRC re-establishment process and restoring the RRC connection with the network. If timer T311 or timer T301 times out, the RRC re-establishment process fails, and the UE enters the RRC idle state.

[0073] Based on current 3GPP discussions, RLF (Event-Based Fault) prediction using AI / ML models can be termed model-based RLF prediction. In networks without model-based RLF prediction, when a UE in an RRC connection state starts its timer T310, the UE must wait for T310 to time out before declaring an RLF and executing link recovery procedures such as RRC re-establishment, which can lead to data interruption. Therefore, when model-based RLF prediction is configured, a feasible way to reduce data interruption is to declare an RLF in advance or initiate an RRC re-establishment process early based on the RLF prediction result obtained by the UE. How to achieve this enhanced RLF monitoring or enhanced RRC re-establishment is the focus of this disclosure. Furthermore, how to avoid data interruption caused by inappropriate early RLF declarations or early RRC re-establishment in the mechanism for implementing early RLF declarations or early RRC re-establishment is also a focus of this disclosure.

[0074] This disclosure proposes solutions to at least some of the problems mentioned above regarding how to achieve RLF detection enhancement or RRC re-establishment enhancement in networks that support model-based RLF prediction, and specific implementation methods are given below.

[0075] First, an outline of one embodiment of this disclosure will be described with reference to Figure 2.

[0076] Figure 2 is a schematic flowchart illustrating a method executed by a user equipment in one embodiment.

[0077] As shown in Figure 1, in S202, an RLF prediction result is obtained from the Radio Link Failure (RLF) prediction model, and in S204, it is determined whether an RLF prediction result has been obtained. The RLF prediction model is trained to output an RLF prediction result related to whether an RLF will occur at a future time or within a future time window when the channel measurement result obtained from actual measurement is input. Here, the RLF prediction result represents a probability value indicating the likelihood of an RLF occurring, or a predicted value from which the probability value can be obtained.

[0078] Upon obtaining the RLF prediction result, in S206, it is determined whether at least one of the following conditions is met: First condition: The RLF prediction configuration information received from the network side includes information elements indicating the process of enabling RLF claim based on RLF prediction, RRC re-establishment based on RLF prediction, or MCG failure information based on RLF prediction; Second condition: The time difference between the RLF occurrence time indicated by the RLF prediction result and the current time is not greater than a first threshold value; Third condition: The input corresponding to the RLF prediction result includes a prediction time window representing the future time window, and the current time is within the prediction time window; Fourth condition: The probability value is not lower than a second threshold value; Fifth condition: There is currently a running timer T310 or timer T312; Sixth condition: There are currently one or more triggered measurement events; Seventh condition: After the user equipment sends the RLF prediction result to the network side, it does not receive a handover command or RRC release message from the network side within a given time period.

[0079] If at least one of the above conditions is determined to be met in S206, then in S208, an RLF is declared to have occurred. If none of the above conditions are met, then in S210, an RLF is not declared to have occurred. S210 in Figure 1 is shown for ease of understanding only; when any of the above conditions are not met in S206, the UE may not perform any operation. The same applies to S305, which will be described later.

[0080] The following provides a detailed description of specific examples and embodiments related to this invention. Furthermore, as described above, the examples and embodiments described in this disclosure are illustrative examples provided for easy understanding of the invention and are not intended to limit the invention. In the following embodiments, the order of the steps is merely illustrative and not strictly limited; the implementation steps can also be combined and implemented without limitation.

[0081] Figure 3 is a schematic flowchart illustrating some implementation steps of the model-assisted RLF prediction RLF claim method or RRC re-establishment method. As shown in Figure 3, this implementation includes any one or more of the following steps.

[0082] Optionally, in step S301, the UE receives an RLF prediction configuration from the network side. The RLF prediction configuration includes a first information element, which enables the UE to perform RLF claiming or RRC re-establishment based on RLF prediction.

