Devices and methods of communication
By determining mobility states to activate or deactivate AI models based on temporary predictions, the method optimizes AI model usage, reducing wastage and enhancing network management efficiency.
Patent Information
- Application Number
- PCT/CN2023/086553
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-04-06
- Publication Date
- 2025-11-06
AI Technical Summary
Current AI techniques for managing air interface protocols lack clarity and efficiency in managing AI models, leading to unnecessary usage and resource wastage.
A terminal device determines its mobility state to activate or deactivate AI models based on temporary predictions, transmitting measurement results to a network device and receiving mobility states based on output parameter variations, and reactivating the model when criteria are met.
This approach optimizes AI model usage by avoiding unnecessary operations, enhancing robustness and resource efficiency in network management.
Smart Images

Figure CN2023086553_06112025_PF_FP_ABST
Abstract
Description
DEVICES AND METHODS OF COMMUNICATIONTECHNICAL FIELD
[0001] Embodiments of the present disclosure generally relate to the field of telecommunication, and in particular, to devices and methods of communication for management of an artificial intelligence (AI) model.BACKGROUND
[0002] Currently, AI techniques for an air interface are highly concerned. It has been proposed to assess potential specification impact on protocol aspects related to management of data and an AI model. However, details of management of an AI model are still unclear and need to be further developed.
[0003] SUMMARY
[0004] In general, embodiments of the present disclosure provide methods, devices and computer storage media of communication for management of an AI model.
[0005] In a first aspect, there is provided a terminal device. The terminal device comprises a processor. The processor is configured to cause the terminal device to: determine a mobility state of the terminal device; and cause the mobility state to be used for activation or deactivation of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction.
[0006] In a second aspect, there is provided a terminal device. The terminal device comprises a processor. The processor is configured to cause the terminal device to: transmit, to a network device, measurement results as an input of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction; and receive, from the network device, a mobility state of the terminal device, the mobility state being determined based on a variation of information of an output parameter of the artificial intelligence model over future time instances.
[0007] In a third aspect, there is provided a terminal device. The terminal device comprises a processor. The processor is configured to cause the terminal device to: determine that an artificial intelligence model is deactivated, the artificial intelligence model being based on a temporary prediction; and in accordance with a determination that a criterion for reactivating the artificial intelligence model is fulfilled, cause the artificial intelligence model to be reactivated.
[0008] In a fourth aspect, there is provided a method of communication. The method comprises: determining, at a terminal device, a mobility state of the terminal device; and causing the mobility state to be used for activation or deactivation of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction.
[0009] In a fifth aspect, there is provided a method of communication. The method comprises: transmitting, at a terminal device and to a network device, measurement results as an input of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction; and receiving, from the network device, a mobility state of the terminal device, the mobility state being determined based on a variation of information of an output parameter of the artificial intelligence model over future time instances.
[0010] In a sixth aspect, there is provided a method of communication. The method comprises: determining, at a terminal device, that an artificial intelligence model is deactivated, the artificial intelligence model being based on a temporary prediction; determining that a criterion for reactivating the artificial intelligence model is fulfilled; and causing the artificial intelligence model to be reactivated.
[0011] In a seventh aspect, there is provided a computer readable medium having instructions stored thereon. The instructions, when executed on at least one processor, cause the at least one processor to perform the method according to any of the fourth to sixth aspects of the present disclosure.
[0012] Other features of the present disclosure will become easily comprehensible through the following description.BRIEF DESCRIPTION OF THE DRAWINGS
[0013] Through the more detailed description of some embodiments of the present disclosure in the accompanying drawings, the above and other objects, features and advantages of the present disclosure will become more apparent, wherein:
[0014] FIG. 1 illustrates an example communication network in which some embodiments of the present disclosure can be implemented;
[0015] FIG. 2A illustrates a schematic diagram of an AI model inference and monitoring in which some embodiments of the present disclosure can be implemented;
[0016] FIG. 2B illustrates a schematic diagram of an AI-based beam management scenario in which some embodiments of the present disclosure can be implemented;
[0017] FIG. 3 illustrates a schematic diagram illustrating a process of communication according to embodiments of the present disclosure;
[0018] FIG. 4 illustrates a schematic diagram of an example determination of a mobility state of a terminal device according to embodiments of the present disclosure;
[0019] FIG. 5 illustrates a schematic diagram illustrating another process of communication according to embodiments of the present disclosure;
[0020] FIG. 6 illustrates a schematic diagram illustrating still another process of communication according to embodiments of the present disclosure;
[0021] FIG. 7 illustrates an example method of communication implemented at a terminal device in accordance with some embodiments of the present disclosure;
[0022] FIG. 8 illustrates another example method of communication implemented at a terminal device in accordance with some embodiments of the present disclosure;
[0023] FIG. 9 illustrates still another example method of communication implemented at a terminal device in accordance with some embodiments of the present disclosure; and
[0024] FIG. 10 is a simplified block diagram of a device that is suitable for implementing embodiments of the present disclosure.
[0025] Throughout the drawings, the same or similar reference numerals represent the same or similar element.DETAILED DESCRIPTION
[0026] Principle of the present disclosure will now be described with reference to some 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 limitations as to the scope of the disclosure. The disclosure described herein can be implemented in various manners other than the ones described below.
[0027] 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.
[0028] As used herein, the term ‘terminal device’ refers to any device having wireless or wired communication capabilities. Examples of the terminal device include, but not limited to, user equipment (UE) , personal computers, desktops, mobile phones, cellular phones, smart phones, personal digital assistants (PDAs) , portable computers, tablets, wearable devices, internet of things (IoT) devices, Ultra-reliable and Low Latency Communications (URLLC) devices, Internet of Everything (IoE) devices, machine type communication (MTC) devices, device on vehicle for V2X communication where X means pedestrian, vehicle, or infrastructure / network, devices for Integrated Access and Backhaul (IAB) , Space borne vehicles or Air borne vehicles in Non-terrestrial networks (NTN) including Satellites and High Altitude Platforms (HAPs) encompassing Unmanned Aircraft Systems (UAS) , eXtended Reality (XR) devices including different types of realities such as Augmented Reality (AR) , Mixed Reality (MR) and Virtual Reality (VR) , the unmanned aerial vehicle (UAV) commonly known as a drone which is an aircraft without any human pilot, devices on high speed train (HST) , or image capture devices such as digital cameras, sensors, gaming devices, music storage and playback appliances, or Internet appliances enabling wireless or wired Internet access and browsing and the like. The ‘terminal device’ can further has ‘multicast / broadcast’ feature, to support public safety and mission critical, V2X applications, transparent IPv4 / IPv6 multicast delivery, IPTV, smart TV, radio services, software delivery over wireless, group communications and IoT applications. It may also incorporate one or multiple Subscriber Identity Module (SIM) as known as Multi-SIM. The term “terminal device” can be used interchangeably with a UE, a mobile station, a subscriber station, a mobile terminal, a user terminal or a wireless device.
[0029] The term “network device” may refer to a device which is capable of providing or hosting a cell or coverage where terminal devices can communicate. Examples of an access network device include, but not limited to, a satellite, a unmanned aerial systems (UAS) platform, a Node B (NodeB or NB) , an evolved NodeB (eNodeB or eNB) , a next generation NodeB (gNB) , a transmission reception point (TRP) , a remote radio unit (RRU) , a radio head (RH) , a remote radio head (RRH) , an IAB node, a low power node such as a femto node, a pico node, a reconfigurable intelligent surface (RIS) , and the like.
[0030] The terminal device or the network device may have AI or machine learning (ML) capability. It generally includes a model which has been trained from numerous collected data for a specific function, and can be used to predict some information.
[0031] The terminal or the network device may work on several frequency ranges, e.g. FR1 (410 MHz to 7125 MHz) , FR2 (24.25GHz to 71GHz) , frequency band larger than 100GHz as well as Tera Hertz (THz) . It can further work on licensed / unlicensed / shared spectrum. The terminal device may have more than one connection with the network devices under Multi-Radio Dual Connectivity (MR-DC) application scenario. The terminal device or the network device can work on full duplex, flexible duplex and cross division duplex modes.
[0032] The embodiments of the present disclosure may be performed in test equipment, e.g. signal generator, signal analyzer, spectrum analyzer, network analyzer, test terminal device, test network device, channel emulator.
