WO2023173262A1 - 信息处理方法及装置、通信设备及存储介质 - Google Patents

信息处理方法及装置、通信设备及存储介质 Download PDF

Info

Publication number
WO2023173262A1
WO2023173262A1 PCT/CN2022/080785 CN2022080785W WO2023173262A1 WO 2023173262 A1 WO2023173262 A1 WO 2023173262A1 CN 2022080785 W CN2022080785 W CN 2022080785W WO 2023173262 A1 WO2023173262 A1 WO 2023173262A1
Authority
WO
WIPO (PCT)
Prior art keywords
model
csi
terminal
information
feedback information
Prior art date
Application number
PCT/CN2022/080785
Other languages
English (en)
French (fr)
Inventor
牟勤
乔雪梅
蔡一凡
Original Assignee
北京小米移动软件有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京小米移动软件有限公司 filed Critical 北京小米移动软件有限公司
Priority to PCT/CN2022/080785 priority Critical patent/WO2023173262A1/zh
Priority to CN202280000750.1A priority patent/CN114788317A/zh
Publication of WO2023173262A1 publication Critical patent/WO2023173262A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W8/00—Network data management
    • H04W8/22—Processing or transfer of terminal data, e.g. status or physical capabilities
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/003—Arrangements for allocating sub-channels of the transmission path
    • H04L5/0048—Allocation of pilot signals, i.e. of signals known to the receiver
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/02—Arrangements for optimising operational condition
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/06—Optimizing the usage of the radio link, e.g. header compression, information sizing, discarding information

