Performance monitoring on prediction model

Abstract

Embodiments of the present disclosure relate to performance monitoring on prediction model In an aspect, a terminal device receives, from a network device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model. The terminal device takes one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received or wherein the indication associated with the switch is received and a previous input of the prediction model is measurements; or measurements of the feature, wherein the indication of the second input type is received or wherein the indication associated with the switch is received and a previous input of the prediction model is prediction results.

Description

PERFORMANCE MONITORING ON PREDICTION MODELCROSS-REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of US provisional application No. 63 / 724,675, filed November 25, 2024. The content of which are hereby incorporated by reference in their entirety.FIELD

[0002] Various example embodiments relate to the field of telecommunication and in particular, to a terminal device, a network device, methods, apparatuses and a computer readable medium for performance monitoring on prediction model.BACKGROUND

[0003] A communication network can be seen as a facility that enables communications between two or more communication devices, or provides communication devices access to a data network. A mobile or wireless communication network is one example of a communication network.

[0004] Such communication networks operate in accordance with standards, such as those promulgated by 3GPP (Third Generation Partnership Project) or ETSI (European Telecommunications Standards Institute). Examples of such standards include the so-called 5G (5th Generation) standard or other standards promulgated by 3GPP.SUMMARY

[0005] In general, example embodiments of the present disclosure provide a solution for performance monitoring on prediction model.

[0006] In a first aspect, there is provided a terminal device. The terminal device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the terminal device at least to: receive, from a network device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model; determine a negative result of performance evaluation on the prediction model; based on determining the negative result, determine whether the first condition is met; and based on determining that the first condition is met, perform an action associated with the prediction model.

[0007] In a second aspect, there is provided a network device. The network device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, theinstructions cause the network device at least to: transmit, to a terminal device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model, the first condition is associated with the terminal device performing an action associated with the prediction model; and perform a prediction on the feature based on the prediction model.

[0008] In a third aspect, there is provided a terminal device. The terminal device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the terminal device at least to: determine at least one negative result of performance evaluation on a prediction model for a feature; transmit, to a network device, at least one negative indication for the prediction model; and receive, from the network device, an indication of an action associated with the prediction model.

[0009] In a fourth aspect, there is provided a network device. The network device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the network device at least to: receive, from a terminal device, at least one negative indication for a prediction model for a feature; determine whether a first condition associated with performance monitoring on the prediction model is met; and based on determining that the first condition is met, transmit, to the terminal device, an indication of an action associated with the prediction model.

[0010] In a fifth aspect, there is provided a terminal device. The terminal device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the terminal device at least to: transmit, to a network device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and take one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is transmitted to the network device; measurements of the feature, wherein the second indication is transmitted to the network device; prediction results of the prediction model, wherein the third indication is transmitted to the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is transmitted to the network device and a previous input of the prediction model is prediction results of the prediction model.

[0011] In a sixth aspect, there is provided a network device. The network device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the network device at least to: receive from a terminal device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; andtake one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is received from the terminal device; measurements of the feature, wherein the second indication is received from the terminal device; prediction results of the prediction model, wherein the third indication is received from the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is received from the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0012] In a seventh aspect, there is provided a terminal device. The terminal device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the terminal device at least to: receive from a network device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and take one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

[0013] In an eighth aspect, there is provided a network device. The network device comprises at least one processor and at least one memory storing instructions. When executed by the at least one processor, the instructions cause the network device at least to: transmit to a terminal device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and take one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0014] In a ninth aspect, there is provided a method performed by a terminal device. The method comprises: receiving, from a network device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model; determining a negative result of performance evaluation on the prediction model; based ondetermining the negative result, determining whether the first condition is met; and based on determining that the first condition is met, performing an action associated with the prediction model.

[0015] In a tenth aspect, there is provided a method performed by a network device. The method comprises: transmitting, to a terminal device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model, the first condition is associated with the terminal device performing an action associated with the prediction model; and performing a prediction on the feature based on the prediction model.

[0016] In an eleventh aspect, there is provided a method performed by a terminal device. The method comprises: determining at least one negative result of performance evaluation on a prediction model for a feature; transmitting, to a network device, at least one negative indication for the prediction model; and receiving, from the network device, an indication of an action associated with the prediction model.

[0017] In a twelfth aspect, there is provided a method performed by a network device. The method comprises: receiving, from a terminal device, at least one negative indication for a prediction model for a feature; determining whether a first condition associated with performance monitoring on the prediction model is met; and based on determining that the first condition is met, transmitting, to the terminal device, an indication of an action associated with the prediction model.

[0018] In a thirteenth aspect, there is provided a method performed by a terminal device. The method comprises: transmitting, to a network device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is transmitted to the network device; measurements of the feature, wherein the second indication is transmitted to the network device; prediction results of the prediction model, wherein the third indication is transmitted to the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is transmitted to the network device and a previous input of the prediction model is prediction results of the prediction model.

[0019] In a fourteenth aspect, there is provided a method performed by a network device. The method comprises: receiving from a terminal device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is received from the terminal device; measurements of the feature, wherein the second indication is received from the terminal device; prediction results of the prediction model, wherein the third indication is received from the terminal device and a previous input of the prediction model is measurementsof the feature; or measurements of the feature, wherein the third indication is received from the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0020] In a fifteenth aspect, there is provided a method performed by a terminal device. The method comprises: receiving from a network device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

[0021] In a sixteenth aspect, there is provided a method performed by a network device. The method comprises: transmitting to a terminal device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0022] In a seventeenth aspect, there is provided an apparatus. The apparatus comprises means for receiving, from a network device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model; means for determining a negative result of performance evaluation on the prediction model; means for based on determining the negative result, determining whether the first condition is met; and means for based on determining that the first condition is met, performing an action associated with the prediction model.

[0023] In an eighteenth aspect, there is provided an apparatus. The apparatus comprises means for transmitting, to a terminal device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model, the first condition is associated with the terminal device performing an action associated with the predictionmodel; and means for performing a prediction on the feature based on the prediction model.

[0024] In a nineteenth aspect, there is provided an apparatus. The apparatus comprises means for determining at least one negative result of performance evaluation on a prediction model for a feature; means for transmitting, to a network device, at least one negative indication for the prediction model; and means for receiving, from the network device, an indication of an action associated with the prediction model.

[0025] In a twenties aspect, there is provided an apparatus. The apparatus comprises means for receiving, from a terminal device, at least one negative indication for a prediction model for a feature; means for determining whether a first condition associated with performance monitoring on the prediction model is met; and means for based on determining that the first condition is met, transmitting, to the terminal device, an indication of an action associated with the prediction model.

[0026] In a twenty-first aspect, there is provided an apparatus. The apparatus comprises means for transmitting, to a network device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is transmitted to the network device; measurements of the feature, wherein the second indication is transmitted to the network device; prediction results of the prediction model, wherein the third indication is transmitted to the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is transmitted to the network device and a previous input of the prediction model is prediction results of the prediction model.

[0027] In a twenty-second aspect, there is provided an apparatus. The apparatus comprises means for receiving from a terminal device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is received from the terminal device; measurements of the feature, wherein the second indication is received from the terminal device; prediction results of the prediction model, wherein the third indication is received from the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is received from the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0028] In a twenty-third aspect, there is provided an apparatus. The apparatus comprises means for receiving from a network device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a nextprediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

[0029] In a twenty-fourth aspect, there is provided an apparatus. The apparatus comprises means for transmitting to a terminal device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0030] In a twenty-fifth aspect, there is provided a non -transitory computer readable medium comprising program instructions for causing an apparatus to perform at least the method according to any one of the above ninth to sixteenth aspects.

[0031] In a twenty-sixth aspect, there is provided a computer program product comprising program instructions for performing at least the method according to any one of the above ninth to sixteenth aspects.

[0032] In a twenty-seventh aspect, there is provided a computer program comprising instructions, which, when executed by an apparatus, cause the apparatus at least to perform at least the method according to any one of the above ninth to sixteenth aspects.

[0033] It is to be understood that the summary section is not intended to identify key or essential features of embodiments of the present disclosure, nor is it intended to be used to limit the scope of the present disclosure. Other features of the present disclosure will become easily comprehensible through the following description.BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Some example embodiments will now be described with reference to the accompanying drawings, in which:

[0035] Fig. 1 illustrates an example communication system in which embodiments of the present disclosure may be implemented;

[0036] Fig. 2A illustrates an example signaling chart of a first example process according to some embodiments of the present disclosure;

[0037] Fig. 2B illustrates an example signaling chart of a second example process according to some embodiments of the present disclosure;

[0038] Fig. 2C illustrates an example signaling chart of a third example process according to some embodiments of the present disclosure;

[0039] Fig. 2D illustrates an example signaling chart of a fourth example process according to some embodiments of the present disclosure;

[0040] Fig. 3A illustrates a schematic diagram illustrating an example UE-driven performance monitoring process according to some embodiments of the present disclosure;

[0041] Fig. 3B illustrates a signaling chart of an example UE-driven performance monitoring process according to some embodiments of the present disclosure;

[0042] Fig. 4A illustrates a schematic diagram illustrating an example BS-driven performance monitoring process according to some embodiments of the present disclosure;

[0043] Fig. 4B illustrates a signaling chart of an example BS-driven performance monitoring process according to some embodiments of the present disclosure;

[0044] Fig. 5 illustrates an example diagram of prediction and performance evaluation of a twin model according to some embodiments of the present disclosure;

[0045] Fig. 6 illustrates a schematic diagram illustrating a method implemented at a terminal device according to some embodiments of the present disclosure;

[0046] Fig. 7 illustrates a schematic diagram illustrating a method implemented at a network device according to some embodiments of the present disclosure;

[0047] Fig. 8 illustrates a schematic diagram illustrating a method implemented at a terminal device according to some embodiments of the present disclosure;

[0048] Fig. 9 illustrates a schematic diagram illustrating a method implemented at a network device according to some embodiments of the present disclosure;

[0049] Fig. 10 illustrates a schematic diagram illustrating a method implemented at a terminal device according to some embodiments of the present disclosure;

[0050] Fig. 11 illustrates a schematic diagram illustrating a method implemented at a network deviceaccording to some embodiments of the present disclosure;

[0051] Fig. 12 illustrates a schematic diagram illustrating a method implemented at a terminal device according to some embodiments of the present disclosure;

[0052] Fig. 13 illustrates a schematic diagram illustrating a method implemented at a network device according to some embodiments of the present disclosure;

[0053] Fig. 14 illustrates a simplified block diagram of an apparatus that is suitable for implementing embodiments of the present disclosure; and

[0054] Fig. 15 illustrates a block diagram of an example computer readable medium in accordance with some embodiments of the present disclosure.

[0055] Throughout the drawings, the same or similar reference numerals represent the same or similar element.DETAILED DESCRIPTION

[0056] Principles of the present disclosure will now be described with reference to some example embodiments. It is to be understood that these embodiments are described only for the purpose of illustration and help those skilled in the art to understand and implement the present disclosure, without suggesting any limitation as to the scope of the disclosure. The disclosure described herein can be implemented in various manners other than the ones described below.

[0057] In the following description and claims, unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skills in the art to which this disclosure belongs.

[0058] References in the present disclosure to "one embodiment,” "an embodiment,” "an example embodiment,” and the like indicate that the embodiment described may include a particular feature, structure, or characteristic, but it is not necessary that every embodiment includes the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

[0059] It shall be understood that although the terms "first” and "second” etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and similarly, a second element could be termed a first element, without departing from the scope of example embodiments. As used herein, the term "and / or” includes any and all combinations of one or more of thelisted terms.

[0060] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms "a”, "an” and "the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises”, "comprising”, "has”, "having”, "includes” and / or "including”, when used herein, specify the presence of stated features, elements, and / or components etc., but do not preclude the presence or addition of one or more other features, elements, components and / or combinations thereof. As used herein, "at least one of the following: ” and "at least one of ” and similar wording, where the list of two or more elements are joined by "and” or "or”, mean at least any one of the elements, or at least any two or more of the elements, or at least all the elements.

[0061] As used in this application, the term "circuitry” may refer to one or more or all of the following:(a) hardware-only circuit implementations (such as implementations in only analog and / or digital circuitry) and(b) combinations of hardware circuits and software, such as (as applicable):(I) a combination of analog and / or digital hardware ci rcuit(s) with software / firmware and(ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions) and(c) hardware circuit(s) and or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.

[0062] This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.

[0063] As used herein, the term "communication network” refers to a network following any suitable communication standards, such as Long Term Evolution (LTE), LTE-Advanced (LTE-A), Wideband Code Division Multiple Access (WCDMA), High-Speed Packet Access (HSPA), Narrow Band Internet of Things (NB-loT) and so on. Furthermore, the communications between a terminal device and a network device in the communication network may be performed according to any suitable generation communication protocols,including, but not limited to, the first generation (1 G), the second generation (2G), 2.5G, 2.75G, the third generation (3G), the fourth generation (4G), 4.5G, the fifth generation (5G), the future sixth generation (6G) communication protocols, and / or any other protocols either currently known or to be developed in the future. Embodiments of the present disclosure may be applied in various communication systems. Given the rapid development in communications, there will of course also be future type communication technologies and systems with which the present disclosure may be embodied. It should not be seen as limiting the scope of the present disclosure to only the aforementioned system.

[0064] As used herein, the term "network device” refers to a node in a communication network via which a terminal device accesses the network and receives services therefrom. The network device may refer to a base station (BS) or an access point (AP), for example, a node B (NodeB or N B), an evolved NodeB (eNodeB or eNB), a NR NB (also referred to as a gNB), a Remote Radio Unit (RRU), a radio header (RH), a remote radio head (RRH), a relay, a low power node such as a femto, a pico, and so forth, depending on the applied terminology and technology.

