Channel metric usage in wireless communications
AI and ML models are employed to enhance CSI reporting accuracy and reduce overhead in wireless communication systems by deriving and reconstructing channel metric sets from CSI-RS, addressing the challenges of existing systems.
Patent Information
- Application Number
- PCT/CN2024/106363
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-07-19
- Publication Date
- 2025-07-17
AI Technical Summary
Existing wireless communication systems face challenges in improving the accuracy of channel state information (CSI) reporting and reducing the overhead of CSI feedback.
The use of artificial intelligence (AI) and machine learning (ML) models to derive and reconstruct channel metric sets based on channel state information reference signals (CSI-RS) at both user devices and network devices, enabling more accurate CSI reporting and reducing feedback overhead.
Enhances the accuracy of CSI reporting while minimizing feedback overhead, thereby improving scheduling efficiency in wireless communication systems.
Smart Images

Figure CN2024106363_17072025_PF_FP_ABST
Abstract
Description
CHANNEL METRIC USAGE IN WIRELESS COMMUNICATIONSTECHNICAL FIELD
[0001] This document is directed generally to channel metric usage in wireless communications.BACKGROUND
[0002] In wireless communication, a base station transmits a channel state information reference signal (CSI-RS) to a user device. In response, the user device derives the channel state information (CSI) , and reports the CSI to the base station. In response, the base station performs scheduling based on the CSI. Ways to improve the accuracy of the CSI and / or ways to reduce overhead of CSI feedback may be desirable.SUMMARY
[0003] This document relates to methods, systems, apparatuses and devices for wireless communication. In some implementations, a method for wireless communication includes: receiving, by a user device, a channel state information reference signal (CSI-RS) ; determining, by the user device, a measured channel metric based on the CSI-RS; and deriving, by the user device, at least one channel metric set based on the measured channel metric, each channel metric set comprising one or more channel metrics.
[0004] In some other implementations, a method for wireless communication includes: transmitting, by a network device, a channel state information reference signal (CSI-RS) ; receiving, by the network device, at least one channel information (CSI) set derived based on the CSI-RS; and reconstructing, by the network device, at least one channel metric set based on the at least one CSI set, each channel metric set comprising one or more channel metrics.
[0005] In some other implementations, a device, such as a network device, is disclosed. The device may include one or more processors and one or more memories, wherein the one or more processors are configured to read computer code from the one or more memories to implement any of the methods above.
[0006] In yet some other implementations, a computer program product is disclosed. The computer program product may include a non-transitory computer-readable program medium with computer code stored thereupon, the computer code, when executed by one or more processors, causing the one or more processors to implement any of the methods above.
[0007] The above and other aspects and their implementations are described in greater detail in the drawings, the descriptions, and the claims.BRIEF DESCRIPTION OF THE DRAWINGS
[0008] FIG. 1 shows a block diagram of an example of a wireless communication system.
[0009] FIG. 2 shows a flow chart of a method for wireless communication.
[0010] FIG. 3 shows a flow chart of another method for wireless communication.
[0011] FIG. 4 shows a diagram of an example of an artificial intelligence (AI) and / or machine learning (ML) model at a user device deriving a channel state information (CSI) set based on a channel metric set input to the AI and / or ML model.
[0012] FIG. 5 shows a diagram of another example of an AI and / or ML model at a user device deriving a CSI set based on a channel metric set input to the AI and / or ML model.
[0013] FIG. 6 shows a diagram of an example of an AI and / or ML model at a network device deriving or reconstructing a channel metric set based on a CSI set input to the AI and / or ML model.
[0014] FIG. 7 shows a diagram of another example of an AI and / or ML at a network device deriving or reconstructing a channel metric set based on a CSI set input to the AI and / or ML model.
[0015] FIG. 8 show a diagram of an example of an AI and / or ML model at a user device deriving multiple CSI sets based on multiple channel metric sets input to the AI and / or ML model.
[0016] FIG. 9 shows a diagram of another example of an AI and / or ML model at a user device deriving multiple CSI sets based on multiple channel metric sets input to the AI and / or ML model.
[0017] FIG. 10 shows a diagram of an example of an AI and / or ML model at a network device deriving or reconstructing multiple channel metric sets based on multiple CSI sets input to the AI and / or ML model.
[0018] FIG. 11 shows a diagram of another example of an AI and / or ML model at a network device deriving or reconstructing multiple channel metric sets based on multiple CSI sets input to the AI and / or ML model.
[0019] FIG. 12 shows a diagram of another example of an AI and / or ML model at a user device deriving multiple CSI sets based on multiple channel metric sets input to the AI and / or ML model.
[0020] FIG. 13 shows a diagram of another example of an AI and / or ML model at a network device deriving or reconstructing multiple channel metric sets based on multiple CSI sets input to the AI and / or ML model.
[0021] FIG. 14 shows a diagram of another example of an AI and / or ML model at the user device deriving multiple CSI sets based on multiple channel metric sets input to the AI and / or ML model.
[0022] FIG. 15 shows a diagram of another example of an AI and / or ML model at a network device deriving or reconstructing multiple channel metric sets based on multiple CSI sets input to the AI / and / or ML model.
[0023] FIG. 16 shows a diagram of an example of an AI and / or ML model at a user device deriving multiple CSI sets based on multiple channel metric sets.
[0024] FIG. 17 shows a diagram of an example of an AI / ML model at a user device deriving multiple CSI sets based on multiple channel metric sets.
[0025] FIG. 18 shows a diagram of an example of an AI / ML model at a user device deriving multiple CSI sets based on multiple channel metric sets.
[0026] FIG. 19 shows a diagram of an example of an AI / ML model at a user device deriving multiple CSI sets based on multiple channel metric sets.
[0027] FIG. 20 shows a diagram of an example of an AI / ML model at a user device deriving multiple CSI sets based on multiple channel metric sets.
[0028] FIG. 21 shows a diagram of an example of an AI / ML model at a user device deriving multiple CSI sets based on multiple channel metric sets.DETAILED DESCRIPTION
[0029] The present description describes various embodiments of systems, apparatuses, devices, and methods for wireless communications related to channel metrics.
[0030] Fig. 1 shows a diagram of an example wireless communication system 100 including a plurality of communication nodes (or just nodes) that are configured to wirelessly communicate with each other. In general, the communication nodes include at least one user device 102 and at least one network device 104. The example wireless communication system 100 in Fig. 1 is shown as including two user devices 102, including a first user device 102 (1) and a second user device 102 (2) , and one network device 104. However, various other examples of the wireless communication system 100 that include any of various combinations of one or more user devices 102 and / or one or more network devices 104 may be possible.
[0031] In general, a user device as described herein, such as the user device 102, may include a single electronic device or apparatus, or multiple (e.g., a network of) electronic devices or apparatuses, capable of communicating wirelessly over a network. A user device may comprise or otherwise be referred to as a user terminal, a user terminal device, or a user equipment (UE) . Additionally, a user device may be or include, but not limited to, a mobile device (such as a mobile phone, a smart phone, a smart watch, a tablet, a laptop computer, vehicle or other vessel (human, motor, or engine-powered, such as an automobile, a plane, a train, a ship, or a bicycle as non-limiting examples) or a fixed or stationary device, (such as a desktop computer or other computing device that is not ordinarily moved for long periods of time, such as appliances, other relatively heavy devices including Internet of things (IoT) , or computing devices used in commercial or industrial environments, as non-limiting examples) . In various embodiments, a user device 102 may include transceiver circuitry 106 coupled to an antenna 108 to effect wireless communication with the network device 104. The transceiver circuitry 106 may also be coupled to a processor 110, which may also be coupled to a memory 112 or other storage device. The memory 112 may store therein instructions or code that, when read and executed by the processor 110, cause the processor 110 to implement various ones of the methods described herein.
[0032] Additionally, in general, a network device as described herein, such as the network device 104, may include a single electronic device or apparatus, or multiple (e.g., a network of) electronic devices or apparatuses, and may comprise one or more wireless access nodes, base stations, or other wireless network access points capable of communicating wirelessly over a network with one or more user devices and / or with one or more other network devices 104. For example, the network device 104 may comprise a 4G LTE base station, a 5G NR base station, a 5G central-unit base station, a 5G distributed-unit base station, a next generation Node B (gNB) , an enhanced Node B (eNB) , or other similar or next-generation (e.g., 6G) base stations, in various embodiments. A network device 104 may include transceiver circuitry 114 coupled to an antenna 116, which may include an antenna tower 118 in various approaches, to effect wireless communication with the user device 102 or another network device 104. The transceiver circuitry 114 may also be coupled to one or more processors 120, which may also be coupled to a memory 122 or other storage device. The memory 122 may store therein instructions or code that, when read and executed by the processor 120, cause the processor 120 to implement one or more of the methods described herein.
[0033] In various embodiments, two communication nodes in the wireless system 100-such as a user device 102 and a network device 104, two user devices 102 without a network device 104, or two network devices 104 without a user device 102-may be configured to wirelessly communicate with each other in or over a mobile network and / or a wireless access network according to one or more standards and / or specifications. In general, the standards and / or specifications may define the rules or procedures under which the communication nodes can wirelessly communicate, which, in various embodiments, may include those for communicating in millimeter (mm) -Wave bands, and / or with multi-antenna schemes and beamforming functions. In addition or alternatively, the standards and / or specifications are those that define a radio access technology and / or a cellular technology, such as Fourth Generation (4G) Long Term Evolution (LTE) , Fifth Generation (5G) New Radio (NR) , or New Radio Unlicensed (NR-U) , as non-limiting examples.
[0034] Additionally, in the wireless system 100, the communication nodes are configured to wirelessly communicate signals between each other. In general, a communication in the wireless system 100 between two communication nodes can be or include a transmission or a reception, and is generally both simultaneously, depending on the perspective of a particular node in the communication. For example, for a given communication between a first node and a second node where the first node is transmitting a signal to the second node and the second node is receiving the signal from the first node, the first node may be referred to as a source or transmitting node or device, the second node may be referred to as a destination or receiving node or device, and the communication may be considered a transmission for the first node and a reception for the second node. Of course, since communication nodes in a wireless system 100 can both send and receive signals, a single communication node may be both a transmitting / source node and a receiving / destination node simultaneously or switch between being a source / transmitting node and a destination / receiving node.
[0035] Also, particular signals can be characterized or defined as either an uplink (UL) signal, a downlink (DL) signal, or a sidelink (SL) signal. An uplink signal is a signal transmitted from a user device 102 to a network device 104. A downlink signal is a signal transmitted from a network device 104 to a user device 102. A sidelink signal is a signal transmitted from a one user device 102 to another user device 102, or a signal transmitted from one network device 104 to another network device 104. Also, for sidelink transmissions, a first / source user device 102 directly transmits a sidelink signal to a second / destination user device 102 without any forwarding of the sidelink signal to a network device 104.
[0036] Additionally, signals communicated between communication nodes in the system 100 may be characterized or defined as a data signal or a control signal. In general, a data signal is a signal that includes or carries data, such multimedia data (e.g., voice and / or image data) , and a control signal is a signal that carries control information that configures the communication nodes in certain ways in order to communicate with each other, or otherwise controls how the communication nodes communicate data signals with each other. Also, certain signals may be defined or characterized by combinations of data / control and uplink / downlink / sidelink, including uplink control signals, uplink data signals, downlink control signals, downlink data signals, sidelink control signals, and sidelink data signals.
[0037] For at least some specifications, such as 5G NR, data and control signals are transmitted and / or carried on physical channels. Generally, a physical channel corresponds to a set of time-frequency resources used for transmission of a signal. Different types of physical channels may be used to transmit different types of signals. For example, physical data channels (or just data channels) , also herein called traffic channels, are used to transmit data signals, and physical control channels (or just control channels) are used to transmit control signals. Example types of traffic channels (or physical data channels) include, but are not limited to, a physical downlink shared channel (PDSCH) used to communicate downlink data signals, a physical uplink shared channel (PUSCH) used to communicate uplink data signals, and a physical sidelink shared channel (PSSCH) used to communicate sidelink data signals. In addition, example types of physical control channels include, but are not limited to, a physical downlink control channel (PDCCH) used to communicate downlink control signals, a physical uplink control channel (PUCCH) used to communicate uplink control signals, and a physical sidelink control channel (PSCCH) used to communicate sidelink control signals. As used herein for simplicity, unless specified otherwise, a particular type of physical channel is also used to refer to a signal that is transmitted on that particular type of physical channel, and / or a transmission on that particular type of transmission. As an example illustration, a PDSCH refers to the physical downlink shared channel itself, a downlink data signal transmitted on the PDSCH, or a downlink data transmission. Accordingly, a communication node transmitting or receiving a PDSCH means that the communication node is transmitting or receiving a signal on a PDSCH.
[0038] Additionally, for at least some specifications, such as 5G NR, and / or for at least some types of control signals, a control signal that a communication node transmits may include control information comprising the information necessary to enable transmission of one or more data signals between communication nodes, and / or to schedule one or more data channels (or one or more transmissions on data channels) . For example, such control information may include the information necessary for proper reception, decoding, and demodulation of a data signals received on physical data channels during a data transmission, and / or for uplink scheduling grants that inform the user device about the resources and transport format to use for uplink data transmissions. In some embodiments, the control information includes downlink control information (DCI) that is transmitted in the downlink direction from a network device 104 to a user device 102. In other embodiments, the control information includes uplink control information (UCI) that is transmitted in the uplink direction from a user device 102 to a network device 104, or sidelink control information (SCI) that is transmitted in the sidelink direction from one user device 102 (1) to another user device 102 (2) .
[0039] Additionally, as used herein, the term “time instance” means the same as, is equivalent to, or includes at least one of: a slot, a sub-slot, a symbol, a sub-symbol, a frame, a sub-frame, a transmission occasion, an occasion, or a unit of time (e.g., a millisecond a microsecond as non-limiting examples) .
[0040] Additionally, in some implementations, a spatial filter may be either a user device (UE) -side spatial filter or a network (gNB) -side spatial filter, and / or the spatial filter is also called a spatial-domain filter.
[0041] Additionally, in some implementations, a “channel metric” includes at least one of the following: a raw channel matrix, a precoding matrix (e.g., one or more eigenvectors derived from a raw channel) , an eigenvector, a precoder, a coefficient matrix (such as but not limited to an eType II codebook coefficient matrix W2) or other channel metric. In addition or alternatively, in some implementations, a raw channel matrix and / or a precoding matrix be in one or more (e.g., a combination) of: a spatial domain, a frequency domain, a time domain, and / or a projection in one or more (e.g., a combination of) an angular domain, a delay domain, and / or a doppler domain.
[0042] Additionally, as used herein, an “UL channel” may include a PUCCH or a PUSCH.
[0043] Additionally, as used herein, an “DL channel” may include a PDCCH or a PDSCH.
[0044] Additionally, as used herein, an “UL RS” may be or include at least one of: a SRS, a PRACH, or a demodulation reference signal (DMRS) (e.g., a DMRS for a PUSCH or a PUCCH) .
