Method, performed in a network node and a UE, for handling channel state information prediction functionality

Abstract

Embodiments herein relate to, for example, a method performed by a UE (10) for handling communication in a wireless communications network. The UE (10) receives a configuration indication from a network node (120), wherein the configuration indication indicates content and / or a format of one or more performance monitoring results5 associated with a CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results;10 a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting. The UE (10) obtains one or more performance monitoring results for the CSI15 prediction functionality based on the received configuration indication; and reports the one or more performance monitoring results to the network node (120).

Description

[0001] METHOD, PERFORMED IN A NETWORK NODE AND A UE, FOR HANDLING CHANNEL

[0002] STATE INFORMATION PREDICTION FUNCTIONALITY

[0003] TECHNICAL FIELD

[0004] Embodiments herein relate to a user equipment (UE), a network node and

[0005] 5 methods performed therein for communication. Furthermore, a computer program and a computer readable storage medium are also provided herein. In particular, embodiments herein relate to handling communication, such as handling channel state information prediction functionality, in a wireless communications network.

[0006] BACKGROUND

[0007] In a typical wireless communications network, UEs, also known as wireless communication devices, mobile stations, stations (STA) and / or wireless devices, communicate via for example a Radio Access Network (RAN) with one or more core networks (CN). The RAN covers a geographical area which is divided into service areas5 or cell areas, with each service area or cell area being served by radio network node such as an access node e.g. a Wi-Fi access point or a radio base station (RBS), which in some networks may also be called, for example, a NodeB, a gNodeB, or an eNodeB. The service area or cell area is a geographical area where radio coverage is provided by the radio network node. The radio network node operates on radio frequencies to 0 communicate over an air interface with the UEs within range of the radio network node. The radio network node communicates over a downlink (DL) to the UE and the UE communicates over an uplink (UL) to the radio network node.

[0008] A Universal Mobile Telecommunications System (UMTS) is a third generation telecommunications network, which evolved from the second generation (2G) Global System for Mobile Communications (GSM). The UMTS terrestrial radio access network (UTRAN) is essentially a RAN using wideband code division multiple access (WCDMA) and / or High-Speed Packet Access (HSPA) for communication with user equipment. In a forum known as the Third Generation Partnership Project (3GPP), telecommunications suppliers propose and agree upon standards for present and future generation networks0 and UTRAN specifically, and investigate enhanced data rate and radio capacity. In some RANs, e.g. as in UMTS, several radio network nodes may be connected, e.g., by landlines or microwave, to a controller node, such as a radio network controller (RNC) or a base station controller (BSC), which supervises and coordinates various activities of the plural radio network nodes connected thereto. The RNCs are typically connected to one or more core networks.

[0009] Specifications for the Evolved Packet System (EPS) have been completed within the 3GPP and this work continues in the coming 3GPP releases, such as 5G, for example New Radio (NR), and beyond networks. The EPS comprises the Evolved Universal Terrestrial Radio Access Network (E-UTRAN), also known as the Long-Term Evolution (LTE) radio access network, and the Evolved Packet Core (EPC), also known as System Architecture Evolution (SAE) core network. E-UTRAN / LTE is a 3GPP radio access technology wherein the radio network nodes are directly connected to the EPC core network. As such, the Radio Access Network (RAN) of an EPS has an essentially “flat” architecture comprising radio network nodes connected directly to one or more core networks.

[0010] With the 5G technologies such as NR, focus is on a set of features such as the use of very many transmit- and receive-antenna elements that makes it possible to utilize beamforming, such as transmit-side and receive-side beamforming. Transmit-side beamforming means that the transmitter can amplify the transmitted signals in a selected direction or directions, while suppressing the transmitted signals in other directions. Similarly, on the receive-side, a receiver can amplify signals from a selected direction or directions, while suppressing unwanted signals from other directions.

[0011] Artificial Intelligence (Al) and Machine Learning (ML) have been investigated, both in academia and industry, as promising tools to optimize the design of the air interface in wireless communication networks. Example use cases include using autoencoders for Channel State Information (CSI) compression to reduce the feedback overhead and improve channel prediction accuracy; using deep neural networks for classifying Line-of- Sight (LOS) and Non-LOS (NLOS) conditions to enhance the positioning accuracy; and using reinforcement learning for beam selection at the network side and / or the User Equipment (UE) side to reduce the signaling overhead and beam alignment latency; and using deep reinforcement learning to learn an optimal precoding policy for complex Multiple Input Multiple Output (MIMO) precoding problems.

[0012] In the 3rd Generation Partnership Project (3GPP) New Radio (NR) standardization work for release 18 (Rel. 18), a study item (SI) on AI / ML for the NR air interface was started in May 2022. It has been agreed that the works will continue in Rel. 19. The works will explore the benefits of augmenting the air interface with features enabling improved support of AI / ML-based algorithms for enhanced performance and / or reduced complexity / overhead. Through studying and specifying a few selected use cases, such as CSI feedback, beam management, and positioning, the works aim to design the mechanisms to accommodate AI / ML into the 3GPP standard.

[0013] An important part of Al development and operation is the lifecycle management (LCM) of the AI / ML model, e.g., model training, model deployment, model inference, model monitoring, model updating, and AI / ML functionality.

[0014] In NR Rel-18 AI / ML for NR air interface study item, the LCM procedure is studied for the case that an AI / ML model has a model identity (ID) with associated information and / or for the case that a given functionality is provided by some AI / ML operations.

[0015] Two types of LCM operations were studied in NR Rel-18, functionality-based LCM, and model-l D-based LCM.

[0016] Functionality based LCM: Functionality refers to an AI / ML-enabled Feature and / or feature group (FG) enabled by configuration(s), where configuration(s) is(are) supported based on conditions indicated by UE capability. Correspondingly, functionality-based LCM operates based on, at least, one configuration of an AI / ML-enabled Feature / FG or specific configurations of an AI / ML-enabled Feature / FG. In functionality-based LCM, the network indicates activation / deactivation / fallback / switching of AI / ML functionality via 3GPP signaling, e.g., radio resource control (RRC), medium access control (MAC)- control element (CE), and / or downlink control information (DCI). Models may not be identified at the Network, and UE may perform model-level LCM. Whether and how much awareness / interaction network (NW) should have about model-level LCM requires further study. For functionality identification, there may be either one or more than one functionalities defined within an AI / ML-enabled feature, whereby AI / ML-enabled Feature refers to a Feature where AI / ML may be used.

[0017] In model-l D-based LCM, ML models are identified at the Network, and Network / UE may activate / deactivate / select / switch individual AI / ML models via model ID. A model may be associated with specific configurations / conditions associated with the UE capability of an AI / ML-enabled Feature / FG and additional conditions (e.g., scenarios, sites, and datasets) as determined / identified between the UE side and NW side. An AI / ML model identified by a model ID may be logical, and how it maps to physical AI / ML model(s) may be up to implementation. Fig. 1 shows a functional framework for AI / ML for NR Air Interface that can be used for studying model LCM aspects for different Al for physical (PHY) use cases. The general framework consists of the following:

[0018] Data collection is a function that provides input data to the model training, management, and inference functions. o Training data: Data needed as input for the AI / ML model training function. o Monitoring data: Data needed as input for the management of AI / ML models or AI / ML functionalities. o Inference data: Data needed as input for the AI / ML Inference function. Model training is a function that performs AI / ML model training, validation, and testing which may generate model performance metrics that can be used as part of the model testing procedure. The model training function is also responsible for data preparation, e.g., data pre-processing and cleaning, formatting, and transformation, based on training data delivered by a data collection function if required. o Trained / Updated model: In case of having a model storage function, this is used to deliver trained, validated, and tested AI / ML models to the model storage function, or to deliver an updated version of a model to the model storage function.

[0019] Management is a function that oversees the operation, e.g., selection, (de)activation, switching, fallback, and monitoring, e.g., performance, of AI / ML models or AI / ML functionalities. This function is also responsible for making decisions to ensure the proper inference operation based on data received from the data collection function and the Inference function. o Management instruction: Information needed as input to manage the Inference function. Concerning information may include selection / (de)activation / switching of AI / ML models or AI / ML-based functionalities, fallback to non-AI / ML operation, i.e. , not relying on inference process, etc. o Model transfer / Delivery request: Used to request model(s) to the model storage function. o Performance feedback / Retraining request: Information needed as input for the model training function, e.g., for model (re)training or updating purposes. Inference is a function that provides outputs from the process of applying AI / ML models or AI / ML functionalities, using the data that is provided by the data collection function, i.e. , inference data, as an input. The Inference function is also responsible for data preparation, e.g., data pre-processing and cleaning, formatting, and transformation, based on inference data delivered by a data collection function, if required. o Inference output: Data used by the management function to monitor the performance of AI / ML models or AI / ML functionalities. model storage is a function responsible for storing trained / updated models that can be used to perform the Inference function. o Note: The model storage function in Figure 1 is only intended as a reference point, if any, when applicable for protocol terminations, model transfer / delivery, and related processes. It should be stressed that its purpose does not encompass restricting the actual storage locations of models. Therefore, the specification impact of all data / information / instruction flows, i.e., the arrows in Fig. 1 , to / from this function should be studied case by case. o Model transfer / delivery: Used to deliver an AI / ML model to the Inference function.

[0020] Signaling procedures for model and functionality life cycle management.