[0065] The term "terminal device” refers to any end device that may be capable of wireless communication. By way of example rather than limitation, a terminal device may also be referred to as a communication device, user equipment (UE), a Subscriber Station (SS), a Portable Subscriber Station, a Mobile Station (MS), or an Access Terminal (AT). The terminal device may include, but not limited to, a mobile phone, a cellular phone, a smart phone, voice over IP (VoIP) phones, wireless local loop phones, a tablet, a wearable terminal device, a personal digital assistant (PDA), portable computers, desktop computer, image capture terminal devices such as digital cameras, gaming terminal devices, music storage and playback appliances, vehiclemounted wireless terminal devices, wireless endpoints, mobile stations, laptop-embedded equipment (LEE), laptop-mounted equipment (LME), USB dongles, smart devices, wireless customer-premises equipment (CPE), an Internet of Things (loT) device, a watch or other wearable, a head-mounted display (HMD), a vehicle, a drone, a medical device and applications (e.g., remote surgery), an industrial device and applications (e.g., a robot and / or other wireless devices operating in an industrial and / or an automated processing chain contexts), a consumer electronics device, a device operating on commercial and / or industrial wireless networks, and the like. In the following description, the terms "terminal device”, "communication device”, "terminal”, "user equipment” and "UE” may be used interchangeably.

[0066] Principles and embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Reference is first made to Fig. 1 , which illustrates an example communication system 100 in which embodiments of the present disclosure may be implemented. As shown in Fig. 1 , the environment 100, which may be a part of a communication network, comprises a terminal device 110 and a network device. The network device 120 may be a gNB, or a NetWork (NW) or a TRP, which schedules a first bandwidth part (BWP) or a first cell 130. The terminal device 110 is capable of connecting and communicating in an UL or DL with the network device 120 as long as the terminal device 110 is locatedwithin the corresponding cells (e.g., the first cell 130) of the network device 120.

[0067] In communication systems, an UL refers to a link in a direction from a terminal device 110 to a network device 120, and a DL refers to a link in a direction from the network device 120 to the terminal device 110. The network device 120 may transmit scheduling information scheduling an uplink transmission to the terminal device 110, and the terminal device 110 may transmit a uplink transmission or a plurality of repetitions of the uplink transmission to the network device 120.

[0068] Communications in the communication system 100 may be implemented according to any proper communication protocol(s), comprising, but not limited to, cellular communication protocols of the first generation (1G), the second generation (2G), the third generation (3G), the fourth generation (4G), the fifth generation (5G) and the sixth generation (6G) and on the like, wireless local network communication protocols such as Institute for Electrical and Electronics Engineers (IEEE) 802.11 and the like, and / or any other protocols currently known or to be developed in the future. Moreover, the communication may utilize any proper wireless communication technology, comprising but not limited to: Code Division Multiple Access (CDMA), Frequency Division Multiple Access (FDMA), Time Division Multiple Access (TDMA), Frequency Division Duplex (FDD), Time Division Duplex (TDD), Multiple-Input Multiple-Output (MIMO), Orthogonal Frequency Division Multiple (OFDM), Discrete Fourier Transform spread OFDM (DFT-s-OFDM) and / or any other technologies currently known or to be developed in the future.

[0069] It is to be understood that the numbers of devices (i.e., the terminal device 110 and the network device 120) and their connection relationships and types shown in Fig. 1 are only for the purpose of illustration without suggesting any limitation. For example, the communication system 100 may include any suitable numbers of devices adapted for implementing embodiments of the present disclosure. For example, while Fig. 1 depicts the terminal device 110 as a mobile phone, the terminal device 110 may be any types of user equipment.

[0070] For channel status information (CSI) prediction using UE-sided model, at least three types of performance monitoring are proposed. For Type 1 performance monitoring, enhancements on the following items are needed: definition / configuration of performance metric; definition of threshold criterion, if configured; definition / configuration and report of monitoring output, and corresponding report mechanism. For Type 2 performance monitoring, enhancements on the following items are needed: definition / configuration and report of ground truth CSI, and corresponding report mechanism. For Type 3 performance monitoring, enhancements on the following items are needed: definition / configuration and report of performance metric, and corresponding report mechanism. For the boundary between Type 3 and Type I performance monitoring, the difference is whether UE reports performance metric or performance monitoring output to the network, respectively. The monitoring output is determined based on the performance metric, and additionally, baseline and / or threshold criterion if configured. For all types of performance monitoring, the network needsto indicate to the UE of the decision regarding the monitoring action. As used herein, the terms "CSI prediction” and "channel prediction” may be used interchangeably.

[0071] Atwin model scheme, i.e., having identical models at gNB side and UE side, may be beneficial and possible. There are different implementation options for the twin model scheme. In the first option, identical models are specified in the specifications and used by both UE and gNB vendors. In the second option, identical models are proprietary models and are shared between UE and network vendors. In the third option, if a company is both UE and network vendor, even no sharing will be required.

[0072] Besides, CSI prediction is crucial for various network tasks such as CSI feedback enhancement, channel aging, power control, call handover and positioning. With regarding to channel prediction, AI / ML- based CSI prediction introduces performance improvement with increased complexity over non-AI / ML based CSI prediction. AI / ML-based CSI prediction is under study for the sake of saving signaling and CSI feedback overheads under the paradigm of Al Native 5G. Considering the use case of CSI feedback enhancement, auto-encoders have been used. However, the performance is enhanced at a cost of high computational complexity.

[0073] As an alternative, a CSI prediction use case based on twin models has been proposed. In a nutshell, one CSI prediction model is deployed at the gNB and another CSI prediction model is at the UE. The feedback from the UE is evaluated with respect to prediction. As a toy example, if the prediction at the UE is perfect then feedback is unnecessary. In a nutshell, the acquired channel at the gNB can be represented by HNB= HNB— Q(HtE— HtE), wherein Hwsand HfEdenote results of identical predictions at the gNB and the UE, respectively, and H^Edenotes the corresponding measurements at the UE. Therefore, in such use cases, it is indispensable to maintain a prediction model at the gNB and the UE.

[0074] To monitor the performance of the Al model at the gNB, the UE either reports the metric (e.g., squared generalized cosine similarity (SGCS)), or reports predicted CSI and measured CSI. However, such approaches might cause an excessive overhead due to reporting of metrics and / or CSI.

[0075] In view of the above, embodiments of the present disclosure provide a scheme of performance monitoring on prediction model based on positive / negative performance indications, which can be implemented either at the gNB or at the UE. It should be understood that, although embodiments of the present disclosure are illustrated with regard to twin AI / ML models for CSI feedback enhancement, the CSI prediction is merely taken as an example use-case, embodiments of the present disclosure may also apply to other use cases. The solution will be described in detail with reference to Figs. 2A to 5 below.

[0076] Fig. 2A illustrates a signaling chart illustrating a first example process 200A according to some embodiments of the present disclosure. For the purpose of discussion, the process 200A will be described with reference to Fig. 1 . The process 200A may involve the terminal device 110 and the network device 120. It would be appreciated that although the process 200A has been described in the communicationenvironment 100 of Fig. 1 , this process may be likewise applied to other communication scenarios.

[0077] As shown in Fig. 2A, the network device 120 transmits (211) a configuration 212 associated with a prediction model for a feature to the terminal device and performs (214) a prediction on the feature based on the prediction model. The configuration 212 includes a first condition associated with performance monitoring on the prediction model. The terminal device 110 receives (213) the configuration 212 from the network device 120. The terminal device 110 determines (215) a negative result of performance evaluation on the prediction model. Based on determining the negative result, the terminal device 110 determines (216) whether the first condition is met. If the first condition is met, the terminal device 110 performs (217) an action associated with the prediction model. In this way, the performance monitoring may be implemented at the UE side based on the network configuration. The UE may perform the performance monitoring based on negative results of performance evaluation on the prediction model, thus reducing the resource overhead for performance monitoring.

[0078] In some embodiments, the first condition may include a first threshold. The terminal device 110 may determine that the first condition is met if a counting number of negative results of the performance evaluation reaches the first threshold. In other words, if the UE determines that occurrences of negative results reaches a predetermined or preconfigured number of times, the UE would perform actions to improve model performance. In this way, the UE does not need to report predicted outputs and measurements for each time instant, thus reducing the resource overhead.

[0079] In some embodiments, the configuration 212 may further include an indication of initializing the counting number of the negative results. The terminal device 110 may initialize the counting number of the negative results based on the indication of initializing the counting number. For example, the terminal device 110 may set the counting number N of the negative results to zero upon receiving the configuration.

[0080] In some embodiments, the terminal device 110 may increment the counting number of the negative results based on determining a negative result of the performance evaluation.

[0081] In some embodiments, the terminal device 110 may re-initialize the counting number of the negative results based on determining a positive result of the performance evaluation. In other words, the terminal device 110 may count the number of consecutive negative results for the performance monitoring.

[0082] In some embodiments, the terminal device 110 may re-initialize the counting number of the negative results if a number of consecutive positive results of the performance evaluation reaches a second threshold. In some implementations, the configuration 212 may further include the second threshold.

[0083] In some embodiments, when performing the action, the terminal device 110 may pause a prediction based on the prediction model. Alternatively or additionally, when performing the action, the terminal device 110 may obtain measurements of the feature and transmit the measurements to the network device 120. Then, the terminal device 110 and the network device 120 may utilize the measurements of the feature toimprove performance of the prediction model. For example, the terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction. Alternatively or additionally, when performing the action, the terminal device 110 may resume the prediction based on the prediction model. Alternatively or additionally, when performing the action, the terminal device 110 may transmit prediction results of the prediction model to the network device 120.

[0084] In some implementations, the configuration 212 may further include a first time duration. For example, the terminal device 110 may pause a prediction based on the prediction model for the first time duration. Alternatively or additionally, the terminal device 110 may transmit, to the network device 120, the obtained measurements of the feature for the first time duration. Alternatively or additionally, the terminal device 110 may resume the prediction based on the prediction model after expiration of the first time duration. In other words, if the first condition associated with performance monitoring on the prediction model is met, the terminal device 110 may pause the prediction for a preconfigured time duration and may transmit measurements of the feature to the network device 120. Alternatively or additionally, the terminal device 110 may transmit, to the network device 120, prediction results of the prediction model and the obtained measurements of the feature for a second time duration after expiration of the first time duration. For example, the configuration 212 may further include the second time duration or a sum of the first time duration and the second time duration. In another example, the terminal device 110 may receive an indication of the second time duration from the network device 120. In a further example, the terminal device 110 may receive an indication of a sum of the first time duration and the second time duration from the network device 120. In some implementations, the indication of time durations may be an indication of time length, or an indication of the number of periodic monitoring instants.

[0085] In some embodiments, when performing the action, the terminal device 110 may obtain measurements of the feature for at least one time instant, or obtain prediction results of the prediction model for the at least one time instant. The terminal device may transmit at least one of the measurements or the prediction results to the network device 120. For example, the terminal device 110 and the network device 120 may utilize the measurements and the prediction results to improve performance of the prediction model.

[0086] In some embodiments, when performing the action, the terminal device 110 may decrease a prediction horizon of the prediction model. In some implementations, the configuration 212 may further include an initial value for the prediction horizon. Alternatively or additionally, the configuration 212 may further include a decrement value of the prediction horizon. The decrement value may be a positive integer smaller than the prediction horizon.

[0087] In some embodiments, when performing the action, the terminal device 110 may switch from a first prediction type to a second prediction type. In some implementations, the first prediction type may be associated with a type-l prediction and the second prediction type may be associated with a type-ll prediction.In some alternative implementations, the first prediction type may be associated with a type-ll prediction and the second prediction type may be associated with a type-l prediction. In some embodiments, the type-l prediction is associated with a type-l codebook, and the type-ll prediction is associated with a type-ll codebook. In other words, if the terminal device 110 determines that the performance of predicting entries of type I codebook is not good, the terminal device 110 may determine to switch to predicting entries of type II codebook. If the terminal device 110 determines that the performance of predicting entries of type II codebook is not good, the terminal device 110 may determine to switch to predicting entries of type I codebook. In some examples, the terminal device 110 may perform prediction on the same prediction model before and after the switching of the prediction type. In some other examples, the terminal device 110 may perform prediction on the different prediction models before and after the switching of the prediction type. The network device 120 may perform the same prediction type switching so that the terminal device 110 and the network device 120 may improve the model performance in the same manner, ensuring the same prediction operations of the twin models at the network side and the UE side.

[0088] In some embodiments, when performing the action, the terminal device 110 may terminate a prediction based on the prediction model, and transmit measurements of the feature to the network device 120. In other words, the terminal device 110 may fallback to legacy reporting, e.g., type I CSI feedback or type II CSI feedback. The prediction model deployed at the terminal device 110 and the network device 120 may be deactivated.

[0089] In some embodiments, the terminal device 110 may determine the action among at least one preconfigured action. In other words, the terminal device 110 may automatically determine which action to perform so as to improve model performance. In some embodiments, the terminal device 110 may transmit an indication of the action to the network device 120. In this way, the terminal device 110 and the network device 120 may improve the model performance in the same manner, ensuring the same prediction operations of the twin models at the network side and the UE side.

[0090] In some embodiments, the terminal device 110 may transmit an indication that the first condition is met to the network device 120. After receiving the indication that the first condition is met, the network device 120 may transmit an indication of the action to the terminal device 110. The terminal device 110 may perform the action based on the received indication. In this way, the terminal device 110 and the network device 120 may improve the model performance in the same manner, ensuring the same prediction operations of the twin models at the network side and the UE side.

[0091] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device 110 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0092] In some implementations, based on determining a positive result of the performance evaluation,the terminal device 110 may transmit a positive indication for the prediction model to the network device 120. Then, the terminal device 110 and the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0093] Alternatively, based on determining a positive result of the performance evaluation, the terminal device 110 may omit transmitting a report of the performance evaluation to the network device 120. If the network device 120 does not receive a report of the performance evaluation from the terminal device 110 at a reporting instant, the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0094] In some embodiments, if the first condition is not met after determining the negative result, the terminal device 110 may take prediction results of the prediction model as an input of the prediction model for a next prediction. If the network device 120 does not receive a report of the performance evaluation from the terminal device 110 at a reporting instant or receives a negative indication, the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0095] In some embodiments, based on determining the negative result, the terminal device 110 may transmit a negative indication for the prediction model to the network device 120. The network device 120 may thus be aware that the performance result is negative. In some implementations, the network device 120 may take at least one best prediction result among a plurality of previous prediction results of the prediction model as the input of the prediction model for the next prediction based on determining not receiving measurements of the feature. The network device 120 may transmit, to the terminal device 110, an index of at least one previous prediction corresponding to the at least one best prediction result. Based on receiving the index of the at least one previous prediction, the terminal device 110 may take at least one prediction result of the prediction model corresponding to the at least one previous prediction as the input of the prediction model for the next prediction.