[0033] In one embodiment, the terminal device may be connected with a first network device and a second network device. One of the first network device and the second network device may be a master node and the other one may be a secondary node. The first network device and the second network device may use different radio access technologies (RATs) . In one embodiment, the first network device may be a first RAT device and the second network device may be a second RAT device. In one embodiment, the first RAT device is eNB and the second RAT device is gNB. Information related with different RATs may be transmitted to the terminal device from at least one of the first network device or the second network device. In one embodiment, first information may be transmitted to the terminal device from the first network device and second information may be transmitted to the terminal device from the second network device directly or via the first network device. In one embodiment, information related with configuration for the terminal device configured by the second network device may be transmitted from the second network device via the first network device. Information related with reconfiguration for the terminal device configured by the second network device may be transmitted to the terminal device from the second network device directly or via the first network device.
[0034] 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. The term ‘includes’ and its variants are to be read as open terms that mean ‘includes, but is not limited to. ’ The term ‘based on’ is to be read as ‘at least in part based on. ’ The term ‘one embodiment’ and ‘an embodiment’ are to be read as ‘at least one embodiment. ’ The term ‘another embodiment’ is to be read as ‘at least one other embodiment. ’ The terms ‘first, ’ ‘second, ’ and the like may refer to different or same objects. Other definitions, explicit and implicit, may be included below.
[0035] In some examples, values, procedures, or apparatus are referred to as ‘best, ’ ‘lowest, ’ ‘highest, ’ ‘minimum, ’ ‘maximum, ’ or the like. It will be appreciated that such descriptions are intended to indicate that a selection among many used functional alternatives can be made, and such selections need not be better, smaller, higher, or otherwise preferable to other selections.
[0036] In the context of the present disclosure, the term “AI” may be interchangeably used with “machine learning (ML) ” or “AI / ML” . The term “AI model” may be interchangeably used with “ML model” or “AI / ML model” .
[0037] In some cases of beam management, historical beam information may be used to predict future beam information in future time instances through an AI model. In some cases of channel state information (CSI) prediction, historical CSI information may be used to predict future CSI information in future time instances through an AI model. Such prediction of future information in future time instances may be called as a temporary prediction.
[0038] Embodiments of the present disclosure provide solutions for management of an AI model, especially an AI model based on a temporary prediction. In one solution, a terminal device determines a mobility state of the terminal device, and causes the mobility state to be used for activation or deactivation of an AI model. In this way, an AI model may be activated or deactivated based on a mobility state of a terminal device. Thus, unnecessary AI model usage may be avoided.
[0039] In another solution, a terminal device transmits, to a network device, measurement results as an input of an AI model, and receives, from the network device, a mobility state of the terminal device. The mobility state is determined based on a variation of information of an output parameter of the AI model over future time instances. In this way, a network device may be facilitated to determine a mobility state of a terminal device for better network management of an AI model.
[0040] In still another solution, upon determination that an AI model is deactivated, a terminal device determines whether a criterion for reactivating the AI model is fulfilled. If the criterion is fulfilled, the terminal device causes the AI model to be reactivated. In this way, robust for the usage of an AI model may be enhanced.
[0041] Principles and implementations of the present disclosure will be described in detail below with reference to the figures.
[0042] EXAMPLE OF COMMUNICATION NETWORK
[0043] FIG. 1 illustrates a schematic diagram of an example communication network 100 in which some embodiments of the present disclosure can be implemented. As shown in FIG. 1, the communication network 100 may include a terminal device 110 and a network device 120. In some embodiments, the network device 120 may provide one or more serving cells (not shown) to serve the terminal device 110.
[0044] The terminal device 110 may have a plurality of beams (not shown) , and the network device 120 may have a plurality of beams (not shown) . A channel (or called as a sub-channel in this case) may be formed between one of the plurality of beams of the terminal device 110 and one of the plurality of beams of the network device 120. The terminal device 110 may transmit information to the network device 120 or receive information from the network device 120 via one or more sub-channels.
[0045] It is to be understood that the number of devices in FIG. 1 is given for the purpose of illustration without suggesting any limitations to the present disclosure. The communication network 100 may include any suitable number of network devices and / or terminal devices and / or other network elements adapted for implementing implementations of the present disclosure.
[0046] As shown in FIG. 1, the terminal device 110 and the network device 120 may communicate with each other via a channel such as a wireless communication channel. The communications in the communication network 100 may conform to any suitable standards including, but not limited to, Global System for Mobile Communications (GSM) , Long Term Evolution (LTE) , LTE-Evolution, LTE-Advanced (LTE-A) , New Radio (NR) , Wideband Code Division Multiple Access (WCDMA) , Code Division Multiple Access (CDMA) , GSM EDGE Radio Access Network (GERAN) , Machine Type Communication (MTC) and the like. The embodiments of the present disclosure may be performed according to any generation communication protocols either currently known or to be developed in the future. Examples of the communication protocols include, but not limited to, the first generation (1G) , the second generation (2G) , 2.5G, 2.75G, the third generation (3G) , the fourth generation (4G) , 4.5G, the fifth generation (5G) communication protocols, 5.5G, 5G-Advanced networks, or the sixth generation (6G) networks.
[0047] In some scenarios, the terminal device 110 may receive, from the network device 120, a configuration indicating beam measurements on a set of downlink (DL) RSs. Then the terminal device 110 may receive the set of DL RSs and perform DL RS measurements on the set of DL RSs. The terminal device 110 may transmit, to the network device 120, a beam report indicating results of the DL RS measurements. The network device 120 may perform DL beam selection based on the beam report. These scenarios may be called as DL BM.
[0048] In some scenarios, the terminal device 110 may receive, from the network device 120, a configuration indicating an uplink (UL) RS transmission on a set of resources. Then the terminal device 110 may transmit a set of UL RSs to the network device 120. The network device 120 may perform UL RS measurements on the set of UL RSs and perform UL beam selection based on results of the UL RS measurements. These scenarios may be called as UL BM.
[0049] In some scenarios, the UL or DL BM (e.g., the UL or DL beam selection) may be performed based on AI model. FIG. 2A illustrates a schematic diagram 200A of an AI model inference and monitoring in which some embodiments of the present disclosure can be implemented. As shown in FIG. 2A, model training may be performed based on training data. An AI model may be deployed or updated by the model training. With an input data set as an input of the AI model, an output data set may be obtained by model inference. This is an AI model inference procedure.
[0050] With reference to FIG. 2A, a model monitoring may be performed to monitor performance of the model inference of the AI model. The model monitoring may be classified into three types: comparison between inference results and ground-truth results; evaluation for system performance (e.g., throughout, block error rate (BLER) , reference signal receiving power (RSRP) , positive acknowledgement (ACK) / NACK) ; distribution detection for the input or output data set.
[0051] FIG. 2B illustrates a schematic diagram 200B illustrating an AI-based beam management scenario in which some embodiments of the present disclosure can be implemented. In this example, a temporary beam prediction is shown. As shown in FIG. 2B, history signal measurements for a set of beams (denoted as Set B) in history or current time instances t-3, t-2, t-1 and t may be used as an input of an AI model 201 for beam prediction. Based on an output of the AI model 201, future signal measurements for another set of beams (denoted as Set A) in future time instances t+1, t+2, t+3…. may be predicted. One or multiple beams may be selected from Set A as a set of predicted beams. In the example of FIG. 2B, the case of one predicted beam in each of future time instances t+1, t+2, t+3…is shown.
[0052] In some embodiments, Set B may be different from Set A, as shown by reference sign 210 of FIG. 2B. In some embodiments, Set B may be a subset of Set A, as shown by reference sign 220 of FIG. 2B. In some embodiments, Set B may be the same as Set A, as shown by reference sign 230 of FIG. 2B.
[0053] It can be seen that using historical beam information to predict beam information in multiple future time instances may reduce beam measurement and report overhead. However, if a terminal device is in a low mobility state or a stationary state, measured beams multiple time instances may be unchanged, and beam information in multiple future time instances may be same. In this case, there is no need for the terminal device to run an AI model for beam prediction.
[0054] In view of this, embodiments of the present disclosure provide solutions of communication for AI model management so as to avoid unnecessary AI model usage. It is to be understood that the AI model for beam management is merely an example, and the solutions according to embodiments of the present disclosure may be applied to any suitable AI models based on temporary prediction, and the present disclosure does not make limitation on a function of an AI model. More details will be described in connection with FIGs. 3 to 4 below.
[0055] EXAMPLE IMPLEMENTATION OF ACTIVATION OR DEACTIVATION OF AI MODEL
[0056] FIG. 3 illustrates a schematic diagram illustrating a process 300 of communication according to embodiments of the present disclosure. For the purpose of discussion, the process 300 will be described with reference to FIG. 1. The process 300 may involve the terminal device 110 and the network device 120 as illustrated in FIG. 1. It is to be understood that the steps and the order of the steps in FIG. 3 are merely for illustration, and not for limitation. For example, the order of the steps may be changed. Some of the steps may be omitted or any other suitable additional steps may be added. It is assumed that an artificial intelligence model based on a temporary prediction is deployed at the terminal device 110 or the network device 120.
