Beam reporting using quantization

Abstract

This application concerns UE-sided beam prediction using an artificial intelligence (AI) / machine learning (ML) model. It has been suggested, in the prior art, that the UE 1110 should predict the top-K best beams and report these 1106 to the network together with probability information associated with each of the best beams, the latter being in a quantised form. The probability information is output from the UE-side AI / ML model. This application is concerned with the quantization of such probability information. This involves a UE transmitting, to a network device, capability information 1101 relating to a quantisation scheme for beam reporting probability. The UE then “obtains” information 1103, 1104 relating to either a quantisation scheme or a quantization setting. This enables the UE to carry out quantisation 1105 of the beam probabilities and to report the quantised probabilities 1106 to a gNB 1120 or other network device. The quantisation scheme is limited to a vector quantization scheme or to use of “statistics of probability values”.

Description

FIELD

[0001] Various example embodiments relate to the field of communication and in particular, to devices, methods, apparatuses and computer readable storage media for beam probability reporting using quantization of the probability information. BACKGROUND

[0002] A communication network can be seen as a facility that enables communications between two or more communication devices, or provides communication devices access to a data network. A mobile or wireless communication network is one example of a communication network.

[0003] Such communication networks operate in according with standards such as those provided by 3GPP (Third Generation Partnership Project) or ETSI (European Telecommunications Standards Institute). Examples of standards are the so-called 5G (5th Generation) standards provided by 3GPP. SUMMARY

[0004] In general, example embodiments of the present disclosure provide a solution for beam reporting, especially, for a terminal device-sided model predicted beam probability reporting using the quantization of the probability information.

[0005] In a first aspect, there is provided a terminal device. The terminal device comprises at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the terminal device at least to: transmit, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtain, at least one of quantization setting or quantization scheme configuration; and transmit, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0006] In a second aspect, there is provided a network device. The network device may comprise: at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the network device at least to: receive, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtain, at least one of quantization setting or quantization scheme configuration; and receive, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0007] In a third aspect, there is provided a method. The method may comprise: transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtaining, at least one of quantization setting or quantization scheme configuration; and transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0008] In a fourth aspect, there is provided a method. The method may comprise: receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtaining, at least one of quantization setting or quantization scheme configuration; and receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0009] In a fifth aspect, there is provided an apparatus. The apparatus may comprise: means for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for obtaining, at least one of quantization setting or quantization scheme configuration; and means for transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0010] In a sixth aspect, there is provided an apparatus. The apparatus may comprise: means for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for obtaining, at least one of quantization setting or quantization scheme configuration; and means for receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0011] In a seventh aspect, there is provided a non-transitory computer readable medium comprising program instructions for causing an apparatus to perform at least the method according to third or fourth aspect.

[0012] In an eighth aspect, there is provided a computer program comprising instructions, which, when executed by an apparatus, cause the apparatus at least to perform at least the method according to third or fourth aspect.

[0013] In a ninth aspect, there is provided a terminal device. The terminal device may comprise: transmitting circuitry for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtaining circuitry for obtaining, at least one of quantization setting or quantization scheme configuration; and transmitting circuitry for transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0014] In a tenth aspect, there is provided a network device. The network device may comprise: receiving circuitry for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtaining circuitry for obtaining, at least one of quantization setting or quantization scheme configuration; and receiving circuitry for receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[0015] In an eleventh aspect, there is provided a terminal device. The terminal device comprises at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the terminal device at least to: transmit to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; receive from the network device, at least one of quantization setting or quantization scheme configuration; and transmit, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0016] In a twelfth aspect, there is provided a network device. The network device may comprise: at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the network device at least to: receive, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; transmit, to the terminal device, at least one of quantization setting or quantization scheme configuration; and receive, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0017] In a thirteenth aspect, there is provided a method. The method may comprise: transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; receiving, from the network device, at least one of quantization setting or quantization scheme configuration; and transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0018] In a fourteenth aspect, there is provided a method. The method may comprise: receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; transmitting, to the terminal device, at least one of quantization setting or quantization scheme configuration; and receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0019] In a fifteenth aspect, there is provided an apparatus. The apparatus may comprise: means for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for receiving, from the network device, at least one of quantization setting or quantization scheme configuration; and means for transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0020] In a sixteenth aspect, there is provided an apparatus. The apparatus may comprise: means for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for transmitting, to the terminal device, at least one of quantization setting or quantization scheme configuration; and means for receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0021] In a seventeenth aspect, there is provided a non-transitory computer readable medium comprising program instructions for causing an apparatus to perform at least the method according to thirteenth or fourteenth aspect.

[0022] In an eighteenth aspect, there is provided a computer program comprising instructions, which, when executed by an apparatus, cause the apparatus at least to perform at least the method according to thirteenth or fourteenth aspect.

[0023] In a nineteenth aspect, there is provided a terminal device. The terminal device may comprise: transmitting circuitry for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; receiving circuitry for receiving, from the network device, at least one of quantization setting or quantization scheme configuration; and transmitting circuitry for transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0024] In a twentieth aspect, there is provided a network device. The network device may comprise: receiving circuitry for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; transmitting circuitry for transmitting, to the terminal device, at least one of quantization setting or quantization scheme configuration; and receiving circuitry for receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[0025] It is to be understood that the summary section is not intended to identify key or essential features of embodiments of the present disclosure, nor is it intended to be used to limit the scope of the present disclosure. Other features of the present disclosure will become easily comprehensible through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Some example embodiments will now be described with reference to the accompanying drawings, in which:

[0027] Fig. 1 illustrates an example of a network environment in which some embodiments of the present disclosure may be implemented;

[0028] Fig. 2 illustrates an example signal process for beam probability information reporting using scalar quantization in accordance with some embodiments of the present disclosure;

[0029] Fig. 3 illustrates an example signal process for beam probability information reporting using vector quantization in accordance with some embodiments of the present disclosure;

[0030] Fig. 4 illustrates an example signaling process for probability information reporting in beam prediction using quantization in accordance with some embodiments of the present disclosure;

[0031] Fig. 5 illustrates an example signaling process for probability information reporting in beam prediction using uniform quantization scheme in accordance with some embodiments of the present disclosure;

[0032] Fig. 6 illustrates an example signaling process for probability information reporting in beam prediction using differential quantization with granularity scheme in accordance with some embodiments of the present disclosure;

[0033] Fig. 7 illustrates an example signaling process for probability information reporting in beam prediction using differential quantization with reduced granularity scheme in accordance with some embodiments of the present disclosure;

[0034] Fig. 8 illustrates an example signaling process for probability information reporting in beam prediction using probability concentration quantization scheme in accordance with some embodiments of the present disclosure;

[0035] Fig. 9 illustrates an example of quantizing Top-4 beam probabilities jointly using a vector quantization codebook with 8 codewords in accordance with some embodiment of the present disclosure;

[0036] Fig. 10 illustrates an example of obtaining the vector quantization codebook based on the statistics of the beam’s probabilities of the trained ML model in accordance with some embodiment of the present disclosure;

[0037] Fig. 11 illustrates an example signaling process for probability information reporting in beam prediction using vector quantization having vector quantization codebook obtained by the network device in accordance with some embodiments of the present disclosure;

[0038] Fig. 12 illustrates an example signaling process for probability information reporting in beam prediction using vector quantization having vector quantization codebook obtained by the terminal device in accordance with some embodiments of the present disclosure;

[0039] Fig. 13 illustrates an example signaling process for probability information reporting in beam prediction using vector quantization having vector quantization codebook and probability concentration quantization in accordance with some embodiments of the present disclosure;

[0040] Fig. 14 illustrates a flowchart of an example method implemented at a terminal device in accordance with some embodiments of the present disclosure;

[0041] Fig. 15 illustrates a flowchart of an example method implemented at a network device in accordance with some embodiments of the present disclosure;

[0042] Fig. 16 illustrates a flowchart of an example method implemented at a terminal device in accordance with some embodiments of the present disclosure;

[0043] Fig. 17 illustrates a flowchart of an example method implemented at a network device in accordance with some embodiments of the present disclosure;

[0044] Fig. 18 illustrates a simplified block diagram of a device that is suitable for implementing some embodiments of the present disclosure; and

[0045] Fig. 19 illustrates a block diagram of an example of a computer-readable medium in accordance with some embodiments of the present disclosure.

[0046] Throughout the drawings, the same or similar reference numerals represent the same or similar elements. DETAILED DESCRIPTION

[0047] Principles of the present disclosure will now be described with reference to some example embodiments. It is to be understood that these embodiments are described only for the purpose of illustration and help those skilled in the art to understand and implement the present disclosure, without suggesting any limitation as to the scope of the disclosure. The disclosure described herein may be implemented in various manners other than the ones described below.

[0048] 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.

[0049] 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.

[0050] It shall be understood that although the terms “first” and “second” etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and similarly, a second element could be termed a first element, without departing from the scope of example embodiments. As used herein, the term “and / or” includes any and all combinations of one or more of the listed terms.

[0051] 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 “includes”, “including”, “has”, “having”, “includes” and / or “including”, when used herein, specify the presence of stated features, elements, and / or components etc., but do not preclude the presence or addition of one or more other features, elements, components and / or combinations thereof. As used herein, “at least one of the following: ” and “at least one of ” and similar wording, where the list of two or more elements are joined by “and” or “or”, mean at least any one of the elements, or at least any two or more of the elements, or at least all the elements.

[0052] 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 processors)), 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 microprocessors) 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.

[0053] 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.

[0054] As used herein, the term “communication network” refers to a network following any suitable communication standards, such as Long Term Evolution (LTE), LTE-Advanced (LTE-A), Wideband Code Division Multiple Access (WCDMA), High-Speed Packet Access (HSPA), Narrow Band Internet of Things (NB-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 future fifth generation (5G) , 5G-Advanced, the future sixth generation (6G) communication protocols, and / or any other protocols either currently known or to be developed in the future. Embodiments of the present disclosure may be applied in various communication systems. Given the rapid development in communications, there will of course also be future type communication technologies and systems with which the present disclosure may be embodied. It should not be seen as limiting the scope of the present disclosure to only the aforementioned system.

[0055] As used herein, the term “network device” refers to a node in a communication network via which a terminal device accesses the network and receives services therefrom. The network device may refer to a base station (BS) or an access point (AP), for example, a node B (NodeB or NB), an evolved NodeB (eNodeB or eNB), a NR NB (also referred to as a gNB), a Remote Radio Unit (RRU), a radio header (RH), a remote radio head (RRH), a relay, a low power node such as a femto, a pico, and so forth, depending on the applied terminology and technology.

[0056] 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 electronics device, a device operating on commercial and / or industrial wireless networks, and the like. In the following description, the terms “terminal device”, “communication device”, “terminal”, “user equipment” and “UE” may be used interchangeably.

[0057] With the development of communication technology, in the 3GPP release, the progress of the air interface artificial intelligence (AI) and machine-learning (ML) model inference has been discussed, and one-side model (for example UE-side, NW-side model only) is relatively mature for normative work. In Release 19, it has discussed limiting the scope e.g., assuming off-line training only, UE-sided, or NW-sided model only, based on selective sub use cases that demonstrate sufficient benefit versus complexity / cost during the Rel-18 Air Interface AI / ML study item.

