Prediction-based mobility management

Abstract

Disclosed are methods, apparatuses, and systems for communications. One of the methods includes: receiving, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

Description

PREDICTION-BASED MOBILITY MANAGEMENTCROSS-REFERENCE TO RELATED PATENT APPLICATION

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 748,712, filed on January 23, 2025, entitled “AI / ML-BASED MOBILITY MANAGEMENT,” the entirety of which is incorporated by reference herein.Field

[0002] Apparatuses and methods consistent with the present disclosure relate generally to communications, more specifically, methods, systems, and devices for prediction-based mobility management in communications.Background

[0003] Conventional mobility management in wireless networks primarily relies on reactive handover procedures, where a handover is initiated only after the user equipment (UE) observes signal degradation. This reactive approach introduces delay and increases the risk of late handovers, especially in fast-changing radio environments or high-mobility scenarios. Also, the handover procedure itself involves multiple signaling and coordination steps between the UE and the network, which can further extend the overall handover completion time. To address these limitations, prediction-based mobility management has emerged as an enhanced approach.

[0004] One prediction-based handover approach involves performing all predictive processing at the network side. However, because modern networks must manage predictions for a large number of UEs, this centralized architecture can lead to substantial computational complexity and poor scalability. Delegating prediction tasks to the UE side therefore represents a more practical and efficient solution.

[0005] To perform prediction at the UE side, the UE may utilize current and / or historical radio resource management measurement results as inputs to a prediction model. The resulting prediction outputs may then be reported to the network to assist the network in making handover decisions. However, when the UE performs such predictions using its own prediction models, the network may have no knowledge of the specificity and performance of the models used at the UE. As a consequence, the network cannot guarantee the stability, reliability, and overall performance of the UE-side prediction process. This lack of visibility and control may lead to premature handover behavior or degraded mobility performance. Systems and methods that enable the network to maintain visibility into, and ensure the stability, reliability, and performance of, UE-side prediction processes are desired.Summary

[0006] According to some embodiments of the present disclosure, there is provided a first node for communication. The first node includes a memory storing an instruction; and a processor configured to execute the instruction stored in the memory to: receive, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receive, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; perform, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmit, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0007] According to some embodiments of the present disclosure, there is provided a second node for communication. The second node includes a memory storing an instruction; and a processor configured to execute the instruction stored in the memory to: transmit, to a first node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmit, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receive, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0008] According to some embodiments of the present disclosure, there is provided a method for a first node for communication. The method includes: receiving, from a second node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0009] According to some embodiments of the present disclosure, there is provided a method for a second node for communication. The method includes: transmitting, to a first node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmitting, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receiving, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0010] According to some embodiments of the present disclosure, there is provided a non-transitory computer-readable medium storing instructions that are executable by one or more processors of a first node for communication, to perform a method. The method includes: receiving, from a second node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0011] According to some embodiments of the present disclosure, there is provided a non-transitory computer-readable medium storing instructions that are executable by one or more processors of a second node for communication, to perform a method. The method includes: transmitting, to a first node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmitting, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receiving, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0012] FIG. 1 is a schematic diagram illustrating a communication system in which only the network has prediction models, consistent with some embodiments of the present disclosure.

[0013] FIG. 2 is a schematic diagram illustrating a communication system in which the UE has prediction models, consistent with some embodiments of the present disclosure.

[0014] FIG. 3 is a schematic diagram illustrating a procedure for RRM measurement prediction at the UE, consistent with some embodiments of the present disclosure.

[0015] FIG. 4 is a schematic diagram illustrating a procedure for determining performance of the prediction model for the RRM measurement prediction at the UE, consistent with some embodiments of the present disclosure.

[0016] FIG. 5 is a schematic diagram illustrating a procedure for RRM measurement event prediction at the UE, consistent with some embodiments of the present disclosure.

[0017] FIG. 6 is a schematic diagram illustrating a procedure for determining performance of the prediction model for the RRM measurement event prediction at the UE, consistent with some embodiments of the present disclosure.

[0018] FIG. 7 is a schematic diagram illustrating a procedure for radio link failure prediction or handover failure prediction at the UE, consistent with some embodiments of the present disclosure.

[0019] FIG. 8 is a schematic diagram illustrating a procedure for determining performance of the prediction model for radio link failure prediction or handover failure prediction at the UE, consistent with some embodiments of the present disclosure.

[0020] FIG. 9 is a flow chart illustrating a method for a first node for communication, consistent with some embodiments of the present disclosure.

[0021] FIG. 10 is a flow chart illustrating a method for a second node for communication, consistent with some embodiments of the present disclosure.

[0022] FIG. 11 is a block diagram of a node, consistent with some embodiments of the present disclosure.DETAILED DESCRIPTION

[0023] The implementations set forth in the following description of exemplary embodiments do not represent all implementations consistent with the present disclosure. Instead, they are merely examples of systems, apparatuses, and methods consistent with aspects related to the present disclosure as recited in the appended claims.

[0024] In the present disclosure, the term “node” is used as a general term that includes, but is not limited to, UE, one or more vehicles, one or more vehicle mounted modules, and one or more network infrastructure nodes such as base stations, core networks, roadside units, repeaters, transponders, wireless routers, controllers, access points, and sub-systems thereof. In the present disclosure, the term “network” and the term “network node” are used interchangeably. In the present disclosure, the terms “model” and “models” are used interchangeably. For example, a statement indicating that a network provides a prediction model to a UE may also encompass scenarios in which the network provides one or more prediction models to the UE. In the present disclosure, the terms “RRM measurement event” and “measurement event” are used interchangeably.

[0025] Handover procedure includes a process of transferring an ongoing wireless service and management of a UE from one base station (cell) to another base station (cell). As described in the 3rd Generation Partnership Project (3GPP) New Radio (NR) specifications, six events (Events A1-A6) may be considered as triggering events for intra radio access technology (RAT) handover. The intra RAT handover described in the present disclosure may include a handover in which the serving cell and the target cell are of the same RAT. The six events are discussed below.

[0026] Event A1 indicates an event in which received signal quality of the serving cell becomes better than a predefined threshold. The received signal quality metrics may include reference signal received power (RSRP), reference signal received quality (RSRQ), signal to interference and noise power ratio (SINR), and received signal strength indicator (RSSI). Event A2 indicates an event in which received signal quality of the serving cell becomes worse than a threshold. Event A3 indicates an event in which received signal quality of a neighbor cell becomes offset better than that of the serving cell. Event A4 indicates that received signal quality of a neighbor cell becomes better than a threshold. Event A5 indicates received signal quality of the serving cell becomes worse than threshold 1 and received signal quality of a neighbor cell becomes better than threshold 2. The threshold 1 and the threshold 2 may be the same or different from each other. Event A6 indicates an event in which received signal quality of a neighbor cell becomes offset better than that of the serving cell.

[0027] In addition to Events A1 to A6, two events (Events B1 and B2) may be considered as triggering events for the inter RAT handover. The inter RAT handover described in the present disclosure may include a handover in which the serving cell and the target cell are of the different RATs. Event B1 indicates an event in which received signal quality of an inter RAT neighbor cell becomes better than a threshold. Event B2 indicates an event in which received signal quality of the serving cell becomes worse than threshold 1 and received signal quality of the inter RAT neighbor cell becomes better than threshold 2. The threshold 1 and the threshold 2 may be the same or different from each other.

[0028] To determine whether these events will happen, a UE may perform received signal quality measurement, also known as RRM measurement. Taking the measurement of RSRP as an example, the RSRP measurement may be the Layer-1 filtered measurement, in which the UE measures RSRP on a demodulation reference signal (DMRS) conveyed by the physical broadcast channel (PBCH) or the channel state information (CSI) reference signal (CSI-RS). The UE may further report the measurement results via CSI feedback. The reporting range of the Layer-1 filtered measurement result may be -140 dBm to -40 dBm with 1 dBm resolution. The purpose of Layer-1 filtered measurement may be for beam management and link adaptation.

[0029] The RSRP measurement may also be the Layer-3 filtered measurement, in which the UE measures RSRP on the DMRS conveyed by the PBCH or the CSI-RS. The UE may further report the measurement results via a radio resource control (RRC) message called the measurement report. The reporting range of the Layer-3 filtered measurement result may be -156 dBm to -30 dBm with 1 dBm resolution. The purpose of Layer-3 filtered measurement may be for handover decision.

[0030] The existing handover procedure may suffer from disadvantages. For example, according to the above-noted handover triggering events, a handover procedure may be triggered only after a UE has knowledge that a current wireless service of the UE may be degraded. To determine whether a handover-triggering event will occur, the UE typically performs signal quality measurements of the serving cell. A handover is generally initiated only after the UE detects that the current signal quality has degraded below a predefined threshold--a process that inherently introduces latency due to the time required for measurement and evaluation. In addition, the handover procedure itself involves multiple signaling and coordination steps between the UE and the network, which can further extend the overall handover completion time. At least some embodiments of the present disclosure address the above-noted disadvantages by providing prediction to assist the operation of handover.

[0031] The prediction can be applied at least in three different aspects and there can be three different types of predictions. The first type of prediction is RRM measurement prediction, in which the UE may predict the quality of future-received signals such as RSRP. The UE may perform RRM measurement prediction based on the present and / or past measurement results of the received signal quality. On the other hand, based on the reports of the present and / or past measurement results of the received signal quality, the network may also predict the quality of future received signals. An example of the network is a gNB or a core network. The second type of prediction is measurement event prediction, in which the UE may predict the future occurrence of the handover events, for example, Events A1-A6 or Events B1-B2 described above. The third type of prediction is radio link failure or handover failure prediction, in which the UE may predict the future occurrences of radio link failure or handover failure.

[0032] The above-noted predictions have advantages. For example, as per existing handover events listed above, the handover procedure may be triggered when the received signal quality degrades. In other words, the handover procedure may be triggered when the performance of communications degrades. The prediction may trigger the handover procedure before the serving cell’s signal quality deteriorates, thereby mitigating potential communication degradation. Also, to accurately capture the handover occasions, the UE typically performs frequent measurements and reports of the received signal quality, which introduces substantial measurement overhead. However, by leveraging predictive mechanisms, the UE can reduce the frequency of real-time measurements while still providing predicted signal quality reports. As a result, the overall measurement overhead can be significantly reduced without compromising handover accuracy.