[0057] As shown in FIG. 3, the terminal device 110 may determine 310 a mobility state of the terminal device 110. In some embodiments, the terminal device 110 may determine 311 the mobility state based on a predetermined criterion. It is to be understood that the predetermined criterion may be any suitable criteria existing or to be developed in future.
[0058] In some embodiments, if the terminal device 110 fulfills a relaxed measurement criterion for low mobility, the terminal device 110 may determine that the terminal device 110 is in a low mobility state. For example, a low mobility criterion below may be used.
[0059] The relaxed measurement criterion for UE with low mobility is fulfilled when:
[0060] - (SrxlevRef –Srxlev) < SSearchDeltaP,
[0061] Where:
[0062] - Srxlev = current Srxlev value of the serving cell (dB) .
[0063] - SrxlevRef = reference Srxlev value of the serving cell (dB) , set as follows:
[0064] - After selecting or reselecting a new cell, or
[0065] - If (Srxlev -SrxlevRef) > 0, or
[0066] - If the relaxed measurement criterion has not been met for TSearchDeltaP:
[0067] - The UE shall set the value of SrxlevRef to the current Srxlev value of the serving cell.
[0068] In some embodiments, if the terminal device 110 fulfills a relaxed measurement criterion for stationary, the terminal device 110 may determine that the terminal device 110 is in a stationary state. For example, a stationary criterion below may be used.
[0069] The relaxed measurement criterion for a stationary UE is fulfilled when:
[0070] - (SrxlevRefStationary –Srxlev) < SSearchDeltaP-Stationary,
[0071] Where:
[0072] - Srxlev = current Srxlev value of the serving cell (dB) .
[0073] - SrxlevRefStationary = reference Srxlev value of the serving cell (dB) , set as follows:
[0074] - After selecting or reselecting a new cell, or
[0075] - If (Srxlev -SrxlevRefStationary) > 0, or
[0076] - If the relaxed measurement criterion has not been met for TSearchDeltaP-Stationary:
[0077] - The UE shall set the value of SrxlevRefStationary to the current Srxlev value of the serving cell.
[0078] In some embodiments, if a difference of signal strength between RS measurements within a period of time (for convenience, also referred to as a first period of time herein) is below a threshold difference, the terminal device 110 may determine that the terminal device 110 is in the low mobility state or the stationary state. For example, the terminal device 110 may determine that the terminal device 110 is in the low mobility state or the stationary state if the terminal device 110 fulfills equation (1) within a period of time T below.
[0079] |RSRP1 –RSRP2| ≤ A (1)
[0080] where RSRP1 and RSRP2 denote two RS measurements derived during T, and A denotes the threshold difference.
[0081] For determination of the difference of signal strength, in some embodiments, the terminal device 110 may determine the difference of signal strength based on RS measurement in a first time point and RS measurement in a second time point later than the first time point. In some embodiments, the terminal device 110 may determine, as the difference of signal strength based on current RS measurement and a reference RS measurement. In some embodiments, the terminal device 110 may periodically set the latest measured signal strength as the reference RS measurement. In some embodiments, the terminal device 110 may set, as the reference RS measurement, the latest measured signal strength when the terminal device 110 performs a cell change (e.g., a cell reselection or a handover) . In some embodiments, the signal strength may be in a cell level. In some embodiments, the signal strength may be in a beam level. In some embodiments, the signal strength may be layer 1 (L1) signal strength. In some embodiments, the signal strength may be layer 3 (L3) signal strength.
[0082] In some embodiments, if number of cell reselections during a period of time (for convenience, also referred to as a second period of time herein) is above a threshold number (for convenience, also referred to as a first threshold number herein) , the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state.
[0083] For example, a state detection criterion may be defined as below.
[0084] Normal-mobility state criteria:
[0085] - If number of cell reselections during time period TCRmax is less than NCR_M.
[0086] Medium-mobility state criteria:
[0087] - If number of cell reselections during time period TCRmax is greater than or equal to NCR_M but less than or equal to NCR_H.
[0088] High-mobility state criteria:
[0089] - If number of cell reselections during time period TCRmax is greater than NCR_H.
[0090] In some embodiments, if the normal-mobility state criteria are fulfilled, when the terminal device 110 enters a RRC connected state, the terminal device 110 may determine that the terminal device 110 is in the low mobility state. In some embodiments, if the medium-mobility state criteria are fulfilled, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state. In some embodiments, if the high-mobility state criteria are fulfilled, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state. It is to be understood that the state detection criterion may be used in any other suitable ways to determine the mobility state.
[0091] In some embodiments, if number of handover times within a period of time (for convenience, also referred to as a third period of time herein) is above a threshold number (for convenience, also referred to as a second threshold number herein) , the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state. For example, the terminal device 110 may record the number of handover times. If the number of handover times is lower than or equal to a threshold number B within a period of time T, the terminal device 110 is in the low mobility state or the stationary state. If the number of handover times is larger than B within a period of time T, the terminal device 110 is not in the low mobility state and the stationary state.
[0092] In some embodiments, if a property of the terminal device 110 indicates low mobility, the terminal device 110 may determine that the terminal device 110 is in the low mobility state. If the property of the terminal device 110 indicates stationary, the terminal device 110 may determine that the terminal device 110 is in the stationary state.
[0093] In some embodiments, the terminal device 110 may determine the property of the terminal device 110 based on at least one of a type of the terminal device 110, a type of a service associated with the terminal device 110, or time period information. For example, if the terminal device 110 is a fixed device such as a roadside sensor or the like, the terminal device 110 may determine that the property of the terminal device 110 indicates stationary. In another example, if the terminal device 110 is an alarm system with a long alarm period, the terminal device 110 may determine that the property of the terminal device 110 indicates low mobility. In another example, if the terminal device 110 is a wearable device and it is a night time, the terminal device 110 may determine that the property of the terminal device 110 indicates low mobility. It is to be understood that any other suitable ways are also feasible for determination of the property.
[0094] In some embodiments, the terminal device 110 may receive, from the network device 120, an indication of the property of the terminal device 110, and determine the property of the terminal device 110 based on the indication.
[0095] In some embodiments, if a distribution variation of an input parameter of the AI model within a period of time (for convenience, also referred to as a fourth period of time herein) is below a threshold distribution variation, the terminal device 110 may determine that the terminal device 110 is in the low mobility state or the stationary state. In other words, if a distribution (e.g., value or range) of the input parameter is not changed significantly during the period of time, the terminal device 110 is in the low mobility state or the stationary state. It is to be understood that the present disclosure does not limit the input parameter, and the input parameter may be any suitable measured parameter.
[0096] For example, the input parameter may comprise historical information for CSI, e.g., a square generalized cosine similar (SGCS) , normalized mean square error (NMSE) , rank indication (RI) , channel quality indicator (CQI) , precoding matrix indicator (PMI) or compressed bits. In another example, the input parameter may comprise historical information for positioning, e.g., channel impulse response (CIR) , power delay profile (PDP) , positioning reference signal (PRS) , time difference of arrival (TDoA) , angle of arrival (AoA) , non-line of sight (NLOS) or line of sight (LOS) indicator.
[0097] For illustration, an example procedure may be described as below.
[0098] Upon reception of the configuration for AI model for BM (i.e., AI model is deployed at UE side) , the UE shall:
[0099] 1> evaluate the mobility state by at least one of the following rules:
[0100] - the legacy low mobility criterion;
[0101] - the legacy stationary criterion;
[0102] - the difference of signal strength between two measured RS within one period;
[0103] - the legacy mobility state for cell reselection;
[0104] - the number of handover times within one period;
[0105] - the property of low mobility or stationary;
[0106] - the AI input distribution compared with the previous one.
[0107] Continue to refer to FIG. 3, in some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may determine 312 the mobility state based on an output of the AI model. In some embodiments, the terminal device 110 may determine an output of the AI model. The output comprises information of an output parameter in future time instances. The terminal device 110 may determine the mobility state of the terminal device 110 based on a variation of the information over the future time instances.
[0108] In some embodiments, the information of the output parameter may comprise indexes and values of the output parameter in the future time instances. In some embodiments, if indexes of the output parameter in the future time instances are unchanged and differences of values of the output parameter in the future time instances are below (e.g., lower than or equal to) a threshold difference, the terminal device 110 may determine that the terminal device 110 is in the low mobility state or the stationary state.
[0109] In some embodiments, if the indexes of the output parameter in the future time instances are unchanged and the differences of values of the output parameter in the future time instances are above (e.g., greater than or equal to) the threshold difference, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state.