[0058] In an agreement on UE-sided model AI / ML for beam management, for beam information on predicted Top K beam(s) among a set of beams and probability information of predicted Top K beam(s) among a set of beams, the quantization method of probability information is still for further study, in which probability information is the probability of the beam to be the Top 1 or Top K beams. As it can be seen, one major issue in above agreement was to define quantization method applied to report probability information of Top-1 or Top-K predicted beams. Although it is advantageous to report confidence / probability information of Top-1 or Top-K beams, there is no detailed quantization approaches proposed which can be applied within the existing channel states information (CSI) measurement and report framework with minimum uplink control information (UCI) overhead impact while applying a UE sided model.

[0059] In view of the above, some embodiments of the present disclosure propose a solution for beam reporting, specifically for a terminal device-sided model predicted beam probability reporting using the quantization of the probability information. In some example embodiments of the present disclosure, the terminal device transmits, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. The terminal device obtains, at least one of quantization setting or quantization scheme configuration. The terminal device transmits, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam. In this way, enhanced methods for configuration, determination and indication of reporting probability information with reduced overhead may be obtained.

[0060] For AI / ML enhancements related to beam management, following two sub-use cases have been identified in Rel-18: bam prediction in the spatial domain (BM-Casel); and beam prediction in the time domain (BM-Case2). The scope of spatial beam prediction (BM-Casel) is to predict the best downlink (DL) transmission (Tx) beam and / or DL Tx / reception (Rx) beam pairs in different spatial locations. Conversely, time-domain beam predictions (BM-Case2) aim to predict the best DL Tx beam and / or DL Tx / Rx beam pairs beam to use for next time instants. The primary motivation is to support a reduced overhead and lower beam measurements and reporting latency. Based on the evaluation, the benefits and gains were verified based on given metrics, and they could be supported by single-sided models and consider supporting the necessary / recommended life cycle management (LCM) components for selected sub use cases.

[0061] Fig. 1 illustrates an example of a network environment in which some embodiments of the present disclosure may be implemented. The environment 100, which may be a part of a communication network, includes terminal devices and network devices. As illustrated in Fig. 1, the communication network 100 may include a terminal device 110 (for example, a user equipment, UE). The communication network 100 may further include a network device 120. As shown in Fig. 1, the network device 120 may provide a plurality of DL Tx beams, for example, Tx beams #0 to Tx Beams #M. As illustrated in Fig. 1, the network device 120 and the terminal device 110 both support a ML model for predicting the best DL Tx beam and / or DL Tx / Rx beam pairs beam. In the scenario of single-sided model, such as UE-sided model, the terminal device 110 may predict the best reception beams#0, #1, and #2 for BM-case 1 (spatial beam prediction) and may also predict the best DL Tx beam and / or DL Tx / Rx beam pairs beam to use for next time instants for BM-case 2(time domain prediction).

[0062] It is to be understood that the number of network devices and terminal devices is given only for the purpose of illustration without suggesting any limitations. It should be understood that each network device may serve a plurality of cells, and only one cell is shown for each network device for the purpose of illustration. The system 100 may include any suitable number of network devices and / or terminal devices adapted for implementing embodiments of the present disclosure. Although not shown, it would be appreciated that one or more terminal devices may be located in the environment 100.

[0063] Communications in the network environment 100 may be implemented according to any proper communication protocol(s), comprising, but not limited to, the third generation (3G), the fourth generation (4G), the fifth generation (5G), 5G-Advanced, the sixth generation (6G) or beyond, 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: multiple-input multipleoutput (MIMO), orthogonal frequency division multiplexing (OFDM), time division multiplexing (TDM), frequency division multiplexing (FDM), code division multiplexing (CDM), Bluetooth, ZigBee, and machine type communication (MTC), enhanced mobile broadband (eMBB), massive machine type communication (mMTC), ultra-reliable low latency communication (URLLC), carrier aggregation (CA), dual connection (DC), and new radio unlicensed (NR-U) technologies.

[0064] It is to be understood that the number of devices and their connection relationships and types shown in Fig. 1 are for illustrative purposes only without suggesting any limitation. The communication system 100 may include any suitable number of devices adapted for implementing embodiments of the present disclosure.

[0065] Hereinafter, the various technical aspects of the CSI framework configuration in the technical specification will be introduced, such as Layer 1-reference signal received power (Ll-RSRP) reporting and UCI bit sequence generation. For CSI reporting framework capability, it describes the capability of the UE to support CSI reporting. It includes parameters defining the maximum number of periodic / aperiodic CSI reports that can be configured per CC (Component Carrier), per BWP (Bandwidth Part) and per beam. Moreover, it specifies the concurrent CSI reports per CC that the UE can measure and process, including periodic, semi-persistent and aperiodic CSI, including beam reports.

[0066] For CSI report configuration, it describes the configuration parameters used to set up periodic, aperiodic or semi-persistent CSI reports sent on the Physical Uplink Control Channel (PUCCH) or Physical Uplink Shared Channel (PUSCH) for a particular cell or triggered by downlink control information (DCI). It includes fields such as report quantity, frequency domain configuration, time domain behavior and channel measurement resource allocation that affect how the UE perform reports based on different configurations.

[0067] For LI RSRP reporting, it defines how the UE calculates and reports Ll-RSRP. It covers configurations involving CSI-RS resources, Synchronization Signal / Physical Broadcast Channel (SS / PBCH) block resources or both, detailing limitations on the number of CSI-RS resource sets and resources within those sets. It also explains how Ll-RSRP is quantized and reported based on different scenarios, considering group-based reporting, differential reporting, and channel measurement timing with respect to SS / PBCH or nonzero power channel state information reference signal (NZP CSI-RS).

[0068] For UCI bit sequence generation, it deals with the generation of UCI bit sequences for uplink transmission. It defines the specific order or mapping of CSI fields within a report for different reporting scenarios such as CRI / RSRP, SSBRI / RSRP or Capability Index reporting. It provides details on the structure of the CSI reports, including CRI, RSRP and Capability Index, for transmission within the UCI. These descriptions are integral for defining how CSI is handled, reported, and utilized in the communication system by UE. Each section covers specific technical aspects, configurations and procedures related to CSI reporting, LI RSRP calculation and UCI bit sequence generation, which are critical for establishing and maintaining the communication link in cellular networks, enabling efficient use of CSI for data transmission and reception.

[0069] An agreement has been made for example in RAN1#116 concerning reporting content for UE-sided model beam prediction. In this agreement, for UE-sided model, at least for BM-Casel, for content in the report of inference results, it supports the report for beam information on predicted Top K beam(s) among a set of beams, probability information of predicted Top K beam(s) among a set of beams, RSRP of predicted Top K beam(s) among a set of beams, and confidence information of the RSRP, in which the probability information is the probability of the beam to be the Top 1 or Top K beams.

[0070] The probability information of predicted Top K beam(s) can be defined, for example, as the probability of each beam in Set A to be the Top-1 or Top-K beam(s) as described in the specification. This information can be used by the gNB to assess the quality and reliability of the prediction reported by the UE and may have several uses, e.g. TCI activation among others. Therefore, the probability information and the confidence information are useful for the network to determine whether to perform TCI switching based on the beam prediction result or not. The probability information could also be useful since the probability can reflect beam prediction accuracy in some extend. For predicted RSRP report, confidence / probability information may be helpful for NW to decide whether / how to use the reported RSRP.

[0071] Therefore, in RAN1#116b, it has mentioned the need about the motivation introducing a probability information (probability of the best beam(s) ID or predicted RSRP). The probability information may be used to determine the Top-K predicted beams among the most likely best beams predicted by the ML model for a second refinement beam selection and / or for other purposes (model monitoring, switching, etc). An (subsequent) issue with reporting predicted LI- RSRPs corresponding or beam ID to DL Tx beam(s) is that, reporting many Top-K predicted beams might impose additional UCI overhead as reporting predicted LI-RSRP or beam IDs with very low probability information (e.g., 1%) totally wastes UCI resources leading to possibly decreasing the performance (and coverage) of the PUSCH. A larger number of bits for unnecessary information limit indeed CSI report resources.

[0072] Moreover, reporting all the predicted beams might impose additional overhead on the gNB while monitoring performance is performed as it requires to receive all the predicted beams probability information and ask UE to continue / switch the model. Given that, Top-K predicted beams with highest probability information can be exploited from the output of the ML model at the UE, number of predicted Ll-RSRP reports can be decreased. In view of the foregoing, one major issue in above agreement was to define quantization method applied to report probability information of Top-1 or Top-K predicted beams.

[0073] However, means for configuration, indication and determination of probability information of Top-K beams using the existing CSI measurement and reporting applying compression approaches (e.g., bit quantization) have not been specified / discussed in details yet in 3GPP. That is to say, the quantization method of probability information is still for further study, and for AI / ML model inference at the UE-side, if the probability information of predicted Top-K beams would be reported, they need to be quantized with overhead efficient manner. Therefore, there is a need for proposing methods to specify quantized probability information for Top-1 up to Top-K predicted beams reported to NW.

[0074] Hereinafter, enhanced methods for configuration, determination and indication of reporting probability information obtained by the model output enabling ML-based CSI report with reduced and minimum overhead will be descried with reference to Fig. 2 to Fig. 13. It introduces two different embodiment families each comprises of multiple options. The first embodiment family focuses on a scalar quantization of the probability information, the example of which will be described with reference to options Ito 7 hereinafter; and the second embodiment family focuses on a vector quantization of the probability information, the example of which will be described with reference to options 8 to 9 hereinafter. These two embodiment families focus on reporting Top-K predicted beams confidence / probability information available at the output of the model in addition to e.g., predicted Ll-RSRP or Beam ID or as independent output by applying bit quantization approaches.

[0075] Hereinafter, an example signal process 200 for beam probability information reporting using scalar quantization in accordance with some embodiments of the present disclosure will be described with reference to Fig. 2, for example, in accordance with first embodiment family in which probability information reporting in beam prediction is performed using scalar quantization scheme. For the purpose of discussion, the process 200 may be described with reference to Fig. 1. The process 200 may involve the terminal device 201, and the network device 202. It would be appreciated that although the process 200 has been described in the communication environment 100 of Fig. 1, this process may be likewise applied to other communication scenarios with similar issues.

[0076] As shown in Fig. 2, the terminal device 201 transmits (205) capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. The network device 202 receives (210) the capability on supported quantization scheme. The network device 202 transmits (215) at least one of quantization setting or quantization scheme configuration to the terminal device 201, and the terminal device receives (220) at least one of quantization setting or quantization scheme configuration. The terminal device 201 transmits (225) report for at least one predicted beam and quantized probability information associated with the at least one predicted beam. The probability information is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam. The network device 202 receives (230) the report for at least one predicted beam and the probability information associated with the at least one predicted beams.

[0077] In some embodiments, the quantization setting is received via a higher-layer signaling; and the quantization scheme configuration is received via a channel state information (CSI) report configuration. In some embodiments, the quantization settings may be configured via CSIReportConfig or may be received as association information using higher-level signaling (e.g., using System Information, SI). Therefore, in some embodiments, the quantization settings may be part of the quantization scheme configuration. In some embodiments, a higher-layer signaling association information is received via at least one of the following: a message; a system information; or a downlink control information (DCI).

[0078] In some embodiments, the probability information is quantized by the terminal device by using the scalar quantization scheme; and the terminal device may map the at least one predicted beam each having a probability value within a certain range to at least one quantized level; and quantize the probability information associated with the at least one predicted beam based on the at least one quantized level.