[0021] The signaling procedures for different scenarios for model-l D-based management and / or functionality-based management are exemplified below for UE-side AI / ML models, where a UE-side AI / ML model is defined as an AI / ML model whose inference is at the UE. The procedures include scenarios for which the management decision is taken by the network or by the UE. For network-side decision, this can be either network-initiated, or UE-initiated and requested to the network. While for UE-side decision, this can be either event-triggered as configured by the network and where the UE’s decision is reported to the network, or UE-autonomous, with or without UE’s decision being reported to the network.

[0022] Decision by the network.

[0023] Network-initiated

[0024] Fig. 2 shows a network-initiated AI / ML management. The case where the LCM decision is taken and initiated by the network is depicted in Fig. 2.

[0025] Note: The management instruction may be a result of model / functionality performance monitoring at the network.

[0026] Note: The management instruction may include information about the model or functionality.

[0027] UE-initiated and requested to the network

[0028] Fig. 3 shows a UE-initiated AI / ML management

[0029] The case where the LCM decision is taken by the network but where the request is initiated by the UE is depicted in Fig. 3.

[0030] Note: The management request may be a result of model / functionality monitoring at the UE.

[0031] Note: In response to the management request, the network may send a management instruction to the UE.

[0032] Note: The Management Request may include information about the model or functionality. Note: The network may accept or reject the Management Request from the UE.

[0033] Note: The Management Request may include information related to model / functionality performance metrics.

[0034] Note: The Management Instruction may include information about the model or functionality.

[0035] Decision by the UE:

[0036] Event-triggered as configured by the network, UE’s decision is reported to the network Fig. 4 shows an event-triggered procedure as configured by the network

[0037] The case where the LCM decision is taken by the UE according to prior network configuration is depicted in Fig. 4.

[0038] Note: Use case-specific events / conditions may be configured by the network for event- triggered AI / ML management at the UE.

[0039] Note: UE may send a management decision report to the network following event- triggered AI / ML management at the UE.

[0040] Note: The management decision report may include information about the model or functionality.

[0041] UE-autonomous, wherein UE’s decision is reported to the network. Fig. 5 shows a UE autonomous scenario wherein the UE decision reported to the network The case where the LCM decision can autonomously be taken by the UE is depicted in Fig. 5.

[0042] Note: The UE may be configured to send a management decision report to the network upon performing a model / functionality management decision.

[0043] UE-autonomous, UE’s decision is not reported to the network.

[0044] For the case where the LCM decision can autonomously be taken by the UE and where the decision is not reported to the network, the AI / ML management is transparent from a network perspective.

[0045] Time domain CSI prediction at UE.

[0046] In 3GPP NR Rel-18, channel measurement resource (CMR) enhancement for Type II CSI prediction at UE, a.k.a., Enhanced Type II predicted precoding matrix indicator (PMI), has been introduced, see the measurement part in Fig. 6, where a burst of K e {4, 8, 12} same CSI-RS resources are configured to the UE in a single CSI- reference signal (RS) resource set. The burst of CSI-RS resources is aperiodically (AP) triggered using a single downlink control information (DCI). The K CSI-RS resources are used for the UE to extract time domain channel properties of the channel, based on which a future CSI can be predicted. Alternatively, NW may also configure a legacy periodic (P) or semi-persistent (SP) CSI-RS resource. The CSI-RS resources are uniformly spaced in time, separated by m e {1, 2} slots, within the resource set.

[0047] For the Rel-18 Type II predicted PMI enhancement, a UE can be configured by gNB to report predicted PMIs for N4e {1, 2, 4, 8} time slots, see the Rel-18 Type II PMI part in Fig. 6. Note that the prediction herein is relative to the CSI-RS reference resource. The predicted N4PMIs are supposed to reflect the channels with d e {l,m} slots separation, starting from 8 e {-ncs / ref 0,1,2} slots into the future relative to the CSI-RS reference resource. For AP CSI-RS burst, m e {1,2}, while for P / SP CSI-RS, m is the CSI- RS periodicity. The spacing d between the N4PMIs and offset 8 relative to the CSI-RS reference resource can be configured by the gNB via RRC signaling. The N4PMIs are compressed in a beam-frequency-Doppler domain, and the compressed PMI is reported to the gNB in a single CSI report. Fig. 6 shows an illustration of CMR enhancement for Rel-18 Type II CSI prediction at UE. 3GPP NR Rel-18 time-domain CSI prediction using UE-side AI / ML model:

[0048] In Rel-18, Al-based UE-side CSI prediction was introduced as an Al for PHY use case, and being studied in the study item on AI / ML for the NR air interface. One or more AI / ML models can be trained and deployed at a UE for the Al-based CSI-prediction feature. During model inference, a UE is configured by the gNB to measure a set of historical CSI-RSs and then report a predicted CSI for one or multiple future-time instances using its AI / ML model(s).

[0049] Model performance monitoring:

[0050] There are several methods for model monitoring. Monitoring based on intermediate KPIs, e.g., inference accuracy, requires collecting new ground-truth data similar / identical to the training data, which is very accurate but has a high cost due to the potentially large measurement / reporting overhead. Monitoring based on data distribution of input / output data does not require any additional signaling overhead but is less accurate than monitoring based on inference accuracy since one does not retrieve the ground truth. Similarly, monitoring based on system performance does not require any additional signaling overhead, however, it can be challenging to identify whether the rootcause for bad system performance is due to an inaccurate model or due to some other malfunctioning procedure or hardware. Monitoring based on data distribution can in contrast identify a potential problem in the model by detecting that the dataset observed during inference is not the same as during training. However, it is not trivial to define conditions and measurable data-distribution based KPIs for sounding a model failure alarm with a good trade-off between model failure detection reliability and accuracy, e.g., low false alarm rate, low missed detection rate and low latency.

[0051] To ensure reliable / accurate model performance monitoring results, sufficient monitoring data samples will be collected and used to derive the performance monitoring results. Examples of model-output / inference accuracy based performance monitoring results include intermediate KPI per monitoring data sample, intermediate KPI statistics associated to a monitoring data set, the percentage of monitoring data samples within a monitoring dataset for which the intermediate KPI fulfills a certain condition, a flag indicating whether the model is functioning ok or not. Examples of data drift based performance monitoring results include: monitoring data statistics, the difference between the monitoring data statistics and the data statistics obtained in the model training stage, and a flag indicating whether a data drift is detected or not. When implementing model monitoring, the monitoring method can be selected based on UE service requirements. For example, a UE with mobile broadband (MBB) could start with a low-cost solution, e.g., system performance based, if problems are observed / predicted, then activate an inference accuracy based monitoring method associated with higher complexity. High-complexity and signaling overhead monitoring may be required for certain UE service requirements, such as for emergency localization use cases or UEs with ultra reliable low latency communication (URLLC) connections.

[0052] For CSI prediction using UE side AI / ML model use case studied in Rel-18 AI / ML for NR air interface study item, at least the following aspects have been proposed by companies on performance monitoring for functionality-based LCM:

[0053] - Type 1 : o UE calculates the performance metric(s) o UE reports performance monitoring output that facilitates functionality fallback decision at the network

[0054] ■ Performance monitoring output details can be further defined

[0055] ■ NW may configure threshold criterion to facilitate UE side performance monitoring (if needed). o NW makes decision(s) of functionality fallback operation (fallback mechanism to legacy CSI reporting).

[0056] - Type 2: o UE reports predicted CSI and / or the corresponding ground-truth o NW calculates the performance metrics. o NW makes decision(s) of functionality fallback operation (fallback mechanism to legacy CSI reporting).

[0057] - Type 3: o UE calculates the performance metric(s) o UE reports performance metric(s) to the NW o NW makes decision(s) of functionality fallback operation (fallback mechanism to legacy CSI reporting).

[0058] Functionality selection / activation / deactivation / switching as defined for other UE side use cases can be reused, if applicable.

[0059] Configuration and procedure for performance monitoring

[0060] CSI-RS configuration for performance monitoring Performance metric including at least intermediate key performance indicator (KPI), e.g., squared generalized cosine similarity (SGCS) or normalized mean square error (NMSE),

[0061] UE report, including periodic / semi-persistent / aperiodic reporting, and event driven report

[0062] Note: UE may make decision within the same functionality on model selection, activation, deactivation, switching operation transparent to the NW.

[0063] SUMMARY

[0064] As part of developing embodiments herein one or more problems were first identified.

[0065] For monitoring the performance of a UE-sided AI / ML model, if the performance monitoring is performed at the UE-side, then, a UE can report the model performance monitoring results to the NW, so that the NW takes the UE reported model performance information into account when making model level or functionality level LCM decisions, e.g., fallback to non-AL / ML algorithm, functionality / model switching, etc.. However, there are problems on how to configure a UE to derive a performance metric or performance monitoring result, and how to test / verify the quality / accuracy / reliability of the result reported from a UE.

[0066] It has been proposed for a network node to configure a UE to derive a model performance monitoring result for an AI / ML model, and report the result to the NW. However, for both Al-based and non-AI based CSI prediction schemes, the Al model and the non-AI based algorithm may generate multiple predicted CSI corresponding to multiple future time instances. It is a problem of how to configure a UE to derive a performance metric(s) and / or performance monitoring output(s) for an AI / ML-based or non-AI based CSI prediction model / algorithm, which can output multiple predicted CSI for multiple future time instances, and report the derived performance metric(s) / output(s) to the NW.

[0067] The object of embodiments herein is to provide a mechanism handling communication, such as handling a CSI prediction functionality, in an efficient manner.