[0083] Optionally, the first information element enables the UE to perform RLF claim or RRC re-establishment based on RLF prediction. When the UE is configured with the first information element or the value of the information element is set to "true", the UE can perform RLF claim or RRC re-establishment based on RLF prediction.

[0084] Optionally, the RLF claim based on RLF prediction refers to the UE declaring the detection of RLF or believing that an RLF has occurred based on the obtained RLF prediction result (determining whether to do so). Optionally, the RRC re-establishment based on RLF prediction refers to the UE initiating an RRC re-establishment process based on the obtained RLF prediction result (determining whether to do so). Optionally, the RLF claim or RRC re-establishment based on RLF prediction can be based on the methods described in subsequent embodiments, and is not limited here.

[0085] Optionally, the RLF prediction configuration is configured to the UE by the network side via an RRC message, or more specifically, it is included in model-related configurations. Alternatively, the RLF prediction configuration is configured to the UE within the "otherConfig" information element of the RRC message. This "otherConfig" information element includes configurations related to other configurations, such as overheating assistance configuration, latency budget reporting configuration, DRX reference configuration, maximum bandwidth reference configuration, and RLM relaxation reporting configuration.

[0086] Optionally, in step S302, the UE obtains an RLF prediction result. Optionally, the UE obtains an RLF prediction result through a model-based direct prediction method or an indirect prediction method. The RLF prediction result is the probability that the UE will experience an RLF at a certain time in the future, or the probability that the UE will experience an RLF within a certain time period in the future.

[0087] Optionally, in step S303, the UE determines whether to declare an RLF or initiate an RRC re-establishment procedure based on the acquired RLF prediction result. If the determination in S303 is "yes", S304 is executed, that is, an RLF is declared to have occurred; if the determination is "no", S305 is executed, that is, an RLF is not declared to have occurred.

[0088] Optionally, declare and detect can be used interchangeably; RLF declare and RLF detect can also be used interchangeably, both referring to the UE's perception that an RLF has occurred in RLF-related operations. In this disclosure, time window and prediction window can be used interchangeably.

[0089] Optionally, in one approach, if the UE is configured with the first information element in step S301, the UE performs the determination in step S303 to decide whether to declare an RLF or initiate an RRC re-establishment based on the acquired RLF prediction result. Optionally, if the UE is not configured with the information element, the operation in step S303 is not performed.

[0090] Optionally, in one method two, the UE determines whether to declare an RLF or initiate an RRC re-establishment process based on the acquired RLF prediction result by performing the following: If the UE's RLF prediction occurs at time T1, and the time difference between the current time T0 and T1 is not greater than or less than a first threshold value TH1, then the UE determines to declare an RLF or initiate an RRC re-establishment process. That is, when there is an RLF prediction result on the UE, if the UE determines that the time difference from the predicted RLF occurrence time is not greater than TH1, the UE considers that an RLF has been detected, or the UE initiates an RRC re-establishment process. Optionally, the first threshold value TH1 is configured to the UE by the network side. Optionally, the first threshold value can be configured to the UE through the RLF prediction configuration in step S301, and further, the first threshold value can be a first information element.

[0091] Optionally, when using direct RLF prediction, the obtained RLF prediction result is the probability value of RLF occurrence within a time window. In the above judgment, T1 can be interpreted as the upper or lower edge of the time window. For example, if the start time of a time window is Tstart and the end time is Tend, then the time point Tstart is considered as the upper edge of the time window, and the time point Tend is considered as the lower edge of the time window.

[0092] Optionally, in one case, if the UE may have multiple occurrence times corresponding to multiple predicted RLF results (probability values ​​of RLF occurrence) within a prediction window, then in the determination of this method, the UE uses the occurrence time value corresponding to the largest RLF prediction probability value among the multiple RLF prediction results for judgment.

[0093] Optionally, for cases using direct RLF prediction, the UE may also perform the following: if the UE has obtained an RLF prediction result and the current time is within the time window corresponding to the RLF prediction result, then the UE considers that an RLF has been detected, or the UE initiates an RRC re-establishment procedure.