[0033] To perform predictions, in some embodiments of the present disclosure, various prediction models, for example, artificial intelligence and / or machine learning (AI / ML) models may be used. In this case, for example, one or more AI / ML models may be used at the UE side to perform prediction. Alternatively or in addition, one or more AI / ML models may also be used at the network side to perform prediction. When the prediction accuracy is lower than a threshold, a current AI / ML model may not be feasible, and one or more other AI / ML models may be used. The prediction accuracy can be measured with different metrics, such as the average of the difference between the predicted received signal quality and the actual received signal quality.

[0034] FIG. 1 is a schematic diagram illustrating a communication system in which only the network has prediction models, consistent with some embodiments of the present disclosure. Referring to FIG. 1, a communication system 100 includes a UE 102 and a network 104 (e.g., gNB). The network 104 has one or more AI / ML models while the UE 102 does not have an AI / ML model. As such, prediction tasks, such as the RRM measurement prediction, measurement event prediction, radio link failure, and / or handover failure are performed at the network side. The UE 102 may perform RRM measurement and send the RRM measurement report to the network 104. The UE 102 may be one or more UEs served by the network 104. To perform prediction at the network side, the network 104 may utilize the reports of RRM measurement from each of the one or more UEs. Since the prediction is performed solely at the network side, the network needs to predict the RRM measurement and / or the measurement event for each of its serving UEs. As the network typically supports a large number of UEs, executing prediction algorithms centrally at the network side may lead to excessive computational complexity, rendering such an approach impractical at scale. From a practical standpoint, performing prediction at the UE side can reduce the computational burden on the network. By generating its own prediction results, the UE can report only these prediction results rather than transmitting full RRM measurement data. Consequently, the signaling overhead between the UE and the network can also be reduced.

[0035] FIG. 2 is a schematic diagram illustrating a communication system in which the UE has prediction models, consistent with some embodiments of the present disclosure. Referring to FIG. 2, a communication system 200 includes a UE 202 and a network 204 (e.g., gNB). The UE 202 has, for example, one or more AI / ML models for prediction. Accordingly, prediction tasks--such as RRM measurement prediction, measurement event prediction, radio link failure prediction, and / or handover failure prediction--can be performed at the UE side. The UE 202 may be one or more UEs served by the network 204. Under this configuration, performing the prediction tasks at the UE side becomes a practical approach to reduce computational complexity at the network side. To perform the prediction, the UE 202 may perform RRM measurement and utilize its own current and / or historical RRM measurement results as input to the AI / ML model. The UE 202 may also report the prediction results to the network 204 so that the network may make a handover decision.

[0036] However, the network 204 may not be aware of the one or more AI / ML models used for performing the prediction tasks at the UE 202. For example, the network 204 may not have knowledge on whether the one or more AI / ML models used for prediction at the UE 202 can deliver predictions with sufficient accuracy. If these models fail to meet the required prediction performance, the resulting inaccuracies may adversely impact the reliability of the handover decisions made by the network 204. Also, when the communication environment changes, the prediction performance of the AI / ML models may degrade. In this case, the new AI / ML models should be used to adapt to the new communication environment. To this end, an effective scheme to allow the network to detect the infeasibility of the current AI / ML model to ensure the stability, reliability, and performance of UE-side prediction is desired.

[0037] At least some embodiments of the present disclosure address this issue by providing effective schemes that allow the network to evaluate the prediction accuracy of the prediction models executed at the UE side and to supply updated models when the accuracy does not meet predefined criteria. In operation, the network may transmit a prediction model to the UE, and the UE may use the provided model to perform prediction. In addition, the UE may monitor whether the prediction performance of the current model remains acceptable, for example, by determining whether its prediction results satisfy predefined accuracy thresholds. If the prediction performance becomes unacceptable, the UE may notify the network, and the network may provide a new model to the UE, thereby improving the accuracy of subsequent predictions.

[0038] Although the methods of the present disclosure are explained in the context of handover, the scope of the present disclosure is not so limited. The methods described in the present disclosure can be applied to any prediction-based communications between a UE and a network, or more generally, between two or more nodes. Although the present disclosure generally refers to AI / ML models for predictions, the prediction models described herein are not limited to AI / ML and may include any type of model--whether currently known or developed in the future--that is capable of performing the intended predictive functions. For example, the prediction models may include, but are not limited to, long short-term memory (LSTM), transformer-based prediction models, binary classification HO-trigger models, gated recurrent unit (GRU), multi-layer perception (MLP), convolutional neural network (CNN), k-nearest neighbors (KNN), reinforcement learning (RL), decision trees, random forests, deep neural network (DNN), graph neural network (GNN), Bayesian networks, federated learning aggregators, ensemble model, etc.

[0039] FIG. 3 is a schematic diagram illustrating a procedure for RRM measurement prediction, consistent with some embodiments of the present disclosure. Referring to FIG. 3, a procedure 300 is used for performing RRM measurement prediction at a UE 302. The UE 302 may be one or more UEs served by a network 304 (e.g., gNB).

[0040] Referring to FIG. 3, the procedure 300 may include a step 306 at which the UE 302 receives a request from the network 304. The request may include an instruction to perform an RRM measurement prediction at the UE 302. The request may also include configuration information for the UE 302. The configuration information may also include information for configuring at least one of: an RRM measurement, an RRM measurement prediction, an RRM measurement period, an RRM measurement prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, or a report period of RRM measurement prediction. The configuration may specify radio quality metrics for the RRM measurement prediction or the RRM measurement. The radio quality metrics may include at least one of: RSRP, RSRQ, RSSI, signal-to-noise ratio (SNR), or SINR of at least one of: a serving cell, or one or more neighboring cells. The one or more neighboring cells can be of the same or different RAT as the serving cell. The reference signal received by the UE 302 may be any reference signal, such as DMRS, CSI-RS, synchronization signal block (SSB), and positioning reference signal (PRS), etc.

[0041] The configuration may also include one or more criteria for use by the UE 302 to evaluate whether the performance of the prediction model used for RRM measurement prediction remains acceptable. Such criteria may include, for example, a target prediction-accuracy threshold, an allowable error range, a confidence-level requirement, or other performance indicators defined by the network 304. The UE 302 may compare its locally evaluated prediction-accuracy metrics against the criteria provided by the network 304 to determine whether the current prediction model satisfies the network-configured performance expectations.

[0042] The request received from the network 304 may also include an indication of the operations that the UE 302 is to perform if the performance of the current prediction model is determined to be unacceptable. These subsequent operations may include, for example, transmitting an indication to the network 304 indicating that the current prediction model no longer meets the required performance level, suspending or discontinuing the generation or reporting of RRM measurement prediction results, or initiating a request for a new or updated prediction model. By providing such criteria and operational instructions, the network 304 may ensure that the UE 302 can autonomously evaluate prediction-model performance while maintaining consistent reporting behavior and supporting adaptive prediction-based RRM procedures.

[0043] The procedure 300 may include a step 308 at which the network 304 sends a prediction model (e.g., AI / ML model) for RRM measurement prediction. The prediction model may be any model--whether currently known or developed in the future--that is capable of performing RRM measurement prediction at a UE. For example, the prediction model may be capable of predicting at least one of: RSRP, RSRQ, RSSI, SNR, or SINR values of at least one of: a serving cell, or one or more neighboring cells, at a future time point. By delivering the prediction model explicitly, the network 304 may ensure alignment between UE-side prediction behavior and network expectations, thereby maintaining consistent prediction performance and supporting stable mobility management.

[0044] In some embodiments, the network 304 may send the request and the prediction model for RRM measurement prediction, on a periodic basis. For example, the network 304 may be configured to provide updated criteria, operational parameters, and prediction-model data at predetermined intervals in order to ensure that the UE 302 consistently operates with up-to-date model information and remains aligned with the network’s RRM framework. Such periodic transmission may allow the network 304 to proactively refresh prediction models in response to evolving radio-environment conditions or system-level performance considerations.

[0045] In some embodiments, the network 304 may send the request information and the prediction model in an event-triggered manner. For example, the network 304 may send the request and the prediction model upon receiving, from the UE 302, an indication that the performance of the currently used prediction model is unacceptable. Such an indication may reflect, for example, that the UE 302 has determined that the prediction accuracy falls below a threshold or fails to meet predetermined criteria provided by the network 304. Upon receipt of such an indication, the network 304 may select or generate an updated prediction model and send it to the UE 302, along with revised instructions or criteria as needed to restore acceptable prediction performance.

[0046] In some embodiments, the order of steps 306 and 308 may be reversed. For example, the network 304 may first transmit the prediction model to the UE 302 and, after the model has been delivered, subsequently transmit the associated configuration information. In some embodiments, the steps 306 and 308 may be combined into one step. For example, the network 304 may transmit the request and the prediction model to the UE 302 at the same time.

[0047] In some embodiments, the procedure 300 may include an additional step (not shown) at which the UE 302 notifies the network 304 that the prediction model for RRM measurement prediction is not applicable at the UE 302. (As used in this disclosure, the the terms “applicable” and “acceptable” have the same meaning.) For example, after receiving the prediction model for RRM measurement prediction, the UE 302 may determine whether the prediction model is supported by the UE 302 so that it can be executed at the UE 302. The UE 302 may make the determination based on the parameters and constraints of the UE 302, for example, but not limited to, computational capacity, battery life, implementation complexity, memory constraints, and mobility of the UE 302. Upon determination that the prediction model does not satisfy one or more of the parameters and constraints of the UE 302, the UE 302 may send a notification to the network 304. In response, the network 304 may send a new prediction model that can be supported by the UE 302 and executable at the UE 302.

[0048] The procedure 300 may include a step 310 at which the UE 302 performs RRM measurement prediction using the prediction model provided by the network 304. For example, the UE 302 may use current RRM measurement results and / or historical RRM measurement results as input to the prediction model, and execute the prediction model to estimate RRM measurement values at a future time point or a future time window. For example, the UE 302 may estimate at least one of: RSRP, RSRQ, RSSI, SNR, or SINR values of a serving cell and / or one or more neighboring cells at a future time point or a future time window.