[0068] According to an aspect of embodiments herein the object is achieved by providing a method performed by a UE for handling communication in a wireless communications network. The UE receives a configuration indication from a network node, wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting.

[0069] The UE obtains one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication; and reports the one or more performance monitoring results to the network node.

[0070] According to another aspect of embodiments herein the object is achieved by providing a method performed by a network node for handling communication in a wireless communications network. The network node transmits to a UE, a configuration indication, wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting.

[0071] The network node further receives one or more performance monitoring results from the UE.

[0072] It is furthermore provided herein a computer program product comprising instructions, which, when executed on at least one processor, cause the at least one processor to carry out the methods herein, as performed by the UE and the network node, respectively. It is additionally provided herein a computer-readable storage medium, having stored thereon a computer program product comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the methods herein, as performed by the UE and the network node, respectively.

[0073] The object is further achieved by providing a UE and a network node configured to perform the methods herein. According to an aspect of embodiments herein the object is achieved by providing a UE for handling communication in a wireless communications network. The UE is configured to receive a configuration indication from a network node, wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting.

[0074] The UE is configured to obtain one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication; and to report the one or more performance monitoring results to the network node.

[0075] According to another aspect of embodiments herein the object is achieved by providing a network node for handling communication in a wireless communications network. The network node is configured to transmit to a UE, a configuration indication, wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting.

[0076] The network node is configured to receive one or more performance monitoring results from the UE.

[0077] Embodiments herein disclose a mechanism for a network node to configure the content and / or format of the one or more performance monitoring results associated to a CSI prediction functionality, such as a UE-side computational model, which predicts a value such as a CSI for one or multiple future time instances. The UE derives the one or more performance monitoring results based on the configuration indication and reports the one or more performance monitoring results to the network node.

[0078] Embodiments herein enable the network node to instruct a UE to do performance monitoring for a CSI prediction functionality that performs a prediction, such as CSI predictions, over one or multiple future time instances and to acquire information about the performance monitoring results, such as performance metric and / or monitoring results, from the UE.

[0079] Embodiments herein may further enable the testability of the one or more performance monitoring results such as one or more performance metrics and / or monitoring result output, reported from the UE. Hence, the network node can trust these reported performance monitoring results when making CSI prediction functionality decisions or computational model feature reconfiguration decisions, such as AI / ML model LCM.

[0080] Thereby embodiments herein handle communication, such as handle a CSI prediction functionality, in an efficient manner.

[0081] BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Embodiments will now be described in more detail in relation to the enclosed drawings, in which: Fig. 1 shows an overview according to prior art; Fig. 2 shows an overview according to prior art; Fig. 3 shows an overview according to prior art; Fig. 4 shows an overview according to prior art; Fig. 5 shows an overview according to prior art; Fig. 6 shows an overview according to prior art;

[0083] Fig. 7 shows an overview depicting a wireless communications network according to embodiments herein;

[0084] Fig. 8 shows a combined signaling scheme and flowchart according to embodiments herein;

[0085] Fig. 9 shows a schematic flowchart depicting a method performed by a UE according to embodiments herein;

[0086] Fig. 10 shows a schematic flowchart depicting a method performed by a network node according to embodiments herein;

[0087] Fig. 11 is a block diagram depicting a UE according to embodiments herein; and Fig. 12 is a block diagram depicting a network node according to embodiments herein. DETAILED DESCRIPTION

[0088] Embodiments herein relate to communication networks in general. Fig. 7 is a schematic overview depicting a wireless communications network 1. The wireless communications network 1 comprises one or more RANs and one or more CNs. The wireless communications network 1 may use a number of different technologies, such as Wi-Fi, Long Term Evolution (LTE), LTE-Advanced, NR, Wideband Code Division Multiple Access (WCDMA), Global System for Mobile communications / Enhanced Data rate for GSM Evolution (GSM / EDGE), Worldwide Interoperability for Microwave Access (WiMax), or Ultra Mobile Broadband (UMB), just to mention a few possible implementations.

[0089] In the wireless communications network 1, wireless devices e.g. a user equipment (UE) 10 such as an loT device, an A-loT device, a ZE device, a mobile station, a non-access point (non-AP) STA, a STA, a wireless device and / or a wireless terminal, communicate via one or more Access Networks (AN), e.g. a RAN, to one or more core networks (CN). It should be understood by those skilled in the art that “UE” is a non-limiting term which means any terminal, wireless communication terminal, internet of things (loT) capable device, Machine Type Communication (MTC) device, Device to Device (D2D) terminal, or node e.g. smart phone, laptop, mobile phone, sensor, relay, mobile tablets or even a base station communicating within a cell.

[0090] The wireless communications network 1 comprises a radio network node 12 providing radio coverage over a geographical area, e.g. a first service area, of a first radio access technology (RAT), such as NR, LTE, UMTS, Wi-Fi or similar. The radio network node 12 may be a radio access network node such as radio network controller or an access point such as a wireless local area network (WLAN) access point or an Access Point Station (AP STA), an access controller, a base station, e.g. a radio base station such as a NodeB, an evolved Node B (eNB, eNodeB), a base transceiver station, Access Point Base Station, base station router, a transmission arrangement of a radio base station, a stand-alone access point or any other network unit capable of serving a UE within the service area served by the radio network node 12 depending e.g. on the first radio access technology and terminology used.

[0091] It is herein disclosed a mechanism for a network node 120 (NW) such as the radio network node 12 or a core network node, or a computational model node, to configure the content and / or format of one or more performance monitoring results associated with a CSI prediction functionality such as CSI prediction model / algorithm / feature, for example, a UE-side computational model, which predicts CSI in one or multiple future time instances, for example, an outcome such as a CSI for one or multiple future time instances.

[0092] The configuration signaling may comprise a configuration indication and the configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results. A ground-truth CSI format, e.g., used to derive the intermediate performance metrics. a number of monitoring data samples to use for deriving one or more performance monitoring results. one or more performance monitoring result definitions, such as one or more performance metrics definition, e.g., SGCS and / or NMSE; and a format for performance monitoring result reporting .

[0093] The UE 10 obtains one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication; and reports the one or more performance monitoring results to the network node 120.

[0094] The NW node 120 may be the gNB, e.g. the gNB-centralized unit (CU) or the gNB- distributed unit (DU), the operation, administration, and maintenance (CAM), or a core network node, e.g. the Network Data Analytics Function (NWDAF). In the following the term NW can refer to any of the aforementioned NW nodes.

[0095] • The concept “network” may refer to one of a generic network node, such as the radio network node 12, e.g., a gNB (or the corresponding node in a 6G network), a base station, transmission and reception point (TRP), a unit within the base station to handle at least some ML operation, a relay node, a core network node, a core network node that handle at least some ML operations, or a device supporting device to device (D2D) communication.

[0096] • A computational model may refer to an ML-based model, a configuration of an ML-based model, a non-ML-based functionality, or a configuration of a non- ML-based functionality.

[0097] • The terms “ML-model” and “Al-model” are interchangeable. An AI / ML model can be defined as a functionality or be part of a functionality that is deployed / implemented in a first node. This first node can receive a message from a second node indicating that the functionality is not performing correctly. Further, an AI / ML model can be defined as a feature or part of a feature that is implemented / supported in a first node. This first node can indicate the feature version to a second node. If the ML-model is updated, the feature version maybe changed by the first node.

[0098] Fig. 8 is a combined flowchart and signaling scheme according to some embodiments herein.

[0099] Action 801. The network node 120 transmit a configuration indication to the UE 10. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results. A ground-truth CSI format, e.g., used to derive the intermediate performance metrics. a number of monitoring data samples to use for deriving one or more performance monitoring results. one or more performance monitoring result definitions, such as one or more performance metrics definition, e.g., SGCS and / or NMSE; and a format for performance monitoring result reporting.

[0100] Action 802. The UE 10 may execute the CSI prediction functionality such as a computational model to obtain a value / a prediction of a signaling parameter such as signal strength or similar.

[0101] Action 803. The UE 10 obtains the one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication. The UE 10 may obtain one or more indications of performance metrics of a computational model related to signaling such as a signal strength or quality based on the received configuration.

[0102] Action 804. The UE 10 reports the one or more performance monitoring results to the network node 120, for example, reports the one or more indications to the network node 120.

[0103] Example embodiments of a method performed by the UE 10 for handling communication in the wireless communications network will now be described with reference to a flowchart depicted in Fig. 9. The actions do not have to be taken in the order stated below, but may be taken in any suitable order. Optional actions are marked in dashed boxes.

[0104] Action 901. The UE receives the configuration indication from the network node 120. The configuration indication indicates content and / or a format of one or more performance monitoring results associated with a CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following:

[0105] - A ground-truth label format or a ground-truth CSI format used to derive the one or more performance monitoring results;

[0106] - The number of monitoring data samples to use for deriving one or more performance monitoring results;

[0107] One or more performance monitoring result definitions such as one or more performance metrics definition, e.g., SGCS and / or NMSE; and a format for performance monitoring result reporting, also referred to as performance monitoring result reporting format

[0108] The CSI prediction functionality may be based on one or more Al and / or ML models and / or one or more non-AI and / or ML based CSI prediction algorithms. The format for performance monitoring result reporting may be configured as one or more intermediate KPI per monitoring data sample. The one or more intermediate KPIs may be defined as at least one of the following

[0109] SGCS between all or a part of the CSI prediction algorithm output and an associated ground truth label; and

[0110] NMSE between all or part of the CSI prediction model / algorithm output and the associated ground truth label.