[0094] Optionally, in one case, the size of the prediction window is determined by the UE itself, rather than configured by the network side. In this case, the UE informs the network side of the size of the prediction window used, for example, through an RRC message, or further, through the UE Assistive Information flow to report to the network side.

[0095] Optionally, in another method three, the UE determines whether to declare an RLF or initiate an RRC re-establishment procedure based on the acquired RLF prediction result by performing the following: If the probability of the RLF event occurring in the RLF prediction result obtained by the UE is greater than or not less than a second threshold value TH2, then the UE determines to declare an RLF or initiate an RRC re-establishment procedure. That is, in this case, when the probability of the RLF event occurring is high, the UE considers that an RLF has been detected, or the UE initiates an RRC re-establishment procedure. Optionally, the second threshold value TH2 is configured to the UE by the network side. Optionally, the second threshold value can be configured to the UE through the RLF prediction configuration in step S301, and further, the second threshold value can be a first information element. Optionally, the second threshold value is configured in the form of a percentage. Optionally, in one case, the UE may have multiple predicted RLF results (probability values ​​of RLF occurrence) within a prediction window, then in this method of determination, the UE uses the largest RLF prediction probability value among the multiple RLF prediction results for judgment.

[0096] Optionally, in another method four, the UE determines whether to declare an RLF or initiate an RRC re-establishment procedure based on the acquired RLF prediction result as follows: If the UE has an acquired RLF prediction result and there is currently a running timer T310 or timer T312, then the UE determines to declare an RLF or initiate an RRC re-establishment procedure; that is, the UE considers an RLF to have been detected. Timers T310 and T312 are both timers used for RLF detection. When a physical layer problem of a special cell SpCell is detected (i.e., receiving N310 consecutive out-of-sync indications from the physical layer), the UE starts T310; when the T310 of the primary cell group times out, the UE initiates an RRC re-establishment procedure or an MCG failure information procedure. For the timer T312 configured for the primary cell, when a measurement report with a measurement identifier configured with T312 is triggered, and the corresponding parameter useT312 is configured to true, and the primary cell's T310 is running, the UE starts T312 associated with the primary cell group. Optionally, timers T312 and T310 are associated with the primary cell group.

[0097] Optionally, in another method five, the UE determines whether to declare an RLF or initiate an RRC re-establishment procedure based on the acquired RLF prediction result by performing the following: If the UE has an acquired RLF prediction result and one or more measurement events are triggered (or have been reported to the network side), then the UE determines to declare an RLF or initiate an RRC re-establishment procedure. Optionally, the measurement event is an A2 or A3 measurement event. Optionally, the measurement event is an actual measurement result, i.e., not a predicted measurement result or measurement event output by the model. Optionally, when the measurement event or the measurement reporting configuration, measurement identifier, or measurement object corresponding to the measurement event is enabled to declare an RLF or initiate an RRC re-establishment based on the acquired RLF prediction result, the UE performs the operation in this method. Optionally, the configuration for enabling "the measurement event or the measurement reporting configuration or measurement identifier or measurement object corresponding to the measurement event to claim RLF or initiate RRC re-establishment based on the obtained RLF prediction result" is the aforementioned first information element. In this way, the first information element is configured to distinguish between the measurement reporting configuration (identifier) ​​or the measurement identifier or measurement object (identifier).

[0098] Optionally, in another method six, the operation of the UE determining whether to declare RLF or initiate an RRC re-establishment process based on the acquired RLF prediction result is performed as follows: If the UE has an acquired RLF prediction result, and the UE does not receive a response message from the network side within a certain period after sending the RLF prediction result to the network side, then the UE determines to declare RLF or initiate an RRC re-establishment process. Optionally, the response message is a handover command or an RRC release message. Optionally, the length / value of the time period is configured by the network side. Optionally, the time period is implemented by a timer, for example, when the UE successfully sends the acquired RLF prediction result to the network side, a timer TimerX is started; if TimerX times out, the UE determines to declare RLF or initiate an RRC re-establishment process. That is, the UE waits for a response message from the network side during the operation of TimerX, and stops TimerX when a response message from the network side is received. Optionally, the value of the timer is configured to the UE through an RRC message, such as being included in the model-related configuration; optionally, the configuration of the timer is a first information element. Optionally, the UE sends the obtained RLF prediction results to the network side via an RRC message.