[0096] In some embodiments, based on determining the negative result, the terminal device 110 may transmit measurements of the feature and optionally a negative indication for the prediction model to the network device 120. Then, the terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0097] In some embodiments, once the terminal device 110 transmits measurements of the feature to the network device 120, the terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0098] In some embodiments, the prediction model may be a CSI based twin model. The terminal device 110 may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0099] Fig. 2B illustrates a signaling chart illustrating a second example process 200B according to someembodiments of the present disclosure. For the purpose of discussion, the process 200B will be described with reference to Fig. 1 . The process 200B may involve the terminal device 110 and the network device 120. It would be appreciated that although the process 200B has been described in the communication environment 100 of Fig. 1 , this process may be likewise applied to other communication scenarios.

[0100] As shown in Fig. 2B, the terminal device 110 determines (221) at least one negative result of performance evaluation on a prediction model for a feature. The terminal device 110 transmits (222) at least one negative indication 223 for the prediction model to the network device 120. The network device 120 receives (224) the at least one negative indication 223 for the prediction model from the terminal device 110, and determines (225) whether a first condition associated with performance monitoring on the prediction model is met. If the first condition is met, the network device 120 transmit (226) an indication 227 of an action associated with the prediction model to the terminal device 110. The terminal device 110 receives the indication 227 of the action from the network device 120. In this way, the performance monitoring may be implemented at the gNB side. The gNB may perform the performance monitoring based on negative indications of performance evaluation on the prediction model, thus reducing the resource overhead for performance monitoring.

[0101] In some embodiments, the first condition may include a first threshold. The network device 120 may determine that the first condition is met if a counting number of negative indications of the performance evaluation reaches the first threshold. In other words, if the gNB determines that occurrences of negative indications reaches a predetermined or preconfigured number of times, the gNB would request the UE to perform actions to improve model performance. In this way, the UE does not need to report predicted outputs and measurements for each time instant, thus reducing the resource overhead.

[0102] In some embodiments, the network device 120 may initialize the counting number. For example, the network device 120 may set the counting number N of the negative indications to zero at an initialization phase.

[0103] In some embodiments, the network device 120 may receive a negative indication for the prediction model from the terminal device 110, and increment the counting number of the negative indications.

[0104] In some implementations, based on determining a positive result of the performance evaluation, the terminal device 110 may transmit a positive indication for the prediction model to the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction. The network device 120 may determine that a result of the performance evaluation is positive based on receiving a positive indication for the prediction model from the terminal device 110. Then, the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0105] Alternatively, based on determining a positive result of the performance evaluation, the terminaldevice 110 may omit transmitting a report of the performance evaluation to the network device 120 and may take prediction results of the prediction model as an input of the prediction model for a next prediction. If the network device 120 does not receive a report of the performance evaluation from the terminal device 110 at a reporting instant, the network device 120 may determine that a result of the performance evaluation is positive. Then, the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0106] In some embodiments, the network device 120 may re-initialize the counting number if the network device 120 determines that a result of the performance evaluation is positive. In other words, the terminal device 110 may count the number of consecutive negative indications for the performance monitoring.

[0107] In some embodiments, the network device 120 may re-initialize the counting number if the network device 120 determines that consecutive results of the performance evaluation are positive and a number of the consecutive results reaches a second threshold.

[0108] In some embodiments, the action may include pausing a prediction based on the prediction model; obtaining measurements of the feature. Alternatively or additionally, the action may include transmitting the measurements to the network device 120. Alternatively or additionally, the action may include resuming the prediction based on the prediction model. Alternatively or additionally, the action may include transmitting prediction results of the prediction model to the network device 120. The terminal device 110 may perform the action based on the indication of the action.

[0109] In some implementations, the network device 120 may transmit a configuration associated with the prediction model to the terminal device 110. The configuration may include a first time duration for the terminal device 110 to perform at least one of reporting measurements of the feature or pausing a prediction based on the prediction model. The terminal device 110 may pause a prediction based on the prediction model for the first time duration, and transmit the obtained measurements of the feature for the first time duration to the network device 120. Then the terminal device 110 may resume the prediction based on the prediction model after expiration of the first time duration. The terminal device 110 and the network device 120 may utilize the measurements of the feature to improve performance of the prediction model. For example, the terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0110] Alternatively or additionally, the configuration may include a second time duration for the terminal device 110 to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration. Alternatively or additionally, the configuration may include a sum of the first time duration and the second time duration. The terminal device 110 may transmit, to the network device 120, prediction results of the prediction model and the obtained measurements of the feature for the second time duration after expiration of the first time duration.

[0111] Alternatively or additionally, the network device 120 may further transmit, to the terminal device 110, an indication of a second time duration for the terminal device 110 to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration. Alternatively, the network device 120 may further transmit a sum of the first time duration and the second time duration to the terminal device 110. For example, the network device 120 may determine the second time duration after receiving the measurements of the first time duration. The terminal device 110 may transmit, to the network device 120, prediction results of the prediction model and the obtained measurements of the feature for the second time duration after expiration of the first time duration.

[0112] In some embodiments, the action may include at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device 120, at least one of the measurements or the prediction results. The terminal device 110 may perform the action accordingly. For example, the terminal device 110 and the network device 120 may utilize the measurements and the prediction results to improve performance of the prediction model.

[0113] In some embodiments, the action may include: decreasing a prediction horizon of the prediction model. The terminal device 110 may decrease a prediction horizon of the prediction model based on the indication of the action. In some embodiments, the network device 120 may transmit a configuration associated with the prediction model to the terminal device 110. The configuration may include an initial value for the prediction horizon. Alternatively or additionally, the configuration may further include a decrement value of the prediction horizon. The decrement value may be a positive integer smaller than the prediction horizon.

[0114] In some embodiments, the action may include: switching from a first prediction type to a second prediction type. The terminal device 110 may switch from a first prediction type to a second prediction type based on the indication of the action. In some implementations, the first prediction type may be associated with a type-l prediction and the second prediction type may be associated with a type-ll prediction. In some alternative implementations, the first prediction type may be associated with a type-ll prediction and the second prediction type may be associated with a type-l prediction. In some embodiments, the type-l prediction is associated with a type-l codebook, and the type-ll prediction is associated with a type-ll codebook. For example, if the terminal device 110 is performing a type I prediction and receives an indication of prediction type switching, the terminal device 110 may switch to predicting entries of type II codebook. In another example, if the terminal device 110 is performing a type II prediction and receives an indication of prediction type switching, the terminal device 110 may switch to predicting entries of type I codebook. In some examples, the terminal device 110 may perform prediction on the same prediction model before and after the switching of the prediction type. In some other examples, the terminal device 110 may perform prediction on the different prediction models before and after the switching of the prediction type. The networkdevice 120 may perform the same prediction type switching so that the terminal device 110 and the network device 120 may improve the model performance in the same manner, ensuring the same prediction operations of the twin models at the network side and the UE side.

[0115] In some embodiments, the action may include: terminating a prediction based on the prediction model; and transmitting measurements of the feature to the network device 120. In other words, the terminal device 110 may fallback to legacy reporting, e.g., type I CSI feedback or type II CSI feedback. The prediction model deployed at the terminal device 110 and the network device 120 may be deactivated.

[0116] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device 110 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0117] In some implementations, based on determining a positive result of the performance evaluation, the terminal device 110 may transmit a positive indication for the prediction model to the network device 120. Then, the terminal device 110 and the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0118] Alternatively, based on determining a positive result of the performance evaluation, the terminal device 110 may omit transmitting a report of the performance evaluation to the network device 120. If the network device 120 does not receive a report of the performance evaluation from the terminal device 110 at a reporting instant, the network device 120 may determine that a result of the performance evaluation is positive, and may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0119] In some embodiments, based on determining a negative result of the performance evaluation, the terminal device 110 may transmit a negative indication for the prediction model to the network device 120. If the first condition is not met, the network device 120 may take prediction results of the prediction model as an input of the prediction model for a next prediction. The terminal device 110 may also take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0120] Alternatively, if the network device 120 receives a negative indication for the prediction model from the terminal device 110 and determines that the first condition is not met, the network device 120 may take at least one best prediction result among a plurality of previous prediction results of the prediction model as the input of the prediction model for the next prediction. The network device 120 may transmit an index of at least one previous prediction corresponding to the at least one best prediction result to the terminal device 110. Based on receiving the index of the at least one previous prediction, the terminal device 110 may take at least one prediction result of the prediction model corresponding to the at least one previous prediction as the input of the prediction model for the next prediction.

[0121] In some embodiments, based on determining a negative result of the performance evaluation, theterminal device 110 may transmit, to the network device 120, a negative indication for the prediction model and measurements of the feature. Then, the terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0122] In some embodiments, once the terminal device 110 transmits measurements of the feature to the network device 120, the terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0123] In some embodiments, the prediction model may be a CSI based twin model. The terminal device 110 may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0124] Fig. 2C illustrates a signaling chart illustrating a second example process 200C according to some embodiments of the present disclosure. For the purpose of discussion, the process 200C will be described with reference to Fig. 1 . The process 200C may involve the terminal device 110 and the network device 120. It would be appreciated that although the process 200C has been described in the communication environment 100 of Fig. 1 , this process may be likewise applied to other communication scenarios. The process 200C may be performed in combination with the process 200A in Fig. 2A or with the process 200B in Fig. 2B or may be performed independently with the processes 200A and 200B.

[0125] As shown in Fig. 2C, the terminal device 110 transmits (231), to the network device 120, a first indication 232, a second indication 233 or a third indication 234. The first indication 232 is associated with a first input type for a prediction model for a feature. The second indication 233 is associated with a second input type for a prediction model for a feature. The third indication 234 is associated with a switch between the first input type and the second input type for a prediction model for a feature. The network device 120 receives (235) the first indication 232, the second indication 233 or the third indication 234 from the terminal device 110. In some implementations, the first indication 232 is transmitted to the network device 120, the terminal device 110 takes (237) prediction results of the prediction model as an input of the prediction model for a next prediction, and the network device 120 takes (236) prediction results of the prediction model as an input of the prediction model for a next prediction. Alternatively, the second indication 233 is transmitted to the network device 120, the terminal device 110 takes (237) measurements of the feature as an input of the prediction model for a next prediction, and the network device 120 takes (236) measurements of the feature as an input of the prediction model for a next prediction. Alternatively, the third indication 234 is transmitted to the network device 120 and a previous input of the prediction model is measurements of the feature, the terminal device 110 takes (237) prediction results of the prediction model as an input of the prediction model for a next prediction, and the network device 120 takes (236) prediction results of the prediction model as an input of the prediction model for a next prediction. Alternatively, the third indication 234 is transmitted to the network device 120 and a previous input of the prediction model is prediction results of the prediction model,the terminal device 110 takes (237) measurements of the feature as an input of the prediction model for a next prediction, and the network device 120 takes (236) measurements of the feature as an input of the prediction model for a next prediction. In this way, the terminal device 110 and the network device 120 may perform the same prediction operations of the twin models. The signaling for indicating the input type is designed.

[0126] In some embodiments, the terminal device 110 may determine whether a result of performance evaluation on the prediction model is positive or negative and determine whether to transmit the first indication 232 or the second indication 233 based on the evaluation result.

[0127] In some embodiments, the first indication 232 may include an indication of the first input type. Alternatively or additionally, the first indication 232 may include a positive indication for the prediction model or a negative indication for the prediction model.

[0128] In some implementations, the terminal device 110 determines that a result of performance evaluation on the prediction model is a positive result, and the first indication 232 may include at least one of the following: a positive indication for the prediction model; or an indication of the first input type.

[0129] In some implementations, the result is a negative result, and the first indication 232 may include at least one of the following : the indication of the first input type; or a negative indication for the prediction model. In some implementations, if the result is a negative result, the terminal device 110 may determine whether a first condition associated with performance monitoring on the prediction model is met. If the terminal device 110 determines that the first condition is not met, the terminal device 110 may transmit the first indication 232 to the network device 120.

[0130] In some embodiments, the first indication 232 may include a negative indication for the prediction model. After receiving the first indication 232 including the negative indication, the network device 120 may take at least one best prediction result among a plurality of previous prediction results of the prediction model as the input of the prediction model for the next prediction based on determining not receiving measurements of the feature. The network device 120 may transmit, to the terminal device 110, an index of at least one previous prediction corresponding to the at least one best prediction result. Based on receiving the index of the at least one previous prediction, the terminal device 110 may take at least one prediction result of the prediction model corresponding to the at least one previous prediction as the input of the prediction model for the next prediction.

[0131] In some embodiments, the terminal device 110 determines that a result of performance evaluation on the prediction model is a negative result, and the second indication 233 may include at least one of the following: an indication of the second input type; a negative indication for the prediction model; or the measurements of the feature.

[0132] In some implementations, if the terminal device 110 determines that the result is a negative result,the terminal device 110 may determine whether a first condition associated with performance monitoring on the prediction model is met. If the terminal device 110 determines that the first condition is met, the terminal device 110 may transmit the second indication 233 to the network device 120.

[0133] In some embodiments, when determining whether the first condition is met, the terminal device 110 may determine a counting number of negative results of the performance evaluation. If the terminal device 110 determines that the counting number reaches a first threshold, the terminal device 110 may determine that the first condition is met. If the terminal device 110 determines that the counting number is below a first threshold, the terminal device 110 may determine that the first condition is not met. The terminal device 110 may increment the counting number based on determining a negative result of the performance evaluation. In some implementations, the terminal device 110 may re-initialize the counting number based on determining a positive result of the performance evaluation. Alternatively, the terminal device 110 may re-initialize the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold.