Definitions

  • the present disclosure relates to the field of wireless communication technology but is not limited to the field of wireless communication technology, and in particular, to an information processing method and device, communication equipment and storage medium.
  • Channel state information can describe the current channel environment.
  • the base station transmits channel state information-reference signal (Channel State Information-Reference Signal, CSI-RS).
  • CSI-RS Channel State Information-Reference Signal
  • the terminal evaluates the channel state information and quantitatively feeds it back to The base station, by introducing Channel State Information (CSI) feedback information, can make timely adjustments when sending the channel state information reference signal, thereby reducing the bit error rate at the terminal and obtaining the optimal received signal.
  • CSI Channel State Information
  • Embodiments of the present disclosure provide an information processing method and device, communication equipment, and storage media.
  • the first aspect of the embodiments of the present disclosure provides an information processing method, which is executed by a base station.
  • the method includes:
  • the second aspect of the embodiment of the present disclosure provides an information processing method, which is executed by a terminal.
  • the method includes:
  • Send second information where the second information is used by the base station to determine whether the terminal supports compression of CSI-RS feedback information by at least one AI model.
  • a third aspect of the embodiments of the present disclosure provides an information processing device, wherein the method includes:
  • a determination module used to determine whether the terminal supports channel state information-reference signal CSI-RS feedback information compression of at least one artificial intelligence AI model
  • a configuration module configured to configure CSI-RS according to whether the terminal supports CSI-RS feedback information compression of at least one AI model.
  • a fourth aspect of the embodiment of the present disclosure provides an information processing device, where the device includes:
  • the sending module is configured to send second information, where the second information is used by the base station to determine whether the terminal supports compression of CSI-RS feedback information by at least one AI model.
  • a fifth aspect of the embodiment of the present disclosure provides a communication device, including a processor, a transceiver, a memory, and an executable program stored on the memory and capable of being run by the processor, wherein the processor runs the executable program.
  • the program executes the information processing method provided by the first aspect or the second aspect.
  • a sixth aspect of the embodiments of the present disclosure provides a computer storage medium that stores an executable program; after the executable program is executed by a processor, the information provided by the first aspect or the second aspect can be realized Approach.
  • the base station will perform CSI-RS configuration for the corresponding terminal based on whether the terminal supports at least one AI model for compressing CSI-RS feedback information.
  • the generated CSI-RS configuration is consistent with the terminal It is adapted to the AI capabilities, thereby reducing UE measurement anomalies caused by unsuitable CSI-RS configuration.
  • Figure 1 is a schematic structural diagram of a wireless communication system according to an exemplary embodiment
  • Figure 2 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 3 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 4 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 5 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 6 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 7A is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 7B is a schematic flowchart of an information processing method according to an exemplary embodiment.
  • Figure 8 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 9 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 10 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 11 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 12 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 13 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 14 is a schematic flowchart of an information processing method according to an exemplary embodiment
  • Figure 15 is a schematic structural diagram of an information processing device according to an exemplary embodiment
  • Figure 16 is a schematic structural diagram of an information processing device according to an exemplary embodiment
  • Figure 17 is a schematic structural diagram of a terminal according to an exemplary embodiment
  • Figure 18 is a schematic structural diagram of a communication device according to an exemplary embodiment.
  • first, second, third, etc. may be used to describe various information in the embodiments of the present disclosure, the information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other.
  • first information may also be called second information, and similarly, the second information may also be called first information.
  • word “if” as used herein may be interpreted as "when” or "when” or "in response to determining.”
  • FIG. 1 shows a schematic structural diagram of a wireless communication system provided by an embodiment of the present disclosure.
  • the wireless communication system is a communication system based on cellular mobile communication technology.
  • the wireless communication system may include: several terminals 11 and several access devices 12.
  • the terminal 11 may be a device that provides voice and/or data connectivity to the user.
  • Terminal 11 can communicate with one or more core networks via a Radio Access Network (RAN).
  • RAN Radio Access Network
  • Terminal 11 can be an Internet of Things terminal, such as a sensor device, a mobile phone (or "cellular" phone) and a device with The computer of the Internet of Things terminal, for example, can be a fixed, portable, pocket-sized, handheld, computer-built-in or vehicle-mounted device.
  • station STA
  • subscriber unit subscriber unit
  • subscriber station subscriber station
  • mobile station mobile station
  • remote station remote station
  • access terminal remote terminal
  • user terminal user agent, user device, or user equipment (terminal).
  • the terminal 11 may be a device of an unmanned aerial vehicle.
  • the terminal 11 may also be a vehicle-mounted device, for example, it may be an on-board computer with a wireless communication function, or a wireless communication device connected to an external on-board computer.
  • the terminal 11 may also be a roadside device, for example, it may be a streetlight, a signal light or other roadside device with wireless communication function.
  • the access device 12 may be a network-side device in the wireless communication system.
  • the wireless communication system can be the 4th generation mobile communication technology (the 4th generation mobile communication, 4G) system, also known as the Long Term Evolution (LTE) system; or the wireless communication system can also be a 5G system, Also called new radio (NR) system or 5G NR system.
  • the wireless communication system may also be a next-generation system of the 5G system.
  • the access network in the 5G system can be called NG-RAN (New Generation-Radio Access Network). Or, MTC system.
  • the access device 12 may be an evolved access device (eNB) used in the 4G system.
  • the access device 12 may also be an access device (gNB) using a centralized distributed architecture in the 5G system.
  • eNB evolved access device
  • gNB access device
  • the access device 12 adopts a centralized distributed architecture it usually includes a centralized unit (central unit, CU) and at least two distributed units (distributed unit, DU).
  • the centralized unit is equipped with a protocol stack including the Packet Data Convergence Protocol (PDCP) layer, the Radio Link Control protocol (Radio Link Control, RLC) layer, and the Media Access Control (Media Access Control, MAC) layer; distributed
  • PDCP Packet Data Convergence Protocol
  • RLC Radio Link Control
  • MAC Media Access Control
  • the unit is provided with a physical (Physical, PHY) layer protocol stack, and the embodiment of the present disclosure does not limit the specific implementation of the access device 12.
  • a wireless connection can be established between the access device 12 and the terminal 11 through a wireless air interface.
  • the wireless air interface is a wireless air interface based on the fourth generation mobile communication network technology (4G) standard; or the wireless air interface is a wireless air interface based on the fifth generation mobile communication network technology (5G) standard, such as