[0079] In some embodiments, the supported quantization scheme comprises uniform quantization in which the probability information associated with the at least one predicted beam is uniformly quantized based on a number of quantized bits. The quantization setting or quantization scheme configuration for the uniform quantization comprises: a number of quantized bits; and a number of the at least one predicted beam. The instances for uniform quantization will be described hereinafter with reference to option 1.

[0080] In some embodiments, the supported quantization scheme comprises different granularity quantization in which the probability information associated with a first beam of the at least one predicted beam is quantized based on a number of quantized bits that is different from a number of quantized bits based on which remaining predicted beams among the at least one predicted beam is quantized. The instances for the different granularity quantization scheme will be described with reference to options 2, 3, and 4 hereinafter.

[0081] In some embodiments, the quantization setting or quantization scheme configuration for the different granularity quantization comprises: a first number of quantized bits for the first beam; a second number of quantized bits for the remaining predicted beams; and a third number of the at least one predicted beam. For these embodiments, the second number of quantized bits for the remaining predicted beams may mean that the number of quantized bits for each of the remaining predicted beams is equal to each other.

[0082] In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the first number of quantized bits, and the probability information associated with the remaining predicted beams in the certain quantization range is uniformly quantized with the second number of quantized bits. In these embodiments, the instances of which may be described with reference to option 2 hereinafter, the quantization range is same for all predicted beams, and the number of quantized bits for each of the remaining predicted beams is equal to each other.

[0083] In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the first number of bits, and the probability information for each of the reaming beams in a quantization range from zero to a quantized value of a probability value of a previous beam is uniformly quantized with the second number of quantized bits. In these embodiments, the instances of which may be described with reference to option 3 hereinafter, the quantization range for each of remaining predicted beams is based on the quantized value of the previous beam, and the number of quantized bits for each of the remaining predicted beams is equal to each other.

[0084] In some embodiments, the quantization setting or quantization scheme configuration for the different granularity quantization comprises: a number of quantized bits for a first beam among the at least one predicted beam; a number of quantized bits for each of remaining predicted beams among the at least one predicted beam; and a number of the at least one predicted beam. In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the number of quantized bits for the first beam, and the probability information associated with a target one of the remaining predicted beams in a quantization range from zero to a quantized value of a probability value of a previous beam is quantized with a target number of quantized bits less than a number of quantized bits associated with the previous beam. In these embodiments, the instances of which may be described with reference to option 4 hereinafter, the quantization range for each of remaining predicted beams is based on the quantized value of the previous beam, and the number of quantized bits for each of the remaining predicted beams is different from each other.

[0085] In some embodiments, the probability information associated with the at least one predicted beam is quantized based on statistics of the probability values of the at least one predicted beam; and the supported quantization scheme comprises probability concentration indication for reporting beam probability information, in which quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam. The instances of these embodiments may be described with reference to options 5, 6, and 7 in details hereinafter.

[0086] In some embodiments, the quantization setting or quantization scheme configuration for the probability concentration indication comprises: a number of quantized bits for report; a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; a total probability of the at least one predicted beam, and a number of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; and a number of the at least one predicted beam. In these embodiments, the terminal device 201 further quantizes the probability information associated with the at least one predicted beam at least based on mapping between the probability difference, quantized bits and the total probability. The instances of these embodiments may be described with reference to option 5 in details hereinafter.

[0087] In some embodiments, the quantization setting or quantization scheme configuration for the probability concentration indication: a number of quantized bits for report, and the terminal device 201 reports the at least one predicted beam and an aggregate of probability values of the at least one predicted beam via a field of the number of quantized bits in a CSI report. The instances of these embodiments may be described with reference to option 6 in details hereinafter.

[0088] In some embodiments, the quantization setting or quantization scheme configuration for the probability concentration indication: a first number of quantized bits for reporting probability information of a first beam of the at least one predicted beam; and a second number of quantized bits for reporting probability information of remaining predicted beams of the at least one predicted beam, and the terminal device 201 reports the at least one predicted beam and the probability information of the first beam via a first field of the first number of quantized bits in a CSI report and an aggregate of probability values of remaining predicted beams via a second field of a second number of quantized bits in the CSI report. The instances of these embodiments may be described with reference to option 7 in details hereinafter.

[0089] Hereinafter, an example signal process 300 for beam probability information reporting using quantization in accordance with some embodiments of the present disclosure will be described with reference to Fig. 3, for example, in accordance with second embodiment family in which probability information reporting in beam prediction is performed using vector quantization scheme. For the purpose of discussion, the process 300 may be described with reference to Fig. 1. The process 300 may involve the terminal device 301, and the network device 302. It would be appreciated that although the process 300 has been described in the communication environment 100 of Fig. 1, this process may be likewise applied to other communication scenarios with similar issues.

[0090] As shown in Fig. 3, the terminal device 301 transmits (305) capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. The network device 302 receives (310) the capability on supported quantization scheme. The network device 302 obtains (315) at least one of quantization setting or quantization scheme configuration to the terminal device 301, and the terminal device 301 obtains (320) at least one of quantization setting or quantization scheme configuration. The terminal device 301 transmits (325) report for at least one predicted beam and the quantized probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam. The network device 302 receives (330) the report for at least one predicted beam and the quantized probability information associated with the at least one predicted beams.

[0091] In some embodiments, the supported quantization scheme comprises vector quantization for reporting beam probability information, and the quantization setting or quantization scheme configuration for the vector quantization comprises a vector quantization codebook for beam probability report. For the first embodiment family having scalar quantization, each beam probability is quantized solely based on its own value and the quantization type and properties. However, for these embodiments according to second embodiment family, probabilities for the at least one predicted beam is quantized for a group of beam probabilities, and quantized probability information associated with the at least one predicted beam is based on the vector quantization codebook. The instances of these embodiments will be described with reference to option 8 in details hereinafter.

[0092] In some embodiments, the vector quantization codebook is determined based on the statistics of beam probabilities of a trained machine-learning model. In some embodiments, the vector quantization codebook for beam probability report is determined by the network device. In some embodiments, the vector quantization codebook is determined by the terminal device and is transmitted to the network device along with the capability on supported quantization scheme for reporting beam probability information. In some embodiments, a number of codewords in the vector quantization codebook is based on a total number of bits allocated for beam probability report, for example, the total number of bits allocated for beam probability report is 3, and the number of codewords in the vector quantization codebook is 23.

[0093] In some embodiments, quantized probability information is further based on statistics of probability values of the at least one predicted beam, the supported quantization scheme further comprises probability concentration indication for reporting beam probability information; the probability information associated with the at least one predicted beam is further quantized based on statistics of the probability values of the at least one predicted beam, and quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam. The instances of these embodiments will be described with reference to option 9 in details hereinafter.

[0094] In some embodiments, the network device 302 further selects the quantization setting for reporting beam probability information based on the received capability on supported quantization scheme, and de-quantizes the probability information based on the vector quantization codebook, and de-quantizes the probability concentration information based on received quantized probability concentration information for the at least one predicted beam.

[0095] Hereinafter, some example embodiments of process 400 for probability information reporting in beam prediction using quantization scheme applicable for the first embodiment family and the second embodiment family will be described with reference to Fig. 4.

[0096] At 401, UE 410 may report supporting beam prediction with associated feature groups (e.g., UE capability framework). For example, the UE 410 may indicate capabilities related to model output (for example, output dimension of the model, probability information quantization at the output, etc.). For example, at 401, the UE 410 may reports capability on supported quantization scheme for reporting probability information during interference to the gNB 420.

[0097] At 402, the gNB 420 may select a proper quantization setting for beam probability report. At 403, the gNB 420 transmits quantization setting via higher-layer signaling for example, system information block 1 (SIB1) and remaining minimum system information (RMSI). Then, the UE 410 receives information on mapping between quantized bits and quantization range (as it is called quantization settings). This information may be made available at the UE at least in three ways: hard-coded in the specification; higher-layer signaled, e.g., SIB1, RMSI; and dynamically / semi-statically indicated, e.g., via DCI.

[0098] The following Table 1 depicts an example of quantization settings applied to quantization options described hereinafter using higher-layer signaling. In one implementation, gNB may provide via higher-layer signaling, e.g., in SIB1 or RMSI, at least a flag to indicate probability information reporting with quantization of Top-K predicted beams is supported in the cell, and remaining configuration on the method used for Top-K beams quantized probability information is provided via higher-layer signaling, e.g., in SIB1 or RMSI. Table 1. Example of hard-coded set of bits with corresponding range, where j may or may not be equal to I, k may or may not be equal to m and n may or may not be equal to 0. Range of hard-coded quantization bits Indicated bits for Top-K beams probability information Exact number Reference to otherwise indicated / specified values Nt to Nj 1 J 1 Top-1 beam value Ni to Nk Second Top beam value Nm to Nn n2 Third Top value No to Np n3 Fourth Top value

[0099] At 404, the gNB 420 configures quantization scheme via CSIReportconfig. For example, the gNB 420 configures UE with a quantization scheme and number of bits for quantization within ML-enabled CSIreport.

[00100] For example, the UE 410 indicates its capability on applying a quantization scheme with respect to the probability information of Top-K predicted beams via UE capability framework at 401. The gNB 420 configures or select quantization settings of the UE based on reported capability at 402. The gNB 420 transmits or configures one or multiple quantization settings (which is mapping between quantized bits, K and quantization scheme) (e.g., through higher-layer signalling) with respect to each supported scalar quantization approach received from UE capability (using UE capability framework) at 403 or 404. For example, if the UE 410 indicates supporting three scalar quantization scheme (e.g., option 1, 2, 3 described below), the gNB 420 needs to transmits one setting per each supported scheme and select one of supported approaches as quantization scheme to configure the UE. Therefore, the quantization settings can be configured via CSIReportConfig, for example, at 404 or it can be received as association information using higher-level signaling (e.g., using System Information, SI), for example, at 403.

[00101] At 405, the UE receives RRC configuration and applies configured quantization scheme for reporting Top-K probability information. For example, UE may be configured with at least a ML-enabled CSI reporting configuration that UE is able to report beam prediction results (e.g., predicted Top-l / K beam IDs, predicted Ll-RSRPs). In an example, UE may be configured to report up to Top-K predicted beams. Each ML-enabled CSI reporting configuration may be associated with a quantization scheme used for determining the probability information report associated with the model output.

[00102] When applying configured quantization scheme for reporting Top-K probability information, the UE 410 may derive quantization range, quantized level based on configured quantization scheme and number of bits for quantization, and maps the predicted beams to quantized level to determine probability information corresponding to predicted CRI / RSRP reporting using any of options 1 to 9, which will be described hereinafter. For example, the UE 410 receives configuration including the quantization method associated with one of options 1 to 9, number of bits and K value. The UE 410 then determines quantization level, number of bits per Top beam and other details and obtains each table with respect to each of options 1 to 9.

[00103] For example, for the scalar quantization scheme according to first embodiment family including for example options 1 to 7, each scalar quantization scheme can be performed at the UE by knowing the selected option, K value and number of bits dedicated to Top-1. For example, in the case that K value is 4, if the UE knows the selected method or option, and number of bits for Top-1 beam, it will obtain autonomously number of bits for Top-2, Top-3, andTop-4 accordingly.

[00104] At 406, the UE 410 reports Top-K predicted beams carrying quantized probability information corresponding to each of Top-K beams. At 407, the gNB 420 receives the report and de-quantizes the probability information.

[00105] Hereinafter, option 1 for quantizing probability information in accordance with first embodiment family will be described, which has uniform quantization for probability information.