[0134] In some embodiments, the prediction model may be a CSI based twin model. The terminal device 110 may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0135] Fig. 2D illustrates a signaling chart illustrating a second example process 200D according to some embodiments of the present disclosure. For the purpose of discussion, the process 200D will be described with reference to Fig. 1 . The process 200D may involve the terminal device 110 and the network device 120. It would be appreciated that although the process 200D has been described in the communication environment 100 of Fig. 1 , this process may be likewise applied to other communication scenarios. The process 200D may be performed in combination with the process 200B in Fig. 2B or may be performed independently with the process 200B.

[0136] As shown in Fig. 2D, the network device 120 transmits (241), to the terminal device 110, an indication 242 of a first input type or an indication 243 of a second input type or an indication 244 associated with a switch between the first input type and the second input type for a prediction model for a feature. The terminal device 110 receives (245) the indication 242 of the first input type or the indication 243 of the second input type or the indication 244 associated with the switch from the network device 120. In some implementations, the indication 242 of the first input type is received from the network device 120, the terminal device 110 takes (247) prediction results of the prediction model as an input of the prediction model for a next prediction, and the network device 120 takes (246) prediction results of the prediction model as an input of the prediction model for a next prediction. Alternatively, the indication243 of the second input type is received from the network device 120, the terminal device 110 takes (247) measurements of the feature as an input of the prediction model for a next prediction, and the network device 120 takes (246) measurementsof the feature as an input of the prediction model for a next prediction. Alternatively, the indication 244 associated with a switch between the first input type and the second input type is received from the network device 120 and a previous input of the prediction model is measurements of the feature, the terminal device 110 takes (247) prediction results of the prediction model as an input of the prediction model for a next prediction, and the network device 120 takes (246) prediction results of the prediction model as an input of the prediction model for a next prediction. Alternatively, the indication 244 associated with a switch between the first input type and the second input type is received from the network device 120 and a previous input of the prediction model is prediction results of the prediction model, the terminal device 110 takes (247) measurements of the feature as an input of the prediction model for a next prediction, and the network device 120 takes (246) measurements of the feature as an input of the prediction model for a next prediction. In this way, the terminal device 110 and the network device 120 may perform the same prediction operations of the twin models. The signaling for indicating the input type is designed.

[0137] In some embodiments, the terminal device 110 may determine a negative result of performance evaluation on the prediction model, and transmit a negative indication for the predication model to a network device 120. The terminal device 110 may receive the indication 242 of the first input type or the indication 243 of the second input type after transmitting the negative indication. For example, the network device 120 may receive the negative indication for the predication model from the terminal device 110, and determine whether a first condition associated with performance monitoring on the prediction model is met. If the network device 120 determines that the first condition is not met, the network device 120 may transmit the indication of the first input type to the terminal device 110. If the network device 120 determines that the first condition is met, the network device 120 may transmit the indication of the second input type to the terminal device 110.

[0138] In some implementations, in order to determine whether the first condition is met, the network device 120 may determine a counting number of negative results of the performance evaluation. If the network device 120 determines that the counting number reaches a first threshold, the network device 120 may determine that the first condition is met. If the network device 120 determines that the counting number is below a first threshold, the network device 120 may determine that the first condition is not met. In some embodiments, the network device 120 may increment the counting number based on determining a negative result of the performance evaluation (e.g., once receiving a negative indication of the prediction model).

[0139] In some embodiments, the network device 120 may re-initialize the counting number of the negative results based on determining a positive result of the performance evaluation. In other words, the network device 120 may count the number of consecutive negative results for the performance monitoring. Alternatively, the network device 120 may re-initialize the counting number of the negative results if a number of consecutive positive results of the performance evaluation reaches a second threshold.

[0140] In some embodiments, based on determining the negative result, the terminal device 110 may transmit a negative indication for the prediction model to the network device 120. The network device 120 may thus be aware that the performance result is negative. In some implementations, the network device 120 may take at least one best prediction result among a plurality of previous prediction results of the prediction model as the input of the prediction model for the next prediction based on determining not receiving measurements of the feature. The network device 120 may transmit, to the terminal device 110, an index of at least one previous prediction corresponding to the at least one best prediction result. Based on receiving the index of the at least one previous prediction, the terminal device 110 may take at least one prediction result of the prediction model corresponding to the at least one previous prediction as the input of the prediction model for the next prediction.

[0141] In some embodiments, the network device 120 may transmit the indication 243 of the second input type to the terminal device 110. Based on receiving the indication 243 of the second input type, the terminal device 110 may transmit the measurements of the feature to the network device 120. The terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0142] In some embodiments, the network device 120 may transmit the indication 244 associated with a switch between the first input type and the second input type to the terminal device 110. A previous input of the prediction model is prediction results of the prediction model. Based on receiving the indication 243 of the second input type and based on determining that a previous input of the prediction model is prediction results of the prediction model, the terminal device 110 may transmit the measurements of the feature to the network device 120. The terminal device 110 and the network device 120 may take the measurements of the feature as an input of the prediction model for a next prediction.

[0143] In some embodiments, the prediction model may be a CSI based twin model. The terminal device 110 may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0144] Fig. 3A illustrates a schematic diagram illustrating an example UE-driven performance monitoring process 300 according to some embodiments of the present disclosure. Fig. 3B illustrates a signaling chart of an example UE-driven performance monitoring process 300 according to some embodiments of the present disclosure. The process 300 may involve the terminal device 110 and the network device 120. The process 300 may be regarded as specific examples of the process 200A of Fig. 2A. The same reference numerals are used to denote the elements or components described in Figs. 3A and 3B having the same operations, and detailed description thereof will be omitted.

[0145] In the process 300, at step 301 , the network device 120 may transmit configuration parameters for the performance monitoring to the terminal device 110. The terminal device 110 may utilize the receivedparameters to evaluate and monitor the performance of the prediction model and take decision(s) accordingly. In some embodiments, the configuration may include an indication to count the number N±of negative performance results (e.g., NACKs). The terminal device 110 may initialize the number N±to zero after receiving the configuration and begin counting the number of NACKs. Alternatively or additionally, the configuration may include a threshold NTfllof the number of NACKs to take actions.

[0146] At step 302, at a given time instant t, the terminal device 110 may determine an ACK or a NACK, which is calculated based on the measurements and the prediction model deployed at the terminal device 110. For example, the terminal device 110 may check if the prediction vs. estimation is < threshold (e.g., \HEE— HtE\pRo< T), then the terminal device 110 may generate an ACK; otherwise, the terminal device 110 may generate a NACK.

[0147] If a NACK is generated (step 303), at step 304, the terminal device 110 may increase the counter N±-» N±+ 1. If N±< NTfll(NO at 305), then the terminal device 110 may wait for a set of NACKs to be generated for the sake of reliability. If N±> NTfll(YES at 305), then it means that the prediction at the terminal device 110 is not reliable. Thus, the terminal device 110 can take (possible) actions among others. In some implementations, the terminal device 110 may transmit a request for the action to be performed to the network device 120, and the network device 120 may transmit an indication of the action to the terminal device. Alternatively, the terminal device 110 may automatically determine the action to be performed from multiple preconfigured actions. The terminal device 110 may transmit an indication of the action the network device 120.

[0148] In some implementations, the terminal device 110 may take action 306. For example, after N±reaches NThl, the terminal device 110 may transmit the measured CSI and predicted CSI of the at least one time instance to the network device 120.

[0149] In some implementations, the terminal device 110 may take action 307. For example, after N±reaches NTfllat time instantthe terminal device 110 may pause the prediction based on the prediction model for a time period T, and obtain and report measured CSI for the time period T to the network device 120. The terminal device 110 may resume the prediction based on the prediction model at time instant t2= t±+ T after expiration of the time period T. In some examples, an indication of the time period T may be included in the configuration transmitted to the terminal device 110 at step 301. Alternatively, the terminal device 110 may receive the indication of the time period T from the network device 120 when receiving the indication of performing the action 307. In some embodiments, the terminal device 110 may continue measuring CSI and report the measured CSI and predicted CSI for a time period T* after expiration of the time period T. The terminal device 110 may stop reporting the measured CSI and predicted CSI at time instant t3= + T + T*. In some examples, an indication of the time period T* or an indication ofT + T* may be included in the configuration transmitted to the terminal device 110 at step 301 . Alternatively,the terminal device 110 may receive the indication of the time period T* or the indication of T + T* from the network device 120 when receiving the indication of performing the action 307.

[0150] In some implementations, the terminal device 110 may take action 308. For example, after N±reaches NThl, the terminal device 110 may switch to other functionalities. In a specific example, the terminal device 110 may decrease the prediction horizon of the prediction model, i.e., D = D — i. In some examples, an indication of an initial value of the prediction horizon D for CSI prediction may be included in the configuration transmitted to the terminal device 110 at step 301. In some examples, the decrement step i may be a predefined fixed value, e.g., 1. Alternatively, the decrement step i may be included in the configuration transmitted to the terminal device 110 at step 301. Alternatively, the terminal device 110 may receive the indication of the decrement step i from the network device 120 when receiving the indication of performing the action 308. Alternatively, when performing action 308, the terminal device 110 may switch from type-l prediction to type-ll prediction or vice versa. The network device 120 may take the same action as the terminal device 110.

[0151] In some implementations, the terminal device 110 may take action 309. For example, after N±reaches NThl, the terminal device 110 may fall back to legacy CSI feedback, e.g., type I CSI feedback or type II CSI feedback (meaning that no prediction model is used for CSI prediction).

[0152] When the terminal device 110 takes the action 306 / 307 / 308 / 309, the network device 120 may take corresponding actions. For example, if the terminal device 110 decreases the prediction horizon of the prediction model, the network device 120 may also decrease the prediction horizon of the prediction model. If the terminal device 110 reports measured CSI to the network device 120, both the terminal device 110 and the network device 120 may take the measured CSI as input for subsequent CSI predictions on the twin models. An example implementation of the twin models will be illustrated with reference to Fig. 5.

[0153] In some embodiments, if an ACK is generated (step 310), the terminal device 110 may re-initialize the counter N±to zero and the process ends at step 313.

[0154] Alternatively, if an ACK is generated (step 310), at step 311 , the terminal device 110 may increase the counter N2-» N2+ 1. If N2< NTh2(NO at 312), then the terminal device 110 may wait for a set of ACKs to be generated for the sake of reliability. If N2> NTfl2(YES at 312), then it means that the prediction at the terminal device 110 is reliable, the process ends at step 313 and the counter N±may be re-initialized to zero. In some embodiments, the configuration transmitted at step 301 may include an indication to count the number N2of positive performance results (e.g., ACKs). The terminal device 110 may initialize the number N2to zero after receiving the configuration and count the number of ACKs. Alternatively or additionally, the configuration may include a threshold NTfl2of the number of ACKs to reinitialize the counter of NACKs and / or to end the process.

[0155] Fig. 4A illustrates a schematic diagram illustrating an example BS-driven performance monitoring process 400 according to some embodiments of the present disclosure. Fig. 4B illustrates a signaling chart of an example UE-driven performance monitoring process 400 according to some embodiments of the present disclosure. The process 400 may involve the terminal device 110 and the network device 120. The process 400 may be regarded as specific examples of the process 200B of Fig. 2B. The same reference numerals are used to denote the elements or components described in Figs. 4A and 4B having the same operations, and detailed description thereof will be omitted.

[0156] In the process 400, at step 401 , the network device 120 may determine configuration parameters for the performance monitoring. Examples of the configuration parameters may include at least one of the following: a threshold NTfllof the number of NACKs to take actions, a threshold NTfl2of the number of ACKs to re-initialize the counter of NACKs and / or to end the process, a time period T for pausing the prediction, a time period T* for the terminal device to report the measured CSI and predicted CSI, a sum of the time period T and the time period T*, an initial value of the prediction horizon D for CSI prediction, or a decrement step i of the prediction horizon D . The network device 120 may utilize the received parameters to monitor the performance of the prediction model and take decision(s) accordingly. In some embodiments, the network device 120 may initialize the number N±of negative performance results (e.g., NACKs) to zero and begin counting the number of NACKs received from the terminal device 110.

[0157] In some embodiments, the network device 120 may transmit at least one configuration parameter to the terminal device 110. For example, the configuration parameter(s) transmitted from the network device 120 to the terminal device 110 may include at least one of the following: the time period T, the time period T*, a sum of the time period T and the time period T*, the initial value of the prediction horizon D, or the decrement step i of the prediction horizon D.

[0158] At step 402, at a given time instant t, the terminal device 110 may feedback an ACK or a NACK, which is calculated based on the measurements and the prediction model deployed at the terminal device 110. For example, the terminal device 110 may check if the prediction vs. estimation is < threshold (e.g., \H^E< T), then the terminal device 110 may generate an ACK; otherwise, the terminal device110 may generate a NACK.

[0159] If a NACK is received (step 403), at step 404, the network device 120 may increase the counter — > JV-L + 1. If N < NThl(NO at 405), then the network device 120 may wait for a set of NACKs for the sake of reliability. If N±> NThl(YES at 405), then it means that the prediction at the terminal device 110 is not reliable. Thus, the network device 120 may take (possible) actions among others. In some implementations, the network device 120 may transmit an indication of the action to the terminal device 110.

[0160] In some implementations, the network device 120 may take action 406. For example, after N±reaches NThl, the network device 120 may request the terminal device 110 to report the measured CSI and predicted CSI of the at least one time instance.

[0161] In some implementations, the network device 120 may take action 407. For example, after N±reaches NThlat time instant tl tthe network device 120 may indicate the terminal device 110 to pause the prediction based on the prediction model for a time period T, and to obtain and report measured CSI for the time period T to the network device 120. The terminal device 110 may resume the prediction based on the prediction model at time instant t2=+ T after expiration of the time period T. In some examples, an indication of the time period T may be included in the configuration transmitted to the terminal device 110 at step 401. Alternatively, the terminal device 110 may receive the indication of the time period T from the network device 120 when receiving the indication of performing the action 407. In some embodiments, the terminal device 110 may continue measuring CSI and report the measured CSI and predicted CSI for a time period T* after expiration of the time period T . The terminal device 110 may stop reporting the measured CSI and predicted CSI at time instant t3=+ T + T* . In some examples, an indication of the time period T* or an indication of T + T* may be included in the configuration transmitted to the terminal device 110 at step 401 . Alternatively, the terminal device 110 may receive the indication of the time period T* or the indication of T + T* from the network device 120 when receiving the indication of performing the action 407.