[0045] Additionally, as used herein, a “DL RS” may be or include at least one of: a synchronization signal block (SSB) , a CSI-RS, or DMRS (e.g., a DMRS for a PDSCH or a PDCCH) .
[0046] Additionally, as used herein, an “UL signal” may be or include at least one of: an UL channel or a UL RS (e.g., a SRS, a physical random access channel (PRACH) , a DMRS, a PUSCH or a PUCCH) .
[0047] Additionally, as used herein, a “DL signal” may be or include at least one of: a DL channel or a DL RS (e.g., a SSB, a CSI-RS, a DMRS, a PDSCH, or a PDCCH) .
[0048] Additionally, as used herein, a power control parameter includes at least one of: a target power (also referred to as “P0” ) , a path loss RS, a scaling factor for path loss (also referred to as “alpha” ) , or a closed loop process. Also, as used herein, a path-loss may be or include a couple loss.
[0049] Additionally, as used herein, “high layer signaling” may be or include at least one of radio resource control (RRC) signaling or a medium access control (MAC) control element (CE) . In addition or alternatively, as used herein, “physical (PHY) layer signaling” may be or include a DCI or a UCI.
[0050] Additionally, as used herein, a “DCI” means the same as, or is equivalent to, a “PDCCH” .
[0051] Additionally, as used herein, the term “precoding information” means the same as, is equivalent to, or includes at least one of: a precoding matrix indicator (PMI) , a transmit precoding matrix indicator (TPMI) , precoding, or a beam.
[0052] Additionally, as used herein, the term transmission and reception point (TRP) means the same as, is equivalent to, or includes at least one of: a RS port, a RS port group, a RS resource, or a RS resource set.
[0053] Additionally, as used herein, the term “port group” means the same as, is equivalent to, or includes at least one of: an antenna group or a user device (UE) port group.
[0054] Additionally, as used herein, the term “model” means the same as, is equivalent to, or include a function, a functionality, a functionality module, a function module, a processing method, an information processing method, an implementation, a feature, a feature group, a configuration, a configuration set, a dataset (e.g., for model training) or data-driven algorithms. In addition or alternatively, as used herein, the term “model” is used to refer to a capability of a communication node (e.g., a user device 102) to perform a certain processing or have a certain functionality, a feature, and / or a feature group.
[0055] Additionally, for at least some embodiments, different models may be associated with different configurations (e.g., different radio resource control (RRC) configurations) . In addition or alternatively, model activation may refer to activation of a corresponding configuration for a given communication node (e.g., user device 102) . Similarly, model deactivation may refer to deactivating the corresponding configuration, and / or switching and falling back may refer to switching back and falling back, respectively, to the configuration used without model activation.
[0056] Additionally, aspects described herein may be used or implemented in any of various communication networks, including wireless communication networks, cellular communication networks, mobile communication networks, or the like, including future implementations of such networks, such as 6G mobile communication networks and beyond.
[0057] Fig. 2 shows a flow chart of an example method 200 of wireless communication related to channel metrics. At block 202, a user device 102 receives a channel state information reference signal (CSI-RS) . At block 204, the user device 102 determines a measured channel metric based on the CSI-RS. At block 206, the user device 102 derives at least one channel metric set based on the measured channel metric. Each channel metric set of the at least one channel metric set includes one or more channel metrics.
[0058] Fig. 3 shows a flow chart of another example method 300 of wireless communication related to channel metrics. At block 302, a network device 104 transmits a channel state information reference signal (CSI-RS) . At block 304, the network device 104 receives at least one channel information (CSI) set derived based on the CSI-RS. At block 306, the network device 104 reconstructs at least one channel metric set based on the at least one CSI set. Each channel metric set of the at least one channel metric set includes one or more channel metrics.
[0059] In some implementations of the method 200 and / or the method 300, the at least one channel metric set includes a plurality of channel metric sets.
[0060] In addition or alternatively, in some implementations of the method 200 and / or the method 300, at least one processing unit of the user device 102 derives at least one channel state information (CSI) set based on the at least one channel metric set input to the at least one processing unit. In some of these implementations, the at least one processing unit includes at least one artificial intelligence (AI) and / or machine learning (ML) model and / or uses at least one AI / ML model to derive the at least one CSI set based on the at least one channel metric set input to the at least one processing unit and / or to the at least one AI / ML model.
[0061] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 transmits the at least one CSI set via at least one CSI report.
[0062] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the network device 104 receives the at least one CSI set via at least one CSI report.
[0063] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one CSI report includes multiple CSI reports. In some of these implementations, the at least one CSI set includes one CSI set comprising multiple parts, and the user device 102 transmits and / or the network device 104 receives the multiple parts of the one CSI set in the multiple CSI reports. In some other of these implementations, the at least one CSI set includes multiple CSI sets, and the user device 102 transmits and / or the network device 104 receives the multiple CSI sets in the multiple CSI reports. In still some other of these implementations, the at least one CSI set includes multiple parts of multiple CSI sets, and the user device 102 transmits and / or the network device 104 receives the multiple parts of the multiple CSI sets in the multiple CSI reports.
[0064] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one channel metric set includes a plurality of channel metric sets, and channel metrics are included in the channel metric sets according to consecutive time units of the channel metrics.
[0065] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one channel metric set includes a plurality of channel metric sets, and channel metrics are included in the plurality of channel metric sets according to time intervals of the channel metrics.
[0066] In addition or alternatively, in some implementations of the method 200 and / or the method 300, X is a total number of channel metrics of the at least one channel metric set, Y is a number of channel metrics that the user device 102 inputs to at least one processing unit per time unit, and Z is equal to (X mod Y) . In some of these implementations, the time unit is or includes inference time, and the at least one processing unit uses an artificial intelligence (AI) and / or machine learning (ML) model to process the Y channel metrics per inference time.
[0067] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one channel metric set is derived by the user device 102 by: initially deriving a channel metric set of the at least one channel metric set to comprise Z channel metrics; and in response to Z being less than Y, padding the channel metric set comprising the Z channel metrics to form a padded channel metric set.
[0068] In addition or alternatively, in some implementations of the method 200 and / or the method 300, padding the channel metric set comprises padding a (Y-Z) number of channel metrics to the channel metric set so that the padded channel metric set comprises Y channel metrics. In some of these implementations, padding the (Y-Z) number of channel metrics to the channel metric set includes: copying a channel metric of a last time instance in the at least one channel metric set the (Y-Z) number of times; copying a measured channel metric of a last time instance the (Y-Z) number of times; copying at least one channel metric of a last (Y-Z) number of time instances; copying at least one measured channel metric of a last (Y-Z) number of time instances; copying at least one measured channel metric of a first (Y-Z) number of time instances; padding according to a preset scheme; or copying at least one channel metric of a last y number of time instances and / or a next z number of time instances, where y+z = Y-Z.
[0069] In addition or alternatively, in some implementations of the method 200 and / or the method 300, at least one channel metric set is derived by the user device 102 by initially deriving multiple channel metric sets that each comprise less than Y channel metrics, and padding each of the multiple channel metric sets. In some of these implementations, wherein padding the (Y-Z) number of channel metrics to the multiple channel metric sets includes overlapping at least one channel metric of y time instances between two adjacent sets of the multiple channel metric sets.
[0070] In addition or alternatively, in some implementations of the method 200 and / or the method 300, at least one channel metric set that is padded includes the measured channel metric.
[0071] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 inputs a padded channel metric set to the at least one processing unit to derive at least one channel state information (CSI) set. In some of these implementations, the at least one processing unit includes at least one AI and / or ML model and / or uses the at least one AL and / or ML model to derive the at least one CSI set based on the padded channel metric set.
[0072] In addition or alternatively, in some implementations of the method 200 and / or the method 300, a channel metric set of the at least one channel metric set comprises Z channel metrics, and the user device 102 omits the channel metric set comprising the Z channel metrics from being input to at least one processing unit to derive at least one channel state information (CSI) set. In some of these implementations, the user device 102 omits the channel metric set in response to Z being less than Y. In addition or alternatively, in some of these implementations, the Z channel metrics of the omitted channel metric set includes a channel metric at a first slot, a last slot, or a middle slot among a plurality of slots of X channel metrics of the at least one channel metric set. In addition or alternatively, in some of these implementations, the at least one processing unit includes at least one AI and / or ML model or uses at least one AI and / or ML model to derive the at least one CSI set based on the at least one channel metric set.
[0073] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 derives the at least one channel metric set by initially deriving a channel metric set of the at least one channel metric set to comprise Z channel metrics, and: in response to Z being less than a threshold, the user device 102 omits the channel metric set comprising the Z channel metrics from being input to at least one processing unit to derive at least one channel state information (CSI) set; and in response to Z being greater than or equal to the threshold, the user device pads 102 the channel metric set comprising the Z channel metrics to form a padded channel metric set comprising Y channel metrics, and inputs the padded channel metric set to the at least one processing unit to derive the at least one CSI set. In some of these implementations, the threshold comprises or is one of: indicated by the network device 104 via physical (PHY) layer signaling or signaling of a layer higher than the PHY layer; indicated by the user device 102 to the network device 104; defined by a specification or protocol according to which the user device and the network device are configured to communicate and / or operate; or determined by an artificial intelligence (AI) / machine learning (ML) model, wherein different AI / ML models are configured with different threshold values.
[0074] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 indicates one or more indexes of one or more padding channel metrics padded to one or more of the at least one channel metric set.
[0075] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the network device 104 receives an indication of one or more indexes of one or more padding channel metrics padded to one or more of the at least one channel metric set.
[0076] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one CSI set that the network device 104 receives is derived based on one or more padding channel metrics, and the network device 104 performs scheduling based on only those channel metrics of the at least one reconstructed channel metric set that do not correspond to one or more indexes of the one or more padding channel metrics.
[0077] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the one or more indexes are indicated using: a bitmap that indicates which channel metrics of the at least one channel metric set are padding channel metrics; a one-level indicator that indicates which channel metrics of the at least one channel metric set are padding channel metrics; or a two-level indicator that indicates which of the at least one channel metric set is a padded channel metric set and which channel metrics of the at least one channel metric set are padding channel metrics. In some of these implementations utilizing the two-level indicator, the two-level indicator includes a bitmap.
[0078] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 pads one or more channel metrics to one or more of the at least one channel metric sets, and the one or more channel metrics that are padded corresponds to: one or more indexes of a last (Y-Z) number of channel metrics in a last channel metric set of the at least one channel metric set; or one or more indexes of a first (Y-Z) number of channel metrics in a first channel metric set of the at least one channel metric set.
[0079] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 indicates one or more indexes of one or more omitted channel metrics of the at least one channel metric set, the one or more omitted channel metrics omitted from being input to at least one processing unit to derive at least one channel state information (CSI) set. In some of these implementations, the at least one processing unit includes an AI and / or ML model and / or uses an AI and / or ML model to derive the at least one CSI set based on the at least one channel metric set.
[0080] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the network device 104 receives an indication of one or more indexes of one or more omitted channel metrics omitted by the user device 102 from being input to at least one processing unit to derive the at least one channel state information (CSI) set. In some of these implementations, the at least one processing unit includes an AI and / or ML model and / or uses an AI and / or ML mode to derive the at least one CSI set based on the at least one channel metric set.
[0081] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the one or more indexes are indicated using: a bitmap that indicates which channel metrics of the at least one channel metric set are omitted channel metrics; or a one-level indicator that indicates which channel metrics of the at least one channel metric set are omitted channel metrics.
[0082] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one CSI set includes a plurality of CSIs having indexes that increase with decreasing priority of the plurality of CSIs.
[0083] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the at least one CSI set includes a plurality of CSI sets having indexes that increase with decreasing priority of the plurality of CSI sets.
[0084] In addition or alternatively, in some implementations of the method 200 and / or the method 300, priorities of the at least one CSI set decreases with increasing time instances of the at least one CSI set.
[0085] In addition or alternatively, in some implementations of the method 200 and / or the method 300, priorities of the at least one CSI set decrease with increasing layer indexes of the at least one CSI set.
[0086] In addition or alternatively, in some implementations of the method 200 and / or the method 300, time domain information associated with the at least one channel metric set has higher priority than layer information associated with the at least one channel metric set, and priorities of the at least one channel metric set decreases with increasing time instances and / or increasing layer indexes.
[0087] In addition or alternatively, in some implementations of the method 200 and / or the method 300, layer information associated with the at least one metric set has higher priority than time-domain information associated with the at least one channel metric set, and priorities of the at least one channel metric set decreases with increasing time instances and / or increasing layer indexes.
[0088] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 transmits and / or the network device 104 receives the at least one CSI set in a first CSI part and a second CSI part, where the at least one CSI set comprises a plurality of CSI sets, and where a first CSI set of the plurality of CSI sets is communicated in the first CSI part and a second CSI set of the plurality of CSI sets is communicated in the second CSI part. In some of these implementations, the first CSI set supports repetition or re-transmission, and the second CSI set supports transmission only once. In addition or alternatively, in some of these implementations, the first CSI set uses a lower modulation order than a modulation order used for the second CSI set. In addition or alternatively, in some of these implementations, the first CSI set uses a lower coding rate than a coding rate used for the second set of CSI. In addition or alternatively, in some of these implementations, the first CSI set uses a lower modulation and coding scheme (MCS) than a MCS used for the second CSI set. In addition or alternatively, in some of these implementations, the first CSI set uses a higher transmission power than a transmission power used for the second CSI set.
[0089] In addition or alternatively, in some implementations of the method 200 and / or the method 300, wherein the at least one CSI set includes a plurality of CSIs for a plurality of time instances, and the plurality of CSIs are mapped to one CSI report according to: a first criterion that a preset number of highest priority CSIs of the plurality of CSIs are included in the first CSI part, and a remainder of the plurality of CSIs are included in the second CSI part; or a second criterion that all of the plurality of CSIs are included in the second CSI part, where a preset number of highest priority CSIs are included in a first group of the second CSI part, and a remainder of the plurality of CSIs are included in a second group of the second CSI part and / or a third group of the second CSI part.
[0090] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 transmits and / or the network device 104 receives the at least one CSI set in at least one CSI report in a fixed time duration.
[0091] In addition or alternatively, in some implementations of the method 200 and / or the method 300, the user device 102 transmits and / or the network device 104 receives different numbers of CSIs and / or different numbers of CSI sets of the at least one CSI set in a plurality of CSI reports.
[0092] Other methods and / or other implementations of the method 200 and / or the method 300 are possible, including but not limited to those that combine one or more aspects from each of two or more of the methods 200 and 300 and / or those that include fewer than all of the aspects for an above recited implementation of the method 200 and / or 300.
[0093] Further details of actions performed by communication nodes in the wireless communication system 100, any or all of which may be implemented in any of various implementations of the method 200, the method 300, and / or other methods, are now described.
