Channel state information processing for model inference

Abstract

There is provided an apparatus, method, and computer program for causing an apparatus to perform: determining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus; determining, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; and providing the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

Description

DESCRIPTIONCHANNEL STATE INFORMATION PROCESSING FOR MODEL INFERENCEFIELD

[0001] Various example embodiments of the present disclosure generally relate to the field of telecommunication and in particular, to methods, devices, apparatuses and computer programs for channel state information (CSI) processing unit (CPU) reports for model inference.

[0002] BACKGROUND

[0003] A communication network may serve as a facility that enables communications between two or more communication devices or provides communication devices access to a data network. A mobile or wireless communication network is one example of a communication network. A communication device may be provided with a service by an application server.

[0004] As communication networks and services increase in size, complexity and number of users, operations in the communication networks become increasingly complicated. To improve the communication performance, machine learning (ML) / artificial intelligence (Al) technology is proposed to be used in the wireless communication networks. The AI / ML model may be applied to different communication functionality in different scenarios, including but not limited to, beam management (BM), CSI compression, positioning, and the like.

[0005] SUMMARY

[0006] According to a first aspect, there is provided an apparatus comprising means for performing: determining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus; determining, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; and providing the report ion the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

[0007] According to a second aspect, there is provided an apparatus comprising: at least one processor; and at least one memory comprising code that, when executed by the at least one processor, causes the apparatus to perform: determining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus; determining, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; and providing the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

[0008] According to a third aspect, there is provided a method for an apparatus, the method comprising: determining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus; determining, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; and providing the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

[0009] According to a fourth aspect, there is provided an apparatus comprising: determining circuitry for determining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus; determining circuitry for determining, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; and providing circuitry for providing the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

[0010] The following may apply in respect of any (e.g., one or more, including all) of the above first to fourth aspects.

[0011] The apparatus may be caused to perform: determining a second number of second CPUs for performing simultaneous machine learning-related CSI calculations for one or more component carriers supported by the apparatus, wherein the determining whether toprovide the report on the machine learning related CSI calculations to the access node may further comprise determining whether to provide the report on the machine learning related CSI calculations to the access node based additionally on the second number.

[0012] The apparatus may be caused to perform: determining at a first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed based on the second number of second CPUs.

[0013] The determining at the first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed using one or more of the second number of second CPUs may comprise: determining an occupancy of the second number of second CPUs by determining how many of the second number of second CPUs are unallocated for use at the first point of time; and determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed based on the occupancy of the second number of second CPUs.

[0014] The determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs may comprise: determining a first priority associated with the report on the machine learning related CSI calculations based on at least one of a configuration for transmitting the report or a configuration for performing the one or more measurements for the report; and determining at least one second priority associated with one or more reports for other CSI calculations that are to be performed at the first point of time based on the second number of second CPUs; wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs may be performed based on the first and second priorities.

[0015] The first priority may be associated with a predefined number of CPUs to be used for CSI calculations for a corresponding CSI report, and the at least one second priority may be associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

[0016] The apparatus may be caused to perform: determining at a second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs, wherein when it is determined that one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs, the determining whether to provide the report on the machinelearning related CSI calculations to the access node may comprise determining to provide the report on the machine learning related CSI calculations to the access node.

[0017] The first and second point of time may be the same time.

[0018] The first and second point of time may be different times.

[0019] The determining at the second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed using one or more of the first number of first CPUs may comprise: determining an occupancy of the first number of first CPUs by determining how many of the first number of first CPUs are unallocated for use at the first point of time; and determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs based on the occupancy of the first number of first CPUs.

[0020] The determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs may comprise: determining a third priority associated with the report on the machine learning related CSI calculations; and determining at least one fourth priority associated with reports for other CSI calculations that are to be performed at the second point of time using the first number of first CPUs, wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs may be performed based on the first and second priorities.

[0021] The third priority may be associated with a predefined number of CPUs to be used for CSI calculations for corresponding CSI report, and the at least one fourth priority may be associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

[0022] The apparatus may be caused to perform: providing the first number to the access node.

[0023] According to a fifth aspect, there is provided an apparatus comprising means for performing: obtaining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus; determining, based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving one or more CSI reports based on the determining whether the report will be received.

[0024] According to a sixth aspect, there is provided an apparatus c comprising: at least one processor; and at least one memory comprising code that, when executed by the at least one processor, causes the apparatus to perform: obtaining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus; determining, based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving one or more CSI reports based on the determining whether the report will be received.

[0025] According to a seventh aspect, there is provided a method for an apparatus, the method comprising: obtaining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus; determining, based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving one or more CSI reports based on the determining whether the report will be received.

[0026] According to an eighth aspect, there is provided an apparatus comprising: obtaining circuitry for obtaining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus; determining circuitry for determining, based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving circuitry for receiving one or more CSI reports based on the determining whether the report will be received.

[0027] The following may apply in respect of any (e.g., one or more, including all) of the above fifth to eighth aspects.

[0028] The receiving one or more CSI reports may comprise using a result of the determining whether the report will be received for decoding received transmissions.

[0029] The apparatus may be caused to perform: scheduling one or more CSI resources based on the determining whether the report will be received.

[0030] The is provided a computer readable medium. The computer readable medium comprises instructions stored thereon for causing an apparatus to perform at least the method according to the third aspect and / or the seventh aspect.

[0031] It is to be understood that the Summary section is not intended to identify key or essential features of embodiments of the present disclosure, nor is it intended to be used to limit the scope of the present disclosure. Other features of the present disclosure will become easily comprehensible through the following description.

[0032] BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Some example embodiments will now be described with reference to the accompanying drawings, where:

[0034] Figure 1 illustrates an example communication environment in which example embodiments of the present disclosure can be implemented;

[0035] Figures 2 to 3 illustrate example methods that may be performed by apparatus described herein;

[0036] Figures 4 to 5 illustrate example methods that may be performed by apparatus described herein; and

[0037] Figures 6 to 7 illustrate an example apparatus for performing methods described herein.

[0038] Throughout the drawings, the same or similar reference numerals represent the same or similar element.

[0039] DETAILED DESCRIPTION

[0040] In general, the following application relates to the use of AI / ML models for performing CSI reporting.

[0041] In particular, the following relates to situations in which there is a pool of processing resources for CSI reporting that is dedicated for performing AI / ML-based calculations for obtaining one or more inferences for the CSI report. This pool of processing resources for CSI resources is also labelled as “CPU” herein. The present application considers how these AI / ML dedicated CPU may be efficiently utilised, especially when other CPU are provided for non-AI / ML-based CSI reporting.

[0042] Principles of the present disclosure will now be described with reference to some examples. It is to be understood that these embodiments are described only for the purpose of illustration and help those skilled in the art to understand and implement the present disclosure, without suggesting any limitation as to the scope of the disclosure.Embodiments described herein can be implemented in various manners other than the ones described below.

[0043] 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.

[0044] References in the present disclosure to “one embodiment,” “an embodiment,” “an example embodiment,” and the like indicate that the embodiment described may include a particular feature, structure, or characteristic, but it is not necessary that every embodiment includes the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.

[0045] It shall be understood that although the terms “first,” “second,”..., etc. in front of noun(s) and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another and they do not limit the order of the noun(s). For example, a first element could be termed a second element, and similarly, a second element could be termed a first element, without departing from the scope of example embodiments. As used herein, the term “and / or” includes any and all combinations of one or more of the listed terms.

[0046] As used herein, “at least one of the following: ” and “at least one of ” and similar wording, where the list of two or more elements are joined by “and” or “or”, mean at least any one of the elements, or at least any two or more of the elements, or at least all the elements.