[0162] In some implementations, the network device 120 may take action 408. For example, after N±reaches NThl, the network device 120 may switch to other functionalities. In a specific example, the network device 120 may decrease the prediction horizon of the prediction model, i.e., D = D — i. The terminal device 110 may also decrease the prediction horizon of the prediction model. In some examples, an indication of an initial value of the prediction horizon for CSI prediction may be included in the configuration transmitted to the terminal device 110 at step 401 . In some examples, the decrement step i may be a predefined fixed value, e.g., 1. Alternatively, the decrement step i may be included in the configuration transmitted to the terminal device 110 at step 401. Alternatively, the terminal device 110 may receive the indication of the decrement step i from the network device 120 when receiving the indication of performing the action 408. Alternatively, when performing action 408, the network device 120 may switch from type-l prediction to type- II prediction or vice versa. The terminal device 110 may take the same action as the network device 120.

[0163] In some implementations, the network device 120 may take action 409. For example, after N±reaches NThl, the network device 120 may request the terminal device 110 to fall back to legacy CSI feedback, e.g., type I CSI feedback or type II CSI feedback (meaning that no prediction model is used for CSI prediction).

[0164] In some embodiments, when the terminal device 110 reports measured CSI to the network device 120, both the terminal device 110 and the network device 120 may take the measured CSI as input forsubsequent CSI predictions on the twin models. An example implementation of the twin models will be illustrated with reference to Fig. 5.

[0165] In some embodiments, if an ACK is received (step 410), the network device 120 may re-initialize the counter N±to zero and the process ends at step 413.

[0166] Alternatively, if an ACK is received (step 410), at step 411 , the network device 120 may increase the counter N2-» N2+ 1. If N2< NTh2(NO at 412), then the network device 120 may wait for a set of ACKs to be generated for the sake of reliability. If N2> NTfl2(YES at 412), then it means that the prediction at the terminal device 110 is reliable, the process ends at step 413 and the counter N±may be re-initialized to zero.

[0167] Fig. 5 illustrates an example diagram of prediction and performance evaluation of a twin model according to some embodiments of the present disclosure. Fig. 5 shows an example implementation of AI / ML for channel prediction at the BS and the UE as well as how the output of channel prediction is used to calculate ACK / NACK at the UE.

[0168] As shown in Fig. 5, at step 1 , an AI / ML model takes input of d past CSI realizations and predicts D future realizations. Once a CSI is predicted and a corresponding measured CSI is obtained, the UE compares the predicted value, H, against the ground truth H. If the squared Frobenius norm is below a certain threshold T, then an ACK is reported; otherwise, an NACK is reported. At step 1 + n, where n = 1,2, .. , N, the AI / ML model uses its predicted output, H as a next input. In other words, the UE first reports a sequence of past CSI realizations (i.e., measured CSIs) to predict future CSIs. At step 1 + n, the prediction is open loop, i.e., each of the BS and the UE rely on its current prediction to use as input for next prediction and so on. This process repeats until a certain number of NACKs are reported along with latest measured CSI. The BS then uses the latest measured CSI (reported by UE) for the next prediction.

[0169] By performing the performance evaluation based on the prediction result and the ground truth, the UE may determine whether the model performance is positive or negative at a given time instant. The gNB (in the BS-driven implementation) or the UE (in the UE-driven implementation) may perform performance monitoring based on the results of the performance evaluation by the UE, so as to determine whether an action is needed to optimize prediction model.

[0170] With some embodiments of the present disclosure, the same CSI prediction AI / ML model is allocated at the gNB side and at the UE side and an accuracy evaluation is performed at the UE side to determine NACKs or ACKs. The AI / ML model may use measured CSIs as well as predicted CSIs within input feature vector. In this way, the gNB can use predict future CSIs that are based on earlier CSIs until CSI inaccuracy is detected.

[0171] In some embodiments, the gNB may configure the UE with a report configuration for AI / ML-basedCSI prediction inaccuracy e.g. via a RRC configuration. In some embodiments, the gNB may configure the UE with a certain amount of UE measured CSI reports before starting using its own past ML-based predictions to predict future CSIs. In other words, before CSI predictions can be used to predict future CSIs, some samples are needed to get a starting point.

[0172] In some embodiments, the UE may report CSI inaccuracy e.g. on PUCCH when a CSI prediction model is not providing sufficient accuracy. Alternatively, a MAC CE may be harnessed to the CSI inaccuracy reports. The CSI inaccuracy report may refer to a NACK signaling. It should be understood that the reported ACK / NACK may correspond to a previous prediction and measurement, and not the current prediction, since the prediction of time T must be compared with measurement of time T.

[0173] In some embodiments, both the indication of ACK and the indication of NACK are reported. For example, the ACK / NACK may be systematically reported. In some implementations, the NACK / ACK may be indicated by a one-bit flag.

[0174] In some embodiments, the indication of ACK is not necessary, and an ACK may be assumed until NACK is received. For example, aa ACK may be assumed until a NACK is received. If the network device does not receive a NACK at a performance evaluation time instant, it means there is an ACK. This may further reduce the resource overhead.

[0175] In some embodiments, the accuracy thresholds (e.g., T) for NACK triggering may be pre-configured or given in a RRC configuration.

[0176] If CSI inaccuracy occurs (i.e., NACK is received) for a configured number of times, the gNB and / or the UE may take required actions, e.g. the gNB may request certain amount of new measured CSIs to get back on track

[0177] Considering the AI / ML-assisted CSI feedback use case, where a gNB is serving a UE, both the gNB and the UE are equipped with an AI / ML-based channel predictor. The performance monitoring may be implemented based on the ACK / NACK, calculated using the predicted CSI and measurements (i.e., the ground truth). The resource overhead may thus be reduced, and the control / reliability of the AI / ML model may be enhanced since the measurement and performance evaluation may be done more frequently.

[0178] Fig. 6 illustrates a schematic diagram illustrating a method 600 implemented at a terminal device according to some embodiments of the present disclosure. For the purpose of discussion, the method 600 will be described from the perspective of the terminal device 110 as shown in Fig. 1 .

[0179] As shown in Fig. 6, at block 610, the terminal device 110 receives, from a network device, a configuration associated with a prediction model for a feature. The configuration comprises a first condition associated with performance monitoring on the prediction model. At block 620, the terminal device 110 determines a negative result of performance evaluation on the prediction model. At block 630, based ondetermining the negative result, the terminal device 110 determines whether the first condition is met. At block 640, based on determining that the first condition is met, the terminal device 110 performs an action associated with the prediction model.

[0180] In some embodiments, the first condition comprises a first threshold. In some embodiments, in order to determine that the first condition is met, the terminal device may determine that a counting number of negative results of the performance evaluation reaches the first threshold.

[0181] In some embodiments, the configuration further comprises an indication of initializing the counting number. The terminal device may initialize the counting number based on the indication of initializing the counting number.

[0182] In some embodiments, the terminal device may perform the action by at least one of: pausing a prediction based on the prediction model; obtaining measurements of the feature; transmitting, to the network device, the measurements; resuming the prediction based on the prediction model; or transmitting, to the network device, prediction results of the prediction model.

[0183] In some embodiments, the configuration further comprises a first time duration. The terminal device may perform at least one of: pausing a prediction based on the prediction model for the first time duration; transmitting, to the network device, the obtained measurements of the feature for the first time duration; or resuming the prediction based on the prediction model after expiration of the first time duration; or transmitting, to the network device, prediction results of the prediction model and the obtained measurements of the feature for a second time duration after expiration of the first time duration.

[0184] In some embodiments, the configuration further comprises the second time duration or a sum of the first time duration and the second time duration. In some embodiments, the terminal device may receive, from the network device, one of the following: an indication of the second time duration, or an indication of a sum of the first time duration and the second time duration.

[0185] In some embodiments, when performing the action, the terminal device may perform at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device, at least one of the measurements or the prediction results.

[0186] In some embodiments, when performing the action, the terminal device may decrease a prediction horizon of the prediction model. In some embodiments, the configuration further comprises at least one of the following: an initial value for the prediction horizon; or a decrement value of the prediction horizon.

[0187] In some embodiments, when performing the action, the terminal device may switch from a first prediction type to a second prediction type. In some embodiments, the first prediction type is associated with a type-1 prediction and the second prediction type is associated with a type-ll prediction. In someembodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-l prediction. In some embodiments, the type-l prediction is associated with a type-l codebook, and the type-ll prediction is associated with a type-ll codebook.

[0188] In some embodiments, when performing the action, the terminal device may terminate a prediction based on the prediction model; and transmit, to the network device, measurements of the feature.

[0189] In some embodiments, when performing the action, the terminal device may determine the action among at least one preconfigured action. In some embodiments, the terminal device may transmit, to the network device, an indication of the action.

[0190] In some embodiments, in order to perform the action, the terminal device may transmit, to the network device, an indication that the first condition is met; and receive, from the network device, an indication of the action.

[0191] In some embodiments, the terminal device may increment the counting number based on determining the negative result of the performance evaluation.

[0192] In some embodiments, the terminal device may re-initialize the counting number based on determining a positive result of the performance evaluation. In some embodiments, the terminal device may re-initialize the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold. In some embodiments, the configuration further comprises the second threshold.

[0193] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device may transmit, to the network device, a positive indication for the prediction model.

[0194] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device may omit transmitting a report of the performance evaluation to the network device.

[0195] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0196] In some embodiments, based on determining that the first condition is not met, the terminal device may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0197] In some embodiments, based on determining the negative result, the terminal device may transmit, to the network device, at least one of a negative indication for the prediction model or measurements of the feature.

[0198] In some embodiments, the terminal device may transmit, to the network device, the measurements of the feature; and take the measurements of the feature as an input of the prediction model for a next prediction.

[0199] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The terminal device may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0200] Fig. 7 illustrates a schematic diagram illustrating a method 700 implemented at a network device according to some embodiments of the present disclosure. For the purpose of discussion, the method 700 will be described from the perspective of the network device 120 as shown in Fig. 1.

[0201] As shown in Fig. 7, at block 710, the network device 120 transmits, to a terminal device, a configuration associated with a prediction model for a feature. The configuration comprises a first condition associated with performance monitoring on the prediction model, and the first condition is associated with the terminal device performing an action associated with the prediction model. At block 720, the network device 120 performs a prediction on the feature based on the prediction model.

[0202] In some embodiments, the first condition comprises a first threshold of a counting number of negative results of performance evaluation on the prediction model, the first threshold of the counting number is associated with the terminal device performing the action associated with the prediction model.

[0203] In some embodiments, the configuration further comprises at least one of the following: an indication of initializing the counting number; or a second threshold of a number of consecutive positive results of the performance evaluation for re-initializing the counting number.

[0204] In some embodiments, the network device may receive, from the terminal device, an indication that the first condition is met; and transmit, to the terminal device, an indication of the action. In some embodiments, the network device may receive, from the terminal device, an indication of the action.

[0205] In some embodiments, the configuration further comprises at least one of the following: a first time duration for the terminal device to perform at least one of reporting measurements of the feature or pausing a prediction based on the prediction model; a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; a sum of the first time duration and the second time duration; an initial value for a prediction horizon of the prediction model; or a decrement value of a prediction horizon of the prediction model.

[0206] In some embodiments, the configuration further comprises a first time duration for the terminal device to pause a prediction based on the prediction model. The network device may transmit, to the terminal device, one of the following: an indication of a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; or a sum of the first time duration and the second time duration.

[0207] In some embodiments, the action comprises at least one of: pausing a prediction based on the prediction model; obtaining measurements of the feature; transmitting, to the network device, themeasurements; resuming the prediction based on the prediction model; or transmitting, to the network device, measurements of the feature and prediction results of the prediction model.

[0208] In some embodiments, the action comprises at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device, at least one of the measurements or the prediction results.

[0209] In some embodiments, the action comprises: decreasing a prediction horizon of the prediction model; or switching from the first prediction type to a second prediction type.

[0210] In some embodiments, the first prediction type is associated with a type-l prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. In some embodiments, the type-1 prediction is associated with a type-1 codebook, and the type-ll prediction is associated with a type-ll codebook.

[0211] In some embodiments, the action comprises: terminating a prediction based on the prediction model; and transmitting, to the network device, measurements of the feature.

[0212] In some embodiments, the network device may receive, from the terminal device, a positive indication for the prediction model; and take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0213] In some embodiments, based on determining an absence of a report of the performance evaluation from the terminal device at a reporting instant, the network device may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0214] In some embodiments, the network device may receive, from the terminal device, a negative indication for the prediction model in a report of the performance evaluation; and based on determining absence of measurements of the feature in the report, take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0215] In some embodiments, the network device may receive, from the terminal device, a negative indication for the prediction model and measurements of the feature in a report of the performance evaluation.

[0216] In some embodiments, the network device may receive, from the terminal device, the measurements of the feature; and take the measurements of the feature as an input of the prediction model for a next prediction.

[0217] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0218] Fig. 8 illustrates a schematic diagram illustrating a method 800 implemented at a terminal device according to some embodiments of the present disclosure. For the purpose of discussion, the method 800 will be described from the perspective of the terminal device 110 as shown in Fig. 1 .

[0219] As shown in Fig. 8, at block 810, the terminal device 110 determines at least one negative result of performance evaluation on a prediction model for a feature. At block 820, the terminal device 110 transmits, to a network device, at least one negative indication for the prediction model. At block 830, the terminal device 110 receives, from the network device, an indication of an action associated with the prediction model.

[0220] In some embodiments, the terminal device may perform the action by at least one of: pausing a prediction based on the prediction model; obtaining measurements of the feature; transmitting, to the network device, the measurements; resuming the prediction based on the prediction model; or transmitting, to the network device, prediction results of the prediction model.