[0099] The aforementioned methods one through six can be executed in combination and are not mutually exclusive. That is, the UE can perform the operation of determining to declare an RLF or initiating an RRC re-establishment procedure when it determines that at least one of the conditions described in methods one through six is ​​satisfied, or when it determines that all the conditions described in methods one through six are satisfied. For example, the UE performs the operation of determining to declare an RLF or initiating an RRC re-establishment procedure when at least one of the following conditions is satisfied or when all the following conditions are satisfied.

[0100] -When the UE has obtained RLF prediction results;

[0101] -When the first information element is configured;

[0102] One or more measurement events are triggered;

[0103] - There is a running timer T310 or timer T312;

[0104] - The time difference from the predicted RLF occurrence time is no greater than the first threshold value TH1;

[0105] The probability of the occurrence of the RLF event in the RLF prediction result is not less than the second threshold value TH2;

[0106] - No handover command or RRC release message was received from the network side after sending the RLF prediction result to the network side;

[0107] -TimerX timed out.

[0108] Optionally, the order in which the above conditions are judged is not restricted.

[0109] Optionally, when the judgment in the above conditions is based on the measurement event in Method 5, the aforementioned first threshold or second threshold is configured to distinguish between measurement reporting configuration (identifier) ​​or measurement identifier or measurement object (identifier).

[0110] Optionally, in addition to the methods described above for RLF claiming and initiating RRC re-establishment, the method of this disclosure can also be applied to initiating the MCG failure information process. That is, it determines whether to initiate the MCG failure information process based on RLF prediction based on the judgments in methods one to six. If one or more conditions in methods one to six are met, the MCG failure information process based on RLF prediction is initiated. The MCG failure information process is used to inform the network-side UE that it has experienced an MCG failure. In this process, the UE initiates the transmission of the MCG failure information message and starts timer T316.

[0111] Figure 4 is a block diagram illustrating a user equipment 400 according to an embodiment of the present disclosure. As shown in Figure 4, the user equipment 400 includes a processor 401 and a memory 402. The processor 401 may include, for example, a microprocessor, a microcontroller, an embedded processor, etc. The memory 402 may include, for example, volatile memory (such as random access memory, RAM), a hard disk drive (HDD), non-volatile memory (such as flash memory), or other memory. Program instructions are stored on the memory 402. When executed by the processor 401, these instructions can perform the methods described in detail in this disclosure that are executed by the user equipment.

[0112] A program running on a device according to this disclosure may be a program that enables a computer to perform the functions of embodiments of this disclosure by controlling a central processing unit (CPU). The program or the information processed by the program may be temporarily stored in volatile memory (such as random access memory RAM), hard disk drive (HDD), non-volatile memory (such as flash memory), or other memory systems.

[0113] Programs used to implement the functions of the embodiments of this disclosure can be recorded on a computer-readable recording medium. The corresponding functions can be implemented by causing a computer system to read and execute the programs recorded on the recording medium. The term "computer system" here can refer to a computer system embedded in the device, and may include an operating system or hardware (such as peripheral devices). "Computer-readable recording medium" can be a semiconductor recording medium, an optical recording medium, a magnetic recording medium, a short-time dynamic storage program recording medium, or any other computer-readable recording medium.

[0114] Various features or functional modules of the devices used in the above embodiments can be implemented or executed by circuits (e.g., monolithic or multi-chip integrated circuits). Circuits designed to perform the functions described in this specification may include general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination of the above devices. A general-purpose processor may be a microprocessor, or any existing processor, controller, microcontroller, or state machine. The circuits described above may be digital circuits or analog circuits. In cases where advancements in semiconductor technology have led to new integrated circuit technologies that replace existing integrated circuits, one or more embodiments of this disclosure may also be implemented using these new integrated circuit technologies.