[0049] The procedure 300 may include a prediction period 312 during which the UE 302 performs RRM measurement prediction. In some embodiments, the request received from the network 304 at the step 306 may include the beginning time and period (duration) of the RRM measurement prediction, and the UE 302 may perform the RRM measurement prediction based on the beginning time and the period (duration) indicated in the request. In some embodiments, the period of RRM measurement prediction may include a plurality of sample periods of RRM measurement prediction and the UE 302 may perform an RRM measurement prediction at each of the plurality of sample periods of RRM measurement. Based on the predicted RRM measurements, the UE 302 may transmit the corresponding RRM measurement prediction results to the network 304. In some embodiments, the request received from the network 304 may include the report period for reporting the predicted RRM measurements, and the UE 302 may transmit the RRM measurement prediction results to the network 304 based on the report period indicated in the request. In some embodiments, the report period may include a plurality of sample report periods and the UE 302 may transmit the RRM measurement prediction results to the network 304 at each of the plurality of sample report periods. For example, as shown in FIG. 3, a sample report period 314 may be a period of time between two consecutive reports sent to the network 304. Such reporting enables the network 304 to utilize the prediction results when performing radio-resource management operations, including mobility decisions.

[0050] In some embodiments, along with the RRM measurement prediction, the UE 302 may also perform RRM measurement. The RRM measurement may be periodic or aperiodic RRM measurements. In some embodiments, the request received from the network 304 may include the beginning time and period (duration) of the RRM measurement, and the UE 302 may perform the RRM measurement based on the beginning time and the period (duration) indicated in the request. In some embodiments, the period of RRM measurement may include a plurality of sample periods of RRM measurement and the UE 302 may perform an RRM measurement at each of the plurality of sample periods of RRM measurement. The UE 302 may measure at least one of: RSRP, RSRQ, RSSI, SNR, or SINR values of at least one of: a serving cell, or one or more neighboring cells. The RRM measurements performed by the UE 302 may be the Layer-1 measurements and / or Layer-3 measurements. Based on the RRM measurements, the UE 302 may generate the RRM measurement report. The UE 302 then transmits the RRM measurement report to the network 304. The UE 302 may transmit the RRM measurement report periodically or aperiodically. In some embodiments, the request received from the network 304 may include the beginning time and period (duration) of the report, and the UE 302 may transmit the RRM measurement report to the network 304 based on the beginning time and the period (duration) for report indicated in the request. In some embodiments, the report period may include a plurality of sample report periods and the UE 302 may transmit the RRM measurement report to the network 304 at each of the plurality of sample report periods.

[0051] In some embodiments, the sample report period of RRM measurement and the sample report period of RRM measurement prediction are the same. In this case, the sample report period of RRM measurement (or the sample report period of RRM measurement prediction) may be indicated in the same field in the request transmitted from the network 304. In some embodiments, the sample report period of RRM measurement and the sample report period of RRM measurement prediction are different. In this case, the sample report period of RRM measurement and the sample report period of RRM measurement prediction may be indicated in two different fields in the request transmitted from the network 304.

[0052] The procedure 300 may include a step 316 at which the UE 302 transmits an indication to the network 304 to notify the network that the performance of the current prediction model is unacceptable or no longer acceptable. The indication may be generated by the UE 302 based on monitoring a prediction accuracy of the RRM measurement prediction performed by the UE 302. The UE 302 may evaluate the performance of the current prediction model by determining whether the accuracy of its RRM measurement prediction results satisfies one or more predetermined criteria or thresholds. For example, the UE 302 may compare its locally evaluated prediction-accuracy metrics against the criteria provided by the network 304 to determine whether the current prediction model satisfies the network-configured performance expectations. If the UE 302 determines that the prediction-accuracy does not meet the predetermined criteria, the UE 302 may transmit an indication specifying that the performance of the prediction model is unacceptable and / or a request for the network 304 to provide an updated or alternative prediction model. Conversely, when the UE 302 determines that the accuracy of the RRM measurement prediction results is acceptable, the UE 302 may continue operating the current prediction model and may further continue generating and transmitting the corresponding RRM measurement prediction results to the network 304 in accordance with the configured reporting procedures. In some embodiments, upon determining that the performance of the current prediction model is unacceptable, the UE 302 may discontinue the reporting of RRM measurement prediction results to the network 304. The network 304 may, in turn, infer the prediction-model performance of the UE 302 based on the absence of the RRM measurement prediction reports during the configured reporting period, and may take appropriate actions in response.

[0053] The procedure 300 may include a step 318 at which the network 304 sends a new or updated prediction model for the RRM measurement prediction to the UE 302. The new prediction model may be supplied in response to a request from the UE 302, a determination by the network 304 that the current model is outdated or underperforming, or as part of a periodic model-update mechanism configured by the network 304. The UE 302 may store and subsequently utilize the newly received prediction model to continue performing RRM measurement prediction operations. In this way, the network 304 and the UE 302 can cooperatively ensure accuracy of the RRM measurement prediction at the UE, thereby enhancing the overall performance and reliability of prediction-based RRM procedures.

[0054] FIG. 4 is a schematic diagram illustrating a procedure for determining performance of the prediction model for the RRM measurement prediction, consistent with some embodiments of the present disclosure. Referring to FIG. 4, a UE, such as the UE 302 of FIG. 3, may perform a procedure 400 of determining performance of the prediction model. The UE may include a prediction model 402 for performing the RRM measurement prediction. The prediction model 402 may be provided by a network, such as the network 304 of FIG. 3. As shown in FIG. 4, the current and / or historical RRM measurement data collected at the UE are provided to the prediction model 402 as input. Such measurement data may include, for example, currently and / or previously observed RSRP, RSRQ, RSSI, SINR values, or other signal-quality metrics associated with serving and / or one or more neighboring cells. The prediction model 402 processes these input measurements data and generates corresponding RRM measurement prediction results as output. These prediction results represent the model’s estimation of future RRM measurement values and may be subsequently used to support prediction-assisted radio resource management procedures, including handover decisions, at the network side.

[0055] Based on the RRM measurement prediction results, the UE may determine the accuracy of the prediction model 402. As shown in FIG. 4, the RRM measurement prediction results output from the prediction model 402 are provided to a comparison module 404 of the UE. The comparison module 404 may be software, hardware, or a combination of software and hardware. In addition, the current and / or historical RRM measurement data, after a certain delay time, are also provided to the comparison module 404. The comparison module 404 compares the RRM measurement prediction results with the current and / or historical RRM measurement data, determines accuracy metrics (e.g., statistical error metrics) of the prediction. The comparison module may compare the determined accuracy metrics with a certain criterion to determine whether the performance of the prediction model meets the performance expectation. The criterion may be received from the network. If the comparison result indicates that the difference between the computed accuracy metrics and the criterion set by the network exceeds a predetermined threshold (corresponding to ‘Yes’ in FIG. 4), the UE determines that the performance of the prediction model used for the RRM measurement prediction is unacceptable. In response, the UE may transmit an indication to the network to notify the network that the current prediction model fails to meet the required performance level. The delay is needed before the accuracy determination as the true measurement (the ground truth) is known at a time later than the prediction time.

[0056] FIG. 5 is a schematic diagram illustrating a procedure for RRM measurement event prediction, consistent with some embodiments of the present disclosure. Referring to FIG. 5, a procedure 500 is used for performing RRM measurement event prediction at a UE 502. The UE 502 may be one or more UEs served by a network 504 (e.g., gNB).

[0057] Referring to FIG. 5, the procedure 500 may include a step 506 at which the UE 502 receives a request from the network 504. The request may include an instruction to perform an RRM measurement event prediction at the UE 502. The request may also include configuration information for the UE 502. The configuration information may include information for configuring at least one of: an RRM measurement, an RRM measurement event prediction, an RRM measurement period, a sample period of RRM measurement, a report period of RRM measurement, an RRM measurement event prediction period, or a report period of RRM measurement event prediction.

[0058] The configuration may also include one or more criteria for use by the UE 502 to evaluate whether the performance of the prediction model used for RRM measurement event prediction remains acceptable. Such criteria may include, for example, a target prediction-accuracy threshold, an allowable error range, a confidence-level requirement, or other performance indicators defined by the network 504. The UE 502 may compare its locally evaluated prediction-accuracy metrics against the provided criteria to determine whether the current prediction model satisfies the network-configured performance expectations. In addition, the request from the network 504 may further include an indication of the operations that the UE 502 is to perform if the performance of the prediction model is determined to be unacceptable. These subsequent operations may include, for example, transmitting an indication to the network 504 indicating that the prediction model no longer meets the required performance level, suspending or discontinuing the generation or reporting of RRM measurement event prediction results, or initiating a request for a new or updated prediction model. By providing such criteria and operational instructions, the network 504 may ensure that the UE 502 can autonomously evaluate prediction-model performance while maintaining consistent reporting behavior and supporting adaptive prediction-based RRM procedures.

[0059] In some embodiments, the request transmitted from the network 504 may optionally specify one or more measurement events for which the UE 502 is to perform prediction. For example, the network 504 may indicate that the UE 502 is to generate prediction results for specific RRM measurement events, such as Event A2, Event A3, Event B1, or any other relevant event. When such an explicit indication is provided, the UE 502 may configure its prediction procedures accordingly, ensuring that the prediction model and reporting behavior align with the designated event(s). In some embodiments, the request may omit any explicit identification of a target measurement event. In this case, the UE 502 may rely on default prediction behavior, predetermined configuration stored locally, or event-specific parameters previously provided by the network 504. Alternatively, the UE 502 may apply model-specific logic or network-defined priorities to determine which events are applicable for prediction in the absence of explicit network indication.

[0060] In some embodiments, the request transmitted by the network may further specify one or more occasions or resources on which the UE 502 is to report prediction results. Such occasions or resources may include reporting opportunities defined in terms of specific time instances, slot or subframe boundaries, reference-signal occasions, or other uplink or downlink transmission resources available within the radio access network. The network may additionally indicate whether the reporting of RRM measurement event prediction results is to be performed periodically or aperiodically. In the case of periodic reporting, the request may define a reporting interval, periodicity value, or other timing parameter that governs how frequently the UE 502 is to transmit prediction results. The periodicity may be uniform across all prediction models or dynamically adjusted based on network load, service requirements, or prediction accuracy. Alternatively, in the case of aperiodic reporting, the request may specify one or more triggering conditions that cause the UE 502 to transmit prediction results on demand.