[0047] As used herein, unless stated explicitly, performing a step “in response to A” does not indicate that the step is performed immediately after “A” occurs and one or more intervening steps may be included.

[0048] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises”, “comprising”, “has”, “having”, “includes” and / or “including”, when used herein, specify the presence of stated features, elements, and / or components etc., butdo not preclude the presence or addition of one or more other features, elements, components and / or combinations thereof.

[0049] As used in this application, the term “circuitry” may refer to one or more or all of the following:(a) hardware-only circuit implementations (such as implementations in only analog and / or digital circuitry) and(b) combinations of hardware circuits and software, such as (as applicable):(i) a combination of analog and / or digital hardware circuit(s) with software / firmware and(ii) any portions of hardware processor(s) with software (including digital signal processor(s)), software, and memory(ies) that work together to cause an apparatus, such as a mobile phone or server, to perform various functions) and (c) hardware circuit(s) and or processor(s), such as a microprocessor(s) or a portion of a microprocessor(s), that requires software (e.g., firmware) for operation, but the software may not be present when it is not needed for operation.

[0050] This definition of circuitry applies to all uses of this term in this application, including in any claims. As a further example, as used in this application, the term circuitry also covers an implementation of merely a hardware circuit or processor (or multiple processors) or portion of a hardware circuit or processor and its (or their) accompanying software and / or firmware. The term circuitry also covers, for example and if applicable to the particular claim element, a baseband integrated circuit or processor integrated circuit for a mobile device or a similar integrated circuit in server, a cellular network device, or other computing or network device.

[0051] As used herein, the term “communication network” refers to a network following any suitable communication standards, such as New Radio (NR), Long Term Evolution (LTE), LTE-Advanced (LTE-A), Wideband Code Division Multiple Access (WCDMA), High-Speed Packet Access (HSPA), Narrow Band Internet of Things (NB-IoT) and so on. Furthermore, the communications between a terminal device and a network device in the communication network may be performed according to any suitable generation communication protocols, including, but not limited to, the first generation (1G), the second generation (2G), 2.5G, 2.75G, the third generation (3G), the fourth generation (4G), 4.5G, the fifth generation (5G), 5.5G, the sixth generation (6G) communication protocols, and / or any other protocols either currently known or to be developed in the future.Embodiments of the present disclosure may be applied in various communication systems. Given the rapid development in communications, there will of course also be future type communication technologies and systems with which the present disclosure may be embodied. It should not be seen as limiting the scope of the present disclosure to only the aforementioned system.

[0052] As used herein, the term “network device” or “network node” refers to a node in a communication network via which a terminal device accesses the network and receives services therefrom. The network device may refer to a base station (BS), access network node, or an access point (AP), for example, a node B (NodeB or NB), an evolved NodeB (eNodeB or eNB), an NR NB (also referred to as a gNB), a Remote Radio Unit (RRU), a radio header (RH), a remote radio head (RRH), a relay, an Integrated Access and Backhaul (IAB) node, a low power node such as a femto, a pico, a non-terrestrial network (NTN) or non-ground network device such as a satellite network device, a low earth orbit (LEO) satellite and a geosynchronous earth orbit (GEO) satellite, an aircraft network device, and so forth, depending on the applied terminology and technology. In some example embodiments, radio access network (RAN) split architecture comprises a Centralized Unit (CU) and a Distributed Unit (DU) at an IAB donor node. An IAB node comprises a Mobile Terminal (IAB-MT) part that behaves like a UE toward the parent node, and a DU part of an IAB node behaves like a base station toward the next-hop IAB node.

[0053] The term “terminal device” refers to any end device that may be capable of wireless communication. By way of example rather than limitation, a terminal device may also be referred to as a communication device, user equipment (UE), a Subscriber Station (SS), a Portable Subscriber Station, a Mobile Station (MS), or an Access Terminal (AT). The terminal device may include, but not limited to, a mobile phone, a cellular phone, a smart phone, voice over IP (VoIP) phones, wireless local loop phones, a tablet, a wearable terminal device, a personal digital assistant (PDA), portable computers, desktop computer, image capture terminal devices such as digital cameras, gaming terminal devices, music storage and playback appliances, vehicle-mounted wireless terminal devices, wireless endpoints, mobile stations, laptop-embedded equipment (LEE), laptop-mounted equipment (LME), USB dongles, smart devices, wireless customer-premises equipment (CPE), an Internet of Things (loT) device, a watch or other wearable, a head-mounted display (HMD), a vehicle, a drone, a medical device and applications (e.g., remote surgery), an industrial device and applications (e.g., a robot and / or other wireless devices operating in an industrial and / or an automated processing chain contexts), a consumer electronicsdevice, a device operating on commercial and / or industrial wireless networks, and the like. The terminal device may also correspond to a Mobile Termination (MT) part of an I AB node (e.g., a relay node). In the following description, the terms “terminal device”, “communication device”, “terminal”, “user equipment” and “UE” may be used interchangeably.

[0054] Any of a terminal device and / or a network device may be implemented using apparatus as described below in relation to Figure 6.

[0055] As used herein, the term “resource,” “transmission resource,” “resource block,” “physical resource block” (PRB), “uplink resource,” or “downlink resource” may refer to any resource for performing a communication, for example, a communication between a terminal device and a network device, such as a resource in time domain, a resource in frequency domain, a resource in space domain, a resource in code domain, or any other combination of the time, frequency, space and / or code domain resource enabling a communication, and the like. In the following, unless explicitly stated, a resource in both frequency domain and time domain will be used as an example of a transmission resource for describing some example embodiments of the present disclosure. It is noted that example embodiments of the present disclosure are equally applicable to other resources in other domains.

[0056] As used herein, the term “model” is referred to as an association between an input and an output learned from training data, and thus a corresponding output may be generated for a given input after the training. Specifically, the AI / ML model may refer to a data driven algorithm that applies AI / ML techniques to generate a set of outputs (also referred to herein as “inferences”) based on a set of inputs. The generation of the model may be based on a ML technique. The ML techniques may also be referred to as Al techniques. In general, an ML model can be built, which receives input information and makes predictions (e.g., makes inferences) based on the input information. As used herein, a model is equivalent to an AI / ML model, or a data-dri ven / data processing algorithm / procedure.

[0057] The term “Data collection” is referred to as a process of collecting data by the network nodes, management entity, or UE for the purpose of AI / ML model training, data analytics and inference. This process is also referred to herein as “measurements”, or similar.

[0058] The term “Network-side (AI / ML) model” may refer to as an AI / ML Model whose inference is performed entirely at the network. The term “UE-side (AI / ML) model” mayrefer to an AI / ML Model whose inference is performed entirely at the UE. The term “Two-sided (AI / ML) model” is referred to as a paired AI / ML Model(s) over which joint inference is performed, where joint inference comprises AI / ML Inference whose inference is performed jointly across the UE and the network, i.e., the first part of inference is firstly performed by UE and then the remaining part is performed by gNB, or vice versa.

[0059] Figure 1 shows an example communication environment 100 in which example embodiments of the present disclosure can be implemented. In the communication environment 100, a plurality of apparatuses including a first apparatus 110 and a second apparatus 120 communicate with each other.

[0060] In some examples, when the first apparatus 110 is a terminal device and the second apparatus 120 is a network device serving the terminal device, a transmission direction from the second apparatus 120 to the first apparatus 110 is referred to as a downlink (DL), while a transmission direction from the first apparatus 110 to the second apparatus 120 is referred to as an uplink (UL). In DL, the second apparatus 120 is a transmitting (TX) device (or a transmitter) and the first apparatus 110 is a receiving (RX) device (or a receiver). In UL, the first apparatus 110 is a TX device (or a transmitter) and the second apparatus 120 is a RX device (or a receiver).