[0115] Furthermore, this disclosure is not limited to the embodiments described above. Although various examples of the embodiments have been described, this disclosure is not limited thereto. Fixed or non-mobile electronic devices installed indoors or outdoors can be used as terminal devices or communication devices, such as AV equipment, kitchen equipment, cleaning equipment, air conditioners, office equipment, vending machines, and other household appliances.

[0116] As described above, embodiments of this disclosure have been described in detail with reference to the accompanying drawings. However, the specific structure is not limited to the above embodiments, and this disclosure also includes any design modifications that do not depart from the spirit of this disclosure. Furthermore, various modifications can be made to this disclosure within the scope of the claims, and embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included within the technical scope of this disclosure. In addition, components with the same effects described in the above embodiments can be substituted for each other.

Claims

1. A method executed by a user equipment, comprising: The RLF prediction result is obtained from a radio link failure (RLF) prediction model, which has been trained to output an RLF prediction result related to whether an RLF will occur at a future time or within a future time window when given channel measurement results obtained from actual measurements. The RLF prediction result represents a probability value indicating the likelihood of an RLF occurring, or a predicted value from which the probability value can be obtained; and An RLF is declared to have occurred when the RLF prediction result is obtained and at least one of the following conditions is met: The first condition is that the RLF prediction configuration information received from the network side includes information elements indicating the process of enabling RLF prediction-based RLF claims, RRC re-establishment based on RLF prediction, or MCG failure information based on RLF prediction. The second condition is that the time difference between the RLF occurrence time indicated by the RLF prediction result and the current time is not greater than the first threshold value. The third condition is that the input corresponding to the RLF prediction result includes a prediction time window representing the future time window, and the current time is within the prediction time window; The fourth condition is that the probability value is not lower than the second threshold value; The fifth condition is that there is currently a running timer T310 or timer T312; The sixth condition is that one or more measurement events have been triggered. The seventh condition is that after the user equipment sends the RLF prediction result to the network side, it does not receive a handover command or RRC release message from the network side within a given time period.

2. The method according to claim 1, wherein, In the case where multiple RLF prediction results are obtained within the prediction time window, in the second condition, the RLF occurrence time is the RLF measurement time corresponding to the maximum value among the multiple probability values ​​of the multiple RLF prediction results.

3. The method according to claim 1, wherein, If multiple RLF prediction results are obtained within the prediction time window, in the fourth condition, the maximum value among the multiple probability values ​​based on the multiple RLF prediction results is compared with the second threshold value.

4. The method according to claim 1, wherein, In the sixth condition, one or more of the triggered measurement events are triggered based on actual measurement results.

5. The method according to claim 4, wherein, The method further includes: If the measurement event involved in the sixth condition, or the measurement reporting configuration, measurement identifier, or measurement object corresponding to the measurement event, is enabled by the RLF claim based on RLF prediction, the RRC re-establishment based on RLF prediction, or the MCG failure information process based on RLF prediction, the judgment of the sixth condition is performed.

6. The method according to claim 4, wherein, The information elements enable RLF claim based on RLF prediction, RRC re-establishment based on RLF prediction, or MCG failure information process based on RLF prediction for the measurement event, or the measurement reporting configuration, measurement identifier, or measurement object corresponding to the measurement event.

7. The method according to any one of claims 1 to 6, wherein, The method further includes: If the first condition is not met, an RLF will not be declared to have occurred even if other conditions are met.

8. The method according to claim 1, wherein, The second threshold value is configured by the network side, and the second threshold value is represented by the information element; or The first threshold value is configured by the network side, and the first threshold value is represented by the information element.

9. The method according to claim 1, wherein, The prediction time window is determined by the user equipment and reported to the network side.

10. A user equipment, comprising: processor; as well as Memory, which stores instructions The instructions, when executed by the processor, perform the method of any one of claims 1 to 9.

Images