[00106] For option 1, uniform quantization in range [0... 100] is applied, and each probability information report is uniformly quantized based on configured quantization bits. In other words, Top-1 as well as Top-K beams probability information will be quantized uniformly with the same bit width.

[00107] As an example, considering uniform quantization for quantization of the probability reports, a 4-bit uniform quantization in range [0, 100] can be used to quantize the probabilities as shown in the following Table 2. Table 2 -bit uniform quantization range and levels Index i Quantization bits Quantization range Quantized level 0 0000 [0, 6.25) 3.125 1 0001 [6.25, 12.5) 9.375 2 0010 [12.5, 18.75) 15.625 3 0011 [18.75, 25) 21.875 4 0100 [25, 31.25) 28.125 5 0101 [31.25, 37.5) 34.375 6 0110 [37.5,43.75] 40.625 7 0111 [43.75, 50) 46.875 8 1000 [50, 56.25] 53.125 9 1001 [56.25, 62.5) 59.375 10 1010 [62.5, 68.75] 65.625 11 1011 [68.75, 75) 71.875 12 1100 [75, 81.25] 78.125 13 1101 [81.25, 87.5) 84.375 14 1110 [87.5, 93.75) 90.625 15 mi [93.75, 100] 96.875

[00108] If 4-bit uniform quantization as explained in option 1 for probability information reporting of Top-4 beams from Set A beams of size 64 is used, the following table 3 depicts an example of selecting Top-4 beams and dedicated quantized bits. In this example, beams 17, 5 42, 56 and 4 have the higher probabilities among the 64 beams in Set A. Table 3 An example of using 4-bit uniform quantization for probability reporting of Top-K beams from Set A with 64 beams. Beam number Probability Quantized Value Quantized bits 17 52% 53.125% 1000 42 17%o 15.625% 0010 56 13% 15.625% 0010 4 8% 9.375% 0001 SUM 100%

[00109] Therefore, the overhead of probability reporting will be as follows: Top-1 (beam 17): 1*4 = 4 bits; Top-2 (beam 17&42): 2*4 = 8 bits; Top-3 (beam 17&42&56): 3*4 = 12 bits; and Top-4 (beam 17&42&56&4): 4*4 = 16 bits.

[00110] Hereinafter, some example embodiments of process 500 for probability information reporting in beam prediction using uniform quantization scheme will be described with reference to Fig. 5.

[00111] As shown in Fig. 5, at 501, the UE 510 reports capability on uniform quantization for reporting probability information during inference. At 502, the gNB 520 selects quantization setting for beam probability report, for example, the gNB 520 selects uniform quantization setting for beam probability report. At 503, the gNB 520 transmits uniform quantization setting via higher-layer signaling for example, SIB I and RMSI. The uniform quantization setting in step 503 can refer to: number of quantized bits, for example, 4 bits; and K value, for example, 4 bits, and Top-4.

[00112] At 504, the gNB 520 configures uniform quantization Via CSIreportconfig. At 505, the UE 510 receives RRC configuration and applies uniform quantization scheme for reporting Top-K probability information. At 506, the UE 510 reports Top-K predicted beams and Top-K uniformed quantized probability information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 1000, 0010, 0010, and 0001 respectively. At 507, the gNB 520 receives Top-K CSI-RS report and de-quantizes probability information with respect to Top-K beams.

[00113] Hereinafter, option 2 for quantizing probability information of predicted beams in accordance with first embodiment family will be described, which has differential quantization granularity for reporting of probability of the 1st beam and Zr-th beam (1 <Zr<K).

[00114] For option 2, quantization in range [0... 100] is applied, and probability information report corresponding to the Top-1 predicted beam is uniformly quantized with higher number of bits and the rest of predicted beams are uniformly quantized with lower number of bits or bit width (the same number of bits for the rest of beams). As an example, in case of Top-4 predicted beams, the Top-1 beam probability information may get quantized with 4 bits and each of the rest 3 beams may get quantized with 2 bits.

[00115] In this option 2, uniform quantization with different quantization range and number of bits for the first and other reports may be considered. For example, the following quantization settings may be considered as follows: For first probability report: 4-bit uniform quantization in range [0,100]; and For other probability reports: 2-bit uniform quantization in range [0, 100].

[00116] If 2-bit differential quantization as explained in option 2 for probability information reporting of Top-4 beams from Set A beams of size 64 is used, table 4 depicts an example of selecting Top-4 beams and dedicated quantized bits. Table 4 2-bit uniform quantization range and levels Index i Quantization bits Quantization range Quantized level 0 00 [0, 25) 12.5 1 01 [25, 50) 37.5 2 10 [50, 75) 62.5 3 11 [75, 100] 87.5

[00117] As shown above, table 4 depicts granularity of reporting for differential quantization of 4-bit uniform quantization for Top-1 as explained in option 2 and 2-bit uniform quantization with quantization range of [0... 100] for the rest of Top beams, where probability information reporting of Top-4 beams from Set A beams of size 64 is used.

[00118] In one example, for example, shown in table 5 which shows an example of using different quantization granularity for probability reporting of Top-K beams from Set A with 64 beams, probability of beams #17 is quantized with 4-bit quantization, and the probabilities of beams #42, #56 and #4 are quantized with 2-bit quantization, as can be seen from table 5. 5 Table 5 An example of using different cpiantization granularity for probability reporting of Top-K beams from Set A with 64 beams Beam number Probability Quantized Value Quantized bits 17 52% 53.125% 1000 42 17% 12.5% 00 56 13% 12.5% 00 4 8% 12.5% 00 SUM 100%

[00119] Therefore, the overhead of probability reporting may be follows: Top-1 (beam 17): 1*4 = 4 bits; 10 Top-2 (beam 17&42): 1*4 + 1*2 = 6 bits; Top-3 (beam 17&42&56): 1*4 + 2*2 = 8 bits; and Top-4 (beam 17&42&56&4): 1*4 + 3*2 = 10 bits.

[00120] Hereinafter, some example embodiments of process 600 for probability information reporting in beam prediction using differential quantization with granularity will be described 15 with reference to Fig. 6.

[00121] As shown in Fig. 6, at 601, the UE 610 reports capability on differential quantization with granularity for reporting probability information during inference. At 602, the gNB 620 selects quantization setting for beam probability report, for example, the gNB 620 selects differential quantization with granularity setting for beam probability report. At 603, the gNB 620 transmits differential quantization with granularity setting via higher-layer signaling for example, SIB1 andRMSI.

[00122] For the option 2, the differential quantization setting in step 603 can refer to: Number of quantized bits for Top-1 predicted beam, for example, 4bits; Number of quantized bits for the rest Top predicted beam, for example, 2 bits; and K value, for example, Top-4 beams.

[00123] At 604, the gNB 620 configures differential quantization with granularity quantization Via CSIreportconfig. At 605, the UE 610 receives RRC configuration and applies differential quantization with granularity scheme for reporting Top-K probability information. At 606, the UE 610 reports Top-K predicted beams and Top-K differential quantization with granularity probability information, for example, report beams #17, #42, #66, and #4, and the quantized bits for them will be 1000, 00, 00, and 00 respectively. At 607, the gNB 620 receives Top-K CSI-RS report and dequantizes probability information with respect to Top-K beams.

[00124] Hereinafter, option 3 for quantizing probability information of predicted beams in accordance with first embodiment family will be described, which has differential quantization for probability reporting of #-th beam (1 <k <K)

[00125] For option 3, to achieve further overhead reduction, different granularity for probability information can be employed in order to reduce quantization error in previous proposed embodiment via decreasing the quantization range of differential quantization. In other words, uniform quantization in range [0...100] is applied for the beam with highest probability information (e.g., Top-1 beam), and probability information report corresponding to the Top-1 predicted beam is uniformly quantized with higher number of bits. The mth probability report corresponding to a predicted beam probability information (lower predicted RSRP or predicted CRI probability compared to Top-1 beam) is uniformly quantized with lower but the same bit width but in [0, Bm_] ] quantization range, and is the quantized value of the m — 1th probability report. As an example, in case of Top-4 predicted beams, the Top-1 beam probability information may get quantized with 4 bits and quantized value of 53%, based on quantization level of * (i + 0.5), second best predicted beam probability information may get quantized with 2 bits considering quantization range of [0, 53%] and so on. Please note that, i denotes beam index from 1 to K among Top-K beams.

[00126] Although the solution of option 2 reduces significantly the overhead, the quantization error increases significantly as well. To keep the quantization error low of beam k (1<#<K) probability quantization, differential quantization approach can be used. In this case, the quantization range for probability of beam k is reduced to be [0, Bm-i], where Bm i is the quantized value of the beam probability k-1. Thus, for example, rreducing the quantization error by reducing the quantization range of non-first probability quantization can be done as follows: First probability report: 4-bit uniform quantization in range [0,100]; and The m-th (m>l) probability reports: 2-bit uniform quantization in range [0, Bm-i], where Bmu is the quantized value of the (m-l)-th probability report.

[00127] Table 6 shows quantization range and quantized levels of Top-4 beam prediction with granularity of reporting for differential quantization of 4-bit uniform quantization for Top-1 as explained in Option 3, and 2-bit uniform quantization with quantization range of [0, Bm-1], for the rest of Top beams, where probability information reporting of Top-4 beams from Set A beams of size 64 is used. Table 6 2-bit uniform differential quantization range and levels Index i Quantization bits Quantization range Quantized level 0 00 [0, Bm-i / 4) Bm-t / 8 1 01 [Bm^ / 4, Bmfr2) 3*Bm^ / 8 2 10 [Bm-i / 2, 3*Bmn / 4) 5 AV; 8 3 11 [3*Bm-i / 4, Bm-i] 7*Bm-i / 8

[00128] Table 7 illustrates an example of using different quantization granularity for probability reporting of Top-4 beams from Set A with 64 beams. As it can be seen, probability of beams #17 is quantized with 4-bit quantization, and the probabilities of beams #42, #56 and #4 are quantized with 2-bit differential quantization scheme. Table 7 An example of using different quantization granularity for probability reporting of Top-4 beams from Set A with 64 beams Beam number Probability Quantized Value Quantized bits 17 52% 53.125% 1000 42 17% 19.92% 01 56 13% 12.45% 10 4 8% 7.78% 10 • • • • SUM 100%

[00129] As it can be seen from the examples in Table 5 for option 2 and table 7 for option 3, the differential quantization technique can reduce the quantization error significantly.

[00130] Therefore, the overhead of probability reporting may be follows: Top-1 (beam 17): 1*4 = 4 bits; 5 Top-2 (beam 17&42): 1*4 + 1*2 = 6 bits; Top-3 (beam 17&42&56): 1*4 + 2*2 = 8 bits; and Top-4 (beam 17&42&56&4): 1*4 + 3*2 = 10 bits.

[00131] When calculating quantized level for beam #42, the following table 7-1 will be used. Table 7-1 2-bit uniform differential quantization range and levels for beam #42 Index i Quantization bits Quantization range Quantized level 0 00 [0, 53.124% / 4=13.28) 6.64 1 01 [13.28,26.56) 19.92 2 10 [26.56,39.84) 33.20 3 11 [39.84, 53.125] 46.48

[00132] As shown in table 7-1, the probability 17% for beam #42 is in the range of [13.28, to 26.56), and thus, the quantized level of beam #42 will be 19.92.