[0111] The format of the one or more performance monitoring results may be configured as a statistics of one or more intermediate KPIs over M monitoring data samples, where M>1.

[0112] The statistics of the one or more intermediate KPIs may be represented using at least one of the following formats: a percentile value of the intermediate KPI associated to the M monitoring data samples; and / or a mean and / or variance of the intermediate KPI associated to the M monitoring data samples.

[0113] The format of the one or more performance monitoring results may be configured as a percentage of monitoring data samples for which the associated intermediate KPI and / or the associated statistics of intermediate KPI fulfills one or more conditions.

[0114] The one or more conditions may be either configured as part of the configuration indication or pre-defined in specification.

[0115] The format of the one or more performance monitoring result may be a flag indicating whether a model monitoring metrics for the monitoring data samples fulfill one or more conditions. The model monitoring metrics may be a single intermediate KPI based on a monitoring data sample, or the statistics of one or more intermediate KPIs over M monitoring data samples, or a percentage of the monitoring data samples for which the associated intermediate KPIs and / or the associated statistics of intermediate KPI fulfills one or more conditions.

[0116] The one or more conditions may include that SGCS is above a certain threshold and / or that SGCS of a proportion of the data monitoring samples is above a threshold.

[0117] Action 902. The UE 10 may execute the CSI prediction functionality such as computational model to obtain a value / a prediction performance such as value related to SGCS or NMSE .

[0118] Action 903. The UE 10 obtains the one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication. The performance monitoring metrics and outcome may comprise one or more performance monitoring metrics and / or outcome. A monitoring data sample may comprise at least one of the following: all or part of a CSI prediction algorithm output wherein the CSI prediction algorithm output comprises one or more predicted CSIs for the one or multiple future time instances; and / or a ground truth label associated to all or part of a CSI prediction algorithm output wherein the ground truth label comprises one or more measured CSIs for the one or multiple future time instances.

[0119] The ground-truth label format may be configured as at least one of the following: raw channel measurement associated to the one or multiple future time instances; and processed or quantized channel measurement associated to the one or multiple of future time instances.

[0120] The processed or quantized channel measurement may be represented in the format of a CSI determined from a pre-specified codebook.

[0121] The ground-truth label format may be pre-defined in specification.

[0122] The UE 10 may obtain the one or more intermediate KPIs for all or a subset of MIMO layers and for all or a subset of the future time instances, based on the received configuration indication.

[0123] The single intermediate KPI may be derived using at least one of the following methods: averaging intermediate KPI values for different MIMO layers over one or more future time instances; and / or a weighted sum operation of the intermediate KPI values for different Ml MO layers over one or more future time instances.

[0124] Action 904. The UE 10 reports the one or more performance monitoring results to the network node 120. The UE 10 may report the one or more performance monitoring results by reporting multiple intermediate KPIs based on a monitoring data sample, where each intermediate KPI is associated to a different Ml MO layer and / or a different future time instance. The UE 10 may report the one or more performance monitoring results by reporting a single intermediate KPI based on a monitoring data sample.

[0125] Example embodiments of a method performed by the network node 120, such as the radio network node 12, for handling communication in the wireless communications network will now be described with reference to a flowchart depicted in Fig. 10. The actions do not have to be taken in the order stated below, but may be taken in any suitable order. Optional actions are marked in dashed boxes.

[0126] Action 1001. The network node 120 may determine the configuration indication. The network node 120 may , receive the configuration indication, be preconfigured with configuration indication, and / or estimate configuration indication.

[0127] Action 1002. The network node 120 transmits the configuration indication to the UE 10. The configuration indication indicates the content and / or the format of the one or more performance monitoring results associated with the CSI prediction functionality. The CSI prediction functionality predicts CSI in one or multiple future time instances, for example, a UE-side computational model, which computational model may predict a signaling parameter in one or multiple future time instances. The configuration indication indicates at least one or more of the following:

[0128] - the ground-truth label format used to derive one or more performance monitoring results. For example, Ground-truth CSI format, e.g., used to derive the intermediate performance metrics;

[0129] - the number of monitoring data samples to use for deriving one or more performance monitoring results. one or more performance monitoring result definitions, such as one or more performance metrics definition, e.g., SGCS and / or NMSE; and / or a format for performance monitoring result reporting

[0130] The monitoring data sample may comprise at least one of the following: all or part of a CSI prediction algorithm output wherein the CSI prediction algorithm output comprises one or more predicted CSIs for the one or multiple future time instances; and a ground truth label associated to all or part of a CSI prediction algorithm output wherein the ground truth label comprises one or more measured CSIs for the one or multiple future time instances.

[0131] The format for performance monitoring result reporting may be configured as one or more intermediate KPIs per monitoring data sample.

[0132] The one or more intermediate KPIs are defined as at least one of the following: SGCS between all or a part of the CSI prediction algorithm output and an associated ground truth label; and NMSE between all or part of the CSI prediction model / algorithm output and the associated ground truth label.

[0133] The format of the one or more performance monitoring results may be configured as a statistics of one or more intermediate KPIs over M monitoring data samples, where M >1.

[0134] The statistics of the one or more intermediate KPIs may be represented using at least one of the following formats: a percentile value of the intermediate KPI associated to the M monitoring data samples; and / or a mean and / or variance of the intermediate KPI associated to the M monitoring data samples.

[0135] The format of the one or more performance monitoring results is configured as a percentage of monitoring data samples for which the associated intermediate KPI and / or the associated statistics of intermediate KPI fulfills one or more conditions.

[0136] The one or more conditions may be configured as part of the configuration indication.

[0137] The format of the one or more performance monitoring results may be a flag indicating whether the model monitoring metrics for the monitoring data samples fulfill one or more conditions. The model monitoring metrics may be a single intermediate KPI based on a monitoring data sample, or the statistics of one or more intermediate KPIs over M monitoring data samples, or a percentage of the monitoring data samples for which the associated intermediate KPIs and / or the associated statistics of intermediate KPI fulfills one or more conditions.

[0138] The one or more conditions include that SGCS is above a certain threshold, and / or that SGCS of a proportion of the data monitoring samples is above a threshold. Action 1003. The network node 120 receives the one or more performance monitoring results from the UE 10. For example, the network node 120 may receive a result indication comprising the one or more indications of performance metrics of the computational model based on the received configuration. The network node 120 may receive the one or more performance monitoring results by receiving multiple intermediate KPIs based on a monitoring data sample, where each intermediate KPI is associated to a different Ml MO layer and / or a different future time instance. The network node 120 may receive the one or more performance monitoring results by receiving the single intermediate KPI based on the monitoring data sample.

[0139] Action 1004. The network node 120 may handle the CSI prediction functionality taking the one or more performance monitoring results into account. For example, the network node 120 may handle or process computational model, such as AI / ML model LCM, decisions or computational model feature reconfiguration decisions taking the one or more performance monitoring results into account. The CSI prediction functionality may be based on the one or more Al and / or ML models and / or one or more non-AI and / or ML based CSI prediction algorithms.

[0140] Thus, embodiments may provide one or more of the following advantages:

[0141] - The proposed solution enables the network node 120 to instruct the UE 10 to do performance monitoring for the CSI prediction functionality that performs CSI prediction over multiple future time instances and acquires information about the one or more performance monitoring results such as one or more performance metrics and / or monitoring results from the UE 10.

[0142] - The proposed solution also enables the testability of the one or more performance monitoring results reported from the UE 10, hence, the network node 120 can trust these reported one or more performance monitoring results when making AI / ML model LCM decisions or feature reconfiguration decisions related to the CSI prediction functionality.

[0143] In the text herein, the concept of ‘network (NW)’ and / or a gNB can be understood as a generic network node, gNB, base station, unit within the base station, relay node, core network node, a core network node, or a device supporting D2D communication. The node may be deployed in a 5G network, or a 6G network. Moreover, although the term AI / ML model uses a single form, it should be well understood that it should not prevent the implementation of more than one AI / ML model. The UE may either be configured or autonomously switch between the models depending on certain conditions and / or proprietary implementations.

[0144] A network node signals performance monitoring configuration for a CSI prediction feature / functionality / model / algorithm to a UE, based on which the UE derives one or more performance monitoring results, and reports the derived one or more performance monitoring results to the NW node.

[0145] The CSI prediction functionality, may be referred to as CSI prediction feature / model / algorithm, is implemented at the UE 10, and it can be an Al-based scheme, e.g., a neural network model trained by using a large dataset and then deployed at the UE, or non-AI based scheme, e.g., an Auto-Regression based algorithm, or a Kalman-filer based algorithm. The CSI prediction functionality can be used to predict raw channel, or quantity that depends on raw channel, e.g., eigenvectors of the channel.

[0146] The network node 120 may be the gNB, e.g. the gNB-CU or the gNB-DU, the CAM, or a core network node, e.g. the NWDAF. In the following the term NW can refer to any of the aforementioned NW nodes.

[0147] Some embodiments herein propose methods on how to configure the one or more contents and / or formats of one or more performance monitoring results associated to a UE-side CSI prediction model / algorithm / functionality / feature.