[0061] In some examples, multiple input multiple output (MIMO) is supported in the communication environment 100. For example, the second apparatus 120 and the first apparatus 110 may communicate with each other via different beams to enable a directional communication. The first apparatus 110 may be configured with at least one beam (corresponding to at least one reference signal (RS)) used as reference for receiving / transmitting data and control channels. For example, the first apparatus 110 may have one or more physical downlink control channel (PDCCH) channels and one or more physical downlink shared channel (PDSCH) channels. The one or more PSCCH channels and / or one or more PDSCH channels may be configured to be received on one or more DL beams. UE may be capable of beamforming (i.e., it may be capable of forming UL / DL beams for TX and RX) or it may use omnidirectional transmission and reception.

[0062] As illustrated in Figure 1, the second apparatus 120 transmits downlink transmission to the first apparatus 110 via one or more of beams 140-1, 140-2, . , and 140-K (K being an integer greater than or equal to 1). For purpose of discussion, the beams 140-1, 140-2, , and 140-K are collectively or individually referred to as beam(s) 140.

[0063] Correspondingly, in uplink, the first apparatus 110 may transmit uplink transmission to the second apparatus 120 via one or more beams. As illustrated in Figure 1,the first apparatus 110 transmits uplink transmission to the second apparatus 120 via the beams 130-1, 130-2, , and 130-J (J being an integer greater than or equal to 1). For purpose of discussion, the beams 130-1, 130-2, , and 130-J are collectively or individually referred to as beam(s) 130.

[0064] In the example of Figure 1, the second apparatus 120 has a certain coverage range, which may be called as a serving area or a cell (not shown). The first apparatuses 110 are located in the cell covered by the second apparatus 120. In the communication environment 100, the second apparatus 120 may communicate data and control information to the first apparatus 110 and the first apparatus 110 may also communication data and control information to the second apparatus 120.

[0065] The coverage area or the cell may be covered by one or more beams provided by one or more Transmission or Reception Points (TRPs), for example, TRP#1, ... TRP#X. Each beam may carry an identifier enabling the first apparatus 110 to identify a beam and perform measurements (e.g., received power, reference signal received power (RSRP)) and other relevant measurements associate with specific identifier. Each synchronization signal block (SSB) may be identified based on the identifier carried by SSB block. Furthermore, for downlink measurement signals for beam management SSB beam may be further used to train.

[0066] In some examples, a model functionality such as an AI / ML based functionality may be provided for the first apparatus 110 and / or the second apparatus 120. For example, the second apparatus 120 may provide a plurality of beams for the first apparatuses 110. The model functionality may be an AI / ML based beam management which predicts a beam such as a DL Tx beam for the first apparatus 110. In another example, the model functionality may be an AI / ML based beam management which predicts a DL Tx Rx beam pair for the first apparatus 110. For purpose of illustration, some examples hereinafter will be described with the beam management or beam prediction as the model functionality. However, it is understood that the model functionality may be for other purposes.

[0067] The AI / ML based beam management may be spatial and / or time domain beam prediction. The spatial beam prediction (also referred to as BM-Casel) is to predict one or more best Tx beams or Tx-Rx beam pairs or corresponding reference signal received power (RSRP) values in different spatial locations. The time-domain beam predictions (also referred to as BM-Case2) aim to predict the best Tx beams or Tx-Rx beam pairs to use for next time instants, e.g., beam prediction in the spatial domain (BM-Casel) for nexttime instants. For purpose of illustration, some example embodiments are described where the model functionality is the spatial and / or time domain beam prediction.

[0068] In some examples, one or more models may derive an outcome such as the predicted beam of the beam management. The one or more models may be implemented at the first apparatus 110 (shown as a model 115), or the second apparatus 120 (not shown), or both (not shown). The first apparatus 110 and / or the second apparatus 120 may perform the beam management by running inference or perform training. For purpose of illustration, some example embodiments hereinafter will be described with the model 115 implemented at the first apparatus 110.

[0069] In the example of Figure 1, by using the model 115, the first apparatus 110 may use beam measurement results of M historical time instances to predict future beam(s) of N future time instances, where M is larger than one and N is larger than or equal to one. The beam measurement results of M historical time instances refer to measurement results of M latest measurement instances (represented as, Pi, P2, ..., PM in the following text), which are used as input of the model 115. The output of the model is N predictions for N future time instances (represented as, Fi, F2, ..., F\ in the following text), where each prediction corresponds to one future time instance and may comprise one or more predicted beams.

[0070] It is to be understood that the number of apparatuses and their connections shown in Figure 1 are only for the purpose of illustration without suggesting any limitation. The communication environment 100 may include any suitable number of apparatuses configured to implementing example embodiments of the present disclosure.

[0071] In the following, for purpose of illustration, some examples are described with the first apparatus 110 operating as a terminal device and the second apparatus 120 operating as a network device. However, in some examples, operations described in connection with a terminal device may be implemented at a network device or other device, and operations described in connection with a network device may be implemented at a terminal device or other device.

[0072] Communications in the communication environment 100 may be implemented according to any proper communication protocol(s), comprising, but not limited to, cellular communication protocols, wireless local network communication protocols such as Institute for Electrical and Electronics Engineers (IEEE) 802.11 and the like, and / or any other protocols currently known or to be developed in the future. Moreover, the communication may utilize any proper wireless communication technology, comprising but not limited to: Code Division Multiple Access (CDMA), Frequency Division MultipleAccess (FDMA), Time Division Multiple Access (TDMA), Frequency Division Duplex (FDD), Time Division Duplex (TDD), Multiple-Input Multiple-Output (MIMO), Orthogonal Frequency Division Multiple (OFDM), Discrete Fourier Transform spread OFDM (DFT-s-OFDM) and / or any other technologies currently known or to be developed in the future.

[0073] Current 3 GPP networks deploy a plurality of methods that may be performed by a UE for obtaining one or more measurements that may be used to make a radio resource management decision.

[0074] When a UE performs one or more of these measurements, the measurement results may be reported to a network device using any of a plurality of different report mechanisms, and the network device may make a radio resource management decision based on them. The decision may be, for example, a mobility-related decision, a resource allocation decision, etc.

[0075] The following will consider the use of a channel state information (CSI) report for reporting one or more measurement results from the UE to an access network node.

[0076] A UE may be configured to obtain measurement results for reporting in a CSI report by performing one or more of these measurements on a predefined signal provided by an access network node, where the predefined signal may be, for example, a synchronisation signal block (SSB) and / or a CSI-reference signal (CSI-RS). The signal and / or resources on which the one or more measurements are performed may be defined in a CSI configuration.

[0077] These measurements (and / or values determined by the UE based on these measurements) may be subsequently provided to the access network node by the UE in accordance with a CSI reporting configuration configured at the UE. The CSI reporting configuration may identify, for example, whether the report is provided periodically, aperiodically or semi-persistently, a reporting granularity in the frequency domain, a type of quantity to be measured (e.g., CSI-related quantities and / or layer 1 reference signal received power- (RSRP-) related quantities), time restrictions in the time domain for channel measurements and / or interference measurements, etc.

[0078] The processing and reporting of the measurements (and / or of values determined based on the measurements) may be performed based on CSI processing criteria, which is currently defined in 3 GPP TS 38.214.

[0079] According to this document, a UE is configured to identify a number of CSI calculations that it can perform simultaneously using a predefined set of UE processingresources. The UE is said to have a same number of CSI processing units (also known as CPUs) as the number simultaneous CSI calculations that the UE is able to perform.