[00133] When calculating quantized level for beam #42, the following table 7-2 will be used. Table 7-2 2-bit uniform differential quantization range and levels for beam #56 Index i Quantization bits Quantization range Quantized level 0 00 [0, 19.92 / 4=4.98) 2.49 7 01 [4.98,9.96) 7.47 2 10 [9.96,14.94) 12.45 3 11 [14.94,19.92] 17.43

[00134] As shown in table 7-2, the probability 13% for beam #56 is in the range of [9.96,14.94), and thus, the quantized level of beam #56 will be 12.45. The quantized level for beam #4 may be calculated in the same way.

[00135] The process for probability information reporting in beam prediction using differential quantization scheme in accordance with option 3 may be the same as the process shown in Fig. 6 for option 2, the difference will be at the quantization setting in operation 603. For option 3, the differential quantization setting in step 603 can refer to: Number of quantized bits for Top-1 predicted beam, for example, 4bits; Number of quantized bits for the rest Top predicted beam, for example, 2 bits; and K value, for example, Top-4 beams.

[00136] At 605, the UE 610 reports Top-K predicted beams and Top-K differential quantization with reduced granularity probability information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 1000, 01, 10, and 10 respectively.

[00137] Hereinafter, option 4 for quantizing probability information of predicted beams in accordance with first embodiment family will be described, which has differential quantization with gradual decreasing granularity.

[00138] For option 4, a sequential reduction of the quantization bits of differential quantization can be employed in order to reduce the quantization range of differential quantization. In other words, uniform quantization in range [0... 100] is applied for the beam with highest probability information (e.g., Top-1 beam), and probability information report corresponding to the Top-1 predicted beam is uniformly quantized. If Top-1 beam is quantized with L bits, the mth probability report corresponding to a predicted beam probability information (lower predicted RSRP or predicted CRI probability compared to Top-1 beam) is quantized with lower but sequential and non-equal bit width of (L-m+1) in [0, ] quantization range, where Bm_] is the quantized value of the m — 1th probability report. As an example, in case of Top-4 predicted beams, the Top-1 beam probability information may get quantized with 4 bits and quantized value of 53%, based on —x (i + 0.5), second best predicted beam probability information may get quantized with 3 bits considering quantization range of [0, 53%] and so on.

[00139] In this option 4, gradual reduction of the quantization bits of differential quantization can be considered to ensure lower quantization error for more important beam probabilities. For example, gradual reduction of the quantization granularity can be done as follows: First probability report: 4-bit uniform quantization in range [0,100]; and The m-th (m>l) probability reports: (4-m+l)-bit uniform quantization in range [0, Bm-i], where is the quantized value of the (m-l)-th probability report.

[00140] Table 8 illustrates an example of using different quantization granularity for probability reporting of Top-4 beams from Set A with 64 beams in which gradual decreasing granularity impacted the quantized bits reported for second, third and fourth best beams. As it can be seen, Probability ofbeam#17 is quantized with 4-bit quantization, and the probabilities of beams #42, #56 and #4 are quantized with 3-bit, 2-bit, and 1-bit differential quantization schemes, respectively. Table 8 An example of using different quantization granularity for probability reporting of Top-4 beams from Set A with 64 beams Beam number Probability Quantized Value Quantized bits 17 52% 53.125% 1000 42 17% 16.60% 010 56 13% 14.53% 11 4 8% 10.9% 1 • • • • SUM 100%

[00141] Thus, the quantization error of the second highest beam probability (beam number 42) is reduced in this approach (comparing the values in table 7 and table 8, reducing from 19.92% to 16.60%).

[00142] Furthermore, the overhead of probability reporting may be follows: Top-1 (beam 17): 4 = 4 bits; Top-2 (beam 17&42): 4 + 3 = 7 bits; Top-3 (beam 17&42&56): 4 + 3+2 = 8 bits; and Top-4 (beam 17&42&56&4): 4 + 3 + 2+ 1 = 10 bits. 5

[00143] When calculating quantized level for beam #42, the following table 8-1 will be used, wherein Bm-i=53.125%. Table 8-1 3-bit uniform differential quantization range and levels Index i Quantization bits Quantization range Quantized level 0 000 f0, BmJ8) Bm^ / 16 1 001 fBmn / 8, Bmn / 4) 3*8,^. !6 2 010 [B^ / 4, 3*Bm.1 / 8) 5*Bm.i / 16 3 on [3*Bm-l / 8, B^ / 2) 7*Bm.i / 16 4 100 [Bm-i / 2, 5BnJ8) 9*Bm.i / 16 5 101 [5Bm-i / 8, 3Bm.1 / 4) ll*Bm-i / 16 6 110 f3Bmn / 4, 7Bmn / 8) 13*8,^. !6 7 111 15*Bm.v16

[00144] When calculating quantized level for beam #56, the following table 8-2 will be used, 10 wherein Bm-i=16.60%. Table 8-2 2-bit uniform differential quantization range and levels Index i Quantization bits Quantization range Quantized level 0 00 [0, B^ / 4) Bm-1'8 1 01 IB;„:4. 8,-,,: 2) 3*Bm.1 / 8 2 10 [Bm.i / 2, 3*Bm.i / 4) 5*Bm.1 / 8 3 11 [3*Bm-i / 4, Bm.i] 7*Bm.1 / 8

[00145] Hereinafter, some example embodiments of process 700 for probability information reporting in beam prediction using differential quantization with reduced granularity scheme 15 will be described with reference to Fig. 7.

[00146] As shown in Fig. 7, at 701, the UE 710 reports capability on differential quantization with reduced granularity for reporting probability information during inference. At 702, the gNB 720 selects quantization setting for beam probability report, for example, the gNB 720 selects differential quantization with reduced granularity setting for beam probability report. At 703, the gNB 720 transmits differential quantization with reduced granularity setting via higher-layer signaling for example, SIB1 and RMSI.

[00147] For the option 4, the differential quantization with reduced granularity setting in step 703 can refer to: Number of quantized bits for Top-1 predicted beam, for example, 4 bits; Number of quantized bits for each of the rest Top predicted beam, for example, 3 bits, 2 bits, and 1 bit; and K value, for example, Top-4 beams.

[00148] At 704, the gNB 720 configures differential quantization with reduced granularity quantization Via CSIreportconfig. At 705, the UE 710 receives RRC configuration and applies differential quantization with reduced granularity scheme for reporting Top-K probability information. At 706, the UE 710 reports Top-K predicted beams and Top-K differential quantization with reduced granularity probability information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 1000, 010, 11, and 1 respectively. At 707, the gNB 720 receives Top-K CSI-RS report and de-quantizes probability information with respect to Top-K beams.

[00149] Hereinafter, option 5 for quantizing probability information of predicted beams in accordance with first embodiment family will be described, which has total probability concentration reporting indication and comparison of probability of 1st beam with other beam probabilities.

[00150] For option 5, the UE obtains probability information associated with Top-K beams (probK,probK_1,probK_2,...,probK_K+1) associated with (K, K-l, K-2,...K-K+1) beams and determine if the best beam in Top-K has a significant difference with respect to probability information compared to other beams in Top-K (this term is called standability ratio). Such a relative comparison may be further extended to consider concentration of probabilities among Top-K beams compared to other beams. In other words, the UE may be defined with reporting indicators that provide relative comparison between probability values without reporting absolute probabilities in the report.

[00151] Table 9 may be considered for the relative probability value reporting, where X and Y can be either defined by the spec or configured to the UE. In one example, X = 20% and Y = 50%. Table 9 An example of using probability concentration reporting indication for probability reporting of Top-4 beams Index Reporting bits Pl (probability value of best beam in Top-K) >= X% compared to P2 (probability value of second-best beam in Top-K) Probability concentration on Top-K beams (i.e., Pl + P2 + P3 .. +PK) 0 00 Yes >= Y% 1 01 No > = Y% 2 10 Yes <y% 3 11 No <Y%

[00152] The reporting bits only provide a relative assessment on probability values determined at the UE for predicted beams. For example, Top-4 beam reported from 64 beams (Set A) may be considered, i.e., 1st best beam, 2nd best beam, 3rd best beam, 4th best beam having pl, p2, p3, p4 as the absolute probabilities at the model output may be considered. What matters is to report the following: (1) whether pl stands out from the rest, because the NW can directly use this beam if this is the case, and this can be extended pl, p2 stands out of the rest, etc.., depending on K value; and (2) what is the total probability concentration on p 1 - p2 + p3 + p4 (e.g., above 50%?). If less concentrated, the NW can switch to Top-8 beams reporting, that is to say, the NW will take essential actions.

[00153] Hereinafter, some example embodiments of process 800 for probability information reporting in beam prediction using probability concentration quantization scheme will be described with reference to Fig. 8.

[00154] As shown in Fig. 8, at 801, the UE 810 reports capability on probability concentration indication for reporting probability information during inference. At 802, the gNB 820 selects quantization setting for beam probability report, for example, the gNB 820 selects quantized probability concentration setting for beam probability report. At 803, the gNB 820 transmits probability concentration setting via higher-layer signaling for example, SIB1 and RMSL

[00155] For the option 5, the probability concentration setting in step 803 can refer to: Number of quantized bits for the report, for example, 2 bits for all predicted beams; X and Y, for example, X = 20% and Y = 50%; and K value, for example, Top-4 beams.

[00156] At 804, the gNB 820 configures probability concentration quantization Via CSIreportconfig. At 805, the UE 810 receives RRC configuration and applies probability concentration scheme for reporting Top-K probability information. At 806, the UE 810 reports Top-K predicted beams and Top-K probability concentration probability information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 00, 01, 10, or 11. At 807, the gNB 820 receives Top-K CSI-RS report and de-quantizes probability information with respect to Top-K beams and takes essential actions.

[00157] Hereinafter, option 6 for quantizing probability information of predicted beams in accordance with first embodiment family will be described, which has quantization of the total probability concentration reporting for Top-K beams.

[00158] In this option 6, the aggregate of the probabilities in Top-K beams is reported in an added field of CSI report. Different granularities of quantization can be assumed as described in the following.

[00159] In some embodiments, if the probability in the range 0-100% is quantized (uniformly) with 7 bits, in one example, the Top-4 beams, e.g. beam #7, #2, #6 and #4 have probabilities 52%, 17%, 13% and 8%, respectively. The sum of the probabilities is equal to 90%, which quantized with 7 bits correspond to 1011010.

[00160] Further, in some embodiments, If the probability in the range 0-100% is quantized (uniformly) with 2 bits, in one example, the Top-4 beams, e.g. beam #7, #2, #6 and #4 have probabilities 52%, 17%, 13% and 8%, respectively. The sum of the probabilities is equal to 90%, which quantized with 2 bits correspond to 11.

[00161] It should be noted the probability in the range 0-100% is quantized (uniformly) with any other bits, rather than 7 bits and 2 bits as mentioned above, and the number of the bits may be included in the probability concentration quantization setting for option 6.

[00162] The process for probability information reporting in beam prediction using probability concentration scheme in accordance with option 6 may be the same as the process shown in Fig. 8 for option 5, the difference will be at the quantization setting in operation 803. For option 6, the probability concentration setting in step 803 may refer to: number of quantized bits for the report, for example, 7 bits or 2 bits.

[00163] Hereinafter, option 7 for quantizing probability information of predicted beams in accordance with first embodiment family will be described, which has total probability concentration reporting and probability of the 1st beam.

[00164] In addition to the alternatives presented in options 1 to 4, this other example combines different alternatives for reporting the Top-1 beam probability with one of the methods described in options 1 to 4 in a first field of the CSI report, in addition to reporting the aggregate of the probabilities in Top-K beams (with K>2) in an added second field of the CSI report. Different granularities of quantization can be assumed.