[0061] In some embodiments, the request may also indicate whether the UE 502 is to perform a one-step measurement event prediction or a two-step measurement event prediction. For the two-step measurement event prediction, the UE 502 may perform the RRM measurement prediction first, and then use the results of the RRM measurement prediction to perform the measurement event prediction. In this case, the UE 502 also includes a prediction model for RRM measurement prediction. For the one-step measurement event prediction, the UE 502 may directly perform the measurement event prediction.

[0062] The procedure 500 may include a step 508 at which the network 504 sends a prediction model (e.g., AI / ML model) for RRM measurement event prediction. The prediction model may be any model--whether currently known or developed in the future--that is capable of performing RRM measurement event prediction at a UE. The prediction model is not limited to any particular AI / ML architecture, algorithmic framework, or computational approach. For example, the prediction model may be based on statistical filtering, regression analysis, neural networks, deep-learning architectures, time-series forecasting methods, ensemble models, transformer-based architectures, or any hybrid or proprietary technique capable of generating RRM measurement event prediction results. Future-developed modeling methodologies or enhanced AI / ML frameworks may likewise be employed, provided they are capable of receiving relevant inputs (e.g., RRM measurement data, historical RRM measurement event prediction data, etc.) and generating corresponding prediction outputs. By delivering the prediction model explicitly, the network 504 may ensure alignment between UE-side prediction behavior and network expectations, thereby maintaining consistent prediction performance and supporting stable mobility management.

[0063] In some embodiments, the network 504 may send the prediction model for RRM measurement event prediction, on a periodic basis. For example, the network 504 may provide updated criteria, operational parameters, and prediction-model data at predetermined intervals in order to ensure that the UE 502 consistently operates with up-to-date model information and remains aligned with the network’s prediction-based RRM framework. Such periodic transmission may allow the network 504 to proactively refresh prediction models in response to evolving radio-environment conditions or system-level performance considerations.

[0064] In some embodiments, the network 504 may send the RRM measurement event prediction model in an event-triggered manner. For example, the network 504 may send the prediction model upon receiving, from the UE 502, an indication that the performance of the currently used prediction model is unacceptable. Such an indication may reflect, for example, that the UE 502 has determined that the prediction accuracy falls below a threshold or fails to meet predetermined criteria provided by the network 504. Upon receipt of such an indication, the network 504 may select or generate an updated prediction model and send it to the UE 502, along with revised instructions or criteria as needed to restore acceptable prediction performance.

[0065] In some embodiments, the order of steps 506 and 508 may be reversed. For example, the network 504 may first transmit the prediction model to the UE 502 and, after the model has been delivered, subsequently transmit the associated configuration information. In some embodiments, the steps 506 and 508 may be combined into one step. For example, the network 504 may transmit the request and the prediction model at the same time.

[0066] In some embodiments, the procedure 500 may include an additional step (not shown) at which the UE 502 notifies the network 504 that the prediction model for RRM measurement event prediction is not applicable at the UE 502. For example, after receiving the prediction model for RRM measurement event prediction, the UE 502 may determine whether the prediction model is supported by the UE 502 and can be executed at the UE 502. The UE 502 may make the determination based on the parameters and constraints of the UE 502, for example, but not limited to, computational capacity, battery life, implementation complexity, memory constraints, and mobility of the UE 502. Upon determination that the prediction model does not satisfy one or more of the parameters and constraints of the UE 502, the UE 502 may send a notification to the network 504. In response, the network 504 may send a new prediction model that can be supported by the UE 502 and executable at the UE 502.

[0067] The procedure 500 may include a step 510 at which the UE 502 performs RRM measurements event prediction using the prediction model provided by the network 504. For example, the UE 502 may use current RRM measurement results, historical RRM measurement results, or any combination thereof as input to the prediction model. The input measurements may include, for example, recently observed or time-filtered values of RSRP, RSRQ, RSSI, SINR, or other RRM-related metrics associated with serving and / or one or more neighboring cells. Based on such measurement inputs, the prediction model generates an estimation as to whether a particular RRM measurement event is predicted to occur at a specified future time point or within a future prediction window. The prediction model may analyze temporal patterns, measurement trends, or other characteristics derived from the input data to determine the likelihood and / or timing of the event.

[0068] The procedure 500 may include a period 512 during which the UE 502 performs RRM measurement event prediction. In some embodiments, the request received from the network 504 at the step 506 may include the beginning time and period (duration) of the RRM measurement event prediction, and the UE 502 may perform the RRM measurement event prediction based on the beginning time and the period (duration) indicated in the request. In some embodiments, for a two-step prediction procedure, more than one prediction model may be employed at the UE 502. In the first step, a prediction model may receive, as input, the RRM measurements collected by the UE 502 during the measurement period and may output corresponding RRM measurement prediction results. In the second step, another prediction model may utilize the RRM measurement prediction results generated in the first step as its input and may produce, as output, an RRM measurement event prediction. This second-stage model may determine, for example, whether a specific measurement event (e.g., A2, A3, or B1 event) is expected to occur within a future measurement-event prediction period, based on the predicted RRM measurements rather than solely on historical or currently observed measurements.

[0069] If the request received from the network 504 specifies the particular measurement event(s) to be predicted, the UE 502 evaluates whether the indicated event(s) are expected to occur within the RRM measurement event prediction period 512 and generates the corresponding prediction results. The UE 502 then transmits these prediction results to the network 504 using the reporting occasions and / or resources indicated in the request. If the request received from the network 504 does not specify which measurement event(s) are to be predicted, the UE 502 may determine whether any RRM measurement event is expected to occur within the RRM measurement event prediction period 512. The UE 502 then generates the corresponding prediction result and transmits the result to the network 504 using the reporting occasions and / or resources indicated in the request.

[0070] The procedure 500 may include a step 514 at which the UE 502 transmits an indication to the network 504 to notify the network that the performance of the current prediction model is unacceptable. The indication may be generated by the UE 502 based on monitoring a prediction accuracy of the RRM measurement event prediction performed by the UE 502. In some embodiments, the UE 502 may evaluate the performance of the current prediction model by determining whether the accuracy of its RRM measurement event prediction results satisfies one or more predetermined criteria or thresholds. For example, the UE 502 may compare its locally evaluated prediction-accuracy metrics against the criteria provided by the network 504 to determine whether the current prediction model satisfies the network-configured performance expectations. If the UE 502 determines that the prediction-accuracy does not meet the predetermined criteria, the UE 502 may send an indication specifying that the prediction accuracy is unacceptable and / or a request for the network 504 to provide an updated or alternative prediction model. Conversely, when the UE 502 determines that the accuracy of the RRM measurement prediction results is acceptable, the UE 502 may continue operating the current prediction model and may further continue generating and transmitting the corresponding RRM measurement event prediction results to the network 504 in accordance with the configured reporting procedures.

[0071] The procedure 500 may include a step 516 at which the network 504 sends a new or updated prediction model associated with RRM measurement event prediction to the UE 502. The new prediction model may be supplied in response to a request from the UE 502, a determination by the network 504 that the current model is outdated or underperforming, or as part of a periodic model-update mechanism configured by the network 504. The UE 502 may store and subsequently utilize the newly received prediction model to continue performing RRM measurement event prediction operations. In this way, the network 504 and the UE 502 can cooperatively ensure accuracy of the RRM measurement event prediction at the UE, thereby enhancing the overall performance and reliability of prediction-based RRM procedures.

[0072] FIG. 6 is a schematic diagram illustrating a procedure for determining performance of the prediction model for the RRM measurement event prediction, consistent with some embodiments of the present disclosure. Referring to FIG. 6, a UE, such as the UE 502 of FIG. 5, may perform a procedure 600 of determining performance of the prediction model for RRM measurement event prediction. The UE may include a prediction model 602 for performing the RRM measurement event prediction. The prediction model 602 may be provided by a network, such as the network 504 of FIG. 5. As shown in FIG. 6, the current and / or historical RRM measurement data collected at the UE are provided to the prediction model 602 as input. Such measurement data may include, for example, previously observed RSRP, RSRQ, RSSI, SINR values, or other signal-quality metrics associated with serving and / or one or more neighboring cells. The prediction model 602 processes these input measurements and generates corresponding RRM measurement event prediction results as output. These prediction results may be subsequently used to support prediction-assisted radio resource management procedures at the network side.

[0073] Based on the RRM measurement event prediction, the UE may determine the accuracy of the prediction model 602. For example, as shown in FIG. 6, the RRM measurement event prediction results output from the prediction model 602 is provided to a comparison module 604 of the UE. The comparison module 604 may be software, hardware, or a combination of software and hardware. Also, the current and / or historical RRM measurement data, after a certain delay time, are provided to the comparison module 604. The comparison module 604 may compare the predicted RRM measurement event with the actual observed events and determines accuracy metrics. The comparison module 604 may compare the accuracy metrics with a predetermined criterion. The predetermined criterion may be received from the network. If the comparison result indicates that the difference between the computed accuracy metrics and the criterion set by the network exceeds a predetermined threshold (corresponding to ‘Yes’ in FIG. 6) --the UE determines that the performance of the prediction model used for the RRM measurement event prediction is unacceptable. In response, the UE may transmit an indication to the network to notify the network that the current prediction model fails to meet the required performance level.

[0074] FIG. 7 is a schematic diagram illustrating a procedure for radio link failure prediction and / or handover failure prediction, consistent with some embodiments of the present disclosure. Referring to FIG. 7, a procedure 700 is used for performing radio link failure (RLF) prediction and / or handover (HO) failure prediction at a UE 702. The UE 702 may be one or more UEs served by a network 704 (e.g., gNB).

[0075] Referring to FIG. 7, the procedure 700 may include a step 706 at which the UE 702 receives a request from the network 704. The request may include an instruction to perform an RLF prediction and / or HO failure prediction at the UE 702. The request may also include configuration information for the UE 702. The configuration information may include information for configuring at least one of: an RRM measurement, an RRM measurement period, a sample period of RRM measurement, an RLF prediction period, or an HO failure prediction period. The configuration may also include one or more criteria for use by the UE 702 to evaluate whether the performance of the prediction model used for the RLF prediction or the prediction model used for the HO failure prediction remains acceptable. Such criteria may include, for example, a target prediction-accuracy threshold, an allowable error range, a confidence-level requirement, or other performance indicators defined by the network 704. The UE 702 may compare its locally evaluated prediction-accuracy metrics against the provided criteria to determine whether the current prediction model satisfies the network-configured performance expectations.