[0094] In some implementations, a network device (e.g., a base station) 104 may transmit a channel state information reference signal (CSI-RS) to a user device 102. In response, the user device 102 may measure the CSI-RS and derive channel state information (CSI) . The CSI may be or include a rank indicator (RI) , a precoding matrix indicator (PMI) , a channel quality indicator (CQI) , a layer indicator (LI) , reference signal received power (RSRP) , reference signal received quality (RSRQ) , a reference signal indicator, or a number of non-zero coefficients, as non-limiting examples. In some of these implementations, the user device 102 may derive the CSI using or with at least one processing unit. In particular of these implementations, the at least one processing unit may perform its processing using at least one artificial intelligence (AI) and / or machine learning (ML) model. Upon determining or deriving the CSI, the user device 102 may transmit or report the CSI to the network device (e.g., base station 104. In response, the network device (e.g., base station) 104 may perform scheduling based on the CSI. The scheduling may include, but is not limited to, determining a coding rate and modulation order, and / or multi-user (MU) pairing.
[0095] Also, as described in further detail below, a communication node (e.g., a user device 102 or a network device 104) may perform one or more functions and / or types of processing with at least one processing unit. In any of various implementations, a processing unit may be at least a part or a component of a processor, such as the processor 110 of the user device 102 or the processor 120 of the network device 104 as previously described with reference to Fig. 1. Also, in some implementations, the processing unit may include, or be otherwise configured to perform its one or more functions or types of processing using at least one model. In particular of these implementations, the model is an artificial intelligence (AI) and / or machine learning (ML) (hereafter collectively referred to as “AI / ML” ) model. For example, suppose the processor is configured to perform a function or certain type of processing to determine or derive a first information set based on a second information set. To do so using at least one AI / ML model, the processing unit may input the second information set into the at least one AI / ML model. In response, the at least one AI / ML model may perform processing on the second information set using AI and / or ML technology (e.g., with or using one or more neural networks or other AI and / or ML-based technology) to determine, derive, and / or generate the first information set. Additionally, the at least one AI / ML model may output the first information set, such as to another component of the communication node for further processing within the communication node, and / or for transmission of one or signals carrying the first information set to one or more other communication nodes.
[0096] In addition or alternatively, in some implementations, a network device (e.g., a base station) 104 may transmit a channel state information reference signal (CSI-RS) to a user device 102. Correspondingly, the user device 102 may receive the CSI-RS, and determine or derive at least one measured channel metric based on the CSI-RS. In addition, the user device 102 may determine or derive at least one channel metric set of channel metric based on the measured channel metric. Each of the at least one channel metric set may include one or more channel metrics. In any of various implementations, the number of channel metrics in each channel metric set may be the same as or different from each other. For example, in some implementations, each channel metric set includes M channel metrics, where M is a positive integer number of one or more (i.e., greater than or equal to one) . Additionally, in some implementations, a given channel metric set corresponds to one or more time instances no earlier than a time instance of the measured channel metric. In addition or alternatively, one or more of the at least one channel metric set may be or include the measured channel metric or a predicted channel metric that is predicted based on the measured channel metric.
[0097] In addition or alternatively, the user device 102 may input the at least one channel metric set into at least one processing unit to determine or derive at least one channel state information (CSI) set. That is, the at least one processing unit may determine or derive the at least one CSI set based on the at least one channel metric set that is input into the at least one processing unit. Also, each CSI set may include one or more CSIs. In some of these implementations, the at least one processing unit may include or otherwise utilize at least one AI / ML model to derive the at least one CSI set. For example, the user device 102 and / or the processing unit of the user device 102 may input the at least one channel metric set into the at least one AI / ML model, and in response, the at least one AI / ML model performs processing on the at least one channel metric set to determine or drive the at least one CSI set.
[0098] In addition or alternatively, in some implementations, for each CSI set, one or more time instances of a given CSI set may be the same as or correspond to one or more time instances of a corresponding channel metric set. In addition or alternatively, in some implementations, the user device 102 may transmit the determined or derived at least one CSI set to the network device (e.g., base station) 104. In response to receiving the at least one CSI set, the network device 104 may determine or derive at least one channel metric based on the at least one CSI set transmitted from user device 102. In some implementations, the number of CSIs in a CSI set may be equal to or less than the number of channel metrics in a corresponding or associated channel metric set.
[0099] In addition or alternatively, in some implementations, a user device 102 may determine or derive multiple channel metric sets based on the measured channel metric. As mentioned, each of the multiple channel metric sets may include one or more channel metrics. In any of various embodiments, the number of channel metrics in a given channel metric set may be the same as or different from each other. For example, each channel metric set may include M channel metrics, where M is positive integer number of one or more.
[0100] Additionally, in some implementations including multiple channel metric sets, each channel metric set may be associated with a respective set of one or more time instances, and may be identified with an associated number. In some of these implementations, smaller numbers may correspond to earlier associated time instances. For example, a first channel metric set having a smaller number than a second channel metric set is also associated with earlier time instances than the second channel metric set.
[0101] In addition or alternatively, in some implementations, the user device 102 may input each channel metric set into at least one processing unit to determine or derive at least one CSI set. In some of these implementations, the at least one processing unit may utilize at least one model to determine or derive the at least one CSI set. In particular of these implementations, the at least model includes at least one AI / ML model that determines or derives the at least one CSI set. In addition or alternatively, for each CSI, one or more time instances of a given CSI set may correspond to the same time instances of a corresponding channel metric set. In addition or alternatively, where multiple CSI sets are generated, the time instances of the multiple CSI sets may be indicated or configured in time relative to each other. For example, a second CSI set may correspond to time instances later than the time instances corresponding to a first CSI set.
[0102] In addition or alternatively, in some implementations, the user device 102 may transmit the at least one CSI set to the network device (e.g., base station) 104. In some of these implementations, where multiple CSI sets are transmitted, the multiple CSI sets may be transmitted via one CSI report or multiple CSI reports. In addition, in response to receipt of the at least one CSI set, such as via at least one CSI report, the network device (e.g., base station) 104 may determine or derive at least one (e.g., multiple) channel metric set based on the at least one CSI set.
[0103] Figs. 4-15 show diagrams of various ways that a user device 102 may determine or derive at least one CSI set based on at least one channel metric set, and that a network device 104 may determine, derive, or reconstruct at least one channel metric set based on at least one CSI set. For purposes of example illustrations, each of Figs. 4-15 show the processing to generate the at least one CSI set (user device) or the at least one channel metric set (network device) being performed by at least one AI / ML model. However, in any of various other implementations, the respective processing may be performed by at least one processing unit and / or at least one model, without necessarily the at least one processing unit and / or the at least one model utilizing AI and / or ML technology to carry out the processing. Various ways of a user device to derive at least one CSI set based on at least one channel metric set and / or a network device to derive / reconstruct at least one channel metric set based on at least one CSI set are possible.
[0104] In further detail, in a first way, a user device 102 may derive one CSI set based on at least one channel metric set. In addition or alternatively, in the first way, the network device (e.g., base station) 104 may derive at least one channel metric set corresponding to the at least one CSI set. In some implementations of the first way, one channel metric set may include one or more channel metrics, and one CSI set includes only one CSI. In some other implementations, one channel metric set includes multiple (more than one) channel metrics, and one CSI set includes one or more CSIs, where the number of CSIs in a CSI set is less than or equal to the number of channel metrics in a corresponding channel metric set
[0105] In a second way, the user device 102 may derive multiple CSI sets based on multiple channel metric sets. In addition or alternatively, in the second way, the network device (e.g., base station) 104 may derive multiple channel metric sets corresponding to the multiple CSI sets. In some implementations of the second way, each channel metric set may include multiple (more than one) channel metrics, and each CSI set includes one or more CSIs. In some other implementations of the second way, each channel metric set includes only one channel metric, and each CSI set includes only one CSI.
[0106] Fig. 4 shows a diagram of an example of an AI / ML model at the user device 102 deriving a CSI set based on a channel metric set input to the AI / ML model, in accordance with the first way. As shown in Fig. 4, one channel metric set is input into an AI / ML model at the user device 102 side. In the example in Fig. 4, the channel metrics forming the channel metric set each include a respective precoding matrix for respective slot T+1 to slot T+4. The user device 102 may derive one CSI set for slot T+1 to slot T+4 based on the channel metric set input to AI / ML model. For example, the AI / ML model at the user device 102 side may compress the channel metric set into one CSI set including one CSI for slot T+1 to slot T+4. Additionally, for at least some implementations, upon deriving the CSI set, the user device 102 may transmit and / or report the derived CSI set for slot T+1 to slot T+4 to the network device (e.g., base station) 104.
[0107] Fig. 5 shows a diagram of another example of an AI / ML model at the user device 102 deriving a CSI set based on a channel metric set input to the AI / ML model, in accordance with the first way. As shown in Fig. 5, one channel metric set is input into the AI / ML model at the user device 102 side. The channel metric set includes four channel metrics configured as precoding matrices for slot T+1 to slot T+4. The user device, using the AI / ML model, may derive one CSI set for slot T+1 to slot T+4 based on the channel metric set input to the AI / ML model at the user device 102 side. For example, the user device 102 may compress the channel metric set into one CSI set including two CSIs, including a first CSI for slot T+1 to slot T+2 and a second CSI slot T+3 to slot T+4. In some implementations, the user device 102 may transmit or report the derived CSI set for slot T+1 to slot T+4 to the network device (e.g., base station) 104.
[0108] Fig. 6 shows a diagram of an example of an AI / ML model at the network device (e.g., base station) 104 deriving or reconstructing a channel metric set based on a CSI set input to the AI / ML model, in accordance with the first way, and which may correspond to the implementation in Fig. 4. As shown in Fig. 6, the network device (e.g., base station) 104 receives the CSI set for slot T+1 to slot T+4. In response, the network device 104, such as using an AI / ML model, may derive or reconstruct the channel metrics for slot T+1 to slot T+4 based on the received CSI set. For example, as shown in Fig. 6, the channel metric set includes four channel metrics each including a respective reconstructed precoding matrix for a respective slot T+1 to slot T+4. For at least some implementations, the network device 104 may perform scheduling based on the derived channel metric set.
[0109] Fig. 7 shows a diagram of another example of an AI / ML model at the network device (e.g., base station) 104 deriving or reconstructing a channel metric set based on a CSI set input to the AI / ML model, in accordance with the first way, and which may correspond to the implementation in Fig. 5. The network device 104 may receive the CSI set for slot T+1 to slot T+4. In the example, the CSI set includes two CSIs, including a first CSI for slot T+1 to slot T+2 and a second CSI for slot T+3 to slot T+4. The network device 104, such as using the AI / ML model, may derive the one channel metric set for slot T+1 to slot T+4 based on the CSI set input into the AI / ML model. As shown in Fig. 7, the channel metric set includes four channel metrics, each in the form of a respective reconstructed precoding matrix for a respective slot T+1 to slot T+4. For at least some implementations, the network device 104 may perform scheduling based on the derived channel metric set.
[0110] Fig. 8 show a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets input to the AI / ML model, in accordance with the second way. As shown in Fig. 8, two channel metric sets are input into the AI / ML model at the user device 102 side. Additionally, in the example in Fig. 8, each of the two channel metric sets includes a respective two channel metrics, each in the form of a respective precoding matrix for a respective time slot. For example, the first channel metric set includes precoding matrices for slot T+1 to slot T+2, and the second channel metric set includes precoding matrices for slot T+3 to slot T+4. The user device 102, such as by using the AI / ML model, may derives two CSI sets of CSI for slot T+1 to slot T+4 based on the two channel metric sets input t the AI / ML model. Fore example, the AI / ML model may compress the two channel metric sets into two CSI sets, including a first CSI set for slot T+1 to slot T+2 and a second CSI set for slot T+3 to slot T+4. In some of these implementations, the user device 102 may transmit or report the derived two CSI sets for slot T+1 to slot T+4 to the network device (e.g., base station) 104.
[0111] Fig. 9 shows a diagram of another example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets input to the AI / ML model, in accordance with the second way. As shown in Fig. 9, a plurality of (i.e., four) channel metric sets are input into the AI / ML model to derive a plurality of (i.e., four) CSI sets. Each channel metric set includes a respective precoding matrix for a respective slot T+1 to T+4. For example, a first channel metric set includes a precoding matrix for slot T+1, a second channel metric set includes a precoding matrix for slot T+2, and so on. In turn, the user device 102, such as by using the AI / ML model, derives the four CSI sets, each for a respective one of the T+1 to slot T+4 based on the four channel metric sets. For example, the AI / ML model at the user device 102 side may compress the four channel metric sets into the four CSI sets including the first set of CSI for slot T+1 to the fourth set of CSI for slot T+4. Also, as shown in Fig. 9, each CSI set includes a respective CSI for a respective one of the slots. In addition, in some implementations, the user device 102 may transmit or report the derived CSI sets for slot T+1 to slot T+4 to the network device 104.
[0112] Fig. 10 shows a diagram of an example of an AI / ML model at the network device (e.g., base station) 104 deriving or reconstructing multiple channel metric sets based on multiple CSI sets input to the AI / ML model, in accordance with the second way, and corresponding to the example in Fig. 8. As shown for the example in Fig. 10, the network device (e.g., base station) may receive two CSI sets, including a first CSI set for slot T+1 to slot T+2 and a second CSI set for T+3 to slot T+4. In turn, the network device 104 may derive a first channel metric set for slot T+1 to slot T+4 based on the first and second CSI sets input to the AI / ML model. As shown in Fig. 10, each of the channel metric sets may include two channel metrics in the form of reconstructed precoding matrices for respective slots. For example, the first channel metric set may include two reconstructed precoding matrices for slots T+1 to T+2, and the second channel matrix set may include two reconstructed precoding matrices for slots T+3 to T+4. In addition, in some implementations, the network device 104 may perform scheduling based on the derived channel metric sets.
[0113] Fig. 11 shows a diagram of another example of an AI / ML model at the network device (e.g., base station) 104 deriving or reconstructing multiple channel metric sets based on multiple CSI sets input to the AI / ML model, in accordance with the second way, and corresponding to the example in Fig. 9. As shown in the example in Fig. 11, the network device 104 may receive a plurality of (i.e., four) CSI sets for slot T+1 to slot T+4. Also, a shown in Fig. 11, each CSI set includes a CSI for a slot. For example, a first CSI set includes a first CSI for slot T+1, a second CSI set includes a second CSI for slot T+2, and so on. In turn, the network device 104 may derive a plurality of channel metric sets for slot T+1 to slot T+4. As shown in the example in Fig. 11, each channel metric set includes a respective reconstructed precoding matrix for a respective slot T+1 to T+4. In some implementations, the network device 104 may perform scheduling based on the derived channel metric sets.