[0221] In some embodiments, the terminal device may receive, from the network device, a configuration associated with the prediction model. The configuration comprises a first time duration. The terminal device may perform the action by at least one of: pausing a prediction based on the prediction model for the first time duration; transmitting, to the network device, the obtained measurements of the feature for the first time duration; or resuming the prediction based on the prediction model after expiration of the first time duration; or transmitting, to the network device, prediction results of the prediction model and the obtained measurements of the feature for a second time duration after expiration of the first time duration.

[0222] In some embodiments, the configuration further comprises the second time duration or a sum of the first time duration and the second time duration. In some embodiments, the terminal device may receive, from the network device, one of the following: an indication of the second time duration, or an indication of a sum of the first time duration and the second time duration.

[0223] In some embodiments, the terminal device may perform the action by at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device, at least one of the measurements or the prediction results.

[0224] In some embodiments, the terminal device may perform the action by decreasing a prediction horizon of the prediction model. In some embodiments, the terminal device may receive, from the network device, a configuration associated with the prediction model. The configuration comprises at least one of the following: an initial value for the prediction horizon; or a decrement value of the prediction horizon.

[0225] In some embodiments, the terminal device may perform the action by switching from a first prediction type to a second prediction type. In some embodiments, the first prediction type is associated with a type-l prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. In some embodiments, the type-1 prediction is associated with a type-1 codebook, and the type-ll prediction is associated with a type-ll codebook.

[0226] In some embodiments, the terminal device may perform the action by terminating a predictionbased on the prediction model; and transmit, to the network device, measurements of the feature.

[0227] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device may transmit, to the network device, a positive indication for the prediction model.

[0228] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device may omit transmitting a report of the performance evaluation to the network device.

[0229] In some embodiments, based on determining a positive result of the performance evaluation, the terminal device may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0230] In some embodiments, based on determining a negative result of the performance evaluation, the terminal device may transmit, to the network device, a negative indication for the prediction model; and take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0231] In some embodiments, based on determining a negative result of the performance evaluation, the terminal device may transmit, to the network device, a negative indication for the prediction model and measurements of the feature.

[0232] In some embodiments, the terminal device may transmit, to the network device, the measurements of the feature; and take the measurements of the feature as an input of the prediction model for a next prediction.

[0233] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The terminal device may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0234] Fig. 9 illustrates a schematic diagram illustrating a method 900 implemented at a network device according to some embodiments of the present disclosure. For the purpose of discussion, the method 900 will be described from the perspective of the network device 120 as shown in Fig. 1.

[0235] As shown in Fig. 9, at block 910, the network device 120 receives, from a terminal device, at least one negative indication for a prediction model for a feature. At block 920, the network device 120 determines whether a first condition associated with performance monitoring on the prediction model is met. At block 930, based on determining that the first condition is met, the network device 120 transmits, to the terminal device, an indication of an action associated with the prediction model.

[0236] In some embodiments, the first condition comprises a first threshold.

[0237] In some embodiments, when determining that the first condition is met, the network device may determine that a counting number of the at least one negative indication reaches the first threshold.

[0238] In some embodiments, the network device may initialize the counting number.

[0239] In some embodiments, the action comprises at least one of: pausing a prediction based on the prediction model; obtaining measurements of the feature; transmitting, to the network device, the measurements; resuming the prediction based on the prediction model; or transmitting, to the network device, prediction results of the prediction model.

[0240] In some embodiments, the network device may transmit, to the terminal device, a configuration associated with the prediction model, wherein the configuration comprises at least one of the following: a first time duration for the terminal device to perform at least one of reporting measurements of the feature or pausing a prediction based on the prediction model; a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; or a sum of the first time duration and the second time duration.

[0241] In some embodiments, the network device may transmit, to the terminal device, a configuration associated with the prediction model. The configuration comprises a first time duration for the terminal device to pause a prediction based on the prediction model. In some embodiments, the network device may transmit, to the terminal device, one of the following: an indication of a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; or a sum of the first time duration and the second time duration.

[0242] In some embodiments, the action comprises at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device, at least one of the measurements or the prediction results.

[0243] In some embodiments, the action comprises: decreasing a prediction horizon of the prediction model; or switching from a first prediction type to a second prediction type.

[0244] In some embodiments, the first prediction type is associated with a type-l prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. In some embodiments, the type-1 prediction is associated with a type-1 codebook, and the type-ll prediction is associated with a type-ll codebook.

[0245] In some embodiments, the network device may transmit, to the terminal device, a configuration associated with the prediction model, wherein the configuration comprises at least one of the following: an initial value for the prediction horizon; or a decrement value of the prediction horizon.

[0246] In some embodiments, the action comprises: terminating a prediction based on the prediction model; and transmitting, to the network device, measurements of the feature.

[0247] In some embodiments, the network device may receive, from the terminal device, a negative indication for the prediction model; and increment the counting number.

[0248] In some embodiments, the network device may determine that a result of the performance evaluation is positive based on receiving, from the terminal device, a positive indication for the prediction model.

[0249] In some embodiments, the network device may determine that a result of the performance evaluation is positive based on determining that a report of the performance evaluation from the terminal device is absent at a reporting instant.

[0250] In some embodiments, the network device may re-initialize the counting number based on determining that a result of the performance evaluation is positive.

[0251] In some embodiments, the network device may re-initialize the counting number based on determining that consecutive results of the performance evaluation are positive and a number of the consecutive results reaches a second threshold.

[0252] In some embodiments, based on determining that a result of the performance evaluation is positive, the network device may take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0253] In some embodiments, the network device may receive, from the terminal device, a negative indication for the prediction model in a report of the performance evaluation; and based on determining that the first condition is not met, take prediction results of the prediction model as an input of the prediction model for a next prediction.

[0254] In some embodiments, the network device may receive, from the terminal device, a negative indication for the prediction model and measurements of the feature in a report of the performance evaluation.

[0255] In some embodiments, the network device may receive, from the terminal device, the measurements of the feature; and take the measurements of the feature as an input of the prediction model for a next prediction.

[0256] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0257] Fig. 10 illustrates a schematic diagram illustrating a method 1000 implemented at a terminal device according to some embodiments of the present disclosure. For the purpose of discussion, the method 1000 will be described from the perspective of the terminal device 110 as shown in Fig. 1 .

[0258] As shown in Fig. 10, at block 1010, the terminal device 110 transmits, to a network device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature. At block 1020, the terminal device 110 takes one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is transmitted to the network device; measurements of the feature, wherein the second indication istransmitted to the network device; prediction results of the prediction model, wherein the third indication is transmitted to the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is transmitted to the network device and a previous input of the prediction model is prediction results of the prediction model.

[0259] In some embodiments, the first indication comprises at least one of the following: a positive indication for the prediction model or a negative indication for the prediction model; or an indication of the first input type.

[0260] In some embodiments, the terminal device may determine whether a result of performance evaluation on the prediction model is positive or negative.

[0261] In some embodiments, the result is a positive result, and the first indication comprises at least one of the following: a positive indication for the prediction model; or an indication of the first input type.

[0262] In some embodiments, the result is a negative result, and the first indication comprises at least one of the following: the indication of the first input type; or a negative indication for the prediction model.

[0263] In some embodiments, the terminal device may determine whether a first condition associated with performance monitoring on the prediction model is met based on determining that the result is a negative result; and based on determining that the first condition is not met, transmit one of the following to the network device: the first indication; or the third indication, wherein a previous input of the prediction model is measurements of the feature.

[0264] In some embodiments, the terminal device may receive, from the network device, an index of at least one previous prediction; and take at least one prediction result of the prediction model corresponding to the at least one previous prediction as the input of the prediction model for the next prediction.

[0265] In some embodiments, the result is a negative result, and the second indication comprises at least one of the following: an indication of the second input type; a negative indication for the prediction model; or the measurements of the feature.

[0266] In some embodiments, the terminal device may determine whether a first condition associated with performance monitoring on the prediction model is met based on determining that the result is a negative result; and based on determining that the first condition is met, transmit one of the following to the network device: the second indication; or the third indication, wherein a previous input of the prediction model is prediction results of the prediction model.

[0267] In some embodiments, when determining whether the first condition is met, the terminal device may determine a counting number of negative results of the performance evaluation; and based on determining that the counting number reaches a first threshold, determine that the first condition is met; or based on determining that the counting number is below a first threshold, determine that the first condition isnot met.

[0268] In some embodiments, the terminal device may increment the counting number based on determining a negative result of the performance evaluation.

[0269] In some embodiments, the terminal device may re-initialize the counting number based on determining a positive result of the performance evaluation. In some embodiments, the terminal device may re-initialize the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold.

[0270] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The terminal device may determine whether the result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0271] Fig. 11 illustrates a schematic diagram illustrating a method 1100 implemented at a network device according to some embodiments of the present disclosure. For the purpose of discussion, the method 1100 will be described from the perspective of the network device 120 as shown in Fig. 1.

[0272] As shown in Fig. 11 , at block 1110, the network device 120 receives, from a terminal device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature. At block 1120, the network device 120 takes one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is received from the terminal device; measurements of the feature, wherein the second indication is received from the terminal device; prediction results of the prediction model, wherein the third indication is received from the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is received from the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0273] In some embodiments, the first indication comprises at least one of the following: a positive indication for the prediction model or a negative indication for the prediction model; or an indication of the first input type.

[0274] In some embodiments, the first indication comprises a negative indication for the prediction model. The network device may take at least one best prediction result among a plurality of previous prediction results of the prediction model as the input of the prediction model for the next prediction based on determining not receiving measurements of the feature.

[0275] In some embodiments, the network device may transmit, to the terminal device, an index of at least one previous prediction corresponding to the at least one best prediction result.

[0276] In some embodiments, the second indication comprises at least one of the following: an indicationof the second input type; a negative indication for the prediction model; or the measurements of the feature.

[0277] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0278] Fig. 12 illustrates a schematic diagram illustrating a method 1200 implemented at a terminal device according to some embodiments of the present disclosure. For the purpose of discussion, the method 1200 will be described from the perspective of the terminal device 110 as shown in Fig. 1 .

[0279] As shown in Fig. 12, at block 1210, the terminal device 110 receives, from a network device, an indication of a first input type, an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature. At block 1220, the terminal device 110 takes one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

[0280] In some embodiments, the terminal device 110 may determine a negative result of performance evaluation on the prediction model; and transmit, to a network device, a negative indication for the predication model. The indication of the first input type or the indication of the second input type is received after transmitting the negative indication.

[0281] In some embodiments, the indication of the first input type comprises an index of at least one previous prediction. The prediction results of the prediction model comprise at least one prediction result of the prediction model corresponding to the at least one previous prediction.

[0282] In some embodiments, the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model. The terminal device may transmit the measurements of the feature to the network device.

[0283] In some embodiments, the indication of the second input type is received from the network device. The terminal device may transmit the measurements of the feature to the network device.

[0284] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The terminal device may determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0285] Fig. 13 illustrates a schematic diagram illustrating a method 1300 implemented at a network device according to some embodiments of the present disclosure. For the purpose of discussion, the method 1300will be described from the perspective of the network device 120 as shown in Fig. 1.

[0286] As shown in Fig. 13, at block 1310, the network device 120 transmits, to a terminal device, an indication of a first input type, an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature. At block 1320, the network device 120 takes one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0287] In some embodiments, the network device may receive, from the terminal device, a negative indication for the predication model; determine whether a first condition associated with performance monitoring on the prediction model is met; and based on determining that the first condition is not met, transmit the indication of the first input type to the terminal device; or based on determining that the first condition is met, transmit the indication of the second input type to the terminal device.

[0288] In some embodiments, when determining whether the first condition is met, the network device may determine a counting number of negative results of the performance evaluation; and based on determining that the counting number reaches a first threshold, determine that the first condition is met; or based on determining that the counting number is below a first threshold, determine that the first condition is not met.

[0289] In some embodiments, the network device may increment the counting number based on determining a negative result of the performance evaluation.

[0290] In some embodiments, the network device may re-initialize the counting number based on determining a positive result of the performance evaluation. In some embodiments, the network device may re-initialize the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold.

[0291] In some embodiments, the network device may receive, from the terminal device, a negative indication for the predication model. The prediction results of the prediction model comprise at least one best prediction result among a plurality of previous prediction results of the prediction model. In some embodiments, the indication of the first input type comprises an index of at least one previous prediction corresponding to the at least one best prediction result.

[0292] In some embodiments, the indication of the second input type is transmitted to the terminal device. The network device may receive the measurements of the feature from the terminal device.

[0293] In some embodiments, the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model. The network device may receive the measurements of the feature from the terminal device.

[0294] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0295] In some embodiments, an apparatus capable of performing any of the method 600 (for example, the terminal device 110) may comprise means for performing the respective steps of the method 600. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0296] In some embodiments, the apparatus may include: means for receiving, from a network device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model; means for determining a negative result of performance evaluation on the prediction model; means for based on determining the negative result, determining whether the first condition is met; and means for based on determining that the first condition is met, performing an action associated with the prediction model.

[0297] In some embodiments, the first condition comprises a first threshold. In some embodiments, the means for determining that the first condition is met may include means for determining that a counting number of negative results of the performance evaluation reaches the first threshold.

[0298] In some embodiments, the configuration further comprises an indication of initializing the counting number. The apparatus may further include means for initializing the counting number based on the indication of initializing the counting number.

[0299] In some embodiments, the means for performing the action may include at least one of: means for pausing a prediction based on the prediction model; means for obtaining measurements of the feature; means for transmitting, to the network device, the measurements; means for resuming the prediction based on the prediction model; or means for transmitting, to the network device, prediction results of the prediction model.

[0300] In some embodiments, the configuration further comprises a first time duration. The apparatus may include at least one of: means for pausing a prediction based on the prediction model for the first time duration; means for transmitting, to the network device, the obtained measurements of the feature for the first time duration; or means for resuming the prediction based on the prediction model after expiration of the first time duration; or means for transmitting, to the network device, prediction results of the prediction model and the obtained measurements of the feature for a second time duration after expiration of the first time duration.

[0301] In some embodiments, the configuration further comprises the second time duration or a sum of the first time duration and the second time duration. In some embodiments, the apparatus may include meansfor receiving, from the network device, one of the following: an indication of the second time duration, or an indication of a sum of the first time duration and the second time duration.