[0148] In an embodiment, the one or more performance monitoring results content and / or format configuration includes at least one of the following:

[0149] • The number of monitoring data samples, M, to be collected for constructing the monitoring data set to determine the one or more performance monitoring results, or a minimum percentage over the number of available samples during the performance monitoring interval, needed for determining the one or more performance monitoring results. o A CSI prediction model / algorithm input sample comprises one or more measured CSIs associated to the one or multiple historical time instances, e.g., the measured CSIs on the K CSI-RS resources within the observation window shown in Fig. 6. A CSI prediction model / algorithm output sample comprises one or more predicted CSIs for one or multiple future time instances, e.g., the predicted CSIs on the N4time slots in the prediction window shown in Fig. 6. A ground truth label for a CSI prediction model / algorithm output sample comprises one or more measured CSIs for the corresponding one or multiple future time instances, e.g., the measured CSIs on the N4time slots. o A monitoring data sample can be defined as, e.g.,

[0150] ■ an CSI prediction model / algorithm input sample, e.g., used for deriving input data distribution, or

[0151] ■ an CSI prediction model / algorithm output sample, e.g., used for deriving output data distribution, or

[0152] ■ consists of both a CSI prediction model / algorithm output sample and the associated ground-truth label for the CSI prediction model / algorithm output sample, e.g., used for deriving prediction accuracy in terms of intermediate KPIs like SGCS and / or NMSE, and / or data distribution related KPIs.

[0153] • The ground-truth label format used to derive the one or more performance monitoring results such as intermediate performance metrics, or input / output data distribution related performance metrics. The ground-truth label format for a model / algorithm output sample can be configured by the network node 120 or defined in specification. Examples of ground-truth label format include o Raw channel measurements associated to one or multiple future time instances, such as raw channel measurements associated to each of the predicted future time instances, or .raw channel measurements associated to a subset of the predicted future time instances. o Processed / quantized channel measurements associated to the one or multiple of future time instances such as all or a subset of the predicted future time instances. The processing / quantization can be done by using, e.g.,

[0154] ■ The Rel-16 Type II CSI codebook, or

[0155] ■ The enhanced eTypell CSI formats with new parameter combinations, or

[0156] ■ The Rel-18 Type II codebook (Type II CSI codebook with doppler domain compression)

[0157] ■ Al-based compression / quantization at the output of an Al model deployed at the UE

[0158] • The format for performance monitoring result reporting also referred to as performance monitoring result format. Examples of the format for performance monitoring result reporting may be configured as: o One or more Intermediate KPIs per monitoring data sample ■ As one option, multiple intermediate KPIs are defined for a given monitoring data sample, where the multiple intermediate KPIs associated to different Ml MO layers and / or different prediction future time instances.

[0159] • For example, consider a CSI prediction functionality such as a CSI prediction model / algorithm configured to predict CSI for rank 2 at three different future time instances (t1, t2 and t3), and assume that the intermediate KPI is defined as the SGCS value between the predicted CSI, being a CSI prediction algorithm output, and the associated ground-truth label per layer per prediction time instance, hence, six intermediate KPI values

[0160] { SGCSJ1_t1(n), SGCS_l2_t1(n), SGCS_l1_t2(n), SGCS_l2_t2(n), SGCSJ1_t3(n), SGCSJ2_t3(n)}, can be derived to indicate the CSI prediction performance for monitoring data sample n, where intermediate KPI, SGCS_li_tj(n), is the SGCS value for the predicted CSI for layer i and prediction time instance ].

[0161] • In another example, the UE 10 may be configured with which subset of future time instances for which intermediate KPIs are to be derived by the UE 10. For instance, the UE 10 may be configured with a bit map or a parameter with multiple enumerated values which indicate the subset of future time instances for which intermediate KPIs are to be derived by the UE 10. Considering a CSI prediction functionality configured to predict CSI for rank 2 at three different future times (t1 , t2, and t3), the following parameter tdKPI may be configured: tdKPI-r19 ENUMERATED {n1 ,n13, n123} OPTIONAL - Need R where o the value n1 that the intermediate KPI, e.g., SGCS, is to be derived for the time instance t1. Hence, two intermediate KPI values { SGCSJ1_t1(n), SGCS_l2_t1 (n)}, can be derived to indicate the CSI prediction performance, being an example of the performance monitoring result, for monitoring data sample n, where intermediate KPI, SGCS_li_t1(n), is the SGCS value for the predicted CSI for layer i and prediction time instance t1; o the value n13 that the intermediate KPI (e.g., SGCS) is to be derived for the time instances t1 and t3. hence, four intermediate KPI values { SGCSJ1_t1(n), SGCS_l2_t1(n), SGCS_l1_t3(n), SGCSJ2_t3(n)}, can be derived to indicate the CSI prediction performance for monitoring data sample n, where intermediate KPI, SGCS_li_tj(n), is the SGCS value for the predicted CSI for layer i and prediction time instance ); o the value n123 that the intermediate KPI (e.g., SGCS) is to be derived for the time instances t1, t2 and t3. hence, six intermediate KPI values { SGCSJ1_t1(n), SGCS_l2_t1(n), SGCS_l1_t2(n), SGCS_l2_t2(n), SGCSJ1_t3(n), SGCS_l2_t3(n)}, can be derived to indicate the CSI prediction performance for monitoring data sample n, where intermediate KPI, SGCS_li_tj(n), is the SGCS value for the predicted CSI for layer i and prediction time instance j;

[0162] • To transmit the performance monitoring results, the SGCS value may be quantized. In one example, each time instance has the same quantization threshold. E.g., for a simple 2 bits performance monitoring report, 00, 01 , 10, and 11 may represent SGCS < th1, th1 < SGCS < th2, th2 < SGCS < th3, and SGCS > th3, respectively. In another example, each time instance may have different quantization threshold. E.g., for a case of monitoring of CSI prediction for two time-instances with a simple 2-bit performance monitoring report, 00, 01 , 10, and 11 may represent SGCS < th1_0, th1_0 < SGCS < th2_0, th2_0 < SGCS < th3_0, and SGCS > th3_0, respectively for the first time instance and may represent SGCS < th1_1, th1_1 < SGCS < th2_1 , th2_1 < SGCS < th3_1 , and SGCS > th3_1, respectively for the second time instance. The threshold values may be predefined in the standard or configured by the network node 120. Note that instead of being defined / configured for each time instance, the threshold values may also be predefined / configured based on the absolute distance of the predicted time instance with a certain time reference, e.g., the last measured CSI-RS.

[0163] • Alternatively, the quantization values for SGCS may be predefined in a table in 3GPP specifications and threshold values may not be explicitly defined or configured. E.g., for a simple 2 bits performance monitoring report, 00, 01, 10, and 11 may represent four different SGCS values predefined in 3GPP specifications. The UE 10 quantizes a derived SGCS value to the closest quantized SGCS value predefined in 3GPP specifications.

[0164] • Further, each of the future time instance may have, either via NW configuration or predefined in the standard, different reporting sizes. As an example, consider a CSI prediction model / algorithm configured to predict four different future time instances (t1 , t2, t3, and t4). The reporting size for CSI prediction performance monitoring for those time instances may be b1, b2, b3, and b4. For example, b1, b2, b3, and b4 may be 4, 4, 3, 2 bits respectively. The similar mechanism may also be additionally or optionally be applied for different layers. ■ As another option, the UE 10 may report a single intermediate KPI based on a monitoring data sample. For example, the UE 10 may report a single intermediate KPI that is derived for a monitoring data sample. For example, a set of weights (e.g., w_l_t1 , wJ2_t1, w_l1_t2, wJ2_t2, wJ1_t3, w_l2_t3) can be configured / defined mapping to the intermediate KPI values for different layers / prediction-time-instances, hence, a single intermediate KPI value can be derived for a single monitoring data sample n using weighted sum operation, e.g., SGCS (n)= sum(w_li_tj*SGCS_li_tj(n)), where n is the monitoring sample index, i is the layer index, and j is the prediction time instance index.

[0165] • In a special case, the single intermediated KPI is calculated by average the intermediated KPI values for different Ml MO layers and / or prediction time instances.

[0166] • Different weights can be configured for different Ml MO layers, e.g., when it is more important to check the performance of the lower Ml MO layers. o In a special case, the weights per MIMO layer can be a function of eigenvalues of the channel associated with the MIMO layer, e.g., proportional to the eigenvalues for a given MIMO layer.

[0167] • Different weights can be applied for different time instances, e.g., when it is more important to check the prediction performance of the first few time instances. o In a special case, for a reduced complexity at the UE 10, KPIs for a subset of time instance is computed per monitoring data sample. For example, either the KPI for first time instance or last time instance or only both first- and last-time instances per monitoring data sample is computed for performance monitoring.

[0168] Hence the single intermediate KPI may be derived using at least one of the following methods: averaging intermediate KPI values for different MIMO layers over one or more future time instances; and / or a weighted sum operation of the intermediate KPI values for different MIMO layers over one or more future time instances. o The format of the one or more performance monitoring results may be configured as statistics of one or more intermediate KPIs over M monitoring data samples, where M >1. That is, statistics of intermediate KPIs over M multiple monitoring data samples

[0169] ■ As an example, the model monitoring metrics is defined as the CDF at X percentile(s), e.g., 50% or / and 90%, of the intermediate KPIs associated to the collected M monitoring data samples.

[0170] ■ As another example, the model monitoring metrics is defined as the mean or / and variance of the intermediate KPIs associated to the collected M monitoring data samples. o The format of the one or more performance monitoring results may be configured as a percentage of monitoring data samples for which the associated intermediate KPI and / or the associated statistics of intermediate KPI fulfills one or more conditions. That is, the percentage of the monitoring data samples for which the associated intermediate KPI values fulfills a certain condition or conditions, and the configuration of the condition or conditions.