[0080] Stated differently, a UE is configured to indicate a number of supported simultaneous CSI calculations NCPUwith parameter simultaneousCSI-ReportsPerCC in a component carrier, and simultaneousCSI-ReportsAHCC across all component carriers. If a UE supports NCPUsimultaneous CSI calculations it is said to have NCPUCSI processing units for processing CSI reports.

[0081] When a UE is configured to perform CSI measurement and reporting, the UE may be configured to determine whether there are enough CPUs available to perform this CSI measurement and reporting, based on the available CPUs. This is described in connection with an “occupancy” of the CPUs, where a CPU is considered to be “occupied” when the CPU is being used (or is allocated for use) for performing a measurement for a CSI-report and / or for processing a measurement for a CSI-report.

[0082] For example, when L CPUs are occupied for calculation of CSI reports in a given orthogonal frequency division multiplex (OFDM) symbol, the UE has NCPU— L unoccupied CPUs. If N CSI reports start occupying their respective CPUs on the same OFDM symbol on which NCPU— L CPUs are unoccupied, where each CSI report n = 0, ... , N — 1 corresponds to Ocp[J, the UE is not required to update the N — M requested CSI reports with lowest priority (according to Clause 5.2.5 of 3GPP TS 38.214), where 0 < M < N is the largest value such that n=o ^CPU — ^CPU ~holds.

[0083] It is expected that as AI / ML is increasingly deployed, AI / ML will be used for performing one or more calculations on measurements to be reported to the network.

[0084] As a baseline, it is expected that such AI / ML-related use cases may use a similar framework for reporting measurements as is described for the CSI reporting framework described in 3GPP TS 38.214.

[0085] For example, under the current CSI framework, a UE has a hardware limitation for the total number of CPUs, refer as NCPU, where the total number is considered as the limit for simultaneous CSI calculations. It is also expected that a UE may have a hardware limitation for a total number of CPUs for AI / ML operation, which may be referred to herein as NCPU, ML, where the total number is considered as the limit for simultaneous inference operations.

[0086] The following also considers the case in which, for AI / ML-based CSI reporting, the CPU occupancy is considered separately for measurement-obtaining purposes and formeasurement-reporting purposes. For example, there may be provided two values for number of CPUs (e.g., 0CPU 1(0CPU 2), where measurements (for input at the model) and inference operation (actual inference operation) may have separate CPU counts and different CPU occupation time durations. Such a consideration on two values for CPUs may also be applicable for the other use cases (UE-sided AI / ML based CSI prediction, UE-sided AI / ML based Positioning methods).

[0087] However, the introduction of CPUs that are dedicated for AI / ML-based calculations and the introduction of the two separate values for number of CPUs (OCPU,I< OCPU 2) breaks down the CSI reporting framework in 3GPP TS 38.214. This is because the disclosure in 3GPP TS 38.214 does not consider the required number of CPUs (OCPU,I< OCPU 2) within these hardware limits (NCPU, ML, NCPU ) to support AI / ML operation. This may lead to a lack of efficient utilisation of resources.

[0088] Stated differently, AI / ML-based CPUs may not be accounted for under currently defined CPU numbers as they can be completely separate hardware (although it is understood that this may not always be the case).

[0089] Also, CPU resources that are to be used for inference calculations for ML-based CSI reports may be drastically different compared to CPU resources needed for non-ML-based CSI reports.

[0090] Therefore, current methods can’t always adequately account for resources needed, which leads to inefficient use of UE resources

[0091] The following aims to address one or more of the above-mentioned issues.

[0092] In particular, the following considers an apparatus, such as a UE, that is configured to determine a first number of CPUs that are dedicated for performing simultaneous AI / ML-related CSI calculations for one or more component carriers support by the apparatus. The apparatus may determine, based on the first number of CPUs, whether to provide a report on the AI / ML-related CSI calculations to an access node. For example, the apparatus may determine whether the apparatus has enough capacity available on AI / ML-dedicated resources for performing one or more inferences for reporting a CSI report to a network.

[0093] As an extension to this, the processes that are used for forming an AI-ML-based CSI report may be split into measurement performing and measurement calculating, where the measurement calculating uses an AI / ML model to form one or more inferences for inclusion in a CSI report. In such a scenario, the AI / ML CPUs may be used for performingat least the measurement calculating, while any CPU (e.g., non-AI / ML CPU and / or AI / ML CPU) may be used for the performing the measurement performing.

[0094] Further, in such a scenario, a different number of CPUs may be associated with the inference performance (e.g., the calculation performed on one or more measurements) relative to the number of CPUs associated with the measurement performance. These different numbers of CPUs may be used for better calculating the occupancy, which may lead to more efficient usage of CPUs.

[0095] In more detail, when an AI / ML inference report is prepared by the UE, the UE may occupy a first number of CPU(s) (OCPU x) and a second number of CPU(s) (OCPU 2), where the first number of CPU(s) may be associated with the measurements that the UE may have to consider (e.g., generate input measurements) for inference operation of an AI / ML model and the second number of CPU(s) may be associated with inference operation of the AI / ML model where inference operation produce the inference results for reporting. For OCPu,i and OCPU 2, a UE may assume respective first- and second-time durations as the CPU occupying time.

[0096] For a beam prediction use case (e.g., where the inference output from the AI / ML model is to be comprised in a CSI report), the value of OCPU xmay be considered for performing layer 1 reference signal received power (Ll-RSRP) measurements, where the UE considers CPUs mainly for Ll-RSRP measurements of Set B beams. Further, the OCPu,2 value may reflect CPU usage for AI / ML model inference operation, wherein the measurements (Ll-RSRPs) are used for the inference operation.

[0097] For a CSI prediction use case (e.g., where the inference output from the AI / ML model is to be comprised in a CSI report), the value of OCPU xmay be considered for CSI measurements associated with set of reference signal resources considered for ML model input. Further, the OCPU 2value may reflect a CPU usage for AI / ML model inference operation, wherein the measurements (e.g., a CSI corresponding to set of RS resources) are used for the inference operation.

[0098] It is understood that the presently described methods may be applied to other types of use cases (e.g., to other types of AI / ML models).

[0099] For an AI / ML inference report, the UE occupies the first number of CPU(s) (OCPu,i) from a first number of total CPU limit (NCPU x) and the second number of CPU(s) (OCPU 2) from asecond number of CPU limit (NCPU 2).

[0100] The first number of total CPU limit may be considered to be the total limit that the UE has when supporting simultaneous measurement-related calculations for multiple reports (AI / ML or non-AI / ML).

[0101] The second number of total CPU limit (NCPU 2) may be considered to be the total limit that the UE has when supporting simultaneous inference operations for multiple AI / ML-related reports.

[0102] For AI / ML-enabled CSI reporting, NCPU xmay be set to be equal to the limit on NCPUdefined in TS 38.214, which is the number of supported simultaneous CSI calculations NCPU.

[0103] In contrast, NCPU 2may be a new CPU limit that is dedicated for AI / ML inference operations. NCPU 2may be considered to be the number of supported simultaneous AI / ML-related CSI calculations NCPU ML.

[0104] However, it is understood that there may be examples in which NCPU 2is a subset of NCPU. In such a case, the UE may set the value of NCPU 2either statically, semi-statically, or dynamically, depending on the configuration of the UE. The semi-statically and dynamically NCPU 2values may be set based on current and / or anticipated uses of the remaining NCPU.