[0302] In some embodiments, the means for performing the action may include at least one of means for obtaining measurements of the feature for at least one time instant; means for obtaining prediction results of the prediction model for the at least one time instant; and means for transmitting, to the network device, at least one of the measurements or the prediction results.

[0303] In some embodiments, the means for performing the action may include means for decreasing a prediction horizon of the prediction model. In some embodiments, the configuration further comprises at least one of the following: an initial value for the prediction horizon; or a decrement value of the prediction horizon.

[0304] In some embodiments, the means for performing the action may include means for switching from a first prediction type to a second prediction type. In some embodiments, the first prediction type is associated with a type-l prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. In some embodiments, the type-1 prediction is associated with a type-1 codebook, and the type-ll prediction is associated with a type-ll codebook.

[0305] In some embodiments, the means for performing the action may include means for terminating a prediction based on the prediction model; and transmit, to the network device, measurements of the feature.

[0306] In some embodiments, the means for performing the action may include means for determining the action among at least one preconfigured action. In some embodiments, the apparatus may further include means for transmitting, to the network device, an indication of the action.

[0307] In some embodiments, the means for performing the action may include means for transmitting, to the network device, an indication that the first condition is met; and means for receiving, from the network device, an indication of the action.

[0308] In some embodiments, the apparatus may further include means for incrementing the counting number based on determining the negative result of the performance evaluation.

[0309] In some embodiments, the apparatus may further include means for re-initializing the counting number based on determining a positive result of the performance evaluation. In some embodiments, the apparatus may further include means for re-initializing the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold. In some embodiments, the configuration further comprises the second threshold.

[0310] In some embodiments, the apparatus may further include means for based on determining a positive result of the performance evaluation, transmitting, to the network device, a positive indication for the prediction model.

[0311] In some embodiments, the apparatus may further include means for based on determining a positive result of the performance evaluation, omitting transmitting a report of the performance evaluation to the network device.

[0312] In some embodiments, the apparatus may further include means for based on determining a positive result of the performance evaluation, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0313] In some embodiments, the apparatus may further include means for based on determining that the first condition is not met, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0314] In some embodiments, the apparatus may further include means for based on determining the negative result, transmitting, to the network device, at least one of a negative indication for the prediction model or measurements of the feature.

[0315] In some embodiments, the apparatus may further include means for transmitting, to the network device, the measurements of the feature; and taking the measurements of the feature as an input of the prediction model for a next prediction.

[0316] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The apparatus may further include means for determining whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0317] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 600. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0318] In some embodiments, an apparatus capable of performing any of the method 700 (for example, the network device 120) may comprise means for performing the respective steps of the method 700. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0319] In some embodiments, the apparatus may include: means for transmitting, to a terminal device, a configuration associated with a prediction model for a feature, wherein the configuration comprises a first condition associated with performance monitoring on the prediction model, the first condition is associated with the terminal device performing an action associated with the prediction model; and means for performing a prediction on the feature based on the prediction model.

[0320] In some embodiments, the first condition comprises a first threshold of a counting number of negative results of performance evaluation on the prediction model, the first threshold of the counting numberis associated with the terminal device performing the action associated with the prediction model.

[0321] In some embodiments, the configuration further comprises at least one of the following: an indication of initializing the counting number; or a second threshold of a number of consecutive positive results of the performance evaluation for re-initializing the counting number.

[0322] In some embodiments, the apparatus may further include means for receiving, from the terminal device, an indication that the first condition is met; and means for transmitting, to the terminal device, an indication of the action. In some embodiments, the apparatus may further include means for receiving, from the terminal device, an indication of the action.

[0323] In some embodiments, the configuration further comprises at least one of the following: a first time duration for the terminal device to perform at least one of reporting measurements of the feature or pausing a prediction based on the prediction model; a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; a sum of the first time duration and the second time duration; an initial value for a prediction horizon of the prediction model; or a decrement value of a prediction horizon of the prediction model.

[0324] In some embodiments, the configuration further comprises a first time duration for the terminal device to pause a prediction based on the prediction model, the apparatus may further include means for transmitting, to the terminal device, one of the following: an indication of a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; or a sum of the first time duration and the second time duration.

[0325] In some embodiments, the action comprises at least one of: pausing a prediction based on the prediction model; obtaining measurements of the feature; transmitting, to the network device, the measurements; resuming the prediction based on the prediction model; or transmitting, to the network device, measurements of the feature and prediction results of the prediction model.

[0326] In some embodiments, the action comprises at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device, at least one of the measurements or the prediction results.

[0327] In some embodiments, the action comprises: decreasing a prediction horizon of the prediction model; or switching from the first prediction type to a second prediction type.

[0328] In some embodiments, the first prediction type is associated with a type-1 prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. In some embodiments, the type-1 prediction is associated with a type-1 codebook, and the type-ll prediction is associated with a type-ll codebook.

[0329] In some embodiments, the action comprises: terminating a prediction based on the prediction model; and transmitting, to the network device, measurements of the feature.

[0330] In some embodiments, the apparatus may further include means for receiving, from the terminal device, a positive indication for the prediction model; and means for taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0331] In some embodiments, the apparatus may further include means for based on determining an absence of a report of the performance evaluation from the terminal device at a reporting instant, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0332] In some embodiments, the apparatus may further include means for receiving, from the terminal device, a negative indication for the prediction model in a report of the performance evaluation; and means for based on determining absence of measurements of the feature in the report, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0333] In some embodiments, the apparatus may further include means for receiving, from the terminal device, a negative indication for the prediction model and measurements of the feature in a report of the performance evaluation.

[0334] In some embodiments, the apparatus may further include means for receiving, from the terminal device, the measurements of the feature; and means for taking the measurements of the feature as an input of the prediction model for a next prediction.

[0335] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0336] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 700. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0337] In some embodiments, an apparatus capable of performing any of the method 800 (for example, the terminal device 110) may comprise means for performing the respective steps of the method 800. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0338] In some embodiments, the apparatus may include: means for determining at least one negative result of performance evaluation on a prediction model for a feature; means for transmitting, to a network device, at least one negative indication for the prediction model; and means for receiving, from the network device, an indication of an action associated with the prediction model.

[0339] In some embodiments, the apparatus may further include means for performing the action. The means for performing the action may include at least one of: means for pausing a prediction based on theprediction model; means for obtaining measurements of the feature; means for transmitting, to the network device, the measurements; means for resuming the prediction based on the prediction model; or means for transmitting, to the network device, prediction results of the prediction model.

[0340] In some embodiments, the apparatus may further include means for receiving, from the network device, a configuration associated with the prediction model. The configuration comprises a first time duration. The apparatus may further include means for performing the action. The means for performing the action may include at least one of: means for pausing a prediction based on the prediction model for the first time duration; means for transmitting, to the network device, the obtained measurements of the feature for the first time duration; or means for resuming the prediction based on the prediction model after expiration of the first time duration; or means for transmitting, to the network device, prediction results of the prediction model and the obtained measurements of the feature for a second time duration after expiration of the first time duration.

[0341] In some embodiments, the configuration further comprises the second time duration or a sum of the first time duration and the second time duration.

[0342] In some embodiments, the apparatus may further include means for receiving, from the network device, one of the following: an indication of the second time duration, or an indication of a sum of the first time duration and the second time duration.

[0343] In some embodiments, the apparatus may further include means for performing the action. The means for performing the action may include at least one of: means for obtaining measurements of the feature for at least one time instant; means for obtaining prediction results of the prediction model for the at least one time instant; and means for transmitting, to the network device, at least one of the measurements or the prediction results.

[0344] In some embodiments, the apparatus may further include means for performing the action. The means for performing the action may include means for decreasing a prediction horizon of the prediction model. In some embodiments, the apparatus may further include means for receiving, from the network device, a configuration associated with the prediction model. The configuration comprises at least one of the following: an initial value for the prediction horizon; or a decrement value of the prediction horizon.

[0345] In some embodiments, the apparatus may further include means for performing the action. The means for performing the action may include means for switching from a first prediction type to a second prediction type. In some embodiments, the first prediction type is associated with a type-l prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. In some embodiments, the type-1 prediction is associated with a type-1 codebook, and the type-ll prediction is associated with a type-ll codebook.

[0346] In some embodiments, the apparatus may further include means for performing the action. The means for performing the action may include means for terminating a prediction based on the prediction model; and transmit, to the network device, measurements of the feature.

[0347] In some embodiments, the apparatus may further include means for based on determining a positive result of the performance evaluation, transmitting, to the network device, a positive indication for the prediction model.

[0348] In some embodiments, the apparatus may further include means for based on determining a positive result of the performance evaluation, omitting transmitting a report of the performance evaluation to the network device.

[0349] In some embodiments, the apparatus may further include means for based on determining a positive result of the performance evaluation, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0350] In some embodiments, the apparatus may further include means for based on determining a negative result of the performance evaluation, transmitting, to the network device, a negative indication for the prediction model; and means for taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0351] In some embodiments, the apparatus may further include means for based on determining a negative result of the performance evaluation, transmitting, to the network device, a negative indication for the prediction model and measurements of the feature.

[0352] In some embodiments, the apparatus may further include means for transmitting, to the network device, the measurements of the feature; and means for taking the measurements of the feature as an input of the prediction model for a next prediction.

[0353] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The apparatus may further include means for determining whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0354] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 800. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0355] In some embodiments, an apparatus capable of performing any of the method 900 (for example, the network device 120) may comprise means for performing the respective steps of the method 900. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0356] In some embodiments, the apparatus may include: means for receiving, from a terminal device, at least one negative indication for a prediction model for a feature; means for determining whether a first condition associated with performance monitoring on the prediction model is met; and means for based on determining that the first condition is met, transmitting, to the terminal device, an indication of an action associated with the prediction model.

[0357] In some embodiments, the first condition comprises a first threshold. In some embodiments, means for determining that the first condition is met may include means for determining that a counting number of the at least one negative indication reaches the first threshold. In some embodiments, the apparatus may further include means for initializing the counting number.

[0358] In some embodiments, the action comprises at least one of: pausing a prediction based on the prediction model; obtaining measurements of the feature; transmitting, to the network device, the measurements; resuming the prediction based on the prediction model; or transmitting, to the network device, prediction results of the prediction model.

[0359] In some embodiments, the apparatus may further include means for transmitting, to the terminal device, a configuration associated with the prediction model, wherein the configuration comprises at least one of the following: a first time duration for the terminal device to perform at least one of reporting measurements of the feature or pausing a prediction based on the prediction model; a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; or a sum of the first time duration and the second time duration.

[0360] In some embodiments, the apparatus may further include means for transmitting, to the terminal device, a configuration associated with the prediction model. The configuration comprises a first time duration for the terminal device to pause a prediction based on the prediction model. In some embodiments, the apparatus may further include means for transmitting, to the terminal device, one of the following: an indication of a second time duration for the terminal device to report prediction results of the prediction model and measurements of the feature after expiration of the first time duration; or a sum of the first time duration and the second time duration.

[0361] In some embodiments, the action comprises at least one of: obtaining measurements of the feature for at least one time instant; obtaining prediction results of the prediction model for the at least one time instant; or transmitting, to the network device, at least one of the measurements or the prediction results.

[0362] In some embodiments, the action comprises: decreasing a prediction horizon of the prediction model; or switching from a first prediction type to a second prediction type.

[0363] In some embodiments, the first prediction type is associated with a type-l prediction and the second prediction type is associated with a type-ll prediction. In some embodiments, the first prediction type is associated with a type-ll prediction and the second prediction type is associated with a type-1 prediction. Insome embodiments, the type-l prediction is associated with a type-l codebook, and the type-ll prediction is associated with a type-ll codebook.

[0364] In some embodiments, the apparatus may further include means for transmitting, to the terminal device, a configuration associated with the prediction model, wherein the configuration comprises at least one of the following: an initial value for the prediction horizon; or a decrement value of the prediction horizon.

[0365] In some embodiments, the action comprises: terminating a prediction based on the prediction model; and transmitting, to the network device, measurements of the feature.

[0366] In some embodiments, the apparatus may further include means for receiving, from the terminal device, a negative indication for the prediction model; and increment the counting number.

[0367] In some embodiments, the apparatus may further include means for determining that a result of the performance evaluation is positive based on receiving, from the terminal device, a positive indication for the prediction model.

[0368] In some embodiments, the apparatus may further include means for determining that a result of the performance evaluation is positive based on determining that a report of the performance evaluation from the terminal device is absent at a reporting instant.

[0369] In some embodiments, the apparatus may further include means for re-initializing the counting number based on determining that a result of the performance evaluation is positive.

[0370] In some embodiments, the apparatus may further include means for re-initializing the counting number based on determining that consecutive results of the performance evaluation are positive and a number of the consecutive results reaches a second threshold.

[0371] In some embodiments, the apparatus may further include means for based on determining that a result of the performance evaluation is positive, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0372] In some embodiments, the apparatus may further include means for receiving, from the terminal device, a negative indication for the prediction model in a report of the performance evaluation; and means for based on determining that the first condition is not met, taking prediction results of the prediction model as an input of the prediction model for a next prediction.

[0373] In some embodiments, the apparatus may further include means for receiving, from the terminal device, a negative indication for the prediction model and measurements of the feature in a report of the performance evaluation.

[0374] In some embodiments, the apparatus may further include means for receiving, from the terminal device, the measurements of the feature; and means for taking the measurements of the feature as an input of the prediction model for a next prediction.

[0375] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0376] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 900. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0377] In some embodiments, an apparatus capable of performing any of the method 1000 (for example, the terminal device 110) may comprise means for performing the respective steps of the method 1000. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0378] In some embodiments, the apparatus may include: means for transmitting, to a network device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is transmitted to the network device; measurements of the feature, wherein the second indication is transmitted to the network device; prediction results of the prediction model, wherein the third indication is transmitted to the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is transmitted to the network device and a previous input of the prediction model is prediction results of the prediction model.

[0379] In some embodiments, the first indication comprises at least one of the following: a positive indication for the prediction model or a negative indication for the prediction model; or an indication of the first input type.