[0110] In some embodiments, if the indexes of the output parameter are changed for the future time instances, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state. It is to be understood that the output parameter may comprise a single parameter, e.g., the best beam. Alternatively, the output parameter may comprise a parameter set, e.g., top-K beams. The present disclosure does not limit a format of the output parameter, and the format of the output parameter is dependent on design of the AI model.
[0111] FIG. 4 illustrates a schematic diagram 400 of an example determination of a mobility state of a terminal device according to embodiments of the present disclosure. As shown in FIG. 4, beam measurements of Set B in historical time instances T (x-2) , T (x-1) and T (x) are used as an input of an AI model 410. Based on an output of the AI model 410, the best beam in Set A is predicted for future time instances T (x+1) , T (x+2) and T(x+3) . In the example of FIG. 4, the same beam 420 is predicted as the best beam in Set A for all the future time instance T (x+1) , T (x+2) and T (x+3) , and the same power of the beam 420 of -75dBm is predicted for all the future time instance T (x+1) , T (x+2) and T(x+3) . Thus, the mobility state may be determined as low mobility or stationary.
[0112] For illustration, an example procedure may be described as below.
[0113] Upon applying the AI model for BM (deployed at UE side) , the UE shall:
[0114] 1> based on AI output (e.g., information about set A beam quality) determine its mobility state as following rules:
[0115] UE may derive the RSRP difference among the AI output related beam quality (i.e., Set A Beam) considering the best / top-K beam index and future time instance in sequence; And within one or multiple inference cycles:
[0116] - If the best / top-K beam index is not changed for all future time instance and all RSRP difference is below the RSRP_threshold, consider the current mobility state as low mobility or stationary;
[0117] - If the best / top-K beam index is not changed for all future time instance and any RSRP difference is above the RSRP_threshold, consider the current mobility state as non-low mobility;
[0118] - If best / top-K beam index is changed for any future time instance, consider the current mobility state as non-low mobility.
[0119] Continue to refer to FIG. 3, upon determination of the mobility state of the terminal device 110, the terminal device 110 causes 320 the mobility state to be used for activation or deactivation of the AI model.
[0120] With reference to FIG. 3, in some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may activate or deactivate 321 the AI model based on the mobility state. In some embodiments, the terminal device 110 may activate or deactivate the whole AI model. In some embodiments, the terminal device 110 may activate or deactivate a part of the AI model, e.g., model inference, model training, model monitoring, model updating or data collection.
[0121] In some embodiments, if the terminal device 110 is not in the low mobility state and the stationary state, the terminal device 110 may activate the AI model. If the terminal device 110 is in the low mobility state or the stationary state, the terminal device 110 may deactivate the AI model.
[0122] For illustration, an example procedure may be described as below.
[0123] 1> if UE determine its mobility state is low mobility or stationary:
[0124] 2> deactivate the AI model (could be whole or part of inference, training, monitoring, updating, data collection) (for a timer or multiple inference cycles) ;
[0125] 2> include deactivation cause –low mobility or stationary into the report message;
[0126] 2> report the information of AI model deactivation to network;
[0127] 1> else (i.e., UE is non-low mobility and non-stationary) :
[0128] 2> apply the configuration for the AI model (i.e., activate the AI model) ;
[0129] 2> report the information of AI model activation to network.
[0130] For illustration, another example procedure may be described as below.
[0131] 1> based on AI output (e.g., information about set A beam quality) determine whether to activate or deactivate the AI model as following rules:
[0132] UE may derive the RSRP difference among the AI output related beam quality (i.e., Set A Beam) considering the best / top-K beam index and future time instance order; And within one or multiple inference cycles:
[0133] - If the best / top-K beam index is not changed and all RSRP difference is below the RSRP_threshold, deactivate the AI model for inactivityTimer or multiple inference cycles;
[0134] - If the best / top-K beam index is not changed and any RSRP difference is above the RSRP_threshold, continue applying the AI model;
[0135] - If best / top-K beam index is changed, continue applying the AI model.
[0136] 1> if the AI model is deactivated:
[0137] 2> report the information of AI model deactivation to network.
[0138] Continue to refer to FIG. 3, in some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may report 322 the mobility state to the network device 120. In some embodiments, the terminal device 110 may transmit, in the reporting, a request for AI model activation or deactivation. In some embodiments, the terminal device 110 may transmit, in the reporting, a request for AI model activation or deactivation for a period of time.
[0139] As shown in FIG. 3, the network device 120 may transmit 323, to the terminal device 110, an indication of activation or deactivation of the AI model. Based on the indication, the terminal device 110 may activate or deactivate 324 of the AI model accordingly. That is, if the indication indicates the activation of the AI model, the terminal device 110 may activate the AI model. If the indication indicates the deactivation of the AI model, the terminal device 110 may deactivate the AI model.
[0140] In some embodiments, the terminal device 110 may deactivate the AI model for a period of time. In some embodiments, the period of time may be a predetermined number of inference cycles of the AI model. In some embodiments, if the AI model is deactivated, the terminal device 110 may start a timer based on the period of time. If the timer expires, the terminal device 110 may reactivate the AI model.
[0141] In some embodiments, the terminal device 110 may further transmit, to the network device 120, information of the activation or deactivation of the AI model. In some embodiments, the information of the activation or deactivation may comprise a cause of the activation or deactivation. In some embodiments, the cause may indicate the mobility state of the terminal device 110. In some embodiments, the information of the activation or deactivation may comprise an indication of the activation or deactivation. It is to be understood that the information of the activation or deactivation may comprise any other suitable information.
[0142] For illustration, an example procedure may be described as below.
[0143] 1> if UE determine its mobility state is low mobility or stationary:
[0144] 2> include the request for AI model deactivation into the report message;
[0145] 2> report the information of the current mobility state to network;
[0146] 1> else (i.e., UE is non-low mobility and non-stationary) :
[0147] 2> include the request for AI model activation into the report message;
[0148] 2> report the information of the current mobility state to network;
[0149] 1> if the indication for AI model deactivation is received:
[0150] 2> deactivate the AI model (could be whole or part of inference, training, monitoring, updating, data collection) (for a timer or multiple inference cycles) ;
[0151] 1> else (i.e., if the indication for AI model activation is received) :
[0152] 2> apply the configuration for the AI model (i.e., activate the AI model) .
[0153] Continue to refer to FIG. 3, in some embodiments where the AI model is deployed at the network device 120, the network device 120 may transmit 325, to the terminal device 110, a request for evaluation on the mobility state of the terminal device 110. In some embodiments, the request may comprise information of AI model deployment at the network device 120.
[0154] Based on the request, the terminal device 110 may report 326 the mobility state to the network device 120. The terminal device 110 may determine the mobility state as described in connection with step 310, and thus other details are omitted here for concise. In some embodiments, the terminal device 110 may transmit, in the reporting, a request for AI model activation or deactivation. In some embodiments, the terminal device 110 may transmit, in the reporting, a request for AI model activation or deactivation for a period of time.
[0155] Based on the mobility state reported by the terminal device 110, the network device 120 may activate or deactivate 327 the AI model. In some embodiments, the network device 120 may transmit 328, to the terminal device 110, an indication of the activation or deactivation of the AI model. In some embodiments, if an indication of the activation of the AI model is received from the network device 120, the terminal device 110 may start or restart to report, to the network device 120, measurement results (e.g., historical Set B beam measurements) as an input of the AI model. In some embodiments, the terminal device may report the measurement results for a period of time. In some embodiments, the period of time may be a predetermined number of inference cycles of the AI model. In some embodiments, if the indication of the activation is received, the terminal device 110 may start a timer based on the period of time. If the timer expires, the terminal device 110 may stop reporting the measurement results.
[0156] In some embodiments, if an indication of the deactivation of the AI model is received from the network device 120, the terminal device 110 may stop the reporting of the measurement results. In some embodiments, the terminal device may stop the reporting of the measurement results for a period of time. In some embodiments, the period of time may be a predetermined number of inference cycles of the AI model. In some embodiments, if the indication of the deactivation is received, the terminal device 110 may start a timer based on the period of time. If the timer expires, the terminal device 110 may restart to report the measurement results.
[0157] For illustration, an example procedure may be described as below.
[0158] When deployment of AI model for BM in network side, the UE shall:
[0159] 1> if the information of AI model deployment or the request for mobility state evaluation is received:
[0160] 2> evaluate the mobility state by at least one of the following rules:
[0161] - the legacy low mobility criterion;
[0162] - the legacy stationary criterion;
[0163] - the difference of signal strength between two measured RS within one period;
[0164] - the legacy mobility state for cell reselection;
[0165] - the number of handover times within one period;
[0166] - the property of low mobility or stationary.