[0171] ■ As an example, the condition is defined as the SCGS value for layer 1 is above a threshold 1 , e.g., 0.8, and the SGCS value for layer 2 is above a threshold 2, e.g., 0.7. Then, the UE 10 may report the performance in terms of the percentage of the monitoring data samples, e.g., 90%, that fulfill this condition. Thus, the one or more conditions may include that SGCS per MIMO layer is above a certain threshold.

[0172] ■ As another example, the condition is defined as a SGCS value, e.g., per layer or averaged over layers, larger than a threshold 1, e.g., 0.8, at prediction time t1, and have a SGCS value, e.g., per layer or averaged over layers, larger than a threshold 2, e.g., 0.7, at another prediction time t2, where t2>t1. Then, the UE 10 reports the performance monitoring result in terms of the percentage of the monitoring data samples, e.g., 80%, that fulfill this condition. Thus, the one or more conditions may include that SGCS per layer per future time instance is above a certain threshold, or SGCS per future time instance averaged over layers is above a certain threshold. Statistics of input / output data distribution over M multiple monitoring data samples The format of the one or more performance monitoring results may be a flag indicating whether a model monitoring metrics for the monitoring data samples fulfill one or more conditions. That is, a flag indicating whether the model monitoring metrics for the monitoring data samples fulfill one or more certain conditions, and the configuration of condition parameters. If the one or more conditions is / are satisfied, then the CSI prediction functionality such as a model / algorithm / feature / function is considered to be in normal operation for this monitoring occasion; otherwise, anomaly of operation is detected.

[0173] ■ As an example, if X%, e.g., 90%, of the monitoring data samples have a SGCS value larger than 0.8 for layer 1 and larger than 0.7 for layer 2, then the CSI prediction functionality such as a model / algorithm / feature / function is considered as working normally. Thus, the one or more conditions may include that SGCS per layer of a proportion of the data monitoring samples is above a threshold.

[0174] ■ As another example, if X%, e.g., 90%, of the monitoring data samples have a SGCS value, e.g., per layer or averaged over layers, larger than 0.8 at prediction time t1, and have a SGCS value larger than 0.7 at prediction time t2 (t2>t1), then the CSI prediction functionality such as a model / algorithm / feature / function is considered as working normally. Thus, the one or more conditions may include that SGCS per future time instance of a proportion of the data monitoring samples is above a threshold. A bitmap where the t-th bit indicates whether the model monitoring metrics for the t-th time instance over one or more monitoring data samples fulfill a certain condition, where the condition can be specified or configured by the network. For example, consider a CSI prediction functionality configured to predict CSI for rank r at N4different future time instances and assume that an intermediate KPI, such as SCGS, NMSE, etc., is defined between the predicted CSI and the associated ground-truth label per layer per prediction time instance, resulting in the intermediate KPI values, given by, KPI£,t.(n), which indicates the CSI prediction performance for the j-th layer and the j-th prediction time instance for the n-th monitoring data sample. Accordingly, a N4length bit map, denoted by, b - [<>., . <><„,]. can be formed where each bit can be given by, where a7is a pre-determined threshold and rl, x e cA

[0175] 10, x g cZT and the function f(.) can be used to get statistics of the KPIs over all r layers and over all (or subset of all) monitoring data samples. Further for a specific example, with M monitoring data samples, the bit map at the t-th time instance is given by such that SGCS is the used intermediate KPI and 0 is the eigenvalue associated with the j-th layer at the j-th prediction instance for the n-th monitoring data sample.

[0176] ■ In another related example, the KPIs can be defined over raw predicted channels. Specifically, consider a CSI prediction functionality configured to predict CSI at N4different future time instances and assume that an intermediate KPI (e.g. NMSE) is defined between the predicted CSI and the associated ground-truth label per prediction time instance, resulting in the intermediate KPI values, given by,

[0177] KPIt.(n), which indicates the CSI prediction performance for the j-th prediction time instance for the n-th monitoring data sample. Accordingly, a N4length bit map, denoted by, b - [<>., . <><„,]. can be formed where each bit can be given by, where for a specific example, with M monitoring data samples, the bit map at the t-th time instance is given by such that NMSEtis the NMSE between the predicted CSI and the corresponding ground truth the j-th prediction instance for the n-th monitoring data sample. o Granularity of the one or more performance monitoring results such as performance monitoring metric

[0178] For all the above, the granularity of the performance monitoring metrics can either be subband or wideband. When subband metric is calculated, the subband size could be the same as the CSI reporting subband size, or it could be different compared to the CSI reporting subband size. o Quantization method, e.g., how to quantize the derived intermediate KPI values.

[0179] In one embodiment, the performance monitoring metric value, e.g., SGCS, or NMSE, if reported, is quantized, e.g., uniformly in linear or log scale, then the corresponding codepoint for the quantized value is reported to the network node 120.

[0180] In a dependent embodiment, the quantization is done within an interval other than between 0 and 1, e.g., quantization is applied for SGCS between 0.5 and 0.8.

[0181] Fig. 11 shows a block diagram depicting the UE 10 for handling communication in the wireless communications network.

[0182] The UE 10 may comprise processing circuitry 1101 , e.g. one or more processors, configured to perform the methods herein. The UE 10 and / or the processing circuitry 1101 is configured to receive the configuration indication from the network node 120, wherein the configuration indication indicates content and / or the format of one or more performance monitoring results associated with the CSI prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances. The configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting.

[0183] The UE 10 and / or the processing circuitry 1101 is configured to obtain the one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication; and to report the one or more performance monitoring results to the network node 120.

[0184] The CSI prediction functionality may be based on one or more Al and / or ML models and / or one or more non-AI and / or ML based CSI prediction algorithms.

[0185] The monitoring data sample may comprise at least one of the following: all or part of a CSI prediction algorithm output wherein the CSI prediction algorithm output comprises one or more predicted CSIs for the one or multiple future time instances; and a ground truth label associated to all or part of a CSI prediction algorithm output wherein the ground truth label comprises one or more measured CSIs for the one or multiple future time instances.

[0186] The ground-truth label format may be configured as at least one of the following: raw channel measurement associated to the one or multiple future time instances; and processed or quantized channel measurement associated to the one or multiple of future time instances.

[0187] The processed or quantized channel measurement may be represented in the format of a CSI determined from a pre-specified codebook.

[0188] The ground-truth label format may be pre-defined in specification.

[0189] The format for the performance monitoring results reporting may be configured as one or more intermediate KPIs per monitoring data sample. The one or more intermediate KPIs may be defined as at least one of the following:

[0190] SGCS between all or a part of the CSI prediction algorithm output and an associated ground truth label; and

[0191] NMSE between all or part of the CSI prediction model / algorithm output and the associated ground truth label.

[0192] The UE 10 and / or the processing circuitry 1101 may be configured to obtain the one or more intermediate KPIs for all or a subset of MIMO layers and for all or a subset of the future time instances, based on the received configuration indication.

[0193] The UE 10 and / or the processing circuitry 1101 may be configured to report the one or more performance monitoring results by reporting multiple intermediate KPIs based on a monitoring data sample, where each intermediate KPI is associated to a different MIMO layer and / or a different future time instance.

[0194] The UE 10 and / or the processing circuitry 1101 may be configured to report the one or more performance monitoring results by reporting a single intermediate KPI based on a monitoring data sample.

[0195] The single intermediate KPI may be derived using at least one of the following methods: averaging intermediate KPI values for different MIMO layers over one or more future time instances; and / or a weighted sum operation of the intermediate KPI values for different MIMO layers over one or more future time instances.

[0196] The format of the one or more performance monitoring results may be configured as a statistics of one or more intermediate KPIs over M monitoring data samples, where M >1.

[0197] The statistics of the one or more intermediate KPIs may be represented using at least one of the following formats: a percentile value of the intermediate KPI associated to the M monitoring data samples; and / or a mean and / or variance of the intermediate KPI associated to the M monitoring data samples.

[0198] The format of the one or more performance monitoring results may be configured as a percentage of monitoring data samples for which the associated one or more intermediate KPIs and / or the associated statistics of one or more intermediate KPIs fulfills one or more conditions. The one or more conditions may either be configured as part of the configuration indication or pre-defined in specification.

[0199] The format of the one or more performance monitoring results may be a flag indicating whether a model monitoring metrics for the monitoring data samples fulfill one or more conditions.

[0200] The model monitoring metrics may be one or more intermediate KPIs based on a monitoring data sample, or the statistics of one or more intermediate KPIs over M monitoring data samples, or a percentage of the monitoring data samples for which the associated one or more intermediate KPIs and / or the associated statistics of one or more intermediate KPIs fulfills one or more conditions.

[0201] The one or more conditions may include that the one or more associated intermediate KPIs and / or the associated statistics of the one or more intermediate KPIs is / are above one or more thresholds.

[0202] The one or more conditions include that the associated one or more intermediate KPIs of a proportion of the monitoring data samples is / are above one or more thresholds.