[00165] For example, if the probability in the range 0-100% is quantized (uniformly) with 7 bits, in one example with the probability in the range 0-100% is quantized (uniformly) with 7 bits, the Top-2, Top-3 and Top-4 beam, e.g. beam #2, #6 and #4 have probabilities 17%, 13% and 8%, respectively. The sum of the probabilities is equal to 38%, which is quantized with 7 bits corresponding to 0100110. Thus, if option 3 is used for Top-1 beam, e.g. beam #7 which has probability 52%, the first field is 4-bit length and corresponds to 1000 whereas the second field correspond to 0100110. The total number of bit used is 11 bits.

[00166] For example, if the probability in the range 0-100% is quantized (uniformly) with 2 bits, the Top-2, Top-3 and Top-4 beam, e.g. beams 2, 6 and 4 have probabilities 17%, 13% and 8%, respectively. The sum of the probabilities of beams 2, 6 and 4 is equal to 38%, which is quantized with 2 bits corresponding to 01. Thus, if option 3 is used for Top-1 beam, e.g. beam #7 which has probability 52%, the first field is 4-bit length and corresponds to 1000 whereas the second field correspond to 01. The total number of bit used is 6 bits.

[00167] The process for probability information reporting in beam prediction using probability concentration scheme in accordance with option 7 may be the same as the process shown in Fig. 8 for option 5, the difference will be at the quantization setting in operation 803. For option 7, the probability concentration setting in step 803 can refer to: number of quantized bits for Top 1 predicted beam, for example, 4 bits according to option 3; and number of quantized bits for the rest predicted beams, for example, 2 bits or 7 bits as mentioned above.

[00168] Hereinafter, option 8 for quantizing probability information of predicted beams in accordance with second embodiment family will be described, which has probability information reporting in beam prediction using vector quantization scheme.

[00169] In the second embodiment family, to further reduce the overhead of sharing beam probabilities, vector quantization (VQ) can be used. In scalar quantization (SQ) according to first embodiment family, each beam probability is quantized solely based on its own value and the quantization type and properties. However, in vector quantization, the beam probabilities are quantized for a group of beam probabilities. As the correlation between beam probabilities is taken into account in finding the optimal vector quantization points (codewords), VQ can reduce the overhead of sharing the beam probability sharing. A VQ codebook needs to be obtained and shared between UE and gNB to allow for coordinated quantization by the UE and dequantization by the gNB. The number of codewords in the VQ codebook depends on the total number of bits allocated to reporting Top-K beam’s probabilities. In other words, n-bits for reporting Top-K beam probabilities leads to a VQ codebook with 2n codewords.

[00170] Fig. 9 shows an example of quantizing Top-4 beam probabilities jointly using a vector quantization codebook with 8 codewords in accordance with some embodiment of the present disclosure.

[00171] As shown in Fig. 9, the vector quantization codebook is shared between the gNB and the UE, the vector quantization is used for reporting Top-4 beam’s probabilities, and 3bits (for example, 111 as shown in Fig. 9) is enough to quantize and share Top-4 beam probabilities.

[00172] A vector quantization codebook can be obtained based on the statistics of the beam’s probabilities of the trained ML model. The UE or the gNB can be in charge of obtaining or updating the VQ codebook. It is important to share the obtained / updated VQ codebook with the other entity to ensure correct de-quantization at the receiver. Fig. 10 shows an example for obtaining the vector quantization codebook based on the statistics of the beam’s probabilities of the trained ML model in accordance with some embodiment of the present disclosure. As shown in Fig. 10, the grey spots represent the vector of 1st beam probability and the 2nd beam probability based on the statistics of the beam’s probabilities of the trained ML model, and the black circle represent the codewords in the VQ codebook.

[00173] Table 10 shows an example of VQ codebook that can be used for quantizing jointly the Top-4 beam’s probabilities with 3-bits. The highest beam probability, the 2nd highest beam probability, the 3rd highest beam probability, the 4th highest beam probability as shown in the table is based on statistics of the beam’s probabilities of the trained ML model for per proper transmission and reception point (TRP) pattern. Table 10 An example ofVQ codebookfor quantizing Top-4 beam’s probabilities. Codeword Highest beam probability 0 95.3% 1 88.2% 2 3 74.5% 69.1% 4 | 5 64.4% 1 53.4% 6 42% 7 31.6% 2nd highest beam probability 3rd highest beam probability 2.1% 0.9% 5.5% 8.3% 12% 16.4% | 19.8% 25.6%> 28.4% 3.4% 3.6% 8.4% 11.2%) j 14.9% 18.6% | 21.4% 4th highest beam probability 0.4% 1.1% 2.5% 5.3% 6.7% \9.1% 10.3% | 15.4% Bits allocated to the Codeword 000 001 010 011 100 | 101 110 | 111 5

[00174] Table 11 shows an example of using vector quantization (VQ) for probability reporting of Top-4 beams from Set A with 64 beams. In this example, the VQ codebook shown in table 10 is used for quantization. Table 11 an example of using vector quantization (VO) for probability reporting of Top-4 beams from Set A with 64 beams Beam number Probability Quantized Value Quantized bits 17 52% 53.4% 101 42 17% 19.8% 56 13% 14.9% 4 8% 9.1% • • • • SUM 100%

[00175] As shown in table 10 and 11, the quantized level for beams #17, #42, #56, and #4 is close to the codeword 5, and then the bits 101 allocated to the codeword 5 is used for reporting the Top-4 probability information. Therefore, only three bits is needed for reporting Top-4 beams from Set A with 64 beams. The codebook shown in table 10 has 23 codewords, that is to say, n-bits for reporting Top-K beam probabilities leads to a VQ codebook with 2n codewords. Therefore, the overhead of probability reporting using the vector quantization codebook is as follows: Top-4 (beam 17&42&56&4): 3 bits. It can be seen that VQ quantization approach can reduce the overhead of sharing beam probabilities significantly.

[00176] Hereinafter, some example embodiments of process 1100 for probability information reporting in beam prediction using vector quantization scheme will be described with reference to Fig. 11.

[00177] As shown in Fig. 11, at 1101, the UE 1110 reports capability on vector quantization for reporting probability information during inference. At 1102, the gNB 1120 selects quantization setting for beam probability report, for example, the gNB 1120 selects vector quantization setting including the VQ codebook for beam probability report. At 1103, the gNB 1120 transmits vector quantization setting including the VQ codebook via higher-layer signaling for example, SIB1 and RMSI.

[00178] At 1104, the gNB 1120 configures vector quantization via CSIreportconfig. At 1105, the UE 1110 receives RRC configuration and applies vector quantization scheme for reporting Top-K probability information. At 1106, the UE 1110 reports Top-K predicted beams and Top-K vector quantization probability information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 101. At 1107, the gNB 1120 receives Top-K CSI-RS report and de-quantizes probability information with respect to Top-K beams using the shared VQ codebook.

[00179] Hereinafter, some example embodiments of process 1200 for probability information reporting in beam prediction using vector quantization scheme will be described with reference to Fig. 12. In the embodiment as shown in Fig. 11, the VQ codebook is obtained by the gNB and shared with the UE. however, in some other embodiments as shown in Fig. 12, the VQ codebook can be obtained by the UE and shared with the gNB.

[00180] As shown in Fig. 12, at 1201, the UE 1210 selects vector quantization setting including the VQ codebook for reporting probability information during inference. At 1202, the UE 1210 reports capability on vector quantization for reporting probability information during inference including sharing quantization setting. At 1203, the gNB 1220 configures vector quantization via CSIreportconfig. At 1204, the UE 1210 receives RRC configuration and applies vector quantization scheme for reporting Top-K probability information. At 1205, the UE 1210 reports Top-K predicted beams and Top-K vector quantization probability information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 101. At 1206, the gNB 1220 receives Top-K CSI-RS report and de-quantizes probability information with respect to Top-K beams using the received VQ codebook from the UE.

[00181] Hereinafter, option 9 for quantizing probability information of predicted beams in accordance with second embodiment family will be described, which has probability information reporting in beam prediction using vector quantization scheme, total probability concentration reporting indication, and comparison of probability of 1st beam with other beam probabilities.

[00182] This option 9 allows for sharing the quantized probabilities using a vector quantization and some statistics on the (original / un-quantized) beam probabilities. For example, table 12 shows an example of using probability concentration reporting indication for probability reporting of Top-4 beams besides sharing quantized probabilities using VQ. Table 12 can be considered for the relative probability value reporting, where X and Y can be either defined by the specification or configured to the UE. In one example, X = 50% and Y = 70%. Table 12 an example of using probability concentration reporting indication for probability reporting of Top-4 beams besides sharing quantized probabilities using VQ Index Reporting bits Pl (probability value of best beam in Top-K) >= X% compared to P2 (probability value of second-best beam in Top-K) Probability concentration on Top-Kbeams (i.e., Pl + P2 + P3 ..+ PK) 0 00 Yes >= Y % 1 01 No > = Y% 2 10 Yes <Y% 3 11 No <Y%

[00183] The reporting bits of the shared probability statistics only provide a relative assessment on un-quantized or original probability values determined at the UE for predicted beams. For example, Top-4 beam reported from 64 beams (Set A) may be considered, i.e. 1 st best beam, 2nd best beam, 3rd best beam, 4th best beam having pl, p2, p3, p4 as the absolute probabilities at the model output may be considered. What matters is to report the following: (1) whether pl stands out from the rest, because the NW can directly use this beam if this is the case, and this can be extended pl, p2 stands out of the rest, etc.., depending on K value; and (2) what is the total probability concentration on pl + p2 + p3 + p4 (e.g., above 50%?). If less concentrated, the NW can switch to Top-8 beams reporting, that is to say, the NW will take essential actions.

[00184] Hereinafter, some example embodiments of process 1300 for probability information reporting in beam prediction using vector quantization and probability concentration will be described with reference to Fig. 13.

[00185] As shown in Fig. 13, at 1301, the UE 1310 reports capability on vector quantization and probability concentration indication for reporting probability information during inference. At 1302, the gNB 1320 selects quantization setting for beam probability report, for example, the gNB 1320 selects vector quantization setting including the VQ codebook and quantized probability concentration setting for beam probability report. At 1303, the gNB 1320 transmits vector quantization setting including the VQ codebook via higher-layer signaling for example, SIB1 and RMSI.

[00186] At 1304, the gNB 1320 configures vector quantization and probability concentration via CSIreportconfig. At 1305, the UE 1310 receives RRC configuration and applies vector quantization and probability concentration scheme for reporting Top-K probability information. At 1306, the UE 1310 reports Top-K predicted beams and Top-K vector quantization probability information and Top-K probability concentration information, for example, report beams #17, #42, #56, and #4, and the quantized bits for them will be 101, and the quantized probability concentration for them will be 10. At 1307, the gNB 1320 receives Top-K CSI-RS report and de-quantizes probability information with respect to Top-K beams using the shared VQ codebook and de-quantizes the probability concentration information using the reported quantized probability concentration.

[00187] Fig. 14 illustrates a flowchart of an example method 1400 implemented at a terminal device in accordance with some other embodiments of the present disclosure. For the purpose of discussion, the method 1400 will be described from the perspective of the terminal device 110 with reference to Fig. 1.