[0105] When a UE determines that an AI / ML-based CSI report is to be performed, the UE may be configured to consider at least one ofo Whether it is feasible to accommodate the first number of CPUs within the available CPUs of the first total number of CPUs, and / or o Whether it is feasible to accommodate the second number of CPUs within the available CPUs of the second total number of CPUs,

[0106] When checking the above feasibility to accommodate CPUs, the UE may further consider whether there are any other reports starting to occupy CPU(s) in the first total limit or second total limit.

[0107] When it is feasible to accommodate both the first number of CPUs and the second number of CPUs for an AI / ML inference report, the AI / ML inference report may be reported by the UE. When it is not feasible to accommodate both the first number of CPUs and the second number of CPUs for an AI / ML inference report (which may occur, for example, when either the first number of CPUs or the second number of CPUs cannot be accommodated), the AI / ML inference report may not be reported by the UE.

[0108] The UE may be configured to either determine at a first time whether there are CPU resources available for performing both measurements and inferences, or may perform separate determinations at different times as to whether there are CPU resources available for performing measurements and whether there are CPU resources available for performing measurements. These two options will be further considered below in relation to Figures 2 and 3.

[0109] The following example illustrates this second case. In an example, when the first and second CPUs will be initiated at different times (e.g., such as when the measurements are performed in advance of the inferences performed using those measurements), the feasibility check may be performed at different time instances, and inference reporting decision may be performed at any of these different time instances. For example, the UE may first determine whether there are enough CPUs for performing the measurements at a first time. At a later time, the UE may determine whether there are enough AI / ML CPUs available for obtaining one or more inferences for the CSI report.

[0110] For example, when the first number of CPUs can be accommodated at a first time T1 (which indicates a starting time for CPU occupation for measurements), but the second number of CPUs cannot be accommodated at time T2 (which indicates a starting time for inference operation), the UE may determine that the inference report is not reported, regardless of preparing the input measurements.

[0111] For some UE configurations and / or for some AI / ML use cases, the UE may be configured to consider the second number of CPUs and not the first number of CPUs when determining whether an AI / ML-based CSI report is to be generated. Stated differently, it may be assumed that there are always CPUs available for measurements, and so a separate determination as to whether these resources are available may not be performed. For such cases, when it is feasible to accommodate at least the second number of CPUs for an AI / ML inference report, the AI / ML inference report can be reported by the UE. When it is not feasible to accommodate at least the second number of CPUs for an AI / ML inference report, the AI / ML inference report can be reported by the UE.

[0112] For example, when there are no time domain restrictions configured for channel measurements associated to AI / ML inference report (e.g., the UE is not required to perform new measurements), the UE determines whether the second number of CPUs can be accommodated at time T2. The UE may determine the inference report regardless the number of CPUs that can be accommodated at time Tl.

[0113] The UE may be configured to consider pre-defined priority values of reports (priority values of AI / ML inference report and other reports) when determining which reports can be accommodated within the available CPUs of the first total number of CPUs and within the available CPUs of the second total number of CPUs. These pre-defined values may be defined by 3 GPP. The use of such priority values may be particularly useful when the occupancy of the CPUs is relatively high, as in such a case, information that is expected to be more highly relevant to the network may be prioritised for reporting.

[0114] For example, higher priority values for performing a particular CSI reporting may be associated with a larger number of CPUs to be used than lower priority values. The UE may be configured to select a priority for a CSI report that both fulfils a minimum quality requirement for the purpose of the CSI report, and that also does not exceed the number of CPUs available for CSI reporting.

[0115] For example, pre-defined priority values may be defined based reporting modes (periodic / semi-persistent / aperiodic reporting or event-based reporting), measurement timelines (e.g., measurement duration), prediction timelines (e.g., prediction window), and other factors.

[0116] Priority will be illustrated with respect to at least the following three examples.

[0117] In a first example, a higher priority value may be applied in respect of an aperiodic CSI report compared to the priority applied in respect of a semi-persistent CSI report or periodic CSI report. This may be useful, for example, as an aperiodic report may be considered as being performed in response to a particular event, and so is more likely to be needed to be received quickly than information to be comprised in more regular reporting types. Analogously, a higher priority value may be applied in respect of an event-based report compared to semi-persistent report or periodic report. As an example, higher to lower priority values may be defined in the order of event-based reporting, aperiodic reporting, semi-persistent reporting, periodic reporting.

[0118] In a second example, a CSI report having a longer measurement window may be deprioritized (e.g., assigned a lower priority) than those CSI reports having a shorter measurement window. This may be because measurements that are collected having a longer measurement window may occupy for CPUs for a longer duration, which may mean that fewer CSI reports may be generated.

[0119] In a third example, a CSI report that is formed using a higher N (may be longer prediction window) may be deprioritized over CSI reports having a smaller N (e.g., shorterprediction window). This is because reports formed using a higher N may be considered to be generally more accurate and require fewer CPUs.

[0120] The following illustrates how at least some of the above-mentioned principles may be reflected in 3GPP specifications. In particular, the following considers the both AI / ML-based CPUs and non-AI / ML-based CPUs.

[0121] The UE indicates the number of supported simultaneous CSI calculations NCPUwith parameter simultaneousCSI-ReportsPerCC in a component carrier, and simultaneousCSI-ReportsAHCC across all component carriers. If a UE supports NCPUsimultaneous CSI calculations it is said to have NCPUCSI processing units for processing CSI reports. The UE indicates the number of supported simultaneous AI / ML-r elated CSI calculations NCPU MLwith parameter simultaneousAIMLCSI-ReportsPerCC in a component carrier, and simultaneousAI / MLCSI-ReportsAHCC across all component carriers. If a UE supports NCPU,ML simultaneous CSI calculations it is said to have NCPU,ML CSI processing units for processing AI / ML-related CSI reports.

[0122] For all CSI reports, including AI / ML-related CSI reports, if L CPUs are occupied for calculation of CSI reports in a given OFDM symbol, the UE has NCPU— L unoccupied CPUs. If N CSI reports start occupying their respective CPUs on the same OFDM symbol on which NCPU— L CPUs are unoccupied, where each CSI report n = 0, ... , N — 1 corresponds to Ocp[J, the UE is not required to update the N — M requested CSI reports with lowest priority (according to Clause 5.2.5 of 3GPP TS 38.214), where 0 < M < N is the largest value such that n=o OCPU < NCPU ~ L holds and the first OCPUvalue is assumed for AI / ML-related CSI reports.

[0123] For AI / ML-related CSI reports, if L CPUs are occupied for calculation of CSI reports in a given OFDM symbol, the UE has NCPU,ML ~ L unoccupied CPUs. If NML AI / ML-related CSI reports start occupying their respective CPUs on the same OFDM symbol on which NCPU ML— L CPUs are unoccupied, where each CSI report n =0, ... , NML— 1 corresponds to Oppu ML, the UE is not required to update the NML— Ml requested CSI reports with lowest priority (according to Clause 5.2.5 of 3GPP TS 38.214), where 0 < Ml < NMLis the largest value such that~ L holds.

[0124] The UE is not expected to report CSI report if AI / ML-related CSI reports are not accommodated within both M and Ml .

[0125] A UE is not expected to be configured with an aperiodic CSI trigger state containing more than NCPUReporting Settings. Processing of a CSI report occupies a number of CPUs for a number of symbols as follows:OCPU = 0 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'none' and CSI-RS-Re sourceSet with higher layer parameter trs- Info configuredOCPU = 1 for a CSI report with CSI-ReportConfig with higher layer parameter reportQuantity set to 'cri-RSRP', 'ssb-Index-RSRP', 'cri-SINR', 'ssb-Index-SINR', 'cri- RSRP- Index', 'ssb-Index-RSRP- Index', 'cri-SINR- Index', 'ssb-Index-SINR- Index ' or 'none' (and CSI-RS-Re sourceSet with higher layer parameter trs-Info not configured)report with CSI-ReportConfig that configured to report, with higher layer parameter reportQuantity set to 'cri-RSRP -N', 'ssb-Index-RSRP -N', 'cri-N’, or 'ssb-Index-N’, corresponding to a prediction measurement resource set.