[0203] Thus, the UE 10 and / or the processing circuitry 1101 is configured to receive the configuration indication from the network node 120. The configuration indication indicates content and / or a format of the one or more performance monitoring results associated to a UE-side computational model, which computational model predicts a signaling parameter in one or multiple future time instances. The configuration indication indicates at least one or more of the following:

[0204] Ground-truth CSI format (e.g., used to derive the intermediate performance metrics)

[0205] - The number of data samples to use for deriving a performance monitoring result. One or more performance metrics definition (e.g., SGCS and / or NMSE) Performance monitoring result reporting format

[0206] The UE 10 and / or the processing circuitry 1101 may be configured to execute the computational model to obtain a value / a prediction of a signaling parameter such as signal strength or similar.

[0207] The UE 10 and / or the processing circuitry 1101 is configured to obtain the one or more indications of performance metrics of a computational model related to signaling such as a signal strength or quality based on the received configuration.

[0208] The UE 10 and / or the processing circuitry 1101 is configured to report the one or more indications to the network node 120.

[0209] The UE 10 further comprises a memory 1105. The memory comprises one or more units to be used to store data on, such as indications, computational model, performance metrics, reconfiguration, applications to perform the methods disclosed herein when being executed, and similar. The UE 10 comprises a communication interface 1106 comprising transmitter, receiver, transceiver and / or one or more antennas. Thus, it is herein provided the UE 10 for handling communication in a wireless communications network, wherein the UE 10 comprises processing circuitry and a memory, said memory comprising instructions executable by said processing circuitry whereby said UE 10 is operative to perform any of the methods herein.

[0210] The methods according to the embodiments described herein for the UE 10 are respectively implemented by means of e g. a computer program product 1107 or a computer program product, comprising instructions, i.e. , software code portions, which, when executed on at least one processor, cause the at least one processor to carry out the actions described herein, as performed by the UE 10. The computer program product 1107 may be stored on a computer-readable storage medium 1108, e g. a universal serial bus (USB) stick, a disc or similar. The computer-readable storage medium 1108, having stored thereon the computer program product, may comprise the instructions which, when executed on at least one processor, cause the at least one processor to carry out the actions described herein, as performed by the UE 10. In some embodiments, the computer-readable storage medium may be a non-transitory or transitory computer- readable storage medium.

[0211] Fig. 12 shows a block diagram depicting the network node 120 for handling communication in the wireless communications network.

[0212] The network node 120 may comprise processing circuitry 1201 , e.g. one or more processors, configured to perform the methods herein.

[0213] The network node 120 and / or the processing circuitry 1201 is configured to transmit the configuration indication to the UE 10, wherein the configuration indication indicates content and / or the format of one or more performance monitoring results associated with the CSI prediction functionality. The CSI prediction functionality predicts CSI in one or multiple future time instances and the configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting.

[0214] The network node 120 and / or the processing circuitry 1201 is configured to receive the one or more performance monitoring results from the UE 10.

[0215] The network node 120 and / or the processing circuitry 1201 may be configured to handle the CSI prediction functionality taking the one or more performance monitoring results into account.

[0216] The CSI prediction functionality may be based on one or more Al and / or ML models and / or one or more non-AI and / or ML based CSI prediction algorithms.

[0217] The monitoring data sample may comprise at least one of the following: all or part of a CSI prediction algorithm output wherein the CSI prediction algorithm output comprises one or more predicted CSIs for the one or multiple future time instances; and, a ground truth label associated to all or part of a CSI prediction algorithm output wherein the ground truth label comprises one or more measured CSIs for the one or multiple future time instances.

[0218] The format for performance monitoring result reporting may be configured as one or more intermediate KPIs per monitoring data sample.

[0219] The one or more intermediate KPIs may be defined as at least one of the following:

[0220] SGCS between all or a part of the CSI prediction algorithm output and an associated ground truth label; and

[0221] NMSE between all or part of the CSI prediction model / algorithm output and the associated ground truth label.

[0222] The network node 120 and / or the processing circuitry 1201 may be configured to receive the one or more performance monitoring results by receiving multiple intermediate KPIs based on a monitoring data sample, where each intermediate KPI is associated to a different Ml MO layer and / or a different future time instance.

[0223] The network node 120 and / or the processing circuitry 1201 may be configured to receive the one or more performance monitoring results by receiving a single intermediate KPI based on a monitoring data sample.

[0224] The format of the one or more performance monitoring results may be configured as a statistics of one or more intermediate KPIs over M monitoring data samples, where M>1. The statistics of the one or more intermediate KPIs may be represented using at least one of the following formats: a percentile value of the intermediate KPI associated to the M monitoring data samples; and / or a mean and / or variance of the intermediate KPI associated to the M monitoring data samples.

[0225] The format of the one or more performance monitoring results may be configured as a percentage of monitoring data samples for which the associated intermediate KPI and / or the associated statistics of intermediate KPI fulfills one or more conditions.

[0226] The one or more conditions may be configured as part of the configuration indication.

[0227] The format of the one or more performance monitoring results may be a flag indicating whether a model monitoring metrics for the monitoring data samples fulfill one or more conditions.

[0228] The model monitoring metrics may be a single intermediate KPI based on a monitoring data sample, or the statistics of one or more intermediate KPIs over M monitoring data samples, or a percentage of the monitoring data samples for which the associated intermediate KPIs and / or the associated statistics of intermediate KPI fulfills one or more conditions.

[0229] The one or more conditions may include that the associated one or more intermediate KPIs and / or the associated statistics of one or more intermediate KPI is / are above one or more threshold.

[0230] The one or more conditions may include that the associated one or more intermediate KPIs of a proportion of the monitoring data samples is / are above one or more thresholds.

[0231] Thus, the network node 120 and / or the processing circuitry 1201 is configured to transmit the configuration indication to the UE 10. The configuration indication indicates content and / or a format of the one or more performance monitoring results associated to a UE-side computational model, which computational model predicts a signaling parameter in one or multiple future time instances. The configuration indication indicates at least one or more of the following:

[0232] Ground-truth CSI format (e.g., used to derive the intermediate performance metrics)

[0233] - The number of data samples to use for deriving a performance monitoring result.

[0234] One or more performance metrics definition (e.g., SGCS and / or NMSE) Performance monitoring result reporting format The network node 120 and / or the processing circuitry 1201 may be configured to receive the one or more result indications from the UE 10, which result indication comprises the one or more indications of performance metrics of the computational model related to signaling such as a signal strength or quality based on the received configuration.

[0235] The network node 120 and / or the processing circuitry 1201 may be configured to handle or process computational model, such as AI / ML model LCM, decisions or computational model feature reconfiguration decisions taking the one or more result indications into account.

[0236] The network node 120 further comprises a memory 1205. The memory comprises one or more units to be used to store data on, such as indications, format, content, performance metrics information, computational model, performance metrics, reconfiguration, applications to perform the methods disclosed herein when being executed, and similar. The network node 120 comprises a communication interface 1206 comprising transmitter, receiver, transceiver and / or one or more antennas. Thus, it is herein provided the network node 120 for handling communication in a wireless communications network, wherein the network node 120 comprises processing circuitry and a memory, said memory comprising instructions executable by said processing circuitry whereby said network node 120 is operative to perform any of the methods herein.

[0237] The methods according to the embodiments described herein for the network node 120 are respectively implemented by means of e.g. a computer program product 1207 or a computer program product, comprising instructions, i.e. , software code portions, which, when executed on at least one processor, cause the at least one processor to carry out the actions described herein, as performed by the network node 120. The computer program product 1207 may be stored on a computer-readable storage medium 1208, e.g. a USB stick, a disc or similar. The computer-readable storage medium 1208, having stored thereon the computer program product, may comprise the instructions which, when executed on at least one processor, cause the at least one processor to carry out the actions described herein, as performed by the network node 120. In some embodiments, the computer-readable storage medium may be a non- transitory or transitory computer-readable storage medium.

[0238] As will be readily understood by those familiar with communications design, that functions means or modules may be implemented using digital logic and / or one or more microcontrollers, microprocessors, or other digital hardware. In some embodiments, several or all of the various functions may be implemented together, such as in a single application-specific integrated circuit (ASIC), or in two or more separate devices with appropriate hardware and / or software interfaces between them. Several of the functions may be implemented on a processor shared with other functional components of a radio network node, for example.

[0239] Alternatively, several of the functional elements of the processing means discussed may be provided through the use of dedicated hardware, while others are provided with hardware for executing software, in association with the appropriate software or firmware. Thus, the term “processor” or “controller” as used herein does not exclusively refer to hardware capable of executing software and may implicitly include, without limitation, digital signal processor (DSP) hardware, read-only memory (ROM) for storing software, random-access memory for storing software and / or program or application data, and non-volatile memory. Other hardware, conventional and / or custom, may also be included. Designers of communications receivers will appreciate the cost, performance, and maintenance trade-offs inherent in these design choices.

[0240] In certain embodiments, some or all of the functionality described herein may be provided by processing circuitry executing instructions stored on in memory, which in certain embodiments may be a computer program product in the form of a non- transitory computer-readable storage medium. In alternative embodiments, some or all of the functionality may be provided by the processing circuitry without executing instructions stored on a separate or discrete device-readable storage medium, such as in a hard-wired manner. In any of those particular embodiments, whether executing instructions stored on a non-transitory computer-readable storage medium or not, the processing circuitry can be configured to perform the described functionality. The benefits provided by such functionality are not limited to the processing circuitry alone or to other components of the computing device, but are enjoyed by the computing device as a whole, and / or by end users and a wireless network generally.