[00188] At block 1410, the terminal device transmits, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. At block 1420, the terminal device receives, from the network device, at least one of quantization setting or quantization scheme configuration. At block 1430, the terminal device transmits to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[00189] In some embodiments, the quantization setting is received via a higher-layer signaling; and the quantization scheme configuration is received via a channel state information (CSI) report configuration. In some embodiments, the probability information is quantized by the terminal device by using the scalar quantization scheme; and the terminal device further maps the at least one predicted beam each having a probability value within a certain range to at least one quantized level; and quantizes the probability information associated with the at least one predicted beam based on the at least one quantized level.

[00190] In some embodiments, the supported quantization scheme comprises uniform quantization in which the probability information associated with the at least one predicted beam is uniformly quantized based on a number of quantized bits.

[00191] In some embodiments, the quantization setting for the uniform quantization comprises: a number of quantized bits; and a number of the at least one predicted beam. In some embodiments, the supported quantization scheme comprises different granularity quantization in which the probability information associated with a first beam of the at least one predicted beam is quantized based on a number of quantized bits that is different from a number of quantized bits based on which remaining predicted beams among the at least one predicted beam is quantized.

[00192] In some embodiments, the quantization setting for the different granularity quantization comprises: a first number of quantized bits for the first beam; a second number of quantized bits for the remaining predicted beams; and a third number of the at least one predicted beam.

[00193] In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the first number of quantized bits, and the probability information associated with the remaining predicted beams in the certain quantization range is uniformly quantized with the second number of quantized bits.

[00194] In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the first number of bits, and the probability information for each of the reaming beams in a quantization range from zero to a quantized value of a probability value of a previous beam is uniformly quantized with the second number of quantized bits.

[00195] In some embodiments, the quantization setting for the different granularity quantization comprises: a number of quantized bits for a first beam among the at least one predicted beam; a number of quantized bits for each of remaining predicted beams among the at least one predicted beam; and a number of the at least one predicted beam. In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the number of quantized bits for the first beam, and the probability information associated with a target one of the remaining predicted beams in a quantization range from zero to a quantized value of a probability value of a previous beam is quantized with a target number of quantized bits less than a number of quantized bits associated with the previous beam.

[00196] In some embodiments, the probability information associated with the at least one predicted beam is quantized based on statistics of the probability values of the at least one predicted beam; and the supported quantization scheme comprises probability concentration indication for reporting beam probability information, in which quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam. In some embodiments, the quantization setting for the probability concentration indication comprises: a number of quantized bits for report; a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; a total probability of the at least one predicted beam, and a number of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; and a number of the at least one predicted beam. In some embodiments, the terminal device further quantizes the probability information associated with the at least one predicted beam at least based on mapping between the probability difference, quantized bits and the total probability.

[00197] In some embodiments, the quantization setting for the probability concentration indication: a number of quantized bits for report. In some embodiments, the terminal device reports the at least one predicted beam and an aggregate of probability values of the at least one predicted beam via a field of the number of quantized bits in a CSI report.

[00198] In some embodiments, the quantization setting for the probability concentration indication: a first number of quantized bits for reporting probability information of a first beam of the at least one predicted beam; and a second number of quantized bits for reporting probability information of remaining predicted beams of the at least one predicted beam. In some embodiments, the terminal device reports the at least one predicted beam and the probability information of the first beam via a first field of the first number of quantized bits in a CSI report and an aggregate of probability values of remaining predicted beams via a second field of a second number of quantized bits in the CSI report.

[00199] In some embodiments, a higher-layer signaling association information is received via at least one of the following: a message; a system information; or a downlink control information.

[00200] Fig. 15 illustrates a flowchart of an example method 1500 implemented at a network device in accordance with some other embodiments of the present disclosure. For the purpose of discussion, the method 1500 will be described from the perspective of the network device 120 with reference to Fig. 1.

[00201] At block 1510, the network device receives, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. At block 1520, the network device transmits to the terminal device, at least one of quantization setting or quantization scheme configuration. At block 1530, the network device receives from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[00202] In some embodiments, the quantization setting is transmitted via a higher-layer signaling; and the quantization scheme configuration is transmitted via a channel state information (CSI) report configuration.

[00203] In some embodiments, the probability information quantized by the terminal device is based on the scalar quantization scheme. In some embodiments, the supported quantization scheme comprises uniform quantization in which the probability information associated with the at least one predicted beam is uniformly quantized based on a number of quantized bits.

[00204] In some embodiments, the quantization setting for the uniform quantization comprises: a number of quantized bits; and a number of the at least one predicted beam.

[00205] In some embodiments, the supported quantization scheme comprises different granularity quantization in which the probability information associated with a first beam of the at least one predicted beam is quantized based on a number of quantized bits that is different from a number of quantized bits based on which remaining predicted beams among the at least one predicted beam is quantized.

[00206] In some embodiments, the quantization setting for the different granularity quantization comprises: a first number of quantized bits for the first beam; a second number of quantized bits for the remaining predicted beams; and a third number of the at least one predicted beam. In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the first number of quantized bits, and the probability information associated with the remaining predicted beams in the certain quantization range is uniformly quantized with the second number of quantized bits. In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the first number of bits, and the probability information for each of the reaming beams in a quantization range from zero to a quantized value of a probability value of a previous beam is uniformly quantized with the second number of quantized bits.

[00207] In some embodiments, the quantization setting for the different granularity quantization comprises: a number of quantized bits for a first beam among the at least one predicted beam; a number of quantized bits for each of remaining predicted beams among the at least one predicted beam; and a number of the at least one predicted beam. In some embodiments, the probability information associated with the first beam in a certain quantization range is quantized with the number of quantized bits for the first beam, and the probability information associated with a target one of the remaining predicted beams in a quantization range from zero to a quantized value of a probability value of a previous beam is quantized with a target number of quantized bits less than a number of quantized bits associated with the previous beam.

[00208] In some embodiments, the probability information quantized by the terminal device is based on statistics of the probability values of the at least one predicted beam; and the supported quantization scheme comprises probability concentration indication for reporting beam probability information, in which quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam. In some embodiments, the quantization setting for the probability concentration indication comprises: a number of quantized bits for report; a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; a total probability of the at least one predicted beam, and a number of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; and a number of the at least one predicted beam.

[00209] In some embodiments, the quantized bits for probability concentration information is received by the network device via the number of quantized bits for report. In some embodiments, the quantization setting for the probability concentration indication: a number of quantized bits for report; and an aggregate of probability values of the at least one predicted beam is received by the network device via a field of the number of quantized bits in a CSI report.

[00210] In some embodiments, the quantization setting for the probability concentration indication comprises: a first number of quantized bits for reporting probability information of a first beam of the at least one predicted beam; and a second number of quantized bits for reporting probability information of remaining predicted beams of the at least one predicted beam; and the probability information of the first beam is received by the network device via a first field of the first number of quantized bits in a CSI report, and an aggregate of probability values of remaining predicted beams is received by the network device via a second field of a second number of quantized bits in the CSI report.

[00211] In some embodiments, the network device further selects the quantization setting for reporting beam probability information based on the received capability on supported quantization scheme. In some embodiments, the network device further de-quantizes the probability information based on the quantized probability information.

[00212] In some embodiments, a higher-layer signaling association information is transmitted via at least one of the following: a message; a system information; or a downlink control information.

[00213] Fig. 16 illustrates a flowchart of an example method 1600 implemented at a terminal device in accordance with some other embodiments of the present disclosure. For the purpose of discussion, the method 1600 will be described from the perspective of the terminal device 110 with reference to Fig. 1.

[00214] At block 1610, the terminal device transmits, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. At block 1620, the terminal device obtains, at least one of quantization setting or quantization scheme configuration. At block 1630, the terminal device transmits to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00215] In some embodiments, the probability information is quantized by the terminal device by using the vector quantization scheme, and the supported quantization scheme comprises vector quantization for reporting beam probability information.

[00216] In some embodiments, the quantization setting for the vector quantization comprises a vector quantization codebook for beam probability report, probabilities for the at least one predicted beam is quantized for a group of beam probabilities, and the probability information associated with the at least one predicted beam is quantized based on the vector quantization codebook.

[00217] In some embodiments, the vector quantization codebook for beam probability report is determined based on the statistics of beam probabilities of a trained machine-learning model. In some embodiments, the vector quantization codebook for beam probability report is determined by the terminal device; and the vector quantization codebook is transmitted to the network device along with the capability on supported quantization scheme for reporting beam probability information. In some embodiments, the vector quantization codebook is determined by the network device and is received from the network device. In some embodiments, a number of codewords in the vector quantization codebook is based on a total number of bits allocated for beam probability report.

[00218] In some embodiments, the probability information is further quantized by the terminal device by using statistics of probability values of the at least one predicted beam; and the supported quantization scheme further comprises probability concentration indication for reporting beam probability information, the probability information associated with the at least one predicted beam is further quantized based on statistics of the probability values of the at least one predicted beam, and quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam.

[00219] In some embodiments, the quantization setting for the probability concentration indication comprises: a number of quantized bits for report; a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; a total probability of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; and a number of the at least one predicted beam.

[00220] In some embodiments, the terminal device further quantizes the probability information associated with the at least one predicted beam based on mapping between the probability difference, quantized bits, and the total probability.

[00221] In some embodiments, the quantization setting is received via a higher-layer signaling; and the quantization scheme configuration is received via a channel state information (CSI) report configuration. In some embodiments, a higher-layer signaling association information is received via at least one of the following: a message; a system information; or a downlink control information.

[00222] Fig. 17 illustrates a flowchart of an example method 1700 implemented at a network device in accordance with some other embodiments of the present disclosure. For the purpose of discussion, the method 1700 will be described from the perspective of the network device 120 with reference to Fig. 1.

[00223] At block 1710, the network device receives, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference. At block 1720, the network device obtains at least one of quantization setting or quantization scheme configuration. At block 1730, the network device receives, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00224] In some embodiments, the supported quantization scheme comprises vector quantization for reporting beam probability information. In some embodiments, the quantization setting for the vector quantization comprises a vector quantization codebook for beam probability report, probabilities for the at least one predicted beam is quantized for a group of beam probabilities, and quantized probability information associated with the at least one predicted beam is based on the vector quantization codebook.

[00225] In some embodiments, the vector quantization codebook is determined based on the statistics of beam probabilities of a trained machine-learning model. In some embodiments, the vector quantization codebook for beam probability report is determined by the network device. In some embodiments, the vector quantization codebook is received from the terminal device along with the capability on supported quantization scheme for reporting beam probability information. In some embodiments, a number of codewords in the vector quantization codebook is based on a total number of bits allocated for beam probability report.

[00226] In some embodiments, quantized probability information is further based on statistics of probability values of the at least one predicted beam; and the supported quantization scheme further comprises probability concentration indication for reporting beam probability information; the probability information associated with the at least one predicted beam is further quantized based on statistics of the probability values of the at least one predicted beam, and quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam.

[00227] In some embodiments, the quantization setting for the probability concentration indication comprises: a number of quantized bits for report; a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; a total probability of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; and a number of the at least one predicted beam.

[00228] In some embodiments, the quantization setting is transmitted via a higher-layer signaling; and the quantization scheme configuration is transmitted via a channel state information (CSI) report configuration.

[00229] In some embodiments, the network device further selects the quantization setting for reporting beam probability information based on the received capability on supported quantization scheme. In some embodiments, the network device further de-quantizes the probability information based on the vector quantization codebook. In some embodiments, the network device further de-quantizes the probability concentration information based on received quantized probability concentration information for the at least one predicted beam.

[00230] In some embodiments, a higher-layer signaling association information is transmitted via at least one of the following: a message; a system information; or a downlink control information.