[0167] 1> if UE determine its mobility state is low mobility or stationary:
[0168] 2> include the request for AI model deactivation into the report message;
[0169] 2> report the information of the current mobility state to network;
[0170] 1> else (i.e., UE is non-low mobility and non-stationary) :
[0171] 2> include the request for AI model activation into the report message;
[0172] 2> report the information of the current mobility state to network;
[0173] 1> if the information of AI model deactivation is received:
[0174] 2> stop reporting AI model related beam (e.g., historical Set B beam) (for a timer or multiple inference cycles) ;
[0175] 1> else (i.e., if the information of AI model activation is received) :
[0176] 2> start or restart to report AI model related beam (e.g., historical Set B beam) .
[0177] With the process 300, an AI model may be activated or deactivated based on a mobility state of a terminal device. In this way, unnecessary AI model usage may be avoided.
[0178] EXAMPLE IMPLEMENTATION OF NETWORK DETERMINATION OF MOBILITY STATE
[0179] FIG. 5 illustrates a schematic diagram illustrating another process 500 of communication according to embodiments of the present disclosure. For the purpose of discussion, the process 500 will be described with reference to FIG. 1. The process 500 may involve the terminal device 110 and the network device 120 as illustrated in FIG. 1. It is to be understood that the steps and the order of the steps in FIG. 5 are merely for illustration, and not for limitation. For example, the order of the steps may be changed. Some of the steps may be omitted or any other suitable additional steps may be added. In this example, an AI model based on a temporary prediction is deployed at the network device 120.
[0180] As shown in FIG. 5, the terminal device 110 transmits 510, to the network device 120, measurement results as an input of the AI model. In some embodiments, the network device 120 may transmit, to the terminal device 110, a configuration of measurements. The terminal device 110 may perform the measurements based on the configuration, and transmit results of the measurements to the network device 120.
[0181] The network device 120 determines 520 a mobility state of the terminal device 110 based on an output of the AI model by taking, as an input of the AI model, the measurement results received from the terminal device 110. For example, the network device 120 may determine the mobility state based on a variation of information of an output parameter of the AI model over future time instances. The determination 520 may be carried out in a similar way as that described for the determination 312. It is to be understood that the determination 520 is dependent on network implementation.
[0182] With reference to FIG. 5, the network device 120 transmits 530, to the terminal device 110, information of the mobility state of the terminal device 110. The terminal device 110 may use the mobility state as assistance information.
[0183] Continue to refer to FIG. 5, the network device 120 may transmit 540, to the terminal device 110, an indication of activation or deactivation of the AI model. In some embodiments, the network device 120 may transmit, to the terminal device 110, an indication of activation or deactivation of the AI model for a period of time. In some embodiments, the period of time is a predetermined number of inference cycles of the AI model. In some embodiments, if the AI model is deactivated, the terminal device 110 may start a timer based on the period of time. If the timer expires, the terminal device 110 may determine that the AI model is reactivated.
[0184] With the process 500, a network device may be facilitated to determine a mobility state of a terminal device for better network management of an AI model.
[0185] EXAMPLE IMPLEMENTATION OF REACTIVATION OF AI MODEL
[0186] In some scenarios, mobility based AI model activation or deactivation may have some problems, e.g., fast beam change but UE is still considered as a low mobility UE. In this case, even though the UE is in low mobility, the AI model should be activated instead of deactivation.
[0187] In view of this, embodiments of the present disclosure provide a solution for reactivating an AI model. The solution will be described in connection with FIG. 6.
[0188] FIG. 6 illustrates a schematic diagram illustrating still another process 600 of communication according to embodiments of the present disclosure. For the purpose of discussion, the process 600 will be described with reference to FIG. 1. The process 600 may involve the terminal device 110 and the network device 120 as illustrated in FIG. 1. It is to be understood that the steps and the order of the steps in FIG. 6 are merely for illustration, and not for limitation. For example, the order of the steps may be changed. Some of the steps may be omitted or any other suitable additional steps may be added. In this example, an AI model based on a temporary prediction is deployed at the terminal device 110 or the network device 120.
[0189] As shown in FIG. 6, the terminal device 110 determines 610 that the AI model is deactivated. In some embodiments, the terminal device 110 may determine the deactivation of the AI model based on a mobility state of the terminal device 110. In some embodiments, the terminal device 110 may receive an indication of the deactivation of the AI model. It is to be understood that any other suitable ways are also feasible.
[0190] Upon determination that the AI model is deactivated, the terminal device 110 determines 620 whether a criterion for reactivating the AI model is fulfilled.
[0191] In some embodiments, the criterion may be based on quality (e.g., block error rate (BLER) ) of a link between the terminal device 110 and the network device 120. In some embodiments, if the quality of the link is below threshold quality, the terminal device 110 may determine that the criterion is fulfilled.
[0192] In some embodiments, the criterion may be based on hybrid automatic repeat request (HARQ) feedback. In some embodiments, if the number of negative acknowledgement (NACK) transmissions is above a threshold number, the terminal device 110 may determine that the criterion is fulfilled.
[0193] In some embodiments, the criterion may be based on automatic repeat request (ARQ) feedback. In some embodiments, if the number of radio link control (RLC) retransmissions is above a further threshold number, the terminal device 110 may determine that the criterion is fulfilled.
[0194] In some embodiments, the criterion may be based on a beam failure detection (BFD) procedure. In some embodiments, if a beam link problem is detected, the terminal device 110 may determine that the criterion is fulfilled. In some embodiments, if a beam failure instance is received, the terminal device 110 may determine that the beam link problem is detected.
[0195] In some embodiments, the criterion may be based on a radio link measurement (RLM) . In some embodiments, if a radio link problem is detected, the terminal device 110 may determine that the criterion is fulfilled. In some embodiments, if a predetermined number of consecutive out-of-synchronization indications are received, the terminal device 110 may determine that the radio link problem is detected.
[0196] It is to be understood that the criterion may be based on any combination of the above information or any other suitable information.
[0197] Continue to refer to FIG. 6, upon determination that the criterion is fulfilled, the terminal device 110 causes 630 the AI model to be reactivated. In some embodiments, the terminal device 110 may cause the AI model to be reactivated for a period of time. In some embodiments, the period of time may be a predetermined number of inference cycles of the AI model. In some embodiments, the period of time may be a value of a timer.
[0198] In some embodiments where the AI model is deployed at the terminal device 110, upon determination that the criterion is fulfilled, the terminal device 110 may reactivate 631 the AI model.
[0199] In some embodiments where the AI model is deployed at the network device 120, the terminal device 110 may transmit 632, to the network device 120, a notification for reactivation of the AI model. The network device 120 may reactivate 633 the AI model based on the notification.
[0200] In some embodiments, the terminal device 110 may transmit the notification for reactivation of the AI model for a period of time. In some embodiments, the period of time may be a predetermined number of inference cycles of the AI model. In some embodiments, the period of time may be a value of a timer.
[0201] For example, if UE detects its radio link quality is worse than a threshold of BLER (e.g., > 10%) , the UE may re-activate the AI model regardless of mobility state or inform the network (NW) of re-activation (e.g., activate for a timer or multiple inference cycles) .
[0202] In another example, if the number of NACK transmission is above than a threshold (e.g., > 10) during the time when AI model is deactivated, UE should re-activate the AI model regardless of mobility state or inform the NW of re-activation (e.g., activate for a timer or multiple inference cycles) .
[0203] In another example, if the number of RLC retransmission is above than a threshold (e.g., > 10) during the time when AI model is deactivated, UE should re-activate the AI model regardless of mobility state or inform the NW of re-activation (e.g., activate for a timer or multiple inference cycles) .
[0204] In another example, if UE detect beam link problem (e.g., a beam failure instance is received) , UE should re-activate the AI model regardless of mobility state or inform the NW of re-activation (e.g., activate for a timer or multiple inference cycles) .
[0205] In another example, if UE detect the radio link problem (e.g., receive N310 consecutive “out-of-sync” or T310 is started) , UE should re-activate the AI model regardless of mobility state or inform the NW of re-activation (e.g., activate for a timer or multiple inference cycles) .
[0206] With the process 600, robust for the usage of an AI model may be enhanced. It is to be understood that the processes 300, 500 and 600 may be performed separately or in any suitable combination.
[0207] EXAMPLE IMPLEMENTATION OF METHODS
[0208] Corresponding to the above processes, embodiments of the present disclosure provide methods of communication implemented at a terminal device. These methods will be described below with reference to FIGs. 7 to 9.
[0209] FIG. 7 illustrates an example method 700 of communication implemented at a terminal device in accordance with some embodiments of the present disclosure. For example, the method 700 may be performed at the terminal device 110 as shown in FIG. 1. For the purpose of discussion, in the following, the method 700 will be described with reference to FIG. 1. It is to be understood that the method 700 may include additional blocks not shown and / or may omit some blocks as shown, and the scope of the present disclosure is not limited in this regard.