[0114] Fig. 12 shows a diagram of another example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets input to the AI / ML model, in accordance with the second way. The example in Fig. 12 illustrates an ability to achieve finer granularity of a channel metric, such as in the form of a per-layer eigenvector of a precoding matrix. As shown in Fig. 12, a plurality of (i.e., four) channel metric sets are input into an AI / ML model at the user device 102 side. Each channel metric set may include a respective channel metric in the form of a certain layer eigenvector of a precoding matrix for a slot. For example, the first channel metric set includes a first layer eigenvector of a precoding matrix for slot T+1, the second channel metric set includes a second layer eigenvector of a precoding matrix for slot T+1, the third channel metric set includes a first layer eigenvector of a precoding matrix for slot T+2, and the fourth channel metric set includes a second layer eigenvector of a precoding matrix for slot T+2. In In turn, the user device 102 may derive a plurality of (i.e., four) CSI sets. As shown in Fig. 12, each CSI set includes a certain layer eigenvector for a slot. For example, the first CSI set includes a first layer CSI for slot T+1, the second CSI set includes a second layer CSI for slot T+1, the third CSI set includes a first layer CSI for slot T+2, and the fourth CSI set includes a second layer CSI for slot T+2. Additionally, the AI / ML model may derive the CSI sets based on the channel metric sets. For example, the AI / ML model may compress the four channel metric sets into the four CSI sets. Additionally, in some implementations, the user device 102 may transmit or report the derived four CSI sets for slot T+1 to slot T+2 to the network device 104.
[0115] Fig. 13 shows a diagram of another example of an AI / ML model at the network device (e.g., base station) 104 deriving or reconstructing multiple channel metric sets based on multiple CSI sets, in accordance with the second way and corresponding to Fig. 12. As shown in Fig. 13, the network device 104 may receive a plurality of (i.e., four) CSI sets, with each CSI set including a first or second layer eigenvector of a precoding matrix for slot T+1 or slot T+2. In turn, the network device 104 may derive the channel metric sets for slot T+1 to slot T+2 based on the CSI sets input to the AI / ML model. For example, the first channel metric set includes a reconstructed first layer eigenvector of a precoding matrix for slot T+1, the second channel metric set includes a reconstructed second layer eigenvector of a precoding matrix for slot T+1, the third channel metric set includes a reconstructed first layer eigenvector of a precoding matrix for slot T+2, and the fourth channel metric set includes a reconstructed second layer eigenvector of a precoding matrix for slot T+2. In some implementations, the network device 104 may perform scheduling based on the derived channel metric sets.
[0116] Fig. 14 shows a diagram of another example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets, in accordance with the second way. As shown in Fig. 14, a channel metric set may include the eigenvectors of the same layer for multiple time slots, which may be derived from the precoding matrices. For example, a plurality of (i.e., two) channel metric sets are input into the AI / ML model at the user device 102 side. A first channel metric set includes a first layer eigenvector of a precoding matrix for slot T+1 and a first layer eigenvector of a precoding matrix for slot T+2. A second channel metric set includes a second layer eigenvector of a precoding matrix for slot T+1 and a second layer eigenvector of a precoding matrix for slot T+2. In turn, the user device 102 may derive two CSI sets, including a first CSI set that includes first layer eigenvectors of precoding matrices for slot T+1 to slot T+2 and a second CSI that includes second layer eigenvectors of precoding matrices for slot T+1 to slot T+2, based on the two channel metric sets, respectively. For example, the AI / ML model at the user device 102 side may compress the two channel metric sets into the two CSI sets. In some implementations, the user device 102 may report the derived two CSI sets for slot T+1 to slot T+2 to the network device 104.
[0117] Fig. 15 shows a diagram of another example of an AI / ML model at the network device (e.g., base station) 104 deriving or reconstructing multiple channel metric sets based on multiple CSI sets, in accordance with the second way and corresponding to Fig. 14. As shown in Fig. 15, the network device 104 may receive a plurality of (e.g., two) CSI sets, including a first CSI set that includes a first layer CSI for slot T+1 to slot T+2 and a second CSI set that includes a second layer CSI for slot T+1 to slot T+2. In turn, the network device 104 may derive a plurality of (e.g., two) channel metric sets, with each set including multiple reconstructed first or second layer eigenvectors with precoding matrixes for slots T+1 to T+2. For example, the first channel metric set includes a reconstructed first layer eigenvector of a precoding matrix for slot T+1 to slot T+2, and the second channel metric set includes a reconstructed second layer eigenvector of a precoding matrix for slot T+1 to slot T+2. In some implementations, the network device 104 may perform scheduling based on the derived channel metric set.
[0118] The example implementations in Figs. 4-15 show one AI / ML model at the user device 102 or at the network device 104 being used to derive the at least one CSI set or the at least one channel metric set. In other implementations, multiple AI / ML models are used to derive the multiple CSI sets and / or the multiple channel metric sets.
[0119] In addition or alternatively, in some implementations, the user device 102 may transmit one CSI report to the network device (e.g., base station) 104, where one CSI set is included in the CSI report. In some of these implementations, the one CSI set includes one or more CSIs for Q time instances, where Q is an integer number of one or more. In particular implementations, the Q time instances for one CSI set is after a time point T, which is described in further detail below.
[0120] In addition or alternatively, in some implementations, the user device 102 may transmit one CSI report to the network device 104, where multiple CSI sets are included in the CSI report. To illustrate, using Fig. 8, two CSI sets are included in the CSI report. The first CSI set may include one or more CSIs for Q1 time instance (s) and the second CSI set may include one or more CSIs for Q2 time instance (s) , where Q1 and Q2 are integer numbers larger than 0. In particular of these implementations, Q1 and Q2 are the same. In addition or alternatively, the Q1 time instance (s) for the first CSI set is before a time point T and the Q2 time instance (s) for the second CSI set is after the time point T. The time point T is described in further detail below.
[0121] In addition or alternatively, parts of CSI sets may include in one or more CSI reports in any of various implementations. In some implementations, multiple CSI reports for CSI feedback may be implemented where CSI feedback overhead is high when one CSI report is used to report one or more CSI sets. In addition or alternatively, in some implementations, multiple parts of one CSI set may be included and transmitted in or via multiple CSI reports. In some other implementations, multiple CSI sets of may be included and transmitted in or via multiple CSI reports. In still some other implementations, multiple parts of multiple CSI sets may be included and transmitted in or via multiple CSI reports.
[0122] In addition or alternatively, in some implementations, a user device 102 may transmit multiple CSI reports to the network device (e.g., base station) 104. To illustrate with reference to Fig. 5 for example, one CSI set may include a first CSI part and a second CSI that are reported in a first CSI report and a second CSI report, respectively. In some of these implementations, the first CSI part may include CSI for Q1 time instance (s) and the second CSI part may include CSI for Q2 time instance (s) , where Q1 and Q2 are each integer numbers larger than zero (i.e., one or more) . In some of these implementations, Q1 and Q2 are the same. In other of these implementations, Q1 and Q2 are different from each other. In addition or alternatively, in some of these implementations, the Q1 time instance (s) for the first CSI part is before a time point T and the Q2 time instance (s) for the second CSI part is after the time point T. The time point T is described in further detail below.
[0123] In addition or alternatively, in some implementations, a user device 102 may transmit multiple CSI reports to the network device (e.g., base station) 104. To illustrate with reference to Fig. 8, a plurality of (e.g., two) CSI sets may be included in a first CSI report and a second CSI report, respectively. In some of these implementations, the first CSI set may include CSI for Q1 time instance (s) and the second CSI set may include CSI for Q2 time instance (s) , where Q1 and Q2 are integer numbers larger than zero (i.e., one or more) . In some of these implementations, Q1 and Q2 are the same. In other implementations, Q1 and Q2 are different from each other. In addition or alternatively, in some of these implementations, the Q1 time instance (s) for the first CSI set is before a time point T and the Q2 time instance (s) for the second CSI set is after the time point T. The time point T is described in further detail below.
[0124] In addition or alternatively, in some implementations, the above-described Q1 and Q2 may be configured by the network device 104, such as via high layer configuration, may be indicated by the user device 102 to the base station 104, or may be determined by a model. In some of these latter implementations, different model may be configured with different values for Q1 and / or different values for Q2 from each other. In addition or alternatively, Q1 and Q2 may be the same as or different from each other in any of various implementations.
[0125] In addition or alternatively, in some implementations, two CSI sets of CSI or two CSI parts of one CSI set are included in one or more CSI reports. However, in other implementations, the number or CSI sets and / or the number of CSI parts of one CSI set included in one or more CSI reports may be more than two. That is, in general, one or more CSI sets and / or one or more CSI parts of one CSI set may be included in and / or transmitted in one or more CSI reports. For example, a user device 102 may transmit P CSI reports to the network device 104, where P CSI sets or P CSI parts in one CSI set are included in the P CSI reports, where P is integer number larger than 0. In some of these implementations, a k-th CSI set or a k-th CSI part in one CSI set includes CSI for Qk time instances, where k is an index of the P CSI sets or k is an index of the P CSI parts of one CSI set, and k is an integer number satisfying 1≤k≤P, and where Qk a is positive integer number. In some of these implementations, Qk maintains the same number where k=1, 2, …, P. In addition or alternatively, in some of these implementations, the first Q1 time instances for the first CSI set or the first CSI part in one CSI set is before a time point T and the remaining time instances for the remaining CSI sets or the remainder of the CSI parts of one CSI set is after the time point T. The time point T is described in further detail below.
[0126] Additionally, in the examples with reference to Figs. 4-15, the slots are denoted as T+n (e.g., n = 1, 2, etc. ) , which denotes future time instances of predicted channel metrics. Correspondingly, the term T-n (e.g., n=1, 2, etc. ) may be used to denote past time instances of measured channel metrics.
[0127] Additionally, in any of various implementations, multiple CSI reports may be carried by the same PUCCH or PUSCH transmission, may be carried by different PUCCH or PUSCH, in different time and / or frequency resources.
[0128] In addition or alternatively, the implementations herein are described using CSI-RS. However, reference signals other than CSI-RS, such as demodulation reference signals (DMRS) may be used signals instead in any of various other implementations.
[0129] In addition or alternatively, in some implementations, a processing unit and / or model (e.g., an AI / ML model) may have an input dimension or size, which may be a number of inputs (e.g., channel metrics) that the processing unit or model is configured to receive per time unit (e.g., per inference time) . For example, where a given processing unit or model has an input dimension or size of two, then two channel metrics may be input into the processing unit or model at each time unit, such as each inference time. In addition, in some implementations, a communication node (e.g., user device 102) may derive a total number of channel metrics that is larger than the input dimension or size of the processing unit or model. To illustrate, suppose the user device 102 derives eight channel metrics based on a CSI-RS, and the input dimension of an AI / ML model is two channel metrics per inference time. In such a situation, the eight channel metrics may be divided into four channel metric sets or groups, and the four channel metric sets may be input into the AI / ML model to deriving multiple CSI sets or groups, two at a time.
[0130] Fig. 16 shows a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets. In the example, the input dimension for the AI / ML model may be two channel metrics per inference time. Additionally, in the example in Fig. 16, the total number of channel metrics is four, including a first channel metric in the form of a precoding matrix for slot T+1, a second channel metric in the form of a precoding matrix for slot T+2, a third channel metric in the form of a precoding matrix for slot T+3, and a fourth channel metric in the form of a precoding matrix for slot T+4. With the input dimension of two, the four channel metrics may be organized or arranged into two channel metric sets, including a first channel metric set that includes the first and second channel metrics, and a second channel metric set that includes the third and fourth channel metrics, as shown in Fig. 16.
[0131] In general, suppose that a total number of derived channel metrics is denoted as X, and that the input dimension or size of a processing unit or model (e.g., AI / ML model) . In some implementations where X is larger than Y, the X channel metrics may be arranged or divided into G channel metric sets or groups, where G is and where denotes a round up operation.
[0132] In some implementations, derived channel metrics may be organized or arranged according to consecutive time instances or slots, such that for each of the channel metric sets, the channel metrics in a given channel metric set are for consecutive time instances. To illustrate with reference to Fig. 8, a sequence of time instances / slots T+1, T+2, T+3, and T+4 are consecutive relative to each other. Correspondingly, the first channel metric set includes precoding matrices for consecutive slots T+1 and T+2, and the second channel metric set includes precoding matrices for consecutive slots T+3 and T+4.
[0133] In some other implementations, derived channel metrics may be organized or arranged according to a time interval associated with the channel metrics. For example, a j-th channel metric set may include channel metrics, among the total number X of channel metrics, having indexes of (i-1) ×G+j, where i=1, 2, …, Y and j=1, 2, …, G. To illustrate with reference to Fig. 16 as an example, for four precoding matrices for slots T+1, T+2, T+3, and T+4, and a time interval of two, the first channel metric set includes the precoding matrices for slot T+1 and slot T+3, and the second channel metric set includes the precoding matrices for slot T+2 and slot T+4.
[0134] In addition or alternatively, in some other implementations, channel metrics may be arranged into channel metric sets according to consecutive time instances or slots, and according to a time interval. To illustrate, suppose twelve channel metrics at the user device 102 side (i.e., X=12) . Further suppose Y=4, resulting in the twelve channel metrics being grouped into three channel metric sets, with each channel metric set including four channel metrics. Moreover, the twelve channel metrics may be arranged into the three channel metric sets according to consecutive time instances or slots and according to the same or a common time interval. For example, a first channel set may include a first group of four channel metrics for slots T+1, slot T+2, slot T+7, and slot T+8; a second channel set may include a second group of four channel metrics for slots T+3, T+4, T+9, and T+10, and a third channel metric set may include a third group of four channel metrics for slots T+5, T+6 , slot T+11, and slot T+12.
[0135] In addition or alternatively, in some implementations where each of the channel metric sets includes Y channel metrics, X channel metrics may be grouped or arranged into such that to form a given channel metric set, a first Y / 2 number of channel metrics for the given channel metric set are selected from an earliest Y / 2 number of remaining slots and a second Y / 2 number of channel metrics for the given channel metric set are selected from a latest Y / 2 number of remaining slots. To illustrate for example, suppose a total of four channel metrics (X=4) and a model input of two (Y=2) , yielding two channel metric sets each with two channel metrics. To form a first of the two channel metric sets, a first channel metric set may include a channel metric for an earliest one (Y / 2=1) remaining time slot T+1 and a channel metric for a latest one remaining time slot T+4. In addition to form a second of the two channel metric sets, the second channel metric set may include a channel metric for an earliest one remaining time slot, this time T+2, and a channel metric for a latest one remaining time slot, this time T+3. To further illustrate as another example, suppose a total of twelve channel metrics (X=12) , and a model input of four (Y=4) , yielding three channel metric sets each with four channel metrics. Correspondingly, the first channel metric set may include channel metrics for the two (Y / 2=2) remaining earliest time slots T+1, T+2 and the two remaining latest time slots T+11, T+12; in turn, the second channel metric set may include channel metrics for the two remaining earliest time slots T+3, T+4 and the two remaining latest time slots T+9, T+10; and in turn, the third channel metric set may include channel metrics for the two remaining earliest time slots T+5, T+6 and the two remaining latest time slots T+7, T+8.