[00231] In some embodiments, an apparatus (for example, the terminal device 110) capable of performing the method 1400 may comprise means for performing the respective steps of the method 1400. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[00232] In some embodiments, the apparatus comprises: means for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for receiving, from the network device, at least one of quantization setting or quantization scheme configuration; and means for transmitting to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[00233] In some embodiments, the apparatus further comprises means for performing other steps in some embodiments of the method 1400. In some embodiments, the means comprises at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[00234] In some embodiments, an apparatus (for example, the network device 120) capable of performing the method 1500 may comprise means for performing the respective steps of the method 1500. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[00235] In some embodiments, the apparatus comprises means for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for transmitting to the terminal device, at least one of quantization setting or quantization scheme configuration; and means for receiving from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of scalar quantization scheme or statistics of probability values of the at least one predicted beam.

[00236] In some embodiments, the apparatus further comprises means for performing other steps in some embodiments of the method 1500. In some embodiments, the means comprises at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[00237] In some embodiments, an apparatus (for example, the terminal device 110) capable of performing the method 1600 may comprise means for performing the respective steps of the method 1600. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[00238] In some embodiments, the apparatus comprises: means for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for obtaining, at least one of quantization setting or quantization scheme configuration; and means for transmitting to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00239] In some embodiments, the apparatus further comprises means for performing other steps in some embodiments of the method 1600. In some embodiments, the means comprises at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[00240] In some embodiments, an apparatus (for example, the network device 120) capable of performing the method 1700 may comprise means for performing the respective steps of the method 1700. The means may be implemented in any suitable form. For example, the means may be implemented in a circuitry or software module.

[00241] In some embodiments, the apparatus comprises means for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for obtaining at least one of quantization setting or quantization scheme configuration; and means for receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00242] In some embodiments, the apparatus further comprises means for performing other steps in some embodiments of the method 1700. In some embodiments, the means comprises at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the at least one processor, cause the performance of the apparatus.

[00243] Fig. 18 is a simplified block diagram of a device 1800 that is suitable for implementing embodiments of the present disclosure. The device 1800 may be provided to implement the communication device, for example the terminal device 110, the network device 120 as shown in Fig. 1. As shown, the device 1800 includes one or more processors 1810, one or more memories 1820 coupled to the processor 1810, and one or more communication modules 1840 coupled to the processor 1810.

[00244] The communication module 1840 is for bidirectional communications. The communication module 1840 has at least one antenna to facilitate communication. The communication interface may represent any interface that is necessary for communication with other network devices.

[00245] The processor 1810 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 1800 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.

[00246] The memory 1820 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) 1824, an electrically programmable read only memory (EPROM), a flash memory, a hard disk, a compact disc (CD), a digital video disk (DVD), and other magnetic storage and / or optical storage. Examples of the volatile memories include, but are not limited to, a random access memory (RAM) 1822 and other volatile memories that may not last in the power-down duration.

[00247] A computer program 1830 includes computer executable instructions that are executed by the associated processor 1810. The program 1830 may be stored in the ROM 1824. The processor 1810 may perform any suitable actions and processing by loading the program 1830 into the RAM 1822.

[00248] The embodiments of the present disclosure may be implemented by means of the program so that the device 1800 may perform any process of the disclosure as discussed with reference to Figs. 2 to 8, and 11 to 17. The embodiments of the present disclosure may also be implemented by hardware or by a combination of software and hardware.

[00249] In some embodiments, the program 1830 may be tangibly contained in a computer readable medium which may be included in the device 1800 (such as in the memory 1820) or other storage devices that are accessible by the device 1800. The device 1800 may load the program 1830 from the computer readable medium to the RAM 1822 for execution. The computer readable medium may include any types of tangible non-volatile storage, such as ROM, EPROM, a flash memory, a hard disk, CD, DVD, and the like.

[00250] Fig. 19 illustrates an example of the computer readable medium 900 in form of CD or DVD in accordance with some embodiments of the present disclosure. The computer readable medium has the program 1830 stored thereon. It is noted that although the computer-readable medium 1900 is depicted in form of CD or DVD, the computer-readable medium 1900 may be in any other form suitable for carry or hold the program 1830.

[00251] Generally, various embodiments of the present disclosure may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. While various aspects of embodiments of the present disclosure are illustrated and described as block diagrams, flowcharts, or using some other pictorial representations, it is to be understood that the block, apparatus, system, technique or method described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

[00252] The present disclosure also provides at least one computer program product tangibly stored on a non-transitory computer readable storage medium. The computer program product includes computer-executable instructions, such as those included in program modules, being executed in a device on a target real or virtual processor, to carry out the methods as described above with reference to Fig. 2 to Fig. 8, and Figs. 11 to 17. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, or the like that perform particular tasks or implement particular abstract data types. The functionality of the program modules may be combined or split between program modules as desired in various embodiments. Machine-executable instructions for program modules may be executed within a local or distributed device. In a distributed device, program modules may be located in both local and remote storage media.

[00253] Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[00254] In the context of the present disclosure, the computer program codes or related data may be carried by any suitable carrier to enable the device, apparatus or processor to perform various processes and operations as described above. Examples of the carrier include a signal, computer readable medium, and the like.

[00255] The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable medium may include but not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the computer readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. The term “non-transitory,” as used herein, is a limitation of the medium itself (i.e., tangible, not a signal) as opposed to a limitation on data storage persistency (e.g., RAM vs. ROM).

[00256] Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are contained in the above discussions, these should not be construed as limitations on the scope of the present disclosure, but rather as descriptions of features that may be specific to particular embodiments. Certain features that are described in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, various features that may be described in the context of a single embodiment may also be implemented in multiple embodiments separately or in any suitable sub-combination.

[00257] 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 may be disclosed as example forms of implementing the claims or any of the below embodiments.

[00258] Embodiment 1. A method implemented at a terminal device, comprising: transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtaining, at least one of quantization setting or quantization scheme configuration; and transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00259] Embodiment 2. A method implemented at a network device, comprising: receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; obtaining, at least one of quantization setting or quantization scheme configuration; and receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00260] Embodiment 3. an apparatus comprising: means for transmitting, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for obtaining, at least one of quantization setting or quantization scheme configuration; and means for transmitting, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00261] Embodiment 4. An apparatus comprising: means for receiving, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference; means for obtaining, at least one of quantization setting or quantization scheme configuration; and means for receiving, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

[00262] Embodiment 5. A non-transitoiy computer readable medium comprising program instructions for causing an apparatus to perform at least the method of embodiment 1 or 2.

Claims

1. A terminal device comprising:at least one processor; andat least one memory storing instructions that, when executed by the processor, cause the terminal device at least to:transmit, to a network device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference;obtain, at least one of quantization setting or quantization scheme configuration; andtransmit, to the network device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

2. The terminal device of claim 1, wherein the probability information is quantized by the terminal device by using the vector quantization scheme, andwherein the supported quantization scheme comprises vector quantization for reporting beam probability information.

3. The terminal device of claim 2, wherein the quantization setting for the vector quantization comprises a vector quantization codebook for beam probability report, wherein probabilities for the at least one predicted beam is quantized for a group of beam probabilities, andthe probability information associated with the at least one predicted beam is quantized based on the vector quantization codebook.

4. The terminal device of claim 3, wherein the vector quantization codebook for beam probability report is determined based on the statistics of beam probabilities of a trained machine-learning model.

5. The terminal device of claim 3, wherein the vector quantization codebook for beam probability report is determined by the terminal device; andwherein the vector quantization codebook is transmitted to the network device along with the capability on supported quantization scheme for reporting beam probability information.

6. The terminal device of claim 3, wherein the vector quantization codebook is received from the network device.

7. The terminal device of any of claims 3 to 6, wherein a number of codewords in the vector quantization codebook is based on a total number of bits allocated for beam probability report.

8. The terminal device any of claims 2 to 7, wherein the probability information is further quantized by the terminal device by using statistics of probability values of the at least one predicted beam; andwherein the supported quantization scheme further comprises probability concentration indication for reporting beam probability information,the probability information associated with the at least one predicted beam is further quantized based on statistics of the probability values of the at least one predicted beam, and quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam.

9. The terminal device of claim 8, wherein the quantization setting for the probability concentration indication comprises:a number of quantized bits for report;a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam;a total probability of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; anda number of the at least one predicted beam.

10. The terminal device of claim 9, wherein the terminal device is further caused to:quantize the probability information associated with the at least one predicted beam based on mapping between the probability difference, quantized bits, and the total probability.

11. The terminal device of any of claims 1 to 10, wherein the quantization setting is received via a higher-layer signaling; andthe quantization scheme configuration is received via a channel state information (CSI) report configuration.

12. The terminal device of claim 11, wherein a higher-layer signaling association information is received via at least one of the following:a message;a system information; ora downlink control information.

13. A network device comprising:at least one processor; andat least one memory storing instructions that, when executed by the processor, cause the network device at least to:receive, from a terminal device, capability on supported quantization scheme for reporting beam probability information associated with at least one beam predicted during inference;obtain, at least one of quantization setting or quantization scheme configuration; andreceive, from the terminal device, report for at least one predicted beam and the probability information that is quantized by the terminal device based on at least one of the quantization setting or the quantization scheme configuration by using at least one of vector quantization scheme or statistics of probability values of the at least one predicted beam.

14. The network device of claim 13, wherein the supported quantization scheme comprises vector quantization for reporting beam probability information.

15. The network device of claim 14, wherein the quantization setting for the vector quantization comprises a vector quantization codebook for beam probability report, wherein probabilities for the at least one predicted beam is quantized for a group of beam probabilities, andquantized probability information associated with the at least one predicted beam is based on the vector quantization codebook.

16. The network device of claim 15, wherein the vector quantization codebook is determined based on the statistics of beam probabilities of a trained machine-learning model.

17. The network device of claim 15, wherein the vector quantization codebook for beam probability report is determined by the network device.

18. The network device of claim 15, wherein the vector quantization codebook is received from the terminal device along with the capability on supported quantization scheme for reporting beam probability information.

19. The network device of any of claims 15 to 18, wherein a number of codewords in the vector quantization codebook is based on a total number of bits allocated for beam probability report.

20. The network device any of claims 14 to 19, wherein quantized probability information is further based on statistics of probability values of the at least one predicted beam; andwherein the supported quantization scheme further comprises probability concentration indication for reporting beam probability information;the probability information associated with the at least one predicted beam is further quantized based on statistics of the probability values of the at least one predicted beam, and quantized bits for probability concentration information indicate a relative assessment on probability values of the at least one predicted beam.

21. The network device of claim 20, wherein the quantization setting for the probability concentration indication comprises:a number of quantized bits for report;a probability difference between probability value of a first beam and probability value of a second beam among the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam;a total probability of the at least one predicted beam based on the statistics of the probability values of the at least one predicted beam; anda number of the at least one predicted beam.

22. The network device of any of claims 13 to 21, wherein the quantization setting is transmitted via a higher-layer signaling; andthe quantization scheme configuration is transmitted via a channel state information (CSI) report configuration.

23. The network device of any of claims 13 to 22, wherein the network device is further caused to:select the quantization setting for reporting beam probability information based on the received capability on supported quantization scheme.

24. The network device of any of claims 15 to 23, wherein the network device is further caused to:dequantize the probability information based on the vector quantization codebook.

25. The network device of any of claims 20 to 23, wherein the network device is further caused to:dequantize the probability concentration information based on received quantized probability concentration information for the at least one predicted beam.

Images