[0126] These principles will be further illustrated with reference to the examples of Figures 2 to 5.

[0127] Figure 2 relates to an example in which it is assumed that inference operations are performed at a same time as measurement operations.

[0128] Figure 2 illustrates an example UE method when, during 201, the UE is configured to identify at time T1 that four CSI reports may be generated for reporting. These four CSI reports relate to two AI / ML-based CSI reports (inference report #1 and inference report #2) and two non-AI / ML-based CSI reports (Non-ML report #3 and Non-ML report #4).

[0129] During 202, the UE determines whether the first number of CPUs (e.g., wherein OCpu,i values for inference report #1 and inference report #2, OCPUvalues for Non-ML report #3 and Non-ML report #4) may be accommodated within the available CPUs of NCPU. In the present example, it is determined that for inference report #1, inference report #2 and Non-ML report #3 may be accommodated, but Non-ML report #4 cannot be accommodated. Stated differently, for inference report #1, inference report #2 and Non-ML report #3, the UE method proceeds to 203, while the UE method proceeds to 204 for Non-ML report #4.

[0130] At time Tl, the UE further determines during 203 whether the second number of CPUs (e.g., OCPU 2values for inference report #1 and inference report #2) can be accommodated within the available number of CPUs of NCPU,ML - In the present example, itis determined that inference report #1 can be accommodated but inference report #2 cannot be accommodated. In such a case, the UE method proceeds to 205 for inference report #1 and Non-ML report #3, and the UE method proceeds to 206 for inference report #2, where inference report #2 is abandoned (e.g., no inferences are performed for this second report). Based on this example of Figure 2, non-ML report #3 and inference report #1 are reported as those are the only reports that are accommodated by the number of CPUs.

[0131] Figure 3 relates to an example in which it is assumed that inference operations are performed at a different time as measurement operations.

[0132] Figure 3 illustrates an example UE method when, during 301, the UE is configured to identify at time T1 that four CSI reports may be generated for reporting. These four CSI reports relate to two AI / ML-based CSI reports (inference report #1 and inference report #2) and two non-AI / ML-based CSI reports (Non-ML report #3 and Non-ML report #4).

[0133] During 302, the UE determines whether the first number of CPUs (e.g., wherein OCpu,i values for inference report #1 and inference report #2, OCPUvalues for Non-ML report #3 and Non-ML report #4) may be accommodated within the available CPUs of NCPU. In the present example, it is determined that for inference report #1, inference report #2 and Non-ML report #3 may be accommodated, but Non-ML report #4 cannot be accommodated. Stated differently, for inference report #1, inference report #2 and Non-ML report #3, the UE method proceeds to 303 (in which measurements are performed for these three reports), while the UE method proceeds to 304 for Non-ML report #4 (in which measurements for the non-ML report #4 are not performed).

[0134] For the inference reports, during 305, the first number of CPUs is occupied until time 2, when it is expected that inference operations will start being performed by the second number of CPUs.

[0135] At time T2, the UE further determines during 306 whether the second number of CPUs (e.g., OCPU 2values for inference report #1 and inference report #2) can be accommodated within the available number of CPUs of NCPU ML. In the present example, it is determined that inference report #1 can be accommodated but inference report #2 cannot be accommodated. In such a case, the UE method proceeds to 307 for inference report #1 (in which the inferences for inference report #1 are performed), and the UE method proceeds to 308 for inference report #2 (in which the inferences for the inference report #2 are not performed). Based on this example of Figure 3, non-ML report #3 and inference report #1 are reported as those are the only reports that are accommodated by the number ofCPUs.

[0136] Figures 4 and 5 illustrate example methods that may be performed by apparatus described herein. It is therefore understood that at least one of the features described below in relation to Figures 4 and 5 may find correspondence with at least one feature mentioned above.

[0137] Figure 4 illustrates a method that may be performed by an apparatus. The apparatus may be, for example, a terminal device (e.g., a UE) such as described in connection with Figures 1 and 6, and more generally herein.

[0138] During 401, the apparatus determines a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus. The number of dedicated CPUs may comprise, for example, NCPU ML.

[0139] As discussed above, the dedicated CPUs may comprise at least some hardware that is not currently accounted for by CPUs for non-AI / ML-based CSI reports. As another option, the dedicated CPUs may comprise at least some hardware that is part of the CPUs that are currently used for CSI-based reporting. In this latter case, the number of dedicated CPUs may be set statically, semi-statically, or dynamically, based on a user configuration and an anticipated or current usage of the CPUs.

[0140] During 402, the apparatus determines, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node.

[0141] During 403, the apparatus provides the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

[0142] For example, when it is determined that the report on the ML-related CSI calculations may be provided to the access node, the report is provided to the access node. Conversely, when it is determined that the report on the ML-related CSI calculations may not be provided to the access node, the report is not provided to the access node. In this latter case, the apparatus may abstain from performing ML-related CSI calculations using the dedicated CPUs.

[0143] The example of Figure 4 may use the first CPUs for performing both measurements and calculations for an AI / ML-related CSI report. However, it is also understood that the example of Figure 4 may use the first CPU’s for performing the AI / ML-related CSI calculations and other (second) CPUs for performing the associated measurements for this AI / ML-related CSI calculation.

[0144] For example, for this second case, the apparatus may determine a second number of second CPUs for performing simultaneous machine learning-related CSI calculations for one or more component carriers supported by the apparatus. These one or more component carriers may be the same component carriers as mentioned in Figure 4 or different component carriers. The second CPUs may be performing non-AI / ML-based CSI calculations. The determining whether to provide the report on the machine learning related CSI calculations to the access node may further determine whether to provide the report on the machine learning related CSI calculations to the access node based additionally on the second number.

[0145] The apparatus of Figure 4 may determine at a first point of time (e.g., at a first time) whether one or more measurements for the report on the machine learning related CSI calculations can be processed based on the second number of second CPUs.

[0146] As mentioned above, the apparatus of Figure 4 may additionally consider an occupancy of the CPUs when determining whether the report may be provided. For example, the apparatus may determine at the first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed using one or more of the second number of second CPUs by: determining an occupancy of the second number of second CPUs by determining how many of the second number of second CPUs are unallocated for use at the first point of time; and determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed based on the occupancy of the second number of second CPUs.

[0147] Further as mentioned above, the apparatus of Figure 4 may consider an associated priority for the report when determining whether the report is to be provided. In such a case, the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs may comprise: determining a first priority associated with the report on the machine learning related CSI calculations based on at least one of a configuration for transmitting the report or a configuration for performing the one or more measurements for the report; and determining at least one second priority associated with one or more reports for other CSI calculations that are to be performed at the first point of time based on the second number of second CPUs;, wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs may be performed based on the first and second priorities.

[0148] The first priority may be associated with a predefined number of CPUs to be used for CSI calculations for a corresponding CSI report, and the at least one second priority is associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

[0149] The apparatus may determine whether there are enough dedicated CPUs available for performing an inference. For example, the apparatus may determine at a second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs, wherein when it is determined that one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs, the determining whether to provide the report on the machine learning related CSI calculations to the access node comprises determining to provide the report on the machine learning related CSI calculations to the access node. Conversely, when it is determined that one or more inferences for the report on the machine learning related CSI calculations cannot be processed based on one or more of the first number of first CPUs, the determining whether to provide the report on the machine learning related CSI calculations to the access node comprises determining to not provide the report on the machine learning related CSI calculations to the access node.