[0210] At block 710, the terminal device 110 determines a mobility state of the terminal device.
[0211] In some embodiments, if the terminal device 110 fulfills a relaxed measurement criterion for low mobility, the terminal device 110 may determine that the terminal device 110 is in the low mobility state. In some embodiments, if the terminal device 110 fulfills a relaxed measurement criterion for stationary, the terminal device 110 may determine that the terminal device is in the stationary state.
[0212] In some embodiments, if a difference of signal strength between RS measurements within a first period of time is below a threshold difference, the terminal device 110 may determine that the terminal device 110 is in the low mobility state or the stationary state.
[0213] In some embodiments, if number of cell reselections during a second period of time is above a first threshold number, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state.
[0214] In some embodiments, if number of handover times within a third period of time is above a second threshold number, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state.
[0215] In some embodiments, if a property of the terminal device indicates low mobility or stationary, the terminal device 110 may determine that the terminal device 110 is in the low mobility state or the stationary state. In some embodiments, the terminal device 110 may determine the property of the terminal device 110 based on at least one of a type of the terminal device 110, a type of a service associated with the terminal device 110, or time period information. In some embodiments, the terminal device 110 may receive, from the network device 120, an indication of the property of the terminal device 110.
[0216] In some embodiments, if a distribution variation of an input parameter of the artificial intelligence model within a fourth period of time is below a threshold distribution variation, the terminal device 110 may determine that the terminal device is in the low mobility state or the stationary state.
[0217] In some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may determine an output of the AI model, the output comprising information of an output parameter in future time instances. The terminal device 110 may determine the mobility state of the terminal device 110 based on a variation of the information over the future time instances.
[0218] In some embodiments, the information of the output parameter may comprise indexes and values of the output parameter in the future time instances. In some embodiments, if indexes of the output parameter in the future time instances are unchanged and differences of values of the output parameter in the future time instances are below a threshold difference, the terminal device 110 may determine that the terminal device is in the low mobility state or the stationary state. In some embodiments, if the indexes of the output parameter in the future time instances are unchanged and the differences of values of the output parameter in the future time instances are above the threshold difference, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state. In some embodiments, if the indexes of the output parameter are changed for the future time instances, the terminal device 110 may determine that the terminal device 110 is not in the low mobility state and the stationary state.
[0219] At block 720, the terminal device 110 causes the mobility state to be used for activation or deactivation of an AI model. In some embodiments, the AI model may be based on a temporary prediction. It is to be understood that any other suitable AI models are also feasible and the present disclosure does not limit a function of the AI model.
[0220] In some embodiments where the AI model is deployed at the terminal device 110, if the terminal device 110 is in a low mobility state or a stationary state, the terminal device 110 may deactivate the AI model. If the terminal device 110 is not in the low mobility state and the stationary state, the terminal device 110 may activate the AI model.
[0221] In some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may report the mobility state of the terminal device 110 to the network device 120, and receive an indication of activation or deactivation of the AI model from the network device 120. If the indication of the activation of the artificial intelligence model is received from the network device 120, the terminal device 110 may activate the AI model. If the indication of the deactivation of the AI model is received from the network device 120, the terminal device 110 may deactivate the AI model.
[0222] In some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may deactivate the AI model for a period of time. In some embodiments, the period of time is a predetermined number of inference cycles of the AI model. In some embodiments, if the AI model is deactivated, the terminal device 110 may start a timer based on the period of time. If the timer expires, the terminal device 110 may reactivate the AI model.
[0223] In some embodiments where the AI model is deployed at the terminal device 110, the terminal device 110 may transmit, to the network device 120, information of the activation or deactivation of the AI model. In some embodiments, the information of the activation or deactivation may comprise at least one of the following: an indication of the activation or deactivation; or a cause of the activation or deactivation comprising the mobility state of the terminal device 110.
[0224] In some embodiments where the AI model is deployed at the network device 120, the terminal device 110 may report the mobility state of the terminal device 110 to the network device 120. If an indication of the activation of the AI model is received from the network device 120, the terminal device 110 may report, to the network device 120, measurement results as an input of the AI model. If an indication of the deactivation of the AI model is received from the network device 120, the terminal device 110 may stop the reporting of the measurement results.
[0225] With the method 700, an AI model may be activated or deactivated based on a mobility state of a terminal device. Thus, unnecessary AI model usage may be avoided.
[0226] FIG. 8 illustrates another example method 800 of communication implemented at a terminal device in accordance with some embodiments of the present disclosure. For example, the method 800 may be performed at the terminal device 110 as shown in FIG. 1. For the purpose of discussion, in the following, the method 800 will be described with reference to FIG. 1. It is to be understood that the method 800 may include additional blocks not shown and / or may omit some blocks as shown, and the scope of the present disclosure is not limited in this regard. In this example, an AI model is deployed at the network device 120.
[0227] At block 810, the terminal device 110 transmits, to the network device 120, measurement results as an input of the AI model. In some embodiments, the AI model may be based on a temporary prediction. It is to be understood that any other suitable AI models are also feasible and the present disclosure does not limit a function of the AI model.
[0228] At block 820, the terminal device 110 receives, from the network device 120, information of a mobility state of the terminal device 110. The mobility state is determined by the network device 120 based on a variation of information of an output parameter of the AI model over future time instances.
[0229] In some embodiments, the terminal device 110 may receive, from the network device 120, an indication of activation or deactivation of the AI model. In some embodiments the terminal device 110 may receive, from the network device 120, an indication of activation or deactivation of the AI model for a period of time. In some embodiments, the period of time is a predetermined number of inference cycles of the AI model. In some embodiments, if the AI model is deactivated, the terminal device 110 may start a timer based on the period of time. If the timer expires, the terminal device 110 may determine that the AI model is reactivated.
[0230] With the method 800, a network device may be facilitated to determine a mobility state of a terminal device for better network management of an AI model.
[0231] FIG. 9 illustrates still another example method 900 of communication implemented at a terminal device in accordance with some embodiments of the present disclosure. For example, the method 900 may be performed at the terminal device 110 as shown in FIG. 1. For the purpose of discussion, in the following, the method 900 will be described with reference to FIG. 1. It is to be understood that the method 900 may include additional blocks not shown and / or may omit some blocks as shown, and the scope of the present disclosure is not limited in this regard. In this example, an AI model is deployed at the terminal device 110 or the network device 120.
[0232] At block 910, the terminal device 110 determines that an AI model is deactivated. In some embodiments, the AI model may be based on a temporary prediction. It is to be understood that any other suitable AI models are also feasible and the present disclosure does not limit a function of the AI model.
[0233] At block 920, the terminal device 110 determines whether a criterion for reactivating the AI model is fulfilled. In some embodiments, the criterion may comprise at least one of the following: quality of a link between the terminal device 110 and the network device 120 is below threshold quality; number of NACK transmissions within a period of time since the deactivating of the AI model is above a threshold number; number of RLC retransmissions within a period of time since the deactivating of the AI model is above a further threshold number; a beam link problem is detected; or a radio link problem is detected.
[0234] Upon determination that the criterion is fulfilled, the method 900 proceeds to block 930. At block 930, the terminal device 110 causes the AI model to be reactivated. In some embodiments, the terminal device 110 may cause the AI model to be reactivated for a period of time.
[0235] In some embodiments where the AI model is deployed at the network device 120, the terminal device 110 may transmit, to the network device 120, a notification for reactivation of the AI model. In some embodiments, the terminal device 110 may transmit the notification for reactivation of the AI model for a period of time.
[0236] In some embodiments, the period of time may be a predetermined number of inference cycles of the AI model. In some embodiments, the period of time may be a value of a timer.
[0237] With the method 900, robust for the usage of an AI model may be enhanced.
[0238] It is to be understood that operations of the methods 700 to 900 correspond to that described in connection with FIGs. 3 to 6, and thus other details are not repeated here for concise.
[0239] EXAMPLE IMPLEMENTATION OF DEVICES
[0240] FIG. 10 is a simplified block diagram of a device 1000 that is suitable for implementing embodiments of the present disclosure. The device 1000 can be considered as a further example implementation of the terminal device 110 or the network device 120 as shown in FIG. 1. Accordingly, the device 1000 can be implemented at or as at least a part of the terminal device 110 or the network device 120.