[0076] The request from the network 704 may also include an indication of the operations that the UE 702 is to perform if the performance of the prediction model is determined to be unacceptable. These subsequent operations may include, for example, transmitting an indication to the network 704 indicating that the prediction model no longer meets the required performance level, suspending or discontinuing the generation or reporting of RLF prediction or HO failure prediction results, or initiating a request for a new or updated prediction model. By providing such criteria and operational instructions, the network 704 ensures that the UE 702 can autonomously evaluate prediction-model performance while maintaining consistent reporting behavior and supporting adaptive prediction-based RRM procedures.

[0077] The request transmitted from the network 704 may specify whether the UE 702 is to perform prediction of RLF, HO failure, or both. That is, the request may explicitly identify the failure type(s) to be predicted, enabling the UE 702 to tailor its prediction operations to the designated failure category or categories. Alternatively, the request may omit such an indication, in which case the UE 702 may apply default rules, previously provisioned settings, or model-specific logic to determine whether RLF prediction, HO failure prediction, or both should be executed during the configured prediction period.

[0078] In some embodiments, the request transmitted by the network 704 may specify one or more reporting occasions and / or uplink resources that the UE 702 is to use when transmitting the prediction results. Such reporting occasions or resources may correspond to particular time instances, slot or subframe boundaries, reference-signal occasions, or other uplink transmission opportunities available within the radio access network. The request may also indicate whether the reporting is to be performed periodically or aperiodically. In some embodiments, the request may also indicate whether the UE 702 is to perform a one-step RLF and / or HO failure prediction or a two-step RLF / HO failure prediction. For the two-step RLF and / or HO failure prediction, the UE 702 may perform the RRM measurement prediction first, and then use the results of the RRM measurement prediction to perform the RLF and / or HO failure prediction. In this case, the UE 702 also includes a prediction model for RRM measurement prediction. For the one-step measurement event prediction, the UE 702 may directly perform the RLF and / or HO failure prediction.

[0079] The procedure 700 may include a step 708 at which the network 704 sends a prediction model (e.g., AI / ML model) for RLF / HO failure prediction. The prediction model may be any model--whether currently known or developed in the future--that is capable of performing RLF and / or HO failure prediction at a UE. By delivering the prediction model explicitly, the network 704 can ensure alignment between UE-side prediction behavior and network expectations, thereby maintaining consistent prediction performance and supporting stable mobility management.

[0080] In some embodiments, the network 704 may send the request and the prediction model for RLF and / or HO failure prediction, on a periodic basis. For example, the network 704 may be configured to provide updated criteria, operational parameters, and prediction-model data at predetermined intervals in order to ensure that the UE 702 consistently operates with up-to-date model information and remains aligned with the network’s prediction-based RRM framework. Such periodic transmission may allow the network 704 to proactively refresh prediction models in response to evolving radio-environment conditions, updated training results, or system-level performance considerations.

[0081] In some embodiments, the network 704 may send the request and the prediction model in an event-triggered manner. For example, the network 704 may send the request and the prediction model upon receiving, from the UE 702, an indication that the performance of the currently used prediction model is unacceptable. Such an indication may reflect, for example, that the UE 702 has determined that the prediction accuracy falls below a threshold or fails to meet predetermined criteria provided by the network 704. Upon receipt of such an indication, the network 704 may select or generate an updated prediction model and send it to the UE 702, along with revised instructions or criteria as needed to restore acceptable prediction performance.

[0082] In some embodiments, the order of steps 706 and 708 may be reversed. For example, the network 704 may first transmit the prediction model to the UE 702 and, after the model has been delivered, subsequently transmit the associated configuration information. In some embodiments, the steps 706 and 708 are combined into one step. For example, the network 704 may transmit the prediction model and the request at the same time.

[0083] In some embodiments, the procedure 700 may include an additional step (not shown) at which the UE 702 notifies the network 704 that the prediction model for RRM measurement prediction is not applicable at the UE 702. For example, after receiving the prediction model for RRM measurement prediction, the UE 702 may determine whether the prediction model is supported by the UE 702 and can be executed at the UE 702. The UE 702 may make the determination based on the parameters and constraints of the UE 702, for example, but not limited to, computational capacity, battery life, implementation complexity, memory constraints, and mobility of the UE 702. Upon determination that the prediction model does not satisfy one or more of the parameters and constraints of the UE 702, the UE 702 may send a notification to the network 704. In response, the network 704 may send a new prediction model that can be supported by the UE 702 and executable at the UE 702.

[0084] The procedure 700 may include a step 710 at which the UE 702 performs RLF prediction and / or HO failure prediction using the prediction model provided by the network 704. For example, the UE 702 may use current RRM measurement results and / or historical RRM measurement results as input to the model, and execute the model to estimate occurrence of RLF failure or HO failure at a future time point or a future time window. By executing the network-provided prediction model, the UE 702 generates forecasts of future occurrence of RLF failure and / or HO failure in a manner that aligns with the prediction logic, accuracy requirements, and operational parameters defined by the network.

[0085] In some embodiments, for a two-step prediction procedure, more than one prediction model may be employed at the UE 702. In the first step, a prediction model may receive, as input, the RRM measurements collected by the UE 502 during the measurement period and may output corresponding RRM measurement prediction results. In the second step, another prediction model may utilize the RRM measurement prediction results generated in the first step as its input and may produce, as output, an RLF prediction and / or HO failure prediction.

[0086] The procedure 700 may include a period 712 during which the UE 702 performs RLF and / or HO failure prediction. In some embodiments, the request received from the network 704 at the step 706 may include the beginning time and period (duration) of the RLF and / or HO failure prediction, and the UE 702 may perform the RLF and / or HO failure prediction based on the beginning time and the period (duration) indicated in the request. Based on the predicted RLF and / or HO failure, the UE 702 may transmit the corresponding RLF and / or HO failure prediction results to the network 704.

[0087] In some embodiments, if the request received from the network 704 specifies whether RLF, HO failure, or both are to be predicted, the UE determines whether the indicated failure type is expected to occur within the configured RLF / HO failure prediction period 712. The UE 702 then generates the corresponding prediction result and transmits the result to the network 704 using the reporting occasions and / or resources indicated in the request. In some embodiments, if the request received from the network 704 does not specify whether RLF, HO failure, or both are to be predicted, the UE 702 may determine whether any of these failure types are expected to occur within the configured RLF and / or HO failure prediction period 712. The UE 702 then generates the corresponding prediction result and transmits the result to the network 704 using the reporting occasions and / or uplink resources indicated in the request.

[0088] The procedure 700 may include a step 714 at which the UE 702 transmits an indication to the network 704 to notify the network that the performance of the current prediction model is unacceptable. The indication may be generated by the UE 702 based on monitoring a prediction accuracy of the RLF and / or HO failure prediction performed by the UE 702. In some embodiments, the UE 702 may evaluate the performance of the current prediction model by determining whether the accuracy of its RLF and / or HO failure prediction results satisfy one or more predetermined criteria. For example, the UE 702 may compare its locally evaluated prediction-accuracy metrics against the criteria provided by the network 704 to determine whether the current prediction model satisfies the network-configured performance expectations. If the UE 702 determines that the prediction-accuracy does not meet the predetermined criteria, the UE 702 may transmit an indication to the network 704 specifying that the prediction accuracy is unacceptable and / or a request for the network 704 to provide an updated or alternative prediction model. Conversely, when the UE 702 determines that the accuracy of the RLF and / or HO failure prediction results is acceptable, the UE 702 may continue operating the current prediction model and may further continue generating and transmitting the corresponding RLF and / or HO failure prediction results to the network 704 in accordance with the configured reporting procedures.

[0089] The procedure 700 may include a step 716 at which the network 704 sends a new or updated prediction model associated with RLF and / or HO failure prediction to the UE 702. The new prediction model may be supplied in response to a request from the UE 702, a determination by the network 704 that the current model is outdated or underperforming, or as part of a periodic model-update mechanism configured by the network 704. The UE 702 may store and subsequently utilize the newly received prediction model to continue performing RLF and / or HO failure prediction operations. In this way, the network 704 and the UE 702 can cooperatively ensure accuracy of the RLF and / or HO failure prediction at the UE, thereby enhancing the overall performance and reliability of prediction-based RRM procedures.

[0090] FIG. 8 is a schematic diagram illustrating a procedure for determining performance of the prediction model for the RLF and / or HO failure prediction, consistent with some embodiments of the present disclosure. Referring to FIG. 8, a UE, such as the UE 702 of FIG. 7, may perform a procedure 800 of determining performance of the prediction model for the RLF and / or HO failure prediction. The UE may include one or more prediction model 802 for performing the RLF and / or HO failure prediction. The prediction model 802 may be provided by a network, such as the network 704 of FIG. 7. As shown in FIG. 8, the current and / or historical RRM measurement data collected at the UE are provided to the prediction model 802 as input. Such measurement data may include, for example, previously observed RSRP, RSRQ, RSSI, SINR values, or other signal-quality metrics associated with serving and / or neighboring cells. The prediction model 802 processes these input measurements and generates corresponding RLF and / or HO failure prediction results as output. These prediction results represent the model’s estimation of future RLF and / or HO failure and may be subsequently used to support prediction-assisted radio resource management procedures at the network side.

[0091] Based on the RLF and / or HO failure prediction, the UE may determine the accuracy of the prediction model 802 using the current and / or historical RRM measurement data. For example, as shown in FIG. 8, the RLF and / or HO failure prediction results output from the prediction model 802 are provided to a comparison module 804 of the UE. The comparison module 804 may be software, hardware, or a combination of software and hardware. In addition, the current and / or historical RRM measurement data, after a certain delay time, are provided to the comparison module 804. The comparison module 804 may compare the RLF and / or HO failure prediction results with the actual observed failure events, and determine accuracy metrics. The comparison module 804 may compare the accuracy metrics with a criterion. The criterion may be received from the network. If the comparison result indicates that the difference between the computed accuracy metrics and the criterion set by the network exceeds a predetermined threshold (corresponding to ‘Yes’ in FIG. 8) --the UE determines that the performance of the prediction model used for the RLF and / or HO failure prediction is no longer acceptable. In response, the UE transmits an indication to the network to notify the network that the current prediction model fails to meet the required performance level. The delay is needed before the accuracy determination as the observation of the actual failure occurs at a time later than the prediction time.