[0136] In some implementations, in event that Y is not an even number, then the first Y / 2 number of channel metrics may be subjected to a round up operation and the second Y / 2 number of channel metrics may be subjected to a round down operation, or vice versa, to determine which channel metrics for which time slots are to be arranged into which channel metric sets. For example, suppose X=12 and Y=3. In this example, since Y / 2=1.5, then the first Y / 2 may be subjected to a round-up operation (i.e., 1.5 rounded up is 2) , and in turn channel metrics for the two remaining earliest time slots may be selected for a given channel metric set. Correspondingly, the second Y / 2 may be subjected to a round-down operation (i.e., 1.5 rounded down is 1) , and in turn channel metrics for the one remaining latest time slot may be selected for a given channel metric set. For example, for twelve channel metrics (X=12) for twelve time slots T+1 to T+12, channel metrics for the first two earliest time slots T+1, T+2 and the one latest time slot T+12 may be grouped into the first channel metric set, and so on.
[0137] In addition or alternatively, in any of various implementations, a method used to divide or arrange channel metrics into sets or groups may be configured by the network device (e.g., base station) 104, such as via high layer signaling, may be indicated by the user device 102 to the network device 104, may be model dependent, in that models may be associated with respective models such that different models may be configured with the same or different grouping methods, and / or may be defined by a specification or protocol according to which the communication nodes (including the user device 102 and the network device 104) in the wireless communication system 100 are configured to communication and / or operate.
[0138] In addition or alternatively, in some implementations or situations, a total number of derived channel metrics X may not be an integer multiple of the input dimension Y of the processing unit or model (e.g., an AI / ML model) . Such a situation is illustrated in Fig. 17, which shows a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets. In the example in Fig. 7, the user device 102 may derive or determine a total number of five channel metrics (i.e., X=5) for five slots, including a first precoding matrix for slot T+1, a second precoding matrix for slot T+2, a third precoding matrix for slot T+3, a fourth precoding matrix for slot T+4, and a fifth precoding matrix for slot T+5. In addition, the input dimension of the AI / ML model is two. In such a situation, the five channel metrics cannot evenly be arranged or divided into channel metric sets of two. For example, as shown in Fig. 17, the first and second precoding matrices are grouped into the first channel metric set, and the third and fourth precoding matrices are grouped into the second channel matrix set. However, the fifth precoding matrix cannot be grouped into a third channel metric set with another channel metric since there no other ungrouped channel metrics remaining.
[0139] In this context, suppose an integer value Z = (X mod Y) , such that where a total number X of channel metrics are grouped into channel metric sets of Y channel metrics, there will be Z channel metrics that remain ungrouped, or otherwise stated, if Z>1, there will Z channel metrics grouped into one or more channel metric sets that each do not have Y channel metrics. To illustrate, using the example above with respect to Fig. 17, Z = (5 mod 2) = 1. Correspondingly, the five total channel metrics to be grouped into channel metric sets of two will leave Z=1 channel metric that will remain ungrouped or that will be grouped into a channel metric set that does not have Y=2 channel metrics, as shown in Fig. 17. The following describes ways to form channel groupings in situations where Z > 0 , i.e., where initially grouping X channel metrics into channel metric sets of Y channel metrics leaves one or more ungrouped channel metrics remaining or, in other words, leaves at least one channel metric set that has less than the Y number of channel metrics.
[0140] In some implementations, where initially grouping X channel metrics into channel metric sets each of Y channel metrics leaves Z>0 channel metrics ungrouped or in one or more channel metric sets with less than Y channel metrics, the user device 102 may pad the one or more channel metric sets with less than Y channel metrics with (Y -Z) channel metrics to each of the one or more channel metric sets with less than the Y channel metrics. Herein, a channel metric set that is padded may be referred to as a padded channel metric set. Similarly, a channel metric that is padded to a channel metric set may be referred to herein as a padding channel metric. In some of these implementations, the user device 102 may perform padding in response to a determination that Z is less than Y (Z<Y) . Also, upon being padded, a padded channel metric set may be input to a processing unit or model (e.g., an AI / ML model) in order to one or more CSI sets. To illustrate using Fig. 17, the five (X=5) channel metrics may be initially grouped into two channel metric sets of two (Y=2) , leaving a single (Z=1) fifth channel metric for slot T+5 in a third channel metric set with only one channel metric. Subsequently, the third channel metric set may be padded with one (i.e., Y-Z=1) additional channel metric to form a padded channel metric set that includes two channel metrics.
[0141] The following describes various ways that the padding may be performed.
[0142] In a first way, a communication node (e.g., the user device 102) may copy the channel metric of the last time instance for (Y -Z) times. To illustrate with reference to Fig. 17, a third channel metric set may include the precoding matrix for the last time instance T+5, and one (Y-Z=1) copy of that precoding matrix. The padded channel metric set including double the precoding matrix for slot T+5 In turn, the processing unit or model (e.g., an AI / ML model) may derive a third CSI set for slot T+5 based on the precoding matrix for slot T+5 and its copy. In other implementations of the first way, the channel metric of a time instance other than the last time instance, such as the first time instance T+1, is copied (Y-Z) times. The (Y-Z) channel metric copies may then be padded to the one or more channel metric sets initially having less than Y channel metrics.
[0143] In a second way, a communication node (e.g., the user device 102) may copy a measured channel metric of the last time instance for (Y -Z) times. To illustrate with reference to Fig. 17, suppose a precoding matrix for slot T+5 is derived based on a measured channel metric for slot T. In turn, a third channel metric set initially having only the precoding matrix for slot T+5 may be padded with measured channel metric (precoding matrix) for slot T to form a padded channel metric set including two channel metrics-i.e., the precoding matrix for slot T+5 and the precoding matrix for slot T. In turn, the processing unit or model may derive a CSI set based on the precoding matrix for slot T+5 and slot T. In other implementations of the second way, the communication node (e.g., user device 102) , may copy the measured channel metric of a specific time instance for (Y -Z) times, where the specific time instance is other than the time instance T immediately before the first future time instance T+1, for example, the measured channel metric of the second last time instance (slot T-1) .
[0144] In a third way, a communication node (e.g., the user device 102) may copy the channel metric of the last (Y -Z) time instance (s) , each one time. To illustrate with reference to Fig. 17 for example, Y-Z=1 and so the (Y-Z) last time instance (s) is only T+5. Correspondingly , one copy of the precoding matrix for the last slot T+5 may be padded to the third channel metric set, such that the padded channel metric set includes the precoding matrix for slot T+5 and its copy. As another example, in event that Y-Z=2, then channel metrics for the last two slots (e.g., slots T+4 and T+5) may be copied. In other implementations of the third way, the communication node (e.g., the user device 102) may copy the channel metric (s) of the first (Y -Z) time instance (s) each one time, and pad the (Y-Z) copies to the initial channel metric set. For example, with reference to Fig. 17, the first (Y-Z=1) time instance is only T+1, and so the communication node (e.g., the user device 102) pads the precoding matrix for slot T+1 to the third channel metric set already or initially including the precoding matrix for slot T+5-i.e., the padded channel metric set includes the precoding matrix for slot T+5 and the precoding matrix (or its copy) for the first slot T+1. In still other implementations, channel metric (s) for (Y-Z) time instance (s) except for the last time instance may be copied each one time. For example, where (Y-Z) =1, then the channel metric for slot T+4, instead of for the last slot T+5, may be copied.
[0145] In a fourth way, a communication node (e.g., the user device 102) may copy the measured channel metric of the last (Y -Z) time instance (s) , each one time. To illustrate with reference to Fig. 17 for example, the last (Y-Z=1) time instance is T+5, and the measured precoding matrix for the precoding matrix for slot T+5 is the measured precoding matrix for slot T, i.e., the last or immediately prior slot before the first future slot T+1. Correspondingly, the measured precoding matrix for slot T may be padded to the third channel metric set, which may already or initially include the precoding matrix for slot T+5-i.e., the padded channel metric set includes the precoding matrix for slot T+5 and the measured precoding matrix for slot T. The padded channel metric set may then be used to derive a third CSI set for slot T and slot T+5. In other implementations of the third way, the communication node (e.g., the user device 102) may copy the measured channel metric of the first (Y -Z) past time instance (s) . To illustrate with reference to Fig. 17 as an example, the first (Y-Z=1) past time instances is T-4. Correspondingly, the communication node may pad the measured precoding matrix for slot T-4 to the third channel metric set, which already or initially includes the precoding matrix for slot T+5-i.e., the padded channel metric set includes the precoding matrix for past slot T-4 and the precoding matrix for slot T+5. In turn, the processing unit or model (e.g., an AI / ML model) may derive a third CSI set of CSI for slot T-4 and slot T+5 based on the precoding matrix for past slot T-4 and slot T+5.
[0146] In a fifth way, a communication node (e.g., the user device 102) may pad the channel metric set comprising less than the Y channel metrics according to a preset scheme. Non-limiting examples of a preset scheme include all zero values padding, all one values padding, an average calculation of the channel metrics or measured channel metrics of multiple time instances.
[0147] In case the padded (Y –Z) channel metric is discrete.
[0148] In a sixth way, a communication node (e.g., the user device 102) may copy the channel metrics of the last y time instance (s) and / or the next z time instance (s) , where y+z=Y-Z, and where y and z are each integer numbers larger than 0. To illustrate, Fig. 18 shows a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets. In the example in Fig. 18, as denoted by dotted arrows, the second channel metric set is padded with a last precoding matrix for slot T+3 and a next precoding matrix for slot T+5. Correspondingly, the AI / ML model derives a second CSI set for slots T+3~T+5. In other implementations of the sixth way, a copied channel metric may be padded to multiple channel metric sets. To illustrate, Fig. 19 shows a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets. In the example in Fig. 19, the second channel metric set is padded with a last precoding matrix for slot T+3. Correspondingly, the AI / ML model may derive the second CSI set for slots T+3 to T+5. Additionally, in the example in Fig. 19, the third channel metric set is padded with a last precoding matrix for slot T+5. Correspondingly, the AI / ML model may derive the third CSI set for slot T+5 to T+7.
[0149] In a seventh way, a communication node (e.g., the user device 102) may overlapping the channel metric (s) of y time instance (s) between each two adjacent sets to perform channel metric padding, where y is an integer number larger than zero, and y can be the same or different between each two adjacent sets. To illustrate, Fig. 20 shows a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets. In the example in Fig. 20, slots T+2 and T+3 are overlapping among the first and second channel metric sets. Correspondingly, the channel metrics for slots T+2 and T+3 for the first channel metric set are copied, and the copies are used to pad the second channel metric set, as denoted by the dotted arrows in Fig. 20.
[0150] In addition or alternatively, a channel metric set including the measured channel metric may be padded, such as in accordance with one or more of the above-described seven ways.
[0151] In addition or alternatively, in any of various implementations, padding may be performed according to a combination of two or more of the seven ways.
[0152] In addition or alternatively, in some implementations, upon initially grouping the channel metrics into one or more channel metric sets, the user device 102 may omit Z ungrouped channel metrics or the Z channel metrics grouped into channel metric sets that do not include the Y channel metrics from being used to derive the at least one CSI set. To illustrate with reference to Fig. 17 for example, the channel metric for slot T+5 is not part of a channel metric set that includes two channel metrics. In turn, the user device 102 omits the channel metric for slot T+5 from being used by the AI / ML model to derive the at least one CSI set. In some of these implementations, the user device 102 may omit the Z channel metrics in response to Z being less than Y (i.e., Z<Y) . For example, the user device 102 may expressly determine that Z is less than Y before omitting the Z channel metrics.
[0153] In addition or alternatively, in some implementations, the user device 102 may decide to pad channel metric sets and / or omit channel metrics based on a threshold. For example, in response to Z being less than a threshold, the user device 102 may omit Z channel metrics not initially grouped into a channel metric set including Y channel metrics from being used to derive the at least one CSI set. Additionally, in response to Z being greater than or equal to the threshold, one or more channel metric sets including at least one of the Z channel metrics not grouped into a channel metric group including Y channel metrics may be padded to form a padded channel metric set including Y channel metrics. In turn, the padded channel metric may be input to the processing unit (e.g., the AI / ML model) to derive the at least one CSI set. In some of these implementations, the threshold is, or is based on, the mathematical operation In addition or alternatively, in some of these implementations, the threshold may be configured by the network device 104, such as via high layer signaling or physical (PHY) layer signaling, indicated by the user device 102 to the network device 104, defined by a specification or protocol according to which the communication nodes of the wireless communication system 100 operate and / or communicate, or determined by a model. In some of these latter implementations, different models may be configured with different threshold values.
[0154] As previously described, in some implementations, a user device 102 may pad one or more (such as a (Y-Z) number of) channel metrics to a given channel metric set that includes Z channel metrics, and in turn, may derive a CSI set based on the padded channel metric set that it then reports via a CSI reported, to the network device (e.g., base station) 104. In response, the network device 104 may derive the channel metric set based on the reported CSI set. Further, in some implementations, the user device 102 may indicate, to the network device 104, the one or more indexes of the one or more channel metrics used for the padding. Through receipt of the indexes, the network device 104 is able to know which channel metric sets were padded by the user device 102, and / or which channel metrics were used for the padding (i.e., which of the channel metrics are padding channel metrics) .
[0155] In some implementations, a bitmap may be used to indicate the indexes of the padding channel metrics. To illustrate with reference to Fig. 17, five channel metrics may be initially derived, where first and second channel metrics are grouped into a first channel metric set, third and fourth channel metrics are grouped into a second channel metric set, and a fifth channel metric is grouped into a third channel metric set. In turn, a sixth channel metric is padded to the third metric set so that the third metric set, like the other two channel metric sets, has Y=2 channel metrics, as previously described. As such, the sixth channel metric is a padding channel metric. Correspondingly, in some implementations, the user device 102 may indicate the sixth channel metric as a padding channel metric using a bitmap. Each bit in the bitmap may correspond to a respective one of the channel metrics used to generate the at least one CSI set, such as the three CSI sets in the example in Fig. 17. Moreover, each bit may have an associated bit value to indicate whether or not the bit corresponds to a padding channel metric or not. For example, a first bit value, such as a bit value of ‘0’ may be used to indicate that an associated channel metric is not a padding channel metric, and a second bit value, such as a bit value of ‘1’ may be used to indicate that an associated channel metric is a channel metric. For the example in Fig. 17, since six channel metrics are ultimately used derive the three CSI sets, then a corresponding bitmap may include six bits, such as ‘000001’ , where the first five bits (left to right) each have a ‘0’ bit value to correspond to the first through fifth channel metrics being non-padding channel metrics, and the last or sixth bit having a ‘1’ bit value to indicate that the corresponding sixth channel metric is a padding channel metric.
[0156] In addition or alternatively, in some implementations, indexes of padded channel metric sets may be indicated via a one-level indicator. To illustrate referring to Fig. 17 for example, the sixth channel metric is a padding channel metric that is padded to the third channel metric set, as previously described. Correspondingly, the user device 102 may indicate to the network device 104 a one-level indicator, such as in the form of three-bit value, where the one-level indicator indicates the one or more padding channel metrics. For example, with reference to Fig. 17, the one-level indicator may be a three-bit value, such as ‘110’ , which is a binary representation indicating the sixth channel metric. In addition or alternatively, in some implementations, the size and / or bit values of the one-level indicator may be dynamically changing on conjunction with the total number of channel metrics and / or the number of derived channel sets. In other implementations, size of the one-level indicator can be a fixed number of bits to adapt to different total numbers of channel metrics.