[0241] It will be appreciated that the foregoing description and the accompanying drawings represent non-limiting examples of the methods and apparatus taught herein. As such, the apparatus and techniques taught herein are not limited by the foregoing description and accompanying drawings. Instead, the embodiments herein are limited only by the following claims and their legal equivalents. ABBREVIATIONS

Claims

CLAIMS1. A method performed by a user equipment, UE, (10) for handling communication in a wireless communications network, the method comprising: receiving (901) a configuration indication from a network node (120), wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a channel state information, CSI, prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances, and the configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting; obtaining (903) one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication; and reporting (904) the one or more performance monitoring results to the network node (120).

2. The method of claim 1, wherein the CSI prediction functionality is based on one or more artificial intelligent, Al, and / or machine learning, ML, models and / or one or more non-AI and / or ML based CSI prediction algorithms.

3. The method of any of claims 1-2, wherein a monitoring data sample comprises at least one of the following: all or part of a CSI prediction algorithm output wherein the CSI prediction algorithm output comprises one or more predicted CSIs for the one or multiple future time instances; and a ground truth label associated to all or part of a CSI prediction algorithm output wherein the ground truth label comprises one or more measured CSIs for the one or multiple future time instances.

4. The method of any of claims 1-3, wherein the ground-truth label format is configured as at least one of the following: raw channel measurement associated to the one or multiple future time instances; and processed or quantized channel measurement associated to the one or multiple of future time instances.

5. The method of the claim 4, wherein the processed or quantized channel measurement is represented in the format of a CSI determined from a pre-specified codebook.

6. The method of any of claims 1-5, wherein the ground-truth label format is pre-defined in specification.

7. The method of any of the claims 1-6, wherein the format for performance monitoring results reporting is configured as one or more intermediate key performance indicators, KPI, per monitoring data sample.

8. The method of the claim 7, wherein the one or more intermediate KPIs are defined as at least one of the following: squared generalized cosine similarity, SGCS, between all or a part of the CSI prediction algorithm output and an associated ground truth label; and normalized mean square error, NMSE, between all or part of the CSI prediction model / algorithm output and the associated ground truth label.

9. The method of any of the claims 7- 8, wherein the UE (10) obtains the one or more intermediate KPIs for all or a subset of multiple input multiple output, MIMO, layers and for all or a subset of the future time instances, based on the received configuration indication.

10. The method of any of the claims 7- 9, wherein reporting (904) the one or more performance monitoring results comprises reporting multiple intermediate KPIs based on a monitoring data sample, where each intermediate KPI is associated to a different MIMO layer and / or a different future time instance.

11. The method of any of the claims 7- 9, wherein reporting (904) the one or more performance monitoring results comprises reporting a single intermediate KPI based on a monitoring data sample.

12. The method of claim 11, wherein the single intermediate KPI is derived using at least one of the following methods: averaging intermediate KPI values for different MIMO layers over one or more future time instances; and / or a weighted sum operation of the intermediate KPI values for different MIMO layers over one or more future time instances.

13. The method of any of claims 1-12, wherein the format of the one or more performance monitoring results is configured as a statistics of one or more intermediate key performance indicators, KPI, over M monitoring data samples, where M >1.

14. The method of claim 13, wherein the statistics of the one or more intermediate KPIs is represented using at least one of the following formats: a percentile value of the intermediate KPI associated to the M monitoring data samples; and / or a mean and / or variance of the intermediate KPI associated to the M monitoring data samples.

15. The method of any of claims 1-14, wherein the format of the one or more performance monitoring results is configured as a percentage of monitoring data samples for which the associated one or more intermediate KPIs and / or the associated statistics of one or more intermediate KPIs fulfills one or more conditions.

16. The method of claim 15, wherein the one or more conditions are either configured as part of the configuration indication or pre-defined in specification.

17. The method of any of claims 1-16, wherein the format of the one or more performance monitoring results is a flag indicating whether a model monitoring metrics for the monitoring data samples fulfill one or more conditions.

18. The method of claim 17, wherein the model monitoring metrics is one or more intermediate KPIs based on a monitoring data sample, or the statistics of one or moreintermediate KPIs over M monitoring data samples, or a percentage of the monitoring data samples for which the associated one or more intermediate KPIs and / or the associated statistics of one or more intermediate KPIs fulfills one or more conditions.

19. The method of any of claims 15-18, wherein the one or more conditions include that the one or more associated intermediate KPIs and / or the associated statistics of the one or more intermediate KPIs is / are above one or more thresholds.

20. The method of any of claims 15-19, wherein the one or more conditions include that the associated one or more intermediate KPIs of a proportion of the monitoring data samples is / are above one or more thresholds.

21. A method performed by a network node (120) for handling communication in a wireless communications network, the method comprising: transmitting (1002) a configuration indication to a user equipment, UE, (10), wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a channel state information, CSI, prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances, and the configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting; and receiving (1003) one or more performance monitoring results from the UE (10).

22. The method according to claim 21, further comprising handling (1004) the CSI prediction functionality taking the one or more performance monitoring results into account.

23. The method of any of the claims 21-22, wherein the CSI prediction functionality is based on one or more artificial intelligent, Al, and / or machine learning, ML, models and / or one or more non-AI and / or ML based CSI prediction algorithms.

24. The method of any of claims 21-23, wherein a monitoring data sample comprises at least one of the following: all or part of a CSI prediction algorithm output wherein the CSI prediction algorithm output comprises one or more predicted CSIs for the one or multiple future time instances; and, a ground truth label associated to all or part of a CSI prediction algorithm output wherein the ground truth label comprises one or more measured CSIs for the one or multiple future time instances.

25. The method of any of the claims 21-24, wherein the format for performance monitoring result reporting is configured as one or more intermediate key performance indicators, KPI, per monitoring data sample.

26. The method of the claim 25, wherein the one or more intermediate KPIs are defined as at least one of the following: squared generalized cosine similarity, SGCS, between all or a part of the CSI prediction algorithm output and an associated ground truth label; and normalized mean square error, NMSE, between all or part of the CSI prediction model / algorithm output and the associated ground truth label.

27. The method of any of the claims 25- 26, wherein receiving (1003) the one or more performance monitoring results comprises receiving multiple intermediate KPI based on a monitoring data sample, where each intermediate KPI is associated to a different MIMO layer and / or a different future time instance.

28. The method of any of the claims 25- 27, wherein receiving (1003) the one or more performance monitoring results comprises receiving a single intermediate KPI based on a monitoring data sample.

29. The method of any of claims 25-28, wherein the format of the one or more performance monitoring results is configured as a statistics of one or more intermediate key performance indicators, KPI, over M monitoring data samples, where M>1.

30. The method of claim 29, wherein the statistics of the one or more intermediate KPIs is represented using at least one of the following formats: a percentile value of the intermediate KPI associated to the M monitoring data samples; and / or a mean and / or variance of the intermediate KPI associated to the M monitoring data samples.

31. The method of any of claims 21-30, wherein the format of the one or more performance monitoring results is configured as a percentage of monitoring data samples for which the associated intermediate KPI and / or the associated statistics of intermediate KPI fulfills one or more conditions.

32. The method of claim 31, wherein the one or more conditions are configured as part of the configuration indication.

33. The method of any of claims 21-32, wherein the format of the one or more performance monitoring results is a flag indicating whether a model monitoring metrics for the monitoring data samples fulfill one or more conditions.

34. The method of claim 33, wherein the model monitoring metrics is a single intermediate KPI based on a monitoring data sample, or the statistics of one or more intermediate KPIs over M monitoring data samples, or a percentage of the monitoring data samples for which the associated intermediate KPIs and / or the associated statistics of intermediate KPI fulfills one or more conditions.

35. The method of any of claims 31-34, wherein the one or more conditions include that the associated one or more intermediate KPIs and / or the associated statistics of one or more intermediate KPI is / are above one or more threshold.

36. The method of any of claims 31-35, wherein the one or more conditions include that the associated one or more intermediate KPIs of a proportion of the monitoring data samples is / are above one or more thresholds.

37. A user equipment, UE, (10) for handling communication in a wireless communications network, wherein the UE (10) is configured to:receive a configuration indication from a network node (120), wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a channel state information, CSI, prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances, and the configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting; obtain one or more performance monitoring results for the CSI prediction functionality based on the received configuration indication; and report the one or more performance monitoring results to the network node (120).

38. The UE (10) according to claim 37, wherein the UE is configured to perform the method according to any of the claims 2-20.

39. A network node (120) for handling communication in a wireless communications network, wherein the network node (120) is configured to: transmit a configuration indication to a user equipment, UE, (10), wherein the configuration indication indicates content and / or a format of one or more performance monitoring results associated with a channel state information, CSI, prediction functionality, which CSI prediction functionality predicts CSI in one or multiple future time instances, and the configuration indication indicates at least one or more of the following: a ground-truth label format used to derive one or more performance monitoring results; a number of monitoring data samples to use for deriving one or more performance monitoring results; one or more performance monitoring result definitions; and / or a format for performance monitoring result reporting; and receive one or more performance monitoring results from the UE (10).

40. The network node (120) according to claim 39, wherein the network node (120) is configured to perform the method according to any of the claims 22-36.

41. A computer program product comprising instructions, which, when executed on at least one processor, cause the at least one processor to carry out the method according to any of the claims 1-36, as performed by the network node and the UE, respectively.

42. A computer-readable storage medium, having stored thereon a computer program product comprising instructions which, when executed on at least one processor, cause the at least one processor to carry out the method according to any of the claims 1-36, as performed by the network node and the UE, respectively.

Images