[0092] FIG. 9 is a flow chart showing a method for a first node for communication, consistent with some embodiments of the present disclosure. The first node may include at least one UE, such as the UE 302 of FIG. 3, the UE 502 of FIG. 5, or the UE 702 of FIG. 7.

[0093] Referring to FIG. 9, a method 900 includes a step 902 of receiving, from a second node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction. The second node may include at least one of: a base station, a core network, a road-side unit, a repeater, a transponder, a wireless router, a controller, or an access point. The second node may be a network node (terrestrial or non-terrestrial) to which the first node is connected (e.g., via wireless communication), such as the network 304 of FIG. 3, the network 504 of FIG. 5, or the network 704 of FIG. 7. In some embodiments, the request may include configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction. In some embodiments, one or more metrics for the one or more RRM measurements or the RRM measurement prediction comprise at least one of: RSRP, RSRQ, RSSI, SNR, or SINR values of at least one of: a serving cell, or one or more neighboring cells.

[0094] The method 900 includes a step 904 of receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the RRM measurement event prediction, the radio link failure prediction, or the handover failure prediction. The one or more prediction models may be supplied in response to a request from the UE, a determination by the network that the current model is outdated or underperforming, or as part of a periodic model-update mechanism configured by the network. In some embodiment, the step 904 and the step 902 may be reversed in order of performance and / or concurrently performed.

[0095] In some embodiments, the method 900 may include an additional step (not shown) of notifying the second node that the one or more prediction models are not applicable at the first node. For example, the UE 302 of FIG. 3 may notify the network 304 that the prediction model for RRM measurement prediction is not applicable at the UE 302. The notifying may be based on a determination that the one or more prediction models are not supported by the first node and / or are not executable at the first node. The first node may make the determination based on the parameters and constraints of the first node, for example, but not limited to, computational capacity, battery life, implementation complexity, memory constraints, and mobility of the first node. After notifying the second node, in response, the first node may receive one or more new prediction models from the second node.

[0096] The method 900 includes a step 906 of performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction. For example, as shown in FIG. 3, the UE 302 performs the RRM measurement prediction using a prediction model (e.g., AI / ML model) for RRM measurement prediction. For another example, as shown in FIG. 5, the UE 502 performs the measurement event prediction using a prediction model (e.g., AI / ML model) for measurement event prediction. For another example, as shown in FIG. 7, the UE 702 performs radio link failure prediction and / or handover prediction using a prediction model (e.g., AI / ML model) for radio link failure prediction and / or handover prediction. In some embodiments, the UE 502 may perform both the RRM measurement prediction and the measurement event prediction. In some embodiments, the RRM measurement period includes a plurality of sample periods of RRM measurement, and the method 900 may further include performing an RRM measurement at each of the plurality of sample periods of RRM measurement. In some embodiments, the RRM measurement prediction period includes a plurality of sample periods of RRM measurement prediction, and the method 900 may further include performing an RRM measurement prediction at each of the plurality of sample periods of RRM measurement prediction.

[0097] The method 900 includes a step 908 of transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy. For example, as shown in FIG. 3, at the step 316, the UE 302 transmits, to the network 304, an indication that the performance of the prediction model for RRM measurement prediction is unacceptable. For another example, as shown in FIG. 5, at the step 514, the UE 502 transmits, to the network 504, an indication that the performance of the prediction model for the RRM measurement event prediction is unacceptable. The at least one prediction result may include one or more results of the at least one of: the RRM measurement prediction, a prediction of occurrence of a measurement event, a prediction of occurrence of a radio link failure, or a prediction of occurrence of a handover failure.

[0098] FIG. 10 is a flow chart illustrating a method 1000 for a second node for communication, consistent with some embodiments of the present disclosure. The second node may include at least one of: a base station, a core network, a road-side unit, a repeater, a transponder, a wireless router, a controller, or an access point. The second node may be a network node (terrestrial or non-terrestrial), such as the network 304 of FIG. 3, the network 504 of FIG. 5, or the network 704 of FIG. 7.

[0099] Referring to FIG. 10, the method 1000 includes a step 1002 of transmitting, to a first node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction. The first node may include at least one UE that is connected (e.g., via wireless communication) to the second node. In some embodiments, the request may include configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction. In some embodiments, one or more metrics for the one or more RRM measurements or the RRM measurement prediction comprise at least one of: RSRP, RSRQ, RSSI, SNR, or SINR of at least one of: a serving cell, or one or more neighboring cells.

[0100] The method 1000 includes a step 1004 of transmitting, to the first node, one or more prediction models for at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction. For example, as shown in FIG. 3, the network 304 transmits to the UE 302, one or more prediction models for RRM measurement prediction. For another example, as shown in FIG. 5, the network 504 transmits, to the UE 502, one or more prediction models for RRM measurement event prediction. For another example, as shown in FIG. 7, the network 704 transmits, to the UE 702, one or more prediction models for radio link failure prediction and / or handover failure prediction.

[0101] In some embodiments, the method 1000 may include an additional step (not shown) of receiving, from the first node, a notification indicating that the one or more prediction models are not applicable at the first node. For example, the network 304 of FIG. 4 may receive a notification from the UE 302 that the prediction model for RRM measurement prediction is not applicable at the UE 302. Upon receiving the notification, the second node may send one or more new prediction models to the first node.

[0102] The method 1000 includes a step 1006 of receiving, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy. For example, as shown in FIG. 3, the network 304 receives, from the UE 302, RRM measurement prediction results or an indication that performance of the prediction model is unacceptable. The at least one prediction result may include at least one of: the RRM measurement prediction, a prediction of occurrence of a measurement event, a prediction of occurrence of a radio link failure, or a prediction of occurrence of a handover failure.

[0103] FIG. 11 is a block diagram of a node 1100, consistent with some embodiments of the present disclosure. In some embodiments, the node 1100 may be a UE that receives prediction models from a network and performs corresponding predictions. For example, the node 1100 may be the UE 302 of FIG. 3, the UE 502 of FIG. 5, or the UE 702 of FIG. 7. In some embodiments, the node 1100 may be a network node that serves a UE. For example, the node 1100 may be the network 304 of FIG. 3, the network 504 of FIG. 5, or the network 704 of FIG. 7. In some embodiments, the node 1100 may be a node that performs the method 900 or the method 1000. The node 1100 may take any form, including but not limited to, a computer, a system including at least one computer, a vehicle, a component mounted in a vehicle, a portable computer, a wireless terminal including a mobile phone, a wireless handheld device, or wireless personal device, or any other form.

[0104] Referring to FIG. 11, the node 1100 may include antenna 1102 that may be used for transmission or reception of electromagnetic signals to / from one or more other nodes. The antenna 1102 may include one or more antenna elements and may enable different input-output antenna configurations, for example, multiple input multiple output (MIMO) configuration, multiple input single output (MISO) configuration, and single input multiple output (SIMO) configuration. In some embodiments, the antenna 1102 may include multiple (e.g., tens or hundreds) antenna elements and may enable multi-antenna functions such as beamforming. In some embodiments, the antenna 1102 is a single antenna.

[0105] The node 1100 may include a transceiver 1104 that is coupled to the antenna 1102. The transceiver 1104 may be a wireless transceiver at the node 1100 and may communicate bi-directionally with one or more other nodes. For example, the transceiver 1104 may receive / transmit wireless signals from / to a base station via downlink / uplink communication. The transceiver 1104 may also receive / transmit wireless signals from / to another node unit via sidelink communication. The transceiver 1104 may include a modem to modulate the packets and provide the modulated packets to the antenna 1102 for transmission, and to demodulate packets received from the antenna 1102.

[0106] The node 1100 may include a memory 1106. The memory 1106 may be any type of computer-readable storage medium including volatile or non-volatile memory devices, or a combination thereof. The computer-readable storage medium includes, but is not limited to, non-transitory computer storage media. A non-transitory storage medium may be accessed by a general purpose or special purpose computer. Examples of non-transitory storage medium include, but are not limited to, a portable computer diskette, a hard disk, solid state drive, random access memory (RAM), read-only memory (ROM), an erasable programmable read-only memory (EPROM), electrically erasable programmable ROM (EEPROM), a digital versatile disk (DVD), flash memory, compact disk (CD) ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, etc. A non-transitory medium may be used to carry or store desired program code means (e.g., instructions and / or data structures) and may be accessed by a general-purpose or special-purpose computer, or a general-purpose or special-purpose processor. In some examples, the software / program code may be transmitted from a remote source (e.g., a website, a server, etc.) using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave. In such examples, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are within the scope of the definition of medium. The memory 1106 may also be a cloud-based remote (cloud) memory device. Combinations of the above examples are also within the scope of computer-readable medium.

[0107] The memory 1106 may store information related to identities of node 1100 and the signals and / or data received by antenna 1102. The memory 1106 may also store post-processing signals and / or data. The memory 1106 may also store computer-readable program instructions, mathematical models, and algorithms that are used in signal processing in receiver 1104 and computations in a processor 1108 of the node 1100. For example, the memory 1106 may store prediction models, such as AI / ML models for predictions (e.g., RRM measurement prediction, measurement event prediction, radio link failure prediction, and handover failure prediction, etc.). The memory 1106 may further store computer-readable program instructions for execution by processor 1108 to operate the node 1100 to perform various functions described in this disclosure. In some examples, the memory 1106 may include a basic input / output system (BIOS) which may control basic hardware or software operation such as the interaction with peripheral components or devices.

[0108] The computer-readable program instructions of the present disclosure may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state-setting data, or source code or object code written in any combination of one or more programming languages, including an object-oriented programming language, and conventional procedural programming languages. The computer-readable program instructions may execute entirely on a computing device as a stand-alone software package, or partly on a first computing device and partly on a second computing device remote from the first computing device. In the latter scenario, the second, remote computing device may be connected to the first computing device through any type of network, including a local area network (LAN) or a wide area network (WAN).