[0157] In addition or alternatively, in some implementations, indexes of padded channel metric sets and padding channel metrics may be indicated via a two-level indicator. To illustrate referring to Fig. 17 for example, a two-level indicator may indicate both that the third channel metric set is a padded channel metric set, and that the sixth channel metric is a padding channel metric. An example two-level indicator may be ‘1001’ , where the first two bits with bit values ‘10’ denote that the third channel metric set is a padded channel metric set, and last two bits with bit values ‘01’ denote that the second channel metric in the third channel metric set is a padding channel metric. In other implementations, the two-level indicator may be in the form of a bitmap that has bits with respective bit values to indicate indexes of channel metric sets that are padded channel metric sets, and that has bits with respective bit values to indicate indexes of channel metrics that are padding channel metrics in the padded channel metric sets.
[0158] In addition or alternatively, in some implementations, the user device 102 and / or the network device 104 may perform padding in accordance with one or more of the following padding criteria. In a first criteria, when channel metric padding is performed, one or more channel metrics are padded to a last (Y -Z) channel metric in the last channel metric set. In a second criteria, one or more channel metrics are padding to a first (Y -Z) channel metric in a first channel metric set. Also, in accordance with these implementations, padding indication information may not be reported since the network device 104 and the user device 102 may have the same understanding as the indexes of the padding channel metrics and the padded channel metric sets.
[0159] In addition or alternatively, in some implementations, the user device 102 may indicate one or more indexes of one or more channel metrics omitted from be used to derive the at least one CSI set. In some of these implementations, the user device 102 may also indicate, to the network device 104, one or more indexes of one or more time instances corresponding to one or more channel metrics that are omitted. In response to the indication, the network device 104 may become aware of which channel metrics were omitted from being used to derive the at least one CSI set.
[0160] In some implementations, the indexes of omitted channel metrics may be indicated via a bitmap. To illustrate, Fig. 21 shows a diagram of an example of an AI / ML model at the user device 102 deriving multiple CSI sets based on multiple channel metric sets. In the example in Fig. 21, five channel metric are initially derived. Suppose Y=2, and correspondingly, a third channel metric for slot T+3 is omitted (i.e., not used by the AI / ML model to derive the CSI sets) . Correspondingly, the user device 102 may indicate the omitted and non-omitted channel metrics to the network device 104 using a bitmap. Using the example with reference to Fig. 21, the bitmap may include five bits, each corresponding to a respective one of the five channel metrics. The five bits may each have a first bit value indicating that it is not omitted, or a second bit value indicating that it is omitted. In some implementations, the first bit value is a ‘0’ bit value, and the second bit value is a ‘1’ bit value. Correspondingly, in the example in Fig. 21, bit values of the bitmap may include ‘00100’ , with the third bit having a ‘1’ bit value to indicate that the third channel metric for slot T+3 is omitted from being used to derive the CSI sets. In other implementations, the first bit value is a ‘1’ bit value and the second bit value is ‘0’ bit value. Correspondingly, in the example in Fig. 21, bit values of the bitmap may include ‘11011’ , with the third bit having a ‘0’ bit value to indicate that the third channel metric for slot T+3 is omitted from being used to derive the CSI sets. Various ways of using bits and bit values to indicate channel metrics that are and are not omitted from being used to derive CSI sets are possible.
[0161] In addition or alternatively, in some implementations, indexes of omitted channel metrics may be indicated via a one-level indicator. To illustrate with reference to Fig. 21 for example, a three-bit value may be used to indicate that the third channel metric for slot T+3 is omitted from being used to derive the at least one CSI set. For example, the three-bit value of ‘010’ may indicate that the third channel metric is omitted. In at least some implementations, the size of the one-level indicator may dynamically change with the total number of channel metrics and / or channel metric sets. In other implementations. The size of the one-level indicator is a fixed number of bits to adapt to different total numbers of input channel metrics.
[0162] In addition or alternatively, in some implementations, the user device 102 may transmit to the network device 104 an indication (such as in the form of K bits, where K>0) that indicates a start index and an end index of one or more channel metrics being omitted. In addition or alternatively, in some implementations, the user device 102 may transmit to the network device 104 an indication (such as in the form of I bits, where I>0) that indicates an omission mode or pattern according to which the user device 102 omits one or more channel metrics. Example omission modes or patterns may include: a discrete omission mode or pattern where individual channel metrics are discretely omitted; a consecutive omission mode or pattern where two or more consecutive (such as according to time instances or slots) channel metrics are omitted; or an interval omission mode or pattern where two or more channel metrics are omitted according to a time interval. Other omission modes or patterns are possible. In some implementations for the interval omission mode or pattern, the size of the indicator indicating the interval may be dynamic or fixed. Also, each of the different omission modes or patterns may be indicated by a respective one of different values, such as different bit values. By knowing the start and end indexes and / or the omission mode or pattern that the user device 102 indicates, the network device 104 may determine which of a plurality of channel metrics are omitted.
[0163] To illustrate further with an example, suppose there are initially eight channel metrics derived (i.e., X=8) . Further, suppose the user device 102 determines to omit three channel metrics according to the consecutive omission mode. Specifically, the user device 102 determines to omit three consecutive channel metrics, namely the third, fourth, and fifth channel metrics. To indicate these channel metrics, the indicator may include a first part indicating the mode. For example, the first part may include a two-bit value (e.g., ‘01’ ) that indicates that the omission mode used to determine the omitted channel metrics is the consecutive omission mode. In addition, the indicator may include a second part that indicates a start index corresponding to the third channel metric and an end index corresponding to the fifth channel metric. For example, the second part of the indicator may include the bit value ‘010100’ , where a first sub-part ‘010’ indicates that the start index (or the start channel metric) of the consecutive channel metrics is the third channel metric, and a second sub-part ‘100’ indicates that the end index (or the end channel metric) of the consecutive channel metrics is the fifth channel metric. In particular implementations, the user device 102 may transmit the first and second parts of the indicator in sequence, such that the indicator has the bit sequence ‘01 010100’ .
[0164] As another example, again suppose the user device 102 initially determines eight channel metrics (X=8) . Additionally, suppose the user device 102 determines to omit channel metrics according to an interval, such as an interval of two for purposes of the example, starting with the first of the eight channel metrics. Accordingly, among the eight channel metrics for an interval of two, the first, third, fifth, and seventh channel metrics may be omitted. Correspondingly, as an example, the user device 102 may indicate the omitted channel metrics using an indicator that includes multiple parts, such as in the following example format. A first part of the indicator may be a value, such as a two-bit value (e.g., ‘10’ ) that indicates that the omission mode or pattern used to determine the omitted channel metrics is the interval omission mode or pattern. In addition, a second part of the indicator may be a value (e.g., two-bit value) that indicates a number of the interval (e.g., ‘10’ to indicate an interval of two) . Also, a third part of the indicator may include a value that indicates a start index and an end index. For example, the third part may include the value ‘000110’ , which in turn may include a first sub-part ‘000’ to indicate that the channel metrics being omitted according to the interval starts with the first channel metric, and a second sub-part ‘110’ to indicate that the channel metrics being omitted according to the interval ends with the seventh channel metric. Also, in particular implementations, the user device 102 may transmit the first and second parts of the indicator in sequence, such that the indicator has the bit sequence 10 10 000110.
[0165] In addition or alternatively, the user device 102 and / or the network device 104 may omit channel metrics based one or more of the following criteria in order to determine which channel metric (s) to omit by default. In a first criteria, the last channel metric is omitted by default. To illustrate with reference to Fig. 17 for example, the user device 102 may omit the fifth channel metric for slot T+5 by default, since that is the last channel metric among the five channel metrics. In a second criteria, the first channel metric is omitted by default. To illustrate with reference to Fig. 17 for example, the user device 102 may omit the first channel metric for slot T+1 since that is the first channel metric among the five channel metrics. In a third criteria, the middle channel metric is omitted by default, To illustrate with reference to Fig. 21 for example, the third channel metric for slot T+3 is omitted by default since that the third channel metric is the middle channel metric among the five channel metrics.
[0166] Of note, while the implementations herein are described with respect to CSI generation and reporting, other implementations may perform the same or similar actions but instead for other functionality, such as beam related information generation and reporting, and / or positioning related information generation and reporting, as non-limiting examples.
[0167] Also of note, while the implementations herein are described with reference to the time domain unit (e.g., the channel metrics are for time instance or slots) , other implementations may perform the same or similar actions, but instead for other domain units, such as spatial domain units (e.g., antenna ports) , frequency domain units (e.g., physical resource block, sub-band) , or doppler domain unit, as non-limiting examples.
[0168] Additionally, suppose some implementations utilize Q1 time instances that are earlier than Q2 time instances, where the CSI for the Q1 time instances are more important (e.g., have a higher priority) than the CSI for Q2 for network scheduling. The following implementations may be used to enhance or optimize for the reliability of important CSI transmission, including in the event that CSI collision may happen or in situations where there are not sufficient transmission resources to transmit all CSIs.
[0169] In some implementations, the indexes for parts of a CSI may increase with decreasing priority. To illustrate with reference to Fig. 5 for example, the first CSI part for slots T+1 and T+2 may have a lower index and a higher priority than the second CSI part for slots T+3 and T+4.
[0170] In addition or alternatively, in some implementations, the indexes of CSI sets may increases with decreasing priority. To illustrate with reference to Fig. 8 for example, the first CSI set may have a lower index and a higher priority than the second CSI set.
[0171] In addition or alternatively, in some implementations, time instances for CSI may increase with decreasing priority. To illustrate with reference to Fig. 9 for example, the CSI for slot T+1 has a higher priority than the CSI for slot T+2.
[0172] In addition or alternatively, in some implementations, layer indexes of CSIs may increase with decreasing priority. To illustrate with respect to Fig. 13 for example, the first CSI set including a first-layer CSI for slot T+1 has a higher priority than the second CSI set including a second-layer CSI for slot T+1.
[0173] In addition or alternatively, in some implementations, channel metrics may be prioritized such that time-domain information has a higher priority than layer information, and time instances (e.g., slots) and layer indexes each increase with decreasing priority. To illustrate with respect to Fig. 12 for example, the first channel metric set including a first-layer eigenvector of a precoding matrix for slot T+1 has a higher priority than the second channel metric set including a second-layer eigenvector of a precoding matrix for slot T+1, which has a higher priority than the third channel metric set including a first-layer eigenvector of a precoding matrix for slot T+2, which has a higher priority than a second-layer eigenvector of a precoding matrix for slot T+2.
[0174] In addition or alternatively, in some implementations, channel metrics are prioritized such that layer information has a higher priority than time-domain information, and time instances (e.g., slots) and layer indexes each increase with decreasing priority. To illustrate with respect to Fig. 12 for example, the first channel metric set including a first-layer eigenvector of a precoding matrix for slot T+1 has a higher priority than the third channel metric set including a first-layer eigenvector of a precoding matrix for slot T+2, which has a higher priority than the second channel metric set including a second-layer eigenvector of a precoding matrix for slot T+1, which has a higher priority than the fourth channel metric set including a second-layer eigenvector of a precoding matrix for slot T+2.
[0175] In addition or alternatively, in some implementations, one or more of the following transmission schemes may be utilized to optimize for transmission reliability for high-priority CSI. Under the following schemes, suppose the user device 102 transmits two CSI sets to the network device 104.
[0176] In a first transmission scheme, the user device 102 transmits a first CSI set in a first CSI part (also called a Part 1 CSI) and a second CSI set in a second CSI part (also called a Part 2 CSI) . In at least some implementations of the first transmission scheme, the Part 1 CSI has a higher priority than the Part 2 CSI. Correspondingly, the Part 2 CSI may be omitted from being transmitted instead the Part 1 CSI in event that there are insufficient transmission resource to transmit both the Part 1 CSI and the Part 2 CSI.
[0177] In a second transmission scheme, the first CSI set supports repetition or re-transmission, while the second CSI set only supports transmission once. For example, in event that the first CSI set is not decoded correctly, then the network device 104 may request a re-transmission of the first CSI set from the user device 102. However, in event that the second CSI set is not decoded correctly, then the network device 104 may not be able to request re-transmission of the second CSI set.
[0178] In a third transmission scheme, transmission of the first CSI set uses a modulation order that is lower than a modulation order used for transmission of the second CSI set. In some implementations of the third transmission scheme, lower modulation order may correspond to higher reliability.
[0179] In a fourth transmission scheme, transmission of the first CSI set uses a lower coding rate than that used for transmission of the second CSI set. In some implementations of the fourth transmission scheme, lower coding rate may correspond to higher reliability.
[0180] In a fifth transmission scheme, transmission of the first CSI set may use a lower modulation and coding scheme (MCS) than the MCS used for transmission of the second CSI set. In some implementations of the fifth transmission scheme, lower MCS may correspond to higher reliability.
[0181] In a sixth transmission scheme, transmission of the first CSI set may use a higher transmission power than the transmission power used for transmission of the second CSI set. For example, a target power configured by the network device 104 for transmission of the first CSI set may be larger than the transmission power used for transmission of the second CSI set.
[0182] Similar transmission schemes and / or CSI relationships may be implemented for implementations involving more than two CSI sets.
[0183] In addition or alternatively, in some implementations, CSI reports may be carried on UCIs. In addition or alternatively, in some implementations, multiple CSI sets of CSI contents may be mapped to CSI reports. In some implementations, in a CSI report, a Part 1 CSI and a Part 2 CSI may include the CSI contents for UCI transmission.
[0184] In addition or alternatively, CSI sets may be mapped to CSI reports in accordance with one or more of the following mapping schemes embodiments for UCI mapping.
[0185] In a first mapping scheme, one or more CSIs for K time instances are mapped to one CSI report according to one or more of the following schemes. In a first scheme, CSIs with higher priority are included in Part 1 CSI, and the rest or remainder of the CSIs are included in Part 2 CSI. In some implementations of the first scheme, a first K1 sets of CSI are included in Part 1 CSI and a remainder K2 sets of CSI are included in Part 2 CSI, where K1 and K2 are each integers larger than zero. In some other implementations of the first scheme, the first K1 parts of CSI in one CSI set are included in Part 1 CSI and the rest or remaining K2 parts of CSI in one CSI set are included in Part 2 CSI, where K1 and K2 are each integers larger than zero. In some other implementations of the first scheme, the first K1 time instances of CSI are included in Part 1 CSI and the rest or remaining K2 time instances of CSI are included in Part 2 CSI, where K1 and K2 are each integers larger than zero. In some other implementations of the first scheme, a first L1 layers of CSI in a first K1 sets are included in Part 1 CSI and the rest or remaining L2 layers of CSI in a remaining K2 sets are included in Part 2 CSI, where L1, L2, K1, K2 are each integers, where 0<L1≤TR, 0≤L2<TR, 0<K1≤K, 0≤K2<K, and where TR is the number of receiving antenna ports.