[0241] As shown, the device 1000 includes a processor 1010, a memory 1020 coupled to the processor 1010, a suitable transceiver 1040 coupled to the processor 1010, and a communication interface coupled to the transceiver 1040. The memory 1010 stores at least a part of a program 1030. The transceiver 1040 may be for bidirectional communications or a unidirectional communication based on requirements. The transceiver 1040 may include at least one of a transmitter 1042 or a receiver 1044. The transmitter 1042 and the receiver 1044 may be functional modules or physical entities. The transceiver 1040 has at least one antenna to facilitate communication, though in practice an Access Node mentioned in this application may have several ones. The communication interface may represent any interface that is necessary for communication with other network elements, such as X2 / Xn interface for bidirectional communications between eNBs / gNBs, S1 / NG interface for communication between a Mobility Management Entity (MME) / Access and Mobility Management Function (AMF) / SGW / UPF and the eNB / gNB, Un interface for communication between the eNB / gNB and a relay node (RN) , or Uu interface for communication between the eNB / gNB and a terminal device.
[0242] The program 1030 is assumed to include program instructions that, when executed by the associated processor 1010, enable the device 1000 to operate in accordance with the embodiments of the present disclosure, as discussed herein with reference to FIGs. 1 to 9. The embodiments herein may be implemented by computer software executable by the processor 1010 of the device 1000, or by hardware, or by a combination of software and hardware. The processor 1010 may be configured to implement various embodiments of the present disclosure. Furthermore, a combination of the processor 1010 and memory 1020 may form processing means 1050 adapted to implement various embodiments of the present disclosure.
[0243] The memory 1020 may be of any type suitable to the local technical network and may be implemented using any suitable data storage technology, such as a non-transitory computer readable storage medium, semiconductor based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory and removable memory, as non-limiting examples. While only one memory 1020 is shown in the device 1000, there may be several physically distinct memory modules in the device 1000. The processor 1010 may be of any type suitable to the local technical network, and may include one or more of general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs) and processors based on multicore processor architecture, as non-limiting examples. The device 1000 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.
[0244] In some embodiments, a terminal device comprises a circuitry configured to: determine a mobility state of the terminal device; and cause the mobility state to be used for activation or deactivation of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction.
[0245] In some embodiments, a terminal device comprises a circuitry configured to: transmit, to a network device, measurement results as an input of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction; and receive, from the network device, a mobility state of the terminal device, the mobility state being determined based on a variation of information of an output parameter of the artificial intelligence model over future time instances.
[0246] In some embodiments, a terminal device comprises a circuitry configured to: determine that an artificial intelligence model is deactivated, the artificial intelligence model being based on a temporary prediction; and in accordance with a determination that a criterion for reactivating the artificial intelligence model is fulfilled, cause the artificial intelligence model to be reactivated.
[0247] The term “circuitry” used herein may refer to hardware circuits and / or combinations of hardware circuits and software. For example, the circuitry may be a combination of analog and / or digital hardware circuits with software / firmware. As a further example, the circuitry may be any portions of hardware processors with software including digital signal processor (s) , software, and memory (ies) that work together to cause an apparatus, such as a terminal device or a network device, to perform various functions. In a still further example, the circuitry may be hardware circuits and or processors, such as a microprocessor or a portion of a microprocessor, that requires software / firmware for operation, but the software may not be present when it is not needed for operation. As used herein, the term circuitry also covers an implementation of merely a hardware circuit or processor (s) or a portion of a hardware circuit or processor (s) and its (or their) accompanying software and / or firmware.
[0248] In summary, embodiments of the present disclosure provide the following solutions.
[0249] In one solution, a terminal device comprises a processor configured to cause the terminal device to: determine a mobility state of the terminal device; and cause the mobility state to be used for activation or deactivation of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction.
[0250] In some embodiments, the terminal device is caused to cause the mobility state to be used for the activation or deactivation of the artificial intelligence model by: in accordance with a determination that the terminal device is in a low mobility state or a stationary state, deactivating the artificial intelligence model; and in accordance with a determination that the terminal device is not in the low mobility state and the stationary state, activating the artificial intelligence model.
[0251] In some embodiments, the terminal device is caused to cause the mobility state to be used for the activation or deactivation of the artificial intelligence model by: reporting the mobility state of the terminal device to a network device; in accordance with a determination that an indication of the activation of the artificial intelligence model is received from the network device, activating the artificial intelligence model; and in accordance with a determination that an indication of the deactivation of the artificial intelligence model is received from the network device, deactivating the artificial intelligence model.
[0252] In some embodiments, the terminal device is caused to deactivate the artificial intelligence model by: deactivating the artificial intelligence model for a period of time.
[0253] In some embodiments, the period of time is a predetermined number of inference cycles of the artificial intelligence model.
[0254] In some embodiments, the terminal device is further caused to: in accordance with a determination that the artificial intelligence model is deactivated, start a timer based on the period of time; and in accordance with a determination that the timer expires, reactivate the artificial intelligence model.
[0255] In some embodiments, the terminal device is further caused to: transmit, to a network device, information of the activation or deactivation of the artificial intelligence model.
[0256] In some embodiments, the information of the activation or deactivation comprises at least one of the following: an indication of the activation or deactivation; or a cause of the activation or deactivation comprising the mobility state of the terminal device.
[0257] In some embodiments, the terminal device is caused to cause the mobility state to be used for the activation or deactivation of the artificial intelligence model by: reporting the mobility state of the terminal device to a network device; in accordance with a determination that an indication of the activation of the artificial intelligence model is received from the network device, reporting, to the network device, measurement results as an input of the artificial intelligence model; and in accordance with a determination that an indication of the deactivation of the artificial intelligence model is received from the network device, stopping the reporting of the measurement results.
[0258] In some embodiments, the terminal device is caused to determine the mobility state by at least one of the following: in accordance with a determination that the terminal device fulfills a relaxed measurement criterion for low mobility, determining that the terminal device is in the low mobility state; in accordance with a determination that the terminal device fulfills a relaxed measurement criterion for stationary, determining that the terminal device is in the stationary state; in accordance with a determination that a difference of signal strength between reference signal measurements within a first period of time is below a threshold difference, determining that the terminal device is in the low mobility state or the stationary state; in accordance with a determination that number of cell reselections during a second period of time is above a first threshold number, determining that the terminal device is not in the low mobility state and the stationary state; in accordance with a determination that number of handover times within a third period of time is above a second threshold number, determining that the terminal device is not in the low mobility state and the stationary state; in accordance with a determination that a property of the terminal device indicates low mobility or stationary, determining that the terminal device is in the low mobility state or the stationary state; or in accordance with a determination that a distribution variation of an input parameter of the artificial intelligence model within a fourth period of time is below a threshold distribution variation, determining that the terminal device is in the low mobility state or the stationary state.
[0259] In some embodiments, the terminal device is further caused to determine the property of the terminal device by at least one of the following: determining the property of the terminal device based on at least one of a type of the terminal device, a type of a service associated with the terminal device, or time period information; or receiving, from a network device, an indication of the property of the terminal device.
[0260] In some embodiments, the terminal device is caused to determine the mobility state by: determining an output of the artificial intelligence model, the output comprising information of an output parameter in future time instances; and determining the mobility state of the terminal device based on a variation of the information over the future time instances.
[0261] In some embodiments, the information of the output parameter comprises indexes and values of the output parameter in the future time instances, and the terminal device is caused to determine the mobility state by at least one of the following: in accordance with a determination that indexes of the output parameter in the future time instances are unchanged and differences of values of the output parameter in the future time instances are below a threshold difference, determining that the terminal device is in the low mobility state or the stationary state; in accordance with a determination that the indexes of the output parameter in the future time instances are unchanged and the differences of values of the output parameter in the future time instances are above the threshold difference, determining that the terminal device is not in the low mobility state and the stationary state; or in accordance with a determination that the indexes of the output parameter are changed for the future time instances, determining that the terminal device is not in the low mobility state and the stationary state.
[0262] In another solution, a terminal device comprises a processor configured to cause the terminal device to: transmit, to a network device, measurement results as an input of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction; and receive, from the network device, information of a mobility state of the terminal device, the mobility state being determined based on a variation of information of an output parameter of the artificial intelligence model over future time instances.
[0263] In some embodiments, the terminal device is further caused to: receive, from the network device, an indication of activation or deactivation of the artificial intelligence model; or receive, from the network device, an indication of activation or deactivation of the artificial intelligence model for a period of time.
[0264] In some embodiments, the period of time is a predetermined number of inference cycles of the artificial intelligence model.
[0265] In some embodiments, the terminal device is further caused to: in accordance with a determination that the artificial intelligence model is deactivated, start a timer based on the period of time; and in accordance with a determination that the timer expires, determine that the artificial intelligence model is reactivated.
[0266] In another solution, a terminal device comprises a processor configured to cause the terminal device to: determine that an artificial intelligence model is deactivated, the artificial intelligence model being based on a temporary prediction; and in accordance with a determination that a criterion for reactivating the artificial intelligence model is fulfilled, cause the artificial intelligence model to be reactivated.