[0150] As illustrated by the above examples of Figures 2 and 3, for an AI / ML-based CSI report, the occupancies of the CPUs for performing measurements and the CPUs for performing inferences may be evaluated at either the same time or at different times. Stated differently, the first and second points of time may be the same time, or the first and second point of time may be different times.

[0151] The apparatus may determine at the second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed using one or more of the first number of first CPUs by: determining an occupancy of the first number of first CPUs by determining how many of the first number of first CPUs are unallocated for use at the first point of time; and determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs based on the occupancy of the first number of first CPUs.

[0152] The apparatus may determine whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs comprises by: determining a third priority associated with the report on the machine learning related CSI calculations; and determining at least one fourth priority associatedwith reports for other CSI calculations that are to be performed at the second point of time using the first number of first CPUs, wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs may be performed based on the first and second priorities.

[0153] The third priority may be associated with a predefined number of CPUs to be used for CSI calculations for corresponding CSI report, and the at least one fourth priority is associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

[0154] The apparatus may provide providing the first number to the access node. For example, the apparatus may provide the first number to the access node in a UE capability message. However, it is understood that other signalling methods may be used to provide this information when the first number is semi-static and / or dynamic, as discussed above.

[0155] Figure 5 illustrates a method that may be performed by an interacting apparatus to the apparatus of Figure 4 in some examples. The apparatus of Figure 5 may comprise a network node (e.g., an access network node) such as described above in connection with Figures 1 and 6. For example, the apparatus of Figure 5 may comprise a gNB.

[0156] During 501, the apparatus obtains a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus. The first number of first CPUs may be as described above in connection with Figure 4.

[0157] During 502, the apparatus determines, based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment.

[0158] During 503, the apparatus receives one or more CSI reports based on the determining whether the report will be received.

[0159] It may be useful to provide an apparatus with the functionality of Figure 5 as when the apparatus configured to receive the reports can determine whether the report is going to be provided by the UE, the apparatus can use this information for determining how to decode received signals and / or for making decisions regarding how to configure different UEs for providing CSI reporting.

[0160] The receiving one or more CSI reports may comprise using a result of the determining whether the report will be received for decoding received transmissions.

[0161] The apparatus may schedule one or more CSI resources based on the determining whether the report will be received.

[0162] Figure 6 is a simplified block diagram of a device 1400 that is suitable for implementing example embodiments of the present disclosure. The device 1400 may be provided to implement a communication device, for example, the first apparatus 110 or the second apparatus 120 as shown in Figure 1. As shown, the device 1400 includes one or more processors 1410, one or more memories 1420 coupled to the processor 1410, and one or more communication modules 1440 coupled to the processor 1410.

[0163] The communication module 1440 is for bidirectional communications. The communication module 1440 has one or more communication interfaces to facilitate communication with one or more other modules or devices. The communication interfaces may represent any interface that is necessary for communication with other network elements. In some example embodiments, the communication module 1440 may include at least one antenna.

[0164] The processor 1410 may be of any type suitable to the local technical network and may include one or more of the following: general purpose computers, special purpose computers, microprocessors, digital signal processors (DSPs) and processors based on multicore processor architecture, as non-limiting examples. The device 1400 may have multiple processors, such as an application specific integrated circuit chip that is slaved in time to a clock which synchronizes the main processor.

[0165] The memory 1420 may include one or more non-volatile memories and one or more volatile memories. Examples of the non-volatile memories include, but are not limited to, a Read Only Memory (ROM) 1424, an electrically programmable read only memory (EPROM), a flash memory, a hard disk, a compact disc (CD), a digital video disk (DVD), an optical disk, a laser disk, and other magnetic storage and / or optical storage. Examples of the volatile memories include, but are not limited to, a random-access memory (RAM) 1422 and other volatile memories that will not last in the power-down duration.

[0166] A computer program 1430 includes computer executable instructions that are executed by the associated processor 1410. The instructions of the program 1430 may include instructions for performing operations / acts of some example embodiments of the present disclosure. The program 1430 may be stored in the memory, e.g., the ROM 1424. The processor 1410 may perform any suitable actions and processing by loading the program 1430 into the RAM 1422.

[0167] The example embodiments of the present disclosure may be implemented by means of the program 1430 so that the device 1400 may perform any process of the disclosure as discussed with reference to Figure 2 to Figure 5. The example embodiments of the present disclosure may also be implemented by hardware or by a combination of software and hardware.

[0168] In some example embodiments, the program 1430 may be tangibly contained in a computer readable medium which may be included in the device 1400 (such as in the memory 1420) or other storage devices that are accessible by the device 1400. The device 1400 may load the program 1430 from the computer readable medium to the RAM 1422 for execution. In some example embodiments, the computer readable medium may include any types of non-transitory storage medium, such as ROM, EPROM, a flash memory, a hard disk, CD, DVD, and the like. The term “non-transitory,” as used herein, is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storage persistency (e.g., RAM vs. ROM).

[0169] Figure 7 shows an example of the computer readable medium 1500 which may be in form of CD, DVD or other optical storage disk. The computer readable medium 1500 has the program 1430 stored thereon.

[0170] 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, and other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. Although various aspects of embodiments of the present disclosure are illustrated and described as block diagrams, flowcharts, or using some other pictorial representations, it is to be understood that the block, apparatus, system, technique or method described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

[0171] Some example embodiments of the present disclosure also provide at least one computer program product tangibly stored on a computer readable medium, such as a non-transitory computer readable medium. The computer program product includes computerexecutable instructions, such as those included in program modules, being executed in a device on a target physical or virtual processor, to carry out any of the methods as described above. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, or the like that perform particular tasks orimplement particular abstract data types. The functionality of the program modules may be combined or split between program modules as desired in various embodiments. Machineexecutable 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.

[0172] Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. The program code 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 code, 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.

[0173] In the context of the present disclosure, the computer program code or related data may be carried by any suitable carrier to enable the device, apparatus or processor to perform various processes and operations as described above. Examples of the carrier include a signal, computer readable medium, and the like.

[0174] The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable medium may include but not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the computer readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random-access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0175] Further, although 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, although 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 particularembodiments. Unless explicitly stated, certain features that are described in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, unless explicitly stated, various features that are described in the context of a single embodiment may also be implemented in a plurality of embodiments separately or in any suitable sub-combination.

[0176] Although the present disclosure has been described in languages specific to structural features and / or methodological acts, it is to be understood that the present disclosure defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

Claims1. An apparatus comprising means for performing:determining (401) a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus;determining (402), based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; andproviding (403) the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

2. An apparatus as claimed in claim 1, further comprising means for:determining a second number of second CPUs for performing simultaneous machine learning-related CSI calculations for one or more component carriers supported by the apparatus, wherein the means for determining whether to provide the report on the machine learning related CSI calculations to the access node further comprises means for determining whether to provide the report on the machine learning related CSI calculations to the access node based additionally on the second number.

3. An apparatus as claimed in claim 2, further comprising means for:determining at a first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed based on the second number of second CPUs.

4. An apparatus as claimed in claim 3, wherein the means for determining at the first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed using one or more of the second number of second CPUs comprises means for:determining an occupancy of the second number of second CPUs by determining how many of the second number of second CPUs are unallocated for use at the first point of time; and32determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed based on the occupancy of the second number of second CPUs.