[0186] In a second scheme, all CSIs are included in Part 2 CSI, where the CSIs with higher priority are included in a first group of Part 2 CSI (e.g., Group 0) , and the rest or remainder of CSIs are included in a second group of Part 2 CSI (e.g., Group 1) and / or a third group of Part 2 CSI (e.g., Group 2) . In some implementations of the second scheme, a first K1 CSI sets are included in Part 2 CSI Group 0 and the rest or remaining K2 CSI sets are included in Part 2 CSI Group 1 and / or Group 2, where K1 and K2 are each integers larger than zero. In some other implementations of the second scheme, the first K1 parts of CSI in one CSI set are included in Part 2 CSI Group 0 and the rest or remaining K2 parts of CSI in one CSI set are included in Part 2 CSI Group 1 and / or Group 2, where K1 and K2 are each integers larger than zero. In some other implementations of the second scheme, the first K1 time instances of CSI are included in Part 2 CSI Group 0 and the rest or remaining K2 time instances of CSI are included in Part 2 CSI Group 1 and / or Group 2, where K1 and K2 are each integers larger than zero. In some other implementations of the second scheme, a first L1 layers of CSI in a first K1 CSI sets are included in Part 2 CSI Group 0 and the rest or remaining L2 layers of CSI in a rest or remaining K2 CSI sets are included in Part 2 CSI Group 1 and / or Group 2, where L1, L2, K1, K2 are an integer, where 0<L1≤TR, 0≤L2<TR,0<K1≤K, 0≤K2<K, and where TR is the number of receiving antenna ports.
[0187] In addition or alternatively, in some implementations, the CSI contents included in Part 1 CSI and Part 2 CSI may be one or more elements of a rank indicator (RI) , a channel quality indicator (CQI) , a precoding matrix indicator (PMI) , a layer indicator (LI) , or a model index corresponding to the time instances.
[0188] In addition or alternatively, in some implementations, in event that one CSI report cannot include all CSI for K time instances, all of the CSI for the K time instances may be mapped to U reports, where U is an integer larger than 1., In some of these implementations, the user device 102 may transmit U CSI reports in a fixed time duration, e.g., 10 milliseconds (ms) . Doing so may optimize for time correlation continuity for the network device (e.g., gNB) 104 to recover the channel metrics. In addition or alternatively, in some implementations, different numbers of CSIs or CSI sets are included in U CSI reports, such as according to a user device-side condition. To illustrate for example, four CSIs may be separated into three reports (e.g., 2 CSIs + 1 CSI + 1 CSI) for feedback under the circumstance that the user device-side has no sufficient CSI processing unit (CPU) for processing two CSIs at the second instance, and so a third CSI report is used for processing the last CSI.
[0189] The description and accompanying drawings above provide specific example embodiments and implementations. The described subject matter may, however, be embodied in a variety of different forms and, therefore, covered or claimed subject matter is intended to be construed as not being limited to any example embodiments set forth herein. A reasonably broad scope for claimed or covered subject matter is intended. Among other things, for example, subject matter may be embodied as methods, devices, components, systems, or non-transitory computer-readable media for storing computer codes. Accordingly, embodiments may, for example, take the form of hardware, software, firmware, storage media or any combination thereof. For example, the method embodiments described above may be implemented by components, devices, or systems including memory and processors by executing computer codes stored in the memory.
[0190] Throughout the specification and claims, terms may have nuanced meanings suggested or implied in context beyond an explicitly stated meaning. Likewise, the phrase “in one embodiment / implementation” as used herein does not necessarily refer to the same embodiment and the phrase “in another embodiment / implementation” as used herein does not necessarily refer to a different embodiment. It is intended, for example, that claimed subject matter includes combinations of example embodiments in whole or in part.
[0191] In general, terminology may be understood at least in part from usage in context. For example, terms, such as “and” , “or” , or “and / or, ” as used herein may include a variety of meanings that may depend at least in part on the context in which such terms are used. Typically, “or” if used to associate a list, such as A, B or C, is intended to mean A, B, and C, here used in the inclusive sense, as well as A, B or C, here used in the exclusive sense. In addition, the term “one or more” as used herein, depending at least in part upon context, may be used to describe any feature, structure, or characteristic in a singular sense or may be used to describe combinations of features, structures or characteristics in a plural sense. Similarly, terms, such as “a, ” “an, ” or “the, ” may be understood to convey a singular usage or to convey a plural usage, depending at least in part upon context. In addition, the term “based on” may be understood as not necessarily intended to convey an exclusive set of factors and may, instead, allow for existence of additional factors not necessarily expressly described, again, depending at least in part on context.
[0192] Reference throughout this specification to features, advantages, or similar language does not imply that all of the features and advantages that may be realized with the present solution should be or are included in any single implementation thereof. Rather, language referring to the features and advantages is understood to mean that a specific feature, advantage, or characteristic described in connection with an embodiment is included in at least one embodiment of the present solution. Thus, discussions of the features and advantages, and similar language, throughout the specification may, but do not necessarily, refer to the same embodiment.
[0193] Furthermore, the described features, advantages and characteristics of the present solution may be combined in any suitable manner in one or more embodiments. One of ordinary skill in the relevant art will recognize, in light of the description herein, that the present solution can be practiced without one or more of the specific features or advantages of a particular embodiment. In other instances, additional features and advantages may be recognized in certain embodiments that may not be present in all embodiments of the present solution.
[0194] The subject matter of the disclosure may also relate to or include, among others, the following aspects:
[0195] A first aspect includes a method for wireless communication that comprises: receiving, by a user device, a channel state information reference signal (CSI-RS) ; determining, by the user device, a measured channel metric based on the CSI-RS; and deriving, by the user device, at least one channel metric set based on the measured channel metric, each channel metric set comprising one or more channel metrics.
[0196] A second aspect includes a method for wireless communication that comprises: transmitting, by a network device, a channel state information reference signal (CSI-RS) ; receiving, by the network device, at least one channel information (CSI) set derived based on the CSI-RS; and reconstructing, by the network device, at least one channel metric set based on the at least one CSI set, each channel metric set comprising one or more channel metrics.
[0197] A third aspect includes any of the first or second aspects, and further includes wherein the at least one channel metric set comprises a plurality of channel metric sets.
[0198] A fourth aspect includes any of the first through third aspects, and further includes: deriving, by at least one processing unit of the user device, at least one channel state information (CSI) set based on the at least one channel metric set input to the at least one processing unit.
[0199] A fifth aspect includes any of the first through fourth aspects, and further includes: transmitting, by the user device, the at least one CSI set via at least one CSI report.
[0200] A sixth aspect includes any of the first through fifth aspects, and further includes: receiving, by the network device, the at least one CSI set via at least one CSI report
[0201] A seventh aspect includes any of the fifth or sixth aspects, and further includes wherein the at least one CSI report comprises multiple CSI reports, and wherein: the at least one CSI set comprises one CSI set comprising multiple parts, and the user device transmits and / or the network device receives the multiple parts of the one CSI set in the multiple CSI reports; the at least one CSI set comprises multiple CSI sets, and the user device transmits and / or the network device receives the multiple CSI sets in the multiple CSI reports; or the at least one CSI set comprises multiple parts of multiple CSI sets, and the user device transmits and / or the network device receives the multiple parts of the multiple CSI sets in the multiple CSI reports.
[0202] An eighth aspect includes any of the first through seventh aspects, and further includes wherein the at least one channel metric set comprises a plurality of channel metric sets, and wherein channel metrics are included in the channel metric sets according to consecutive time units of the channel metrics.
[0203] A ninth aspect includes any of the first through eighth aspects, and further includes wherein the at least one channel metric set comprises a plurality of channel metric sets, and where channel metrics are included in the plurality of channel metric sets according to time intervals of the channel metrics.
[0204] A tenth aspect includes any of the first through ninth aspects, and further includes wherein X is a total number of channel metrics of the at least one channel metric set, Y is a number of channel metrics that the user device inputs to at least one processing unit per time unit, and Z is equal to (X mod Y) .
[0205] An eleventh aspect includes the tenth aspect, and further includes wherein the time unit comprises inference time, and wherein the at least one processing unit uses an artificial intelligence (AI) and / or machine learning (ML) model to process the Y channel metrics per inference time.
[0206] A twelfth aspect includes any of the tenth or eleventh aspects, and further includes wherein the at least one channel metric set is derived by the user device by: initially deriving a channel metric set of the at least one channel metric set to comprise Z channel metrics; and in response to Z being less than Y, padding the channel metric set comprising the Z channel metrics to form a padded channel metric set.
[0207] A thirteenth aspect includes the twelfth aspect, and further includes wherein padding the channel metric set comprises padding a (Y-Z) number of channel metrics to the channel metric set so that the padded channel metric set comprises Y channel metrics.
[0208] A fourteenth aspect includes the thirteenth aspect, and further includes wherein padding the (Y-Z) number of channel metrics to the channel metric set comprises: copying a channel metric of a last time instance in the at least one channel metric set the (Y-Z) number of times; copying a measured channel metric of a last time instance the (Y-Z) number of times; copying at least one channel metric of a last (Y-Z) number of time instances; copying at least one measured channel metric of a last (Y-Z) number of time instances; copying at least one measured channel metric of a first (Y-Z) number of time instances; padding according to a preset scheme; or copying at least one channel metric of a last y number of time instances and / or a next z number of time instances, where y+z = Y-Z.
[0209] A fifteenth aspect includes any of the tenth through fourteenth aspects, and further includes wherein at least one channel metric set is derived by the user device by initially deriving multiple channel metric sets that each comprise less than Y channel metrics, and padding each of the multiple channel metric sets.
[0210] A sixteenth aspect includes the fifteenth aspect, and further includes wherein padding the (Y-Z) number of channel metrics to the multiple channel metric sets comprises overlapping at least one channel metric of y time instances between two adjacent sets of the multiple channel metric sets.
[0211] A seventeenth aspect includes any of the twelfth through sixteenth aspects, and further includes wherein one of the one or more of the at least one channel metric set that is padded comprises the measured channel metric.
[0212] An eighteenth aspect includes any of the twelfth through seventeenth aspects, and further includes: inputting, by the user device, the padded channel metric set to the at least one processing unit to derive at least one channel state information (CSI) set.
[0213] A nineteenth aspect includes the tenth aspect, and further includes wherein a channel metric set of the at least one channel metric set comprises Z channel metrics, and wherein the user device omits the channel metric set comprising the Z channel metrics from being input to at least one processing unit to derive at least one channel state information (CSI) set.
[0214] A twentieth aspect includes the nineteenth aspect, and further includes wherein the user device omits the channel metric set in response to Z being less than Y.
[0215] A twenty-first aspect includes any of the nineteenth or twentieth aspects, and further includes wherein the Z channel metrics of the omitted channel metric set comprises a channel metric at a first slot, a last slot, or a middle slot among a plurality of slots of X channel metrics of the at least one channel metric set.
[0216] A twenty-second aspect includes the tenth aspect, and further includes wherein deriving the at least one channel metric set comprises initially deriving a channel metric set of the at least one channel metric set to comprise Z channel metrics, and wherein: in response to Z being less than a threshold, the user device omits the channel metric set comprising the Z channel metrics from being input to at least one processing unit to derive at least one channel state information (CSI) set; and in response to Z being greater than or equal to the threshold, the user device pads the channel metric set comprising the Z channel metrics to form a padded channel metric set comprising Y channel metrics, and inputs the padded channel metric set to the at least one processing unit to derive the at least one CSI set.
[0217] A twenty-third aspect includes the twenty-second aspect, and further includes wherein the threshold comprises or is one of: indicated by a network device via physical (PHY) layer signaling or signaling of a layer higher than the PHY layer; indicated by the user device to the network device; defined by a specification or protocol according to which the user device and the network device are configured to communicate and / or operate; or determined by an artificial intelligence (AI) / machine learning (ML) model, wherein different AI / ML models are configured with different threshold values.
[0218] A twenty-fourth aspect includes any of the first through twenty-third aspects, and further includes: indicating, by the user device, one or more indexes of one or more padding channel metrics padded to one or more of the at least one channel metric set.
[0219] A twenty-fifth aspect includes any of the first through twenty-fourth aspects, and further includes: receiving, by the network device, an indication of one or more indexes of one or more padding channel metrics padded to one or more of the at least one channel metric set.
[0220] A twenty-sixth aspect includes the twenty-fifth aspect, and further includes wherein the at least one CSI set that the network device receives is derived based on the one or more padding channel metrics, and wherein the network device performs scheduling based on only those channel metrics of the at least one reconstructed channel metric set that do not correspond to the one or more indexes of the one or more padding channel metrics that are indicated.
[0221] A twenty-seventh aspect includes any of the twenty-fourth through twenty-sixth aspects, and further includes wherein the one or more indexes are indicated using: a bitmap that indicates which channel metrics of the at least one channel metric set are padding channel metrics; a one-level indicator that indicates which channel metrics of the at least one channel metric set are padding channel metric sets; or a two-level indicator that indicates which of the at least one channel metric set is a padded channel metric set and which channel metrics of the at least one channel metric set are padding channel metrics.
[0222] A twenty-eighth aspect includes the twenty-seventh aspect, and further includes wherein the two-level indicator comprises a bitmap.
[0223] A twenty-ninth aspect includes any of the tenth through twenty-eighth aspects, and further includes wherein the user device pads one or more channel metrics to one or more of the at least one channel metric sets, wherein the one or more channel metrics that are padded corresponds to:one or more indexes of a last (Y-Z) number of channel metrics in a last channel metric set of the at least one channel metric set; or one or more indexes of a first (Y-Z) number of channel metrics in a first channel metric set of the at least one channel metric set.
[0224] A thirtieth aspect includes any of the first through twenty-ninth aspects, and further includes: indicating, by the user device, one or more indexes of one or more omitted channel metrics of the at least one channel metric set, the one or more omitted channel metrics omitted from being input to at least one processing unit to derive at least one channel state information (CSI) set.
[0225] A thirty-first aspect includes any of the first through thirtieth aspects, and further includes: receiving, by the network device, an indication of one or more indexes of one or more omitted channel metrics omitted by the user device from being input to at least one processing unit to derive the at least one channel state information (CSI) set.
[0226] A thirty-second aspect includes any of the thirtieth or thirty-first aspects, and further includes wherein the one or more indexes are indicated using: a bitmap that indicates which channel metrics of the at least one channel metric set are omitted channel metrics; or a one-level indicator that indicates which channel metrics of the at least one channel metric set are omitted channel metrics.
[0227] A thirty-third aspect includes any of the second through thirty-second aspects, and further includes wherein the at least one CSI set comprises a plurality of CSIs having indexes that increase with decreasing priority of the plurality of CSIs.
[0228] A thirty-fourth aspect includes any of the second through thirty-third aspects, and further includes wherein the at least one CSI set comprises a plurality of CSI sets having indexes that increase with decreasing priority of the plurality of CSI sets.
[0229] A thirty-fifth aspect includes any of the second through thirty-fourth aspects, and further includes wherein priorities of the at least one CSI set decreases with increasing time instances of the at least one CSI set.