[0380] In some embodiments, the apparatus may further include: means for determining whether a result of performance evaluation on the prediction model is positive or negative.

[0381] In some embodiments, the result is a positive result, and the first indication comprises at least one of the following: a positive indication for the prediction model; or an indication of the first input type.

[0382] In some embodiments, the result is a negative result, and the first indication comprises at least one of the following: the indication of the first input type; or a negative indication for the prediction model.

[0383] In some embodiments, the apparatus may further include: means for determining whether a first condition associated with performance monitoring on the prediction model is met based on determining that the result is a negative result; and means for based on determining that the first condition is not met, transmitting one of the following to the network device: the first indication; or the third indication, wherein a previous input of the prediction model is measurements of the feature.

[0384] In some embodiments, the apparatus may further include: means for receiving from the network device, an index of at least one previous prediction; and means for taking at least one prediction result of the prediction model corresponding to the at least one previous prediction as the input of the prediction model for the next prediction.

[0385] In some embodiments, the result is a negative result, and the second indication comprises at least one of the following: an indication of the second input type; a negative indication for the prediction model; or the measurements of the feature.

[0386] In some embodiments, the apparatus may further include: means for determining whether a first condition associated with performance monitoring on the prediction model is met based on determining that the result is a negative result; and means for based on determining that the first condition is met, transmitting one of the following to the network device: the second indication; or the third indication, wherein a previous input of the prediction model is prediction results of the prediction model.

[0387] In some embodiments, the means for determining whether the first condition is met may include: means for determining a counting number of negative results of the performance evaluation; and means for based on determining that the counting number reaches a first threshold, determining that the first condition is met; or means for based on determining that the counting number is below a first threshold, determining that the first condition is not met.

[0388] In some embodiments, the apparatus may further include: means for incrementing the counting number based on determining a negative result of the performance evaluation.

[0389] In some embodiments, the apparatus may further include: means for re-initializing the counting number based on determining a positive result of the performance evaluation. In some embodiments, the apparatus may further include: means for re-initializing the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold.

[0390] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The apparatus may further include: means for determining whether the result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0391] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 1000. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0392] In some embodiments, an apparatus capable of performing any of the method 1100 (for example, the network device 120) may comprise means for performing the respective steps of the method 1100. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitryor software module.

[0393] In some embodiments, the apparatus may include: means for receiving from a terminal device, a first indication associated with a first input type, a second indication associated with a second input type or a third indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the first indication is received from the terminal device; measurements of the feature, wherein the second indication is received from the terminal device; prediction results of the prediction model, wherein the third indication is received from the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the third indication is received from the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0394] In some embodiments, the first indication comprises at least one of the following: a positive indication for the prediction model or a negative indication for the prediction model; or an indication of the first input type.

[0395] In some embodiments, the first indication comprises a negative indication for the prediction model. The apparatus may further include: means for taking at least one best prediction result among a plurality of previous prediction results of the prediction model as the input of the prediction model for the next prediction based on determining not receiving measurements of the feature.

[0396] In some embodiments, the apparatus may further include: means for transmitting to the terminal device, an index of at least one previous prediction corresponding to the at least one best prediction result.

[0397] In some embodiments, the second indication comprises at least one of the following: an indication of the second input type; a negative indication for the prediction model; or the measurements of the feature.

[0398] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0399] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 1100. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0400] In some embodiments, an apparatus capable of performing any of the method 1200 (for example, the terminal device 110) may comprise means for performing the respective steps of the method 1200. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0401] In some embodiments, the apparatus may include: means for receiving from a network device, an indication of a first input type, an indication of a second input type or an indication associated with a switchbetween the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

[0402] In some embodiments, the apparatus may further include: means for determining a negative result of performance evaluation on the prediction model; and means for transmitting to a network device, a negative indication for the predication model. The indication of the first input type or the indication of the second input type is received after transmitting the negative indication.

[0403] In some embodiments, the indication of the first input type comprises an index of at least one previous prediction, and the prediction results of the prediction model comprise at least one prediction result of the prediction model corresponding to the at least one previous prediction.

[0404] In some embodiments, the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model. The apparatus may further include: means for transmitting the measurements of the feature to the network device.

[0405] In some embodiments, the indication of the second input type is received from the network device. The apparatus may further include: means for transmitting the measurements of the feature to the network device.

[0406] In some embodiments, the prediction model is a channel status information (CS I) based twin model. The apparatus may further include: means for determining whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

[0407] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 1200. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0408] In some embodiments, an apparatus capable of performing any of the method 1300 (for example, the network device 120) may comprise means for performing the respective steps of the method 1300. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[0409] In some embodiments, the apparatus may include: means for transmitting to a terminal device, anindication of a first input type, an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and means for taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.

[0410] In some embodiments, the apparatus may further include: means for receiving from the terminal device, a negative indication for the predication model; means for determining whether a first condition associated with performance monitoring on the prediction model is met; and means for based on determining that the first condition is not met, transmitting the indication of the first input type to the terminal device; or means for based on determining that the first condition is met, transmitting the indication of the second input type to the terminal device.

[0411] In some embodiments, when determining whether the first condition is met, the apparatus may further include: means for determining a counting number of negative results of the performance evaluation; and means for based on determining that the counting number reaches a first threshold, determining that the first condition is met; or means for based on determining that the counting number is below a first threshold, determining that the first condition is not met.

[0412] In some embodiments, the apparatus may further include: means for incrementing the counting number based on determining a negative result of the performance evaluation.

[0413] In some embodiments, the apparatus may further include: means for re-initializing the counting number based on determining a positive result of the performance evaluation. In some embodiments, the apparatus may further include: means for re-initializing the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold.

[0414] In some embodiments, the apparatus may further include: means for receiving, from the terminal device, a negative indication for the predication model. The prediction results of the prediction model comprise at least one best prediction result among a plurality of previous prediction results of the prediction model.

[0415] In some embodiments, the indication of the first input type comprises an index of at least one previous prediction corresponding to the at least one best prediction result.

[0416] In some embodiments, the indication of the second input type is transmitted to the terminal device. The apparatus may further include: means for receiving the measurements of the feature from the terminaldevice.

[0417] In some embodiments, the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model. The apparatus may further include: means for receiving the measurements of the feature from the terminal device.

[0418] In some embodiments, the prediction model is a channel status information (CS I) based twin model.

[0419] In some embodiments, the apparatus further may include means for performing other steps in some embodiments of the method 1300. In some embodiments, the means may include at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[0420] Fig. 14 is a simplified block diagram of a device 1400 that is suitable for implementing embodiments of the present disclosure. The device 1400 may be provided to implement the communication device, for example the terminal device 110, or the network device 120 as shown in Fig. 1 . As shown, the device 1400 includes one or more processors 1410, one or more memories 1420 coupled to the processor 1410, and one or more communication modules 1440 coupled to the processor 1410.

[0421] The communication module 1440 is for bidirectional communications. The communication module 1440 has at least one antenna to facilitate communication. The communication interface may represent any interface that is necessary for communication with other network elements.

[0422] The processor 1410 may be of any type suitable to the local technical network and may include one or more of the following: general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs) and processors based on multicore processor architecture, as non-limiting examples. The device 1400 may have multiple processors, such as an application specific integrated circuit chip that is slaved in time to a clock which synchronizes the main processor.

[0423] The memory 1420 may include one or more non-volatile memories and one or more volatile memories. Examples of the non-volatile memories include, but are not limited to, a Read Only Memory (ROM) 1424, an electrically programmable read only memory (EPROM), a flash memory, a hard disk, a compact disc (CD), a digital video disk (DVD), and other magnetic storage and / or optical storage. Examples of the volatile memories include, but are not limited to, a random access memory (RAM) 1422 and other volatile memories that will not last in the power-down duration.

[0424] A computer program 1430 includes computer executable instructions that are executed by the associated processor 1410. The program 1430 may be stored in the ROM 1420. The processor 1410 may perform any suitable actions and processing by loading the program 1430 into the RAM 1420.

[0425] The embodiments of the present disclosure may be implemented by means of the program 1430 so that the device 1400 may perform any process of the disclosure as discussed with reference to Figs. 2Ato 13. The embodiments of the present disclosure may also be implemented by hardware or by a combination of software and hardware.

[0426] In some embodiments, the program 1430 may be tangibly contained in a computer readable medium which may be included in the device 1400 (such as in the memory 1420) or other storage devices that are accessible by the device 1400. The device 1400 may load the program 1430 from the computer readable medium to the RAM 1422 for execution. The computer readable medium may include any types of tangible non-volatile storage, such as ROM, EPROM, a flash memory, a hard disk, CD, DVD, and the like. Fig. 15 shows an example of the computer readable medium 1500 in form of CD or DVD. The computer readable medium has the program 1430 stored thereon.

[0427] Generally, various embodiments of the present disclosure may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. While various aspects of embodiments of the present disclosure are illustrated and described as block diagrams, flowcharts, or using some other pictorial representations, it is to be understood that the block, apparatus, system, technique or method described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

[0428] The present disclosure also provides at least one computer program product tangibly stored on a non-transitory computer readable storage medium. The computer program product includes computerexecutable instructions, such as those included in program modules, being executed in a device on a target real or virtual processor, to carry out and of the methods, 600 or 1300 as described above with reference to Figs. 6-13. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, or the like that perform particular tasks or implement particular abstract data types. The functionality of the program modules may be combined or split between program modules as desired in various embodiments. Machine-executable instructions for program modules may be executed within a local or distributed device. In a distributed device, program modules may be located in both local and remote storage media.

[0429] Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partlyon a remote machine or entirely on the remote machine or server.

[0430] In the context of the present disclosure, the computer program codes or related data may be carried by any suitable carrier to enable the device, apparatus or processor to perform various processes and operations as described above. Examples of the carrier include a signal, computer readable medium, and the like.

[0431] The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable medium may include but not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the computer readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. The term "non- transitory,” as used herein, is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storage persistency (e.g., RAM vs. ROM).

[0432] Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are contained in the above discussions, these should not be construed as limitations on the scope of the present disclosure, but rather as descriptions of features that may be specific to particular embodiments. Certain features that are described in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment may also be implemented in multiple embodiments separately or in any suitable sub-combination.

[0433] Although the present disclosure has been described in languages specific to structural features and / or methodological acts, it is to be understood that the present disclosure defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

WHAT IS CLAIMED IS:1 . A terminal device comprising: at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the terminal device at least to: receive, from a network device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and take one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

2. The terminal device of claim 1 , wherein the terminal device is further caused to: determine a negative result of performance evaluation on the prediction model; and transmit, to a network device, a negative indication for the predication model, wherein the indication of the first input type or the indication of the second input type is received after transmitting the negative indication.

3. The terminal device of claim 2, wherein the indication of the first input type comprises an index of at least one previous prediction, and the prediction results of the prediction model comprise at least one prediction result of the prediction model corresponding to the at least one previous prediction.

4. The terminal device of claim 1 or 2, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model, and the terminal device is further caused to: transmit the measurements of the feature to the network device.

625. The terminal device of claim 1 or 2, wherein the indication of the second input type is received from the network device, and the terminal device is further caused to: transmit the measurements of the feature to the network device.

6. The terminal device of claim 2 or 3, wherein the prediction model is a channel status information (CSI) based twin model, and the terminal device is caused to: determine whether a result of the performance evaluation is positive or negative based on measured CSI and predicted CSI of the prediction model.

7. A network device comprising: at least one processor; and at least one memory storing instructions that, when executed by the at least one processor, cause the network device at least to: transmit, to a terminal device, an indication of a first input type, an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and take one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; or measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.

8. The network device of claim 7, wherein the network device is further caused to: receive, from the terminal device, a negative indication for the predication model; determine whether a first condition associated with performance monitoring on the prediction model is met; and based on determining that the first condition is not met, transmit the indication of the first input type to the terminal device; or based on determining that the first condition is met, transmit the indication of the second input typeto the terminal device.

9. The network device of claim 8, wherein the network device is caused to determine whether the first condition is met by: determining a counting number of negative results of the performance evaluation; and based on determining that the counting number reaches a first threshold, determining that the first condition is met; or based on determining that the counting number is below a first threshold, determining that the first condition is not met.

10. The network device of claim 9, wherein the network device is further caused to: increment the counting number based on determining a negative result of the performance evaluation.11 . The network device of claim 9 or 10, wherein the network device is further caused to: re-initialize the counting number based on determining a positive result of the performance evaluation; or re-initialize the counting number based on determining that a number of consecutive positive results of the performance evaluation reaches a second threshold.

12. The network device of any of claims 7-11 , wherein the network device is further caused to: receive, from the terminal device, a negative indication for the predication model; and wherein the prediction results of the prediction model comprise at least one best prediction result among a plurality of previous prediction results of the prediction model.

13. The network device of claim 12, wherein the indication of the first input type comprises an index of at least one previous prediction corresponding to the at least one best prediction result.

14. The network device of any of claims 7-11 , wherein the indication of the second input type is transmitted to the terminal device, and the network device is further caused to: receive the measurements of the feature from the terminal device.

15. The network device of any of claims 7-11 , wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model, and the network device is further caused to: receive the measurements of the feature from the terminal device.6416. The network device of any of claims 7-15, wherein the prediction model is a channel status information (CSI) based twin model.

17. A method performed by a terminal device, comprising: receiving from a network device, an indication of a first input type, or an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is received from the network device; measurements of the feature, wherein the indication of the second input type is received from the network device; prediction results of the prediction model, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is received from the network device and a previous input of the prediction model is prediction results of the prediction model.

18. A method performed by a network device, comprising: transmitting to a terminal device, an indication of a first input type, an indication of a second input type or an indication associated with a switch between the first input type and the second input type for a prediction model for a feature; and taking one of the following as an input of the prediction model for a next prediction: prediction results of the prediction model, wherein the indication of the first input type is transmitted to the terminal device; measurements of the feature, wherein the indication of the second input type is transmitted to the terminal device; prediction results of the prediction model, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is measurements of the feature; or measurements of the feature, wherein the indication associated with the switch is transmitted to the terminal device and a previous input of the prediction model is prediction results of the prediction model.65

Images