[0109] The processor 1108 may include a hardware device with processing capabilities. The processor 1108 may include at least one of a general-purpose processor, a digital signal processor (DSP), a central processing unit (CPU), a graphical processing unit (GPU), a microcontroller, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a programmable logic device, a discrete gate or transistor logic component, a discrete hardware component, or other programmable logic device. Examples of the general-purpose processor include, but are not limited to, a microprocessor, any conventional processor, a controller, a microcontroller, or a state machine. In some embodiments, the processor 1108 may be implemented using a combination of devices (e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration). The processor 1108 may perform predictions (e.g., RRM measurement prediction, measurement event prediction, radio link failure prediction, and handover failure prediction, etc.) using the prediction models stored in the memory 1106. The processor 1108 may further monitor the accuracy of the prediction and determine whether the performance of the prediction models is acceptable. The processor 1108 may receive, from transceiver 1104, downlink / uplink signals or sidelink signals and further process the signals. The processor 1108 may also receive, from transceiver 1104, data packets and further process the packets. In some embodiments, the processor 1108 may be configured to operate a memory using a memory controller. In some embodiments, a memory controller may be integrated into the processor 1108. The processor 1108 may be configured to execute computer-readable instructions stored in a memory (e.g., the memory 1106) to cause the node 1100 to perform various functions.

[0110] The node 1100 may include a global positioning system (GPS) 1110. The GPS 1110 may be used for enabling location-based services or other services based on a geographical position of the node 1100 and / or synchronization among nodes. The GPS 1110 may receive global navigation satellite systems (GNSS) signals from a single satellite or a plurality of satellite signals via the antenna 1102 and provide a geographical position of the node 1100 (e.g., coordinates of the node 1100). In some embodiments, the GPS 1110 is omitted. In some embodiments, a timer is included.

[0111] The node 1100 may include an input / output (I / O) device 1112 that may be used to communicate a result of signal processing and computation to a user or another device. The I / O device 1112 may include a user interface including a display and an input device to transmit a user command to processor 1108. The display may be configured to display a status of signal reception at the node 1100, the data stored at memory 1106, a status of signal processing, and a result of computation, etc. The display may include, but is not limited to, a cathode ray tube (CRT), a liquid crystal display (LCD), a light-emitting diode (LED), a gas plasma display, a touch screen, or other image projection devices for displaying information to a user. The input device may be any type of computer hardware equipment used to receive data and control signals from a user. The input device may include, but is not limited to, a keyboard, a mouse, a scanner, a digital camera, a joystick, a trackball, cursor direction keys, a touchscreen monitor, or audio / video commanders, etc.

[0112] The node 1100 may further include a machine interface 1114, such as an electrical bus that connects the transceiver 1104, the memory 1106, the processor 1108, the GPS 1110, and the I / O device 1112.

[0113] In some embodiments, the node 1100 may be a first node for communication. The processor 1108 may be configured or programmed to execute the instructions stored in the memory 1106 to: receive, from a second node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receive, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; perform, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmit, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0114] In some embodiments, the node 1100 may be a second node for communication. The processor 1108 may be configured or programmed to execute the instructions stored in the memory 1106 to: transmit, to a first node, a request to perform at least one of: an RRM measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmit, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receive, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0115] The embodiments disclosed in the present disclosure can be applied to any radio access technologies, for example, 3GPP 4G technology, 5G technology, or future 3GPP radio technology generations such as 6G, 7G, etc.

[0116] While the examples in this disclosure relate to 3GPP technologies, embodiments described in this disclosure could be used for non-3GPP technologies, for example, IEEE and its 802.11 variants, Wi-Fi, WiMAX, etc.

[0117] As used in this disclosure, use of the term “or” in a list of items indicates an inclusive list. The list of items may be prefaced by a phrase such as “at least one of” or “one or more of.” For example, a list of at least one of A, B, or C includes A or B or C or AB (i.e., A and B) or AC or BC or ABC (i.e., A and B and C). Also, as used in this disclosure, prefacing a list of conditions with the phrase “based on” shall not be construed as “based only on” the set of conditions and rather shall be construed as “based at least in part on” the set of conditions. For example, an outcome described as “based on condition A” may be based on both a condition A and a condition B without departing from the scope of this disclosure.

[0118] In this specification, the terms “comprise,” “include,” or “contain” may be used interchangeably and have the same meaning and are to be construed as inclusive and open-ended. The terms “comprise,” “include,” or “contain” may be used before a list of elements and indicate that at least all of the listed elements within the list exist but other elements that are not in the list may also be present. For example, if A comprises B and C, both {B, C} and {B, C, D} are within the scope of A.

[0119] The present disclosure, in connection with the accompanied drawings, describes example configurations that are not representative of all the examples that may be implemented or all configurations that are within the scope of this disclosure. The term “exemplary” should not be construed as “preferred” or “advantageous compared to other examples” but rather “an illustration, an instance or an example.” By reading this disclosure, including the description of the embodiments and the drawings, it will be appreciated by a person of ordinary skills in the art that the technology disclosed herein may be implemented using alternative embodiments. The person of ordinary skill in the art would appreciate that the embodiments, or certain features of the embodiments described herein, may be combined to arrive at yet other embodiments for practicing the technology described in the present disclosure. Thus, the disclosure is not limited to the examples and designs described herein but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein.

[0120] The flowcharts and block diagrams in the figures illustrate examples of the architecture, functionality, and operation of possible implementations of systems, methods, and devices according to various embodiments. It should be noted that, in some alternative implementations, the functions noted in blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. Likewise, additional steps may be included in such methods, and certain steps may be omitted or combined, in methods consistent with various embodiments.

[0121] It is understood that the described embodiments are not mutually exclusive, and elements, components, materials, or steps described in connection with one example embodiment may be combined with, or eliminated from, other embodiments in suitable ways to accomplish desired design objectives.

[0122] Reference herein to “some embodiments” or “some exemplary embodiments” means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment. The appearance of the phrases “one embodiment” “some embodiments” or “another embodiment” in various places in the present disclosure do not all necessarily refer to the same embodiment, nor are separate or alternative embodiments necessarily mutually exclusive of other embodiments.

[0123] Additionally, the articles “a” and “an” as used in the present disclosure and the appended claims should generally be construed to mean “one or more” unless specified otherwise or clear from context to be directed to a singular form.

[0124] Unless explicitly stated otherwise, each numerical value and range should be interpreted as being approximate as if the word “about” or “approximately” preceded the value of the value or range.

[0125] Although the elements in the following method claims, if any, are recited in a particular sequence, unless the claim recitations otherwise imply a particular sequence for implementing some or all of those elements, those elements are not necessarily intended to be limited to being implemented in that particular sequence.

[0126] It is appreciated that certain features of the present disclosure, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the specification, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable subcombination or as suitable in any other described embodiment of the specification. Certain features described in the context of various embodiments are not essential features of those embodiments, unless noted as such.

[0127] It will be further understood that various modifications, alternatives, and variations in the details, materials, and arrangements of the parts which have been described and illustrated in order to explain the nature of described embodiments may be made by those skilled in the art without departing from the scope. Accordingly, the following claims embrace all such alternatives, modifications, and variations that fall within the terms of the claims.

[0128] Clause 1: A first node for communication, the first node comprising: a memory storing an instruction; and a processor configured to execute the instruction stored in the memory to: receive, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receive, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; perform, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmit, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0129] Clause 2: The first node of clause 1, wherein the first node comprises at least one user equipment (UE).

[0130] Clause 3: The first node of clause 1, wherein the second node comprises a network node to which the first node is connected.

[0131] Clause 4: The first node of clause 1, wherein the request comprises configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction.

[0132] Clause 5: The first node of clause 4, wherein one or more metrics for the one or more RRM measurements or the RRM measurement prediction comprise at least one of: reference signal received power (RSRP), reference signal received quality (RSRQ), received signal strength indicator (RSSI), signal-to-noise ratio (SNR), or signal-to-interference-plus-noise ratio (SINR) of at least one of: a serving cell, or one or more neighboring cells.

[0133] Clause 6: The first node of clause 4, wherein the processor is configured to execute the instruction stored in the memory to: transmit, to the second node, a request for the one or more prediction models and the configuration information.

[0134] Clause 7: The first node of clause 1, wherein the processor is configured to execute the instruction stored in the memory to: receive, from the second node, upon transmission of the indication of failure to identify the prediction model, at least one new prediction model for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction, the at least one new prediction model being different from the one or more prediction models.

[0135] Clause 8: The first node of clause 1, wherein the processor is configured to execute the instruction stored in the memory to: transmit, to the second node, an RRM measurement report, the RRM measurement report being based on one or more RRM measurements performed by the first node.

[0136] Clause 9: The first node of clause 1, wherein the processor is configured to execute the instruction stored in the memory to: determine the prediction accuracy of the prediction model based on one or more RRM measurements performed by the first node.

[0137] Clause 10: The first node of clause 4, wherein the RRM measurement period comprises a plurality of sample periods of RRM measurement, and the processor is configured to execute the instruction stored in the memory to: perform an RRM measurement at each of the plurality of sample periods of RRM measurement.

[0138] Clause 11: The first node of clause 4, wherein the RRM measurement prediction period comprises a plurality of sample periods of RRM measurement prediction, and the processor is configured to execute the instruction stored in the memory to: perform an RRM measurement prediction at each of the plurality of sample periods of RRM measurement prediction.

[0139] Clause 12: The first node of clause 1, wherein the one or more prediction models comprise at least one artificial intelligence or machine learning (AI / ML) algorithm.

[0140] Clause 13: The first node of clause 1, wherein the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction comprises the RRM measurement prediction and the measurement event prediction, and the request comprises an indication whether a two-step measurement event prediction or a one-step measurement event prediction is to be performed, wherein in the two-step measurement event prediction, the RRM measurement prediction is performed first, and a result of the RRM measurement prediction is used for the measurement event prediction, and wherein in the one-step measurement event prediction, the measurement event prediction is performed without first performing the RRM measurement prediction.

[0141] Clause 14: The first node of clause 1, wherein the at least one prediction result comprises one or more results of the at least one of: the RRM measurement prediction, a prediction of occurrence of a measurement event, a prediction of occurrence of a radio link failure, or a prediction of occurrence of a handover failure.