[0230] A thirty-sixth aspect includes any of the second through thirty-fifth aspects, and further includes wherein priorities of the at least one CSI set decrease with increasing layer indexes of the at least one CSI set.
[0231] A thirty-seventh aspect includes any of the first through thirty-sixth aspects, and further includes wherein time domain information associated with the at least one channel metric set has higher priority than layer information associated with the at least one channel metric set, and priorities of the at least one channel metric set decreases with increasing time instances and / or increasing layer indexes.
[0232] A thirty-eighth aspect includes any of the first through thirty-seventh aspects, and further includes wherein layer information associated with the at least one metric set has higher priority than time-domain information associated with the at least one channel metric set, and priorities of the at least one channel metric set decreases with increasing time instances and / or increasing layer indexes.
[0233] A thirty-ninth aspect includes any of the second through thirty-eighth aspects, and further includes wherein the user device transmits and / or the network device receives the at least one CSI set in a first CSI part and a second CSI part, wherein the at least one CSI set comprises a plurality of CSI sets, and wherein a first CSI set of the plurality of CSI sets is communicated in the first CSI part and a second CSI set of the plurality of CSI sets is communicated in the second CSI part.
[0234] A fortieth aspect includes the thirty-ninth aspect, and further includes wherein the first CSI set supports repetition or re-transmission, and the second CSI set supports transmission only once.
[0235] A forty-first aspect includes any off the thirty-ninth or fortieth aspects, and further includes wherein the first CSI set uses a lower modulation order than a modulation order used for the second CSI set.
[0236] A forty-second aspect includes any of the thirty-ninth through forty-first aspects, and further includes wherein the first CSI set uses a lower coding rate than a coding rate used for the second set of CSI.
[0237] A forty-third aspect includes any of the thirty-ninth through forty-second aspects, and further includes wherein the first CSI set uses a lower modulation and coding scheme (MCS) than a MCS used for the second CSI set.
[0238] A forty-fourth aspect includes any of the thirty-ninth through forty-third aspects, and further includes wherein the first CSI set uses a higher transmission power than a transmission power used for the second CSI set.
[0239] A forty-fifth aspect includes any of the thirty-ninth through forty-fourth aspects, and further includes wherein the at least one CSI set comprises a plurality of CSIs for a plurality of time instances, the plurality of CSIs mapped to one CSI report according to: a first criterion that a preset number of highest priority CSIs of the plurality of CSIs are included in the first CSI part, and a remainder of the plurality of CSIs are included in the second CSI part; or a second criterion that all of the plurality of CSIs are included in the second CSI part, wherein a preset number of highest priority CSIs are included in a first group of the second CSI part, and a remainder of the plurality of CSIs are included in a second group of the second CSI part and / or a third group of the second CSI part.
[0240] A forty-sixth aspect includes any of the second through forty-sixth aspects, and further includes wherein the user device transmits and / or the network device receives the at least one CSI set in at least one CSI report in a fixed time duration.
[0241] A forty-seventh aspect includes any of the second through forty-sixth aspects, and further includes wherein the user device transmits and / or the network device receives different numbers of CSIs and / or different numbers of CSI sets of the at least one CSI set in a plurality of CSI reports.
[0242] A forty-eighth aspect includes any of the first through forty-seventh aspect, and further includes wherein the at least one processing unit performs processing using at least one artificial intelligence (AI) and / or machine learning (ML) model.
[0243] A forty-ninth aspect includes a wireless communications apparatus comprising a processor and a memory, wherein the processor is configured to read code from the memory to cause the apparatus to perform any of the first through forty-eighth aspects.
[0244] A fiftieth aspect includes a computer program product comprising a computer-readable program medium comprising code stored thereupon, the code, when executed by a processor, causing the processor to perform any of the first through forty-eighth aspects.
[0245] In addition to the features mentioned in each of the independent aspects enumerated above, some examples may show, alone or in combination, the optional features mentioned in the dependent aspects and / or as disclosed in the description above and shown in the figures.
Claims
1.A method for wireless communication, the method comprising:receiving, by a user device, a channel state information reference signal (CSI-RS) ;determining, by the user device, a measured channel metric based on the CSI-RS; andderiving, by the user device, at least one channel metric set based on the measured channel metric, each channel metric set comprising one or more channel metrics.2.A method for wireless communication, the method comprising:transmitting, by a network device, a channel state information reference signal (CSI-RS) ;receiving, by the network device, at least one channel information (CSI) set derived based on the CSI-RS; andreconstructing, by the network device, at least one channel metric set based on the at least one CSI set, each channel metric set comprising one or more channel metrics.3.The method of any of claims 1 or 2, wherein the at least one channel metric set comprises a plurality of channel metric sets.4.The method of claim 1, further comprising:deriving, by at least one processing unit of the user device, at least one channel state information (CSI) set based on the at least one channel metric set input to the at least one processing unit.5.The method of claim 1, further comprising:transmitting, by the user device, the at least one CSI set via at least one CSI report.6.The method of claim 2, further comprising:receiving, by the network device, the at least one CSI set via at least one CSI report.7.The method of any of claims 5 or 6, wherein the at least one CSI report comprises multiple CSI reports, and wherein:the at least one CSI set comprises one CSI set comprising multiple parts, and the user device transmits and / or the network device receives the multiple parts of the one CSI set in the multiple CSI reports;the at least one CSI set comprises multiple CSI sets, and the user device transmits and / or the network device receives the multiple CSI sets in the multiple CSI reports; orthe at least one CSI set comprises multiple parts of multiple CSI sets, and the user device transmits and / or the network device receives the multiple parts of the multiple CSI sets in the multiple CSI reports.8.The method of any of claims 1 or 2, wherein the at least one channel metric set comprises a plurality of channel metric sets, and wherein channel metrics are included in the channel metric sets according to consecutive time units of the channel metrics.9.The method of any of claims 1 or 2, wherein the at least one channel metric set comprises a plurality of channel metric sets, and where channel metrics are included in the plurality of channel metric sets according to time intervals of the channel metrics.10.The method of any of claims 1 or 2, wherein X is a total number of channel metrics of the at least one channel metric set, Y is a number of channel metrics that the user device inputs to at least one processing unit per time unit, and Z is equal to (X mod Y) .11.The method of claim 10, wherein the time unit comprises inference time, and wherein the at least one processing unit uses an artificial intelligence (AI) and / or machine learning (ML) model to process the Y channel metrics per inference time.12.The method of claim 10, wherein the at least one channel metric set is derived by the user device by:initially deriving a channel metric set of the at least one channel metric set to comprise Z channel metrics; andin response to Z being less than Y, padding the channel metric set comprising the Z channel metrics to form a padded channel metric set.13.The method of claim 12, wherein padding the channel metric set comprises padding a (Y-Z) number of channel metrics to the channel metric set so that the padded channel metric set comprises Y channel metrics.14.The method of claim 13, wherein padding the (Y-Z) number of channel metrics to the channel metric set comprises:copying a channel metric of a last time instance in the at least one channel metric set the (Y-Z) number of times;copying a measured channel metric of a last time instance the (Y-Z) number of times;copying at least one channel metric of a last (Y-Z) number of time instances;copying at least one measured channel metric of a last (Y-Z) number of time instances;copying at least one measured channel metric of a first (Y-Z) number of time instances;padding according to a preset scheme; orcopying at least one channel metric of a last y number of time instances and / or a next z number of time instances, where y+z = Y-Z.15.The method of claim 10, wherein at least one channel metric set is derived by the user device by initially deriving multiple channel metric sets that each comprise less than Y channel metrics, and padding each of the multiple channel metric sets.16.The method of claim 15, wherein padding the (Y-Z) number of channel metrics to the multiple channel metric sets comprises overlapping at least one channel metric of y time instances between two adjacent sets of the multiple channel metric sets.17.The method of claim 12, wherein one of the one or more of the at least one channel metric set that is padded comprises the measured channel metric.18.The method of claim 12, further comprising:inputting, by the user device, the padded channel metric set to the at least one processing unit to derive at least one channel state information (CSI) set.19.The method of claim 10, wherein a channel metric set of the at least one channel metric set comprises Z channel metrics, and wherein the user device omits the channel metric set comprising the Z channel metrics from being input to at least one processing unit to derive at least one channel state information (CSI) set.20.The method of claim 19, wherein the user device omits the channel metric set in response to Z being less than Y.21.The method of claim 19, wherein the Z channel metrics of the omitted channel metric set comprises a channel metric at a first slot, a last slot, or a middle slot among a plurality of slots of X channel metrics of the at least one channel metric set.22.The method of claim 10, wherein deriving the at least one channel metric set comprises initially deriving a channel metric set of the at least one channel metric set to comprise Z channel metrics, and wherein:in response to Z being less than a threshold, the user device omits the channel metric set comprising the Z channel metrics from being input to at least one processing unit to derive at least one channel state information (CSI) set; andin response to Z being greater than or equal to the threshold, the user device pads the channel metric set comprising the Z channel metrics to form a padded channel metric set comprising Y channel metrics, and inputs the padded channel metric set to the at least one processing unit to derive the at least one CSI set.23.The method of claim 22, wherein the threshold comprises or is one of:indicated by a network device via physical (PHY) layer signaling or signaling of a layer higher than the PHY layer;indicated by the user device to the network device;defined by a specification or protocol according to which the user device and the network device are configured to communicate and / or operate; ordetermined by an artificial intelligence (AI) / machine learning (ML) model, wherein different AI / ML models are configured with different threshold values.24.The method of claim 1, further comprising:indicating, by the user device, one or more indexes of one or more padding channel metrics padded to one or more of the at least one channel metric set.25.The method of claim 2, further comprising:receiving, by the network device, an indication of one or more indexes of one or more padding channel metrics padded to one or more of the at least one channel metric set.26.The method of claim 25, wherein the at least one CSI set that the network device receives is derived based on the one or more padding channel metrics, and wherein the network device performs scheduling based on only those channel metrics of the at least one reconstructed channel metric set that do not correspond to the one or more indexes of the one or more padding channel metrics that are indicated.27.The method of any of claims 24 or 25, wherein the one or more indexes are indicated using:a bitmap that indicates which channel metrics of the at least one channel metric set are padding channel metrics;a one-level indicator that indicates which channels metrics of the at least one channel metric set are padding channel metric set; ora two-level indicator that indicates which of the at least one channel metric set is a padded channel metric set and which channel metrics of the at least one channel metric set are padding channel metrics.28.The method of claim 27, wherein the two-level indicator comprises a bitmap.29.The method of claim 10, wherein the user device pads one or more channel metrics to one or more of the at least one channel metric sets, wherein the one or more channel metrics that are padded corresponds to:one or more indexes of a last (Y-Z) number of channel metrics in a last channel metric set of the at least one channel metric set; orone or more indexes of a first (Y-Z) number of channel metrics in a first channel metric set of the at least one channel metric set.30.The method of claim 1, further comprising:indicating, by the user device, one or more indexes of one or more omitted channel metrics of the at least one channel metric set, the one or more omitted channel metrics omitted from being input to at least one processing unit to derive at least one channel state information (CSI) set.31.The method of claim 2, further comprising:receiving, by the network device, an indication of one or more indexes of one or more omitted channel metrics omitted by the user device from being input to at least one processing unit to derive the at least one channel state information (CSI) set.32.The method of any of claims 30 or 31, wherein the one or more indexes are indicated using:a bitmap that indicates which channel metrics of the at least one channel metric set are omitted channel metrics; ora one-level indicator that indicates which channel metrics of the at least one channel metric set are omitted channel metrics.33.The method of any of claims 2 or 4, wherein the at least one CSI set comprises a plurality of CSIs having indexes that increase with decreasing priority of the plurality of CSIs.34.The method of any of claims 2 or 4, wherein the at least one CSI set comprises a plurality of CSI sets having indexes that increase with decreasing priority of the plurality of CSI sets.35.The method of any of claims 2 or 4, wherein priorities of the at least one CSI set decreases with increasing time instances of the at least one CSI set.36.The method of any of claims 2 or 4, wherein priorities of the at least one CSI set decrease with increasing layer indexes of the at least one CSI set.37.The method of claim 1, wherein time domain information associated with the at least one channel metric set has higher priority than layer information associated with the at least one channel metric set, and priorities of the at least one channel metric set decreases with increasing time instances and / or increasing layer indexes.38.The method of claim 1, wherein layer information associated with the at least one metric set has higher priority than time-domain information associated with the at least one channel metric set, and priorities of the at least one channel metric set decreases with increasing time instances and / or increasing layer indexes.39.The method of any of claims 2 or 4, wherein the user device transmits and / or the network device receives the at least one CSI set in a first CSI part and a second CSI part, wherein the at least one CSI set comprises a plurality of CSI sets, and wherein a first CSI set of the plurality of CSI sets is communicated in the first CSI part and a second CSI set of the plurality of CSI sets is communicated in the second CSI part.40.The method of claim 39, wherein the first CSI set supports repetition or re-transmission, and the second CSI set supports transmission only once.41.The method of claim 39, wherein the first CSI set uses a lower modulation order than a modulation order used for the second CSI set.42.The method of claim 39, wherein the first CSI set uses a lower coding rate than a coding rate used for the second set of CSI.43.The method of claim 39, wherein the first CSI set uses a lower modulation and coding scheme (MCS) than a MCS used for the second CSI set.44.The method of claim 39, wherein the first CSI set uses a higher transmission power than a transmission power used for the second CSI set.45.The method of claim 39, wherein the at least one CSI set comprises a plurality of CSIs for a plurality of time instances, the plurality of CSIs mapped to one CSI report according to:a first criterion that a preset number of highest priority CSIs of the plurality of CSIs are included in the first CSI part, and a remainder of the plurality of CSIs are included in the second CSI part; ora second criterion that all of the plurality of CSIs are included in the second CSI part, wherein a preset number of highest priority CSIs are included in a first group of the second CSI part, and a remainder of the plurality of CSIs are included in a second group of the second CSI part and / or a third group of the second CSI part.46.The method of any of claims 2 or 4, wherein the user device transmits and / or the network device receives the at least one CSI set in at least one CSI report in a fixed time duration.47.The method of any of claims 2 or 4, wherein the user device transmits and / or the network device receives different numbers of CSIs and / or different numbers of CSI sets of the at least one CSI set in a plurality of CSI reports.48.The method of any of claims 4, 18, 19, 22, 30, or 31, wherein the at least one processing unit performs processing using at least one artificial intelligence (AI) and / or machine learning (ML) model.49.A wireless communications apparatus comprising a processor and a memory, wherein the processor is configured to read code from the memory to cause the apparatus to perform a method of any of claims 1 to 48.50.A computer program product comprising a computer-readable program medium comprising code stored thereupon, the code, when executed by a processor, causing the processor to perform a method of any of claims 1 to 48.
Citation Information
Patent Citations
Channel status information reporting method, user equipment and base station
CN105309030A
Hybrid channel state feedback
CN113366881A
Method for reporting channel state information and user equipment
CN116915291A
Transmitting channel state information reports
CN117581489A