5. An apparatus as claimed in claim 4, wherein the means for determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs comprises means for:determining a first priority associated with the report on the machine learning related CSI calculations based on at least one of a configuration for transmitting the report or a configuration for performing the one or more measurements for the report; anddetermining at least one second priority associated with one or more reports for other CSI calculations that are to be performed at the first point of time based on the second number of second CPUs;wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs is performed based on the first and second priorities.

6. An apparatus as claimed in any of claims 4 to 5, wherein the first priority is associated with a predefined number of CPUs to be used for CSI calculations for a corresponding CSI report, and the at least one second priority is associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

7. An apparatus as claimed in any preceding claim, wherein the apparatus further comprises means for:determining at a second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs,wherein when it is determined that one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs, the means for determining whether to provide the report on the machine learning related CSI calculations to the access node comprises meansfor determining to provide the report on the machine learning related CSI calculations to the access node.

8. An apparatus as claimed in claim 7 when dependent on any of claims 3 to 6, wherein the first and second point of time are the same time.

9. An apparatus as claimed in claim 7 when dependent on any of claims 3 to 6, wherein the first and second point of time are different times.

10. An apparatus as claim in any of claims 7 to 9, wherein the means for determining at the second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed using one or more of the first number of first CPUs comprises means for:determining an occupancy of the first number of first CPUs by determining how many of the first number of first CPUs are unallocated for use at the first point of time; and determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs based on the occupancy of the first number of first CPUs.

11. An apparatus as claimed in claim 10, wherein the means for determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs comprises means for:determining a third priority associated with the report on the machine learning related CSI calculations; anddetermining at least one fourth priority associated with reports for other CSI calculations that are to be performed at the second point of time using the first number of first CPUs;wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs is performed based on the first and second priorities.

12. An apparatus as claimed in any of claims 10 to 11, wherein the third priority is associated with a predefined number of CPUs to be used for CSI calculations forcorresponding CSI report, and the at least one fourth priority is associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

13. An apparatus as claimed in any preceding claim, further comprising means for providing the first number to the access node.

14. An apparatus comprising means for performing:obtaining (501) a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus;determining (502), based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving (503) one or more CSI reports based on the determining whether the report will be received.

15. An apparatus as claimed in claim 14, wherein the means for receiving one or more CSI reports comprises means for using a result of the determining whether the report will be received for decoding received transmissions.

16. An apparatus as claimed in any of claims 14 to 15, further comprising means for scheduling one or more CSI resources based on the determining whether the report will be received.

17. A method for an apparatus, the method comprising:determining (401) a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus;determining (402), based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; andproviding (403) the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

18. A method for an apparatus, the method comprising:obtaining (501) a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus;determining (502), based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving (503) one or more CSI reports based on the determining whether the report will be received.

19. A computer program comprising instructions which, when the program is executed by a computer of an apparatus, cause the apparatus to carry out:determining (401) a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus;determining (402), based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; andproviding (403) the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

20. A computer program comprising instructions which, when the program is executed by a computer of an apparatus, cause the apparatus to carry out:obtaining (501) a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus;determining (502), based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving (503) one or more CSI reports based on the determining whether the report will be received.

21. An apparatus comprising:at least one processor; andat least one memory comprising code that, when executed by the at least one processor, causes the apparatus to perform:determining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers supported by the apparatus;determining, based on at least the first number, whether to provide a report on the machine learning-related CSI calculations to an access node; andproviding the report on the machine learning related CSI calculations to the access node based on the determining whether to provide the report or not.

22. An apparatus as claimed in claim 21, wherein the apparatus is further caused to perform:determining a second number of second CPUs for performing simultaneous machine learning-related CSI calculations for one or more component carriers supported by the apparatus, wherein the determining whether to provide the report on the machine learning related CSI calculations to the access node further comprises determining whether to provide the report on the machine learning related CSI calculations to the access node based additionally on the second number.

23. An apparatus as claimed in claim 22, wherein the apparatus is further caused to perform:determining at a first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed based on the second number of second CPUs.

24. An apparatus as claimed in claim 23, wherein the determining at the first point of time whether one or more measurements for the report on the machine learning related CSI calculations can be processed using one or more of the second number of second CPUs comprises:determining an occupancy of the second number of second CPUs by determining how many of the second number of second CPUs are unallocated for use at the first point of time; anddetermining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed based on the occupancy of the second number of second CPUs.

25. An apparatus as claimed in claim 24, wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs comprises:determining a first priority associated with the report on the machine learning related CSI calculations based on at least one of a configuration for transmitting the report or a configuration for performing the one or more measurements for the report; anddetermining at least one second priority associated with one or more reports for other CSI calculations that are to be performed at the first point of time based on the second number of second CPUs;wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated second CPUs is performed based on the first and second priorities.

26. An apparatus as claimed in any of claims 24 to 25, wherein the first priority is associated with a predefined number of CPUs to be used for CSI calculations for a corresponding CSI report, and the at least one second priority is associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

27. An apparatus as claimed in any of claims 21 to 26, wherein the apparatus is further caused to perform:determining at a second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more of the first number of first CPUs,wherein when it is determined that one or more inferences for the report on the machine learning related CSI calculations can be processed based on one or more ofthe first number of first CPUs, the determining whether to provide the report on the machine learning related CSI calculations to the access node comprises determining to provide the report on the machine learning related CSI calculations to the access node.

28. An apparatus as claimed in claim 27 when dependent on any of claims 23 to 26, wherein the first and second point of time are the same time.

29. An apparatus as claimed in claim 27 when dependent on any of claims 23 to 26, wherein the first and second point of time are different times.

30. An apparatus as claim in any of claims 27 to 29, wherein the determining at the second point of time whether one or more inferences for the report on the machine learning related CSI calculations can be processed using one or more of the first number of first CPUs comprises:determining an occupancy of the first number of first CPUs by determining how many of the first number of first CPUs are unallocated for use at the first point of time; and determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs based on the occupancy of the first number of first CPUs.

31. An apparatus as claimed in claim 30, wherein the determining whether the one or more inferences for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs comprises:determining a third priority associated with the report on the machine learning related CSI calculations; anddetermining at least one fourth priority associated with reports for other CSI calculations that are to be performed at the second point of time using the first number of first CPUs;wherein the determining whether the one or more measurements for the report on the machine learning related CSI calculations can be processed using the unallocated first CPUs is performed based on the first and second priorities.

32. An apparatus as claimed in any of claims 30 to 31, wherein the third priority is associated with a predefined number of CPUs to be used for CSI calculations forcorresponding CSI report, and the at least one fourth priority is associated within another predefined number of CPUs for CSI calculations for one or more other CSI reports.

33. An apparatus as claimed in any of claims 21 to 32, wherein the apparatus is further caused to perform providing the first number to the access node.

34. An apparatus comprising:at least one processor; andat least one memory comprising code that, when executed by the at least one processor, causes the apparatus to perform:obtaining a first number of first channel state information processing units, CPUs, that are dedicated for performing simultaneous machine learning-related channel state information, CSI, calculations for one or more component carriers that are supported by the apparatus;determining, based on at least the first number, whether a report on the machine learning-related CSI calculations will be received from a user equipment; and receiving one or more CSI reports based on the determining whether the report will be received.

35. An apparatus as claimed in claim 34, wherein the receiving one or more CSI reports comprises using a result of the determining whether the report will be received for decoding received transmissions.

36. An apparatus as claimed in any of claims 34 to 35, wherein the apparatus is further caused to perform scheduling one or more CSI resources based on the determining whether the report will be received.

Images