[0142] Clause 15: The first node of clause 1, wherein the processor is configured to execute the instruction stored in the memory to: notify the second node that the one or more prediction models are not applicable at the first node.

[0143] Clause 16: A second node for communication, the first node comprising: a memory storing an instruction; and a processor configured to execute the instruction stored in the memory to: transmit, to a first node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmit, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receive, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0144] Clause 17: The second node of clause 16, wherein the second node comprises a network node to which the first node is connected.

[0145] Clause 18: The second node of clause 16, wherein the first node comprises at least one user equipment (UE).

[0146] Clause 19: The second node of clause 16, wherein the request comprises configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction.

[0147] Clause 20: The second node of clause 16, wherein the processor is configured to execute the instruction stored in the memory to: transmit, to the first node, in response to receiving the indication of failure to identify the prediction model, at least one new prediction model for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction, the at least one new prediction model being different from the one or more prediction models.

[0148] Clause 21: The second node of clause 16, wherein the processor is configured to execute the instruction stored in the memory to: receive, from the first node, an RRM measurement report, the RRM measurement report being based on one or more RRM measurements performed by the first node.

[0149] Clause 22: The second node of clause 16, wherein the processor is configured to execute the instruction stored in the memory to: receive, from the first node, a notification indicating that the one or more prediction models are not applicable at the first node.

[0150] Clause 23: The second node of clause 16, wherein the one or more prediction models comprise at least one artificial intelligence or machine learning (AI / ML) algorithm.

[0151] Clause 24: A method for a first node for communication, the method comprising: receiving, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0152] Clause 25: The method of clause 24, wherein the first node comprises at least one user equipment (UE).

[0153] Clause 26: The method of clause 24, wherein the second node comprises a network node to which the first node is connected.

[0154] Clause 27: The method of clause 24, wherein the request comprises configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction.

[0155] Clause 28: The method of clause 27, wherein one or more metrics for the one or more RRM measurements or the RRM measurement prediction comprise at least one of: reference signal received power (RSRP), reference signal received quality (RSRQ), received signal strength indicator (RSSI), signal-to-noise ratio (SNR), or signal-to-interference-plus-noise ratio (SINR) of at least one of: a serving cell, or one or more neighboring cells.

[0156] Clause 29: The method of clause 27, further comprising: transmitting, to the second node, a request for the one or more prediction models and the configuration information.

[0157] Clause 30: The method of clause 24, further comprising: receiving, from the second node, upon transmission of the indication of failure to identify the prediction model, at least one new prediction model for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction, the at least one new prediction model being different from the one or more prediction models.

[0158] Clause 31: The method of clause 24, further comprising: transmitting, to the second node, an RRM measurement report, the RRM measurement report being based on one or more RRM measurements performed by the first node.

[0159] Clause 32: The method of clause 24, further comprising: determining the prediction accuracy of the prediction model based on one or more RRM measurements performed by the first node.

[0160] Clause 33: The method of clause 27, wherein the RRM measurement period comprises a plurality of sample periods of RRM measurement, and the method further comprises: performing an RRM measurement at each of the plurality of sample periods of RRM measurement.

[0161] Clause 34: The method of clause 27, wherein the RRM measurement prediction period comprises a plurality of sample periods of RRM measurement prediction, and the method further comprises: performing an RRM measurement prediction at each of the plurality of sample periods of RRM measurement prediction.

[0162] Clause 35: The method of clause 24, wherein the one or more prediction models comprise at least one artificial intelligence or machine learning (AI / ML) algorithm.

[0163] Clause 36: The method of clause 24, wherein the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction comprises the RRM measurement prediction and the measurement event prediction, and the request comprises an indication whether a two-step measurement event prediction or a one-step measurement event prediction is to be performed, wherein in the two-step measurement event prediction, the RRM measurement prediction is performed first, and a result of the RRM measurement prediction is used for the measurement event prediction, and wherein in the one-step measurement event prediction, the measurement event prediction is performed without first performing the RRM measurement prediction.

[0164] Clause 37: The method of clause 24, wherein the at least one prediction result comprises one or more results of the at least one of: the RRM measurement prediction, a prediction of occurrence of a measurement event, a prediction of occurrence of a radio link failure, or a prediction of occurrence of a handover failure.

[0165] Clause 38: The method of clause 24, further comprising: notifying the second node that the one or more prediction models are not applicable at the first node.

[0166] Clause 39: A method for a second node for communication, the method comprising: transmitting, to a first node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmitting, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receiving, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0167] Clause 40: The method of clause 39, wherein the second node comprises a network node to which the first node is connected.

[0168] Clause 41: The method of clause 39, wherein the first node comprises at least one user equipment (UE).

[0169] Clause 42: The method of clause 39, wherein the request comprises configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction.

[0170] Clause 43: The method of clause 39, further comprising: transmitting, to the first node, in response to receiving the indication of failure to identify the prediction model, at least one new prediction model for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction, the at least one new prediction model being different from the one or more prediction models.

[0171] Clause 44: The method of clause 39, further comprising: receiving, from the first node, an RRM measurement report, the RRM measurement report being based on one or more RRM measurements performed by the first node.

[0172] Clause 45: The method of clause 39, wherein the one or more prediction models comprise at least one artificial intelligence or machine learning (AI / ML) algorithm.

[0173] Clause 46: The method of clause 39, further comprising: receiving, from the first node, a notification indicating that the one or more prediction models are not applicable at the first node.

[0174] Clause 47: A non-transitory computer-readable medium storing instructions that are executable by one or more processors of a first node for communication, to perform a method, the method comprising: receiving, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

[0175] Clause 48: A non-transitory computer-readable medium storing instructions that are executable by one or more processors of a second node for communication, to perform a method, the method comprising: transmitting, to a first node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmitting, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receiving, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

Claims

1. A first node for communication, the first node comprising: a memory storing an instruction; and a processor configured to execute the instruction stored in the memory to: receive, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receive, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; perform, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmit, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

2. The first node of claim 1, wherein the first node comprises at least one user equipment (UE).

3. The first node of claim 1, wherein the second node comprises a network node to which the first node is connected.

4. The first node of claim 1, wherein the request comprises configuration information for at least one of: one or more RRM measurements, the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, the handover failure prediction, an RRM measurement period, an RRM measurement prediction period, a measurement event prediction period, a radio link failure prediction period, a handover failure prediction period, a sample period of RRM measurement, a sample period of RRM measurement prediction, a report period of RRM measurement, a report period of RRM measurement prediction, a report period of measurement event prediction, a report period of radio link failure prediction, or a report period of handover failure prediction.

5. The first node of claim 4, wherein one or more metrics for the one or more RRM measurements or the RRM measurement prediction comprise at least one of: reference signal received power (RSRP), reference signal received quality (RSRQ), received signal strength indicator (RSSI), signal-to-noise ratio (SNR), or signal-to-interference-plus-noise ratio (SINR) of at least one of: a serving cell, or one or more neighboring cells.

6. The first node of claim 4, wherein the processor is configured to execute the instruction stored in the memory to: transmit, to the second node, a request for the one or more prediction models and the configuration information.

7. The first node of claim 1, wherein the processor is configured to execute the instruction stored in the memory to: receive, from the second node, upon transmission of the indication of failure to identify the prediction model, at least one new prediction model for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction, the at least one new prediction model being different from the one or more prediction models.

8. The first node of claim 1, wherein the processor is configured to execute the instruction stored in the memory to: transmit, to the second node, an RRM measurement report, the RRM measurement report being based on one or more RRM measurements performed by the first node.

9. The first node of claim 1, wherein the processor is configured to execute the instruction stored in the memory to: determine the prediction accuracy of the prediction model based on one or more RRM measurements performed by the first node.

10. The first node of claim 4, wherein the RRM measurement period comprises a plurality of sample periods of RRM measurement, and the processor is configured to execute the instruction stored in the memory to: perform an RRM measurement at each of the plurality of sample periods of RRM measurement.

11. The first node of claim 4, wherein the RRM measurement prediction period comprises a plurality of sample periods of RRM measurement prediction, and the processor is configured to execute the instruction stored in the memory to: perform an RRM measurement prediction at each of the plurality of sample periods of RRM measurement prediction.

12. The first node of claim 1, wherein the one or more prediction models comprise at least one artificial intelligence or machine learning (AI / ML) algorithm.

13. The first node of claim 1, wherein the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction comprises the RRM measurement prediction and the measurement event prediction, and the request comprises an indication whether a two-step measurement event prediction or a one-step measurement event prediction is to be performed, wherein in the two-step measurement event prediction, the RRM measurement prediction is performed first, and a result of the RRM measurement prediction is used for the measurement event prediction, and wherein in the one-step measurement event prediction, the measurement event prediction is performed without first performing the RRM measurement prediction.

14. The first node of claim 1, wherein the at least one prediction result comprises one or more results of the at least one of: the RRM measurement prediction, a prediction of occurrence of a measurement event, a prediction of occurrence of a radio link failure, or a prediction of occurrence of a handover failure.

15. The first node of claim 1, wherein the processor is configured to execute the instruction stored in the memory to: notify the second node that the one or more prediction models are not applicable at the first node.

16. A second node for communication, the second node comprising: a memory storing an instruction; and a processor configured to execute the instruction stored in the memory to: transmit, to a first node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmit, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receive, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

17. A method for a first node for communication, the method comprising: receiving, from a second node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; receiving, from the second node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; performing, based on the one or more prediction models, the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and transmitting, to the second node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

18. A method for a second node for communication, the method comprising: transmitting, to a first node, a request to perform at least one of: a radio resource management (RRM) measurement prediction, a measurement event prediction, a radio link failure prediction, or a handover failure prediction; transmitting, to the first node, one or more prediction models for the at least one of: the RRM measurement prediction, the measurement event prediction, the radio link failure prediction, or the handover failure prediction; and receiving, from the first node, at least one prediction result, or an indication of failure to identify, among the one or more prediction models, a prediction model having prediction accuracy meeting threshold accuracy.

19. A non-transitory computer-readable medium storing instructions that are executable by one or more processors of a first node for communication, to perform the method of claim 17.

20. A non-transitory computer-readable medium storing instructions that are executable by one or more processors of a second node for communication, to perform the method of claim 18.

Images