[0267] In some embodiments, the criterion comprises at least one of the following: quality of a link between the terminal device and a network device is below threshold quality; number of negative acknowledgment transmissions within a period of time since the deactivating of the artificial intelligence model is above a threshold number; number of radio link control retransmissions within a period of time since the deactivating of the artificial intelligence model is above a further threshold number; a beam link problem is detected; or a radio link problem is detected.
[0268] In some embodiments, the terminal device is caused to cause the artificial intelligence model to be reactivated by: causing the artificial intelligence model to be reactivated for a period of time.
[0269] In some embodiments, the terminal device is caused to cause the artificial intelligence model to be reactivated by: transmitting, to a network device, a notification for reactivation of the artificial intelligence model.
[0270] In some embodiments, the terminal device is caused to transmit the notification by: transmitting the notification for reactivation of the artificial intelligence model for a period of time.
[0271] In some embodiments, the period of time is a predetermined number of inference cycles of the artificial intelligence model or a value of a timer.
[0272] In another solution, a method of communication comprises: determining, at a terminal device, a mobility state of the terminal device; and causing the mobility state to be used for activation or deactivation of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction.
[0273] In another solution, a method of communication comprises: transmitting, at a terminal device and to a network device, measurement results as an input of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction; and receiving, from the network device, a mobility state of the terminal device, the mobility state being determined based on a variation of information of an output parameter of the artificial intelligence model over future time instances.
[0274] In another solution, a method of communication comprises: determining, at a terminal device, that an artificial intelligence model is deactivated, the artificial intelligence model being based on a temporary prediction; and in accordance with a determination that a criterion for reactivating the artificial intelligence model is fulfilled, causing the artificial intelligence model to be reactivated.
[0275] 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 representation, it will be appreciated that the blocks, apparatus, systems, techniques or methods 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.
[0276] 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 computer-executable instructions, such as those included in program modules, being executed in a device on a target real or virtual processor, to carry out the process or method as described above with reference to FIGs. 1 to 9. 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.
[0277] 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 partly on a remote machine or entirely on the remote machine or server.
[0278] The above program code may be embodied on a machine readable medium, which may be any tangible medium that may contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine readable medium may be a machine readable signal medium or a machine readable storage medium. A machine 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 machine 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.
[0279] 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.
[0280] Although the present disclosure has been described in language 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
1.A terminal device, comprising:a processor configured to cause the terminal device to:determine a mobility state of the terminal device; andcause the mobility state to be used for activation or deactivation of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction.2.The terminal device of claim 1, wherein the terminal device is caused to cause the mobility state to be used for the activation or deactivation of the artificial intelligence model by:in accordance with a determination that the terminal device is in a low mobility state or a stationary state, deactivating the artificial intelligence model; andin accordance with a determination that the terminal device is not in the low mobility state and the stationary state, activating the artificial intelligence model.3.The terminal device of claim 1, wherein the terminal device is caused to cause the mobility state to be used for the activation or deactivation of the artificial intelligence model by:reporting the mobility state of the terminal device to a network device;in accordance with a determination that an indication of the activation of the artificial intelligence model is received from the network device, activating the artificial intelligence model; andin accordance with a determination that an indication of the deactivation of the artificial intelligence model is received from the network device, deactivating the artificial intelligence model.4.The terminal device of claim 2 or 3, wherein the terminal device is caused to deactivate the artificial intelligence model by:deactivating the artificial intelligence model for a period of time.5.The terminal device of claim 4, wherein the period of time is a predetermined number of inference cycles of the artificial intelligence model.6.The terminal device of claim 4, wherein the terminal device is further caused to:in accordance with a determination that the artificial intelligence model is deactivated, start a timer based on the period of time; andin accordance with a determination that the timer expires, reactivate the artificial intelligence model.7.The terminal device of claim 2 or 3, wherein the terminal device is further caused to:transmit, to a network device, information of the activation or deactivation of the artificial intelligence model.8.The terminal device of claim 7, wherein the information of the activation or deactivation comprises at least one of the following:an indication of the activation or deactivation; ora cause of the activation or deactivation comprising the mobility state of the terminal device.9.The terminal device of claim 1, wherein the terminal device is caused to cause the mobility state to be used for the activation or deactivation of the artificial intelligence model by:reporting the mobility state of the terminal device to a network device;in accordance with a determination that an indication of the activation of the artificial intelligence model is received from the network device, reporting, to the network device, measurement results as an input of the artificial intelligence model; andin accordance with a determination that an indication of the deactivation of the artificial intelligence model is received from the network device, stopping the reporting of the measurement results.10.The terminal device of claim 1, wherein the terminal device is caused to determine the mobility state by at least one of the following:in accordance with a determination that the terminal device fulfills a relaxed measurement criterion for low mobility, determining that the terminal device is in the low mobility state;in accordance with a determination that the terminal device fulfills a relaxed measurement criterion for stationary, determining that the terminal device is in the stationary state;in accordance with a determination that a difference of signal strength between reference signal measurements within a first period of time is below a threshold difference, determining that the terminal device is in the low mobility state or the stationary state;in accordance with a determination that number of cell reselections during a second period of time is above a first threshold number, determining that the terminal device is not in the low mobility state and the stationary state;in accordance with a determination that number of handover times within a third period of time is above a second threshold number, determining that the terminal device is not in the low mobility state and the stationary state;in accordance with a determination that a property of the terminal device indicates low mobility or stationary, determining that the terminal device is in the low mobility state or the stationary state; orin accordance with a determination that a distribution variation of an input parameter of the artificial intelligence model within a fourth period of time is below a threshold distribution variation, determining that the terminal device is in the low mobility state or the stationary state.11.The terminal device of claim 10, wherein the terminal device is further caused to determine the property of the terminal device by at least one of the following:determining the property of the terminal device based on at least one of a type of the terminal device, a type of a service associated with the terminal device, or time period information; orreceiving, from a network device, an indication of the property of the terminal device.12.The terminal device of claim 1, wherein the terminal device is caused to determine the mobility state by:determining an output of the artificial intelligence model, the output comprising information of an output parameter in future time instances; anddetermining the mobility state of the terminal device based on a variation of the information over the future time instances.13.The terminal device of claim 12, wherein the information of the output parameter comprises indexes and values of the output parameter in the future time instances, and wherein the terminal device is caused to determine the mobility state by at least one of the following:in accordance with a determination that indexes of the output parameter in the future time instances are unchanged and differences of values of the output parameter in the future time instances are below a threshold difference, determining that the terminal device is in the low mobility state or the stationary state;in accordance with a determination that the indexes of the output parameter in the future time instances are unchanged and the differences of values of the output parameter in the future time instances are above the threshold difference, determining that the terminal device is not in the low mobility state and the stationary state; orin accordance with a determination that the indexes of the output parameter are changed for the future time instances, determining that the terminal device is not in the low mobility state and the stationary state.14.A terminal device, comprising:a processor configured to cause the terminal device to:transmit, to a network device, measurement results as an input of an artificial intelligence model, the artificial intelligence model being based on a temporary prediction; andreceive, from the network device, information of a mobility state of the terminal device, the mobility state being determined based on a variation of information of an output parameter of the artificial intelligence model over future time instances.15.The terminal device of claim 14, wherein the terminal device is further caused to:receive, from the network device, an indication of activation or deactivation of the artificial intelligence model; orreceive, from the network device, an indication of activation or deactivation of the artificial intelligence model for a period of time.16.The terminal device of claim 15, wherein the period of time is a predetermined number of inference cycles of the artificial intelligence model.17.A terminal device, comprising:a processor configured to cause the terminal device to:determine that an artificial intelligence model is deactivated, the artificial intelligence model being based on a temporary prediction; andin accordance with a determination that a criterion for reactivating the artificial intelligence model is fulfilled, cause the artificial intelligence model to be reactivated.18.The terminal device of claim 17, wherein the criterion comprises at least one of the following:quality of a link between the terminal device and a network device is below threshold quality;number of negative acknowledgment transmissions within a period of time since the deactivating of the artificial intelligence model is above a threshold number;number of radio link control retransmissions within a period of time since the deactivating of the artificial intelligence model is above a further threshold number;a beam link problem is detected; ora radio link problem is detected.19.The terminal device of claim 17, wherein the terminal device is caused to cause the artificial intelligence model to be reactivated by:causing the artificial intelligence model to be reactivated for a period of time.20.The terminal device of claim 17, wherein the terminal device is caused to cause the artificial intelligence model to be reactivated by:transmitting, to a network device, a notification for reactivation of the artificial intelligence model.