  • the wireless air interface is a new air interface; alternatively, the wireless air interface may also be a wireless air interface based on the next generation mobile communication network technology standard of 5G.
  • an embodiment of the present disclosure provides an information processing method, which is executed by a network side device.
  • the method includes:
  • S110 Determine whether the terminal supports channel state information-reference signal CSI-RS feedback information processing based on at least one artificial intelligence AI model;
  • S120 Configure CSI-RS for the terminal according to whether the terminal supports CSI-RS feedback information processing based on at least one AI model.
  • the network side device may be a base station, which may be an evolved base station (eNB) and/or a next-generation base station (gNB), or a base station of any generation communication system.
  • eNB evolved base station
  • gNB next-generation base station
  • the network-side device is not limited to the base station, but can also be any device in the network, which is not limited here.
  • the base station transmits CSI-RS, and the terminal receives the CSI-RS.
  • the terminal generates CSI-RS feedback information based on its own reception of CSI-RS. That is, the CSI-RS feedback information can be used by the network side device to determine the reception status of the terminal's CSI-RS, for example, whether the CSI-RS on the corresponding transmission resource block is received and/or the terminal's receiving power of the CSI-RS, etc. .
  • the data amount of the CSI-RS feedback information may be relatively large.
  • the terminal may only send part of the CSI-RS feedback information to the network side device.
  • the terminal can use an AI model to process CSI-RS feedback information.
  • the terminal reports processed CSI-RS feedback information.
  • the processed CSI-RS feedback information is processed by the network side device (such as the base station). After corresponding processing, the complete CSI-RS feedback information of the terminal can be obtained.
  • using the AI model to process the CSI-RS feedback information may include: the terminal uses the AI model to compress the CSI-RS feedback information. Based on this, the terminal reports compressed CSI-RS feedback information. After the compressed CSI-RS feedback information is correspondingly decompressed by the base station, the complete CSI-RS feedback information of the terminal can be obtained.
  • an AI model to process CSI-RS feedback information usually requires at least the terminal to have AI computing capabilities. For example, if the terminal contains an AI chip, the terminal has AI computing capabilities. Different AI chips have different AI computing capabilities, and different AI models have different computing power requirements. Therefore, some terminals may support processing of CSI-RS feedback information through all AI models, and some terminals may support only some AI models. Process CSI-RS information, but some terminals do not support using the AI model to process CSI-RS feedback information at all.
  • the AI model is used to compress CSI-RS feedback information.
  • those skilled in the art can understand that using the AI model to process CSI-RS feedback information may also include other methods. The operation will not be described in detail here.
  • network side devices such as base stations will first determine whether the terminal supports CSI-RS feedback information processing based on at least one AI model. According to the determination result, a CSI-RS configuration for the UE is generated. In this way, it can be ensured that the generated CSI-RS configuration is adapted to whether the UE supports the compression of CSI-RS feedback information by the AI model, and unsuitable CSI-RS configurations can be reduced. UE measurement anomalies caused by RS configuration.
  • the generated CSI-RS configuration can be sent to the terminal through RRC message or MAC CE.
  • Embodiments of the present disclosure provide an information processing method, which is executed by a network side device.
  • the method includes:
  • determining a CSI-RS used by the terminal In response to determining that the terminal supports AI model-based CSI-RS feedback information processing, determining a CSI-RS used by the terminal, wherein the CSI-RS corresponds to an AI model supported by the terminal.
  • an embodiment of the present disclosure provides an information processing method, which is executed by a network side device.
  • the method includes:
  • S320 Receive the first information provided by the terminal according to the complexity information
  • S330 According to the first information, determine whether the terminal supports CSI-RS feedback information processing based on at least one AI model;
  • S340 Configure CSI-RS according to whether the terminal processes CSI-RS feedback information based on supporting at least one AI model.
  • using the AI model to process the CSI-RS feedback information may include: the terminal uses the AI model to compress the CSI-RS feedback information. Based on this, the terminal reports compressed CSI-RS feedback information. After the compressed CSI-RS feedback information is correspondingly decompressed by the network side device (such as the base station), the complete CSI-RS feedback information of the terminal can be obtained.
  • the AI model is used to compress CSI-RS feedback information.
  • those skilled in the art can understand that using the AI model to process CSI-RS feedback information may also include other methods. The operation will not be described in detail here.
  • the network side device will send the complexity information of the AI model to the terminal.
  • the terminal can combine its own AI capabilities to determine whether it supports using the corresponding AI model to CSI-RS Feedback information compression.
  • network-side devices and terminals can also be determined through candidate AI models determined by the communication protocol, which will not be described again here.
  • This complexity information can be sent by the base station to the terminal through an RRC message or MAC CE.
  • the first information may be feedback information for various network side devices (such as base stations) to determine whether the terminal supports AI model compressed CSI-RS, or the first information may be used for the terminal to determine which AI model compressed CSI-RS Feedback.
  • the terminal actually determines whether it supports the AI model to compress CSI-RS feedback information and informs the base station.
  • the first information may explicitly indicate whether the terminal supports AI model compressed CSI-RS feedback information or which AI model supports compressed CSI-RS feedback information.
  • the first information may also implicitly indicate whether the terminal supports AI model compressed CSI-RS feedback information and/or supports which AI model compressed CSI-RS feedback information.
  • the first information indicating that the terminal supports CSI-RS feedback information compression of at least one AI model includes at least one of the following parameters:
  • Model identification indicating the AI model supported by the terminal for CSI-RS feedback information compression
  • the CSI calculation duration indicates the duration required for the terminal to use the AI model identified by the model to compress CSI-RS feedback information.
  • the CSI calculation duration can be either the duration itself or a parameter used to determine the CSI calculation duration; for example, it can be the computing power of the AI model or the computing power of the terminal, etc.
  • the model identifier refers to an AI model supported by the terminal.
  • the model identification may also indicate an AI model that is not supported by the terminal.
  • the CSI calculation time can be: the time required for the terminal to use the corresponding AI model to compress CSI-RS feedback information. If the CSI calculation time is greater than the maximum time allowed by the base station, even if the terminal supports the AI model, the network side device may fail due to timeout. The terminal will not be configured to use this AI model to compress CSI-RS feedback information.
  • the first information is equivalent to an implicit indication of whether the terminal supports AI model compressed CSI-RS feedback information. If the first information does not carry any model identifier or the terminal does not send the first information, it is equivalent to indicating that the terminal does not support any AI model compressed CSI-RS feedback information. If the first information carries at least one model identifier, it means that the terminal supports at least one AI model to compress CSI-RS feedback information.
  • the first information indicating that the terminal does not support at least one AI model for CSI-RS information feedback compression indicates that the CSI-RS feedback information adopts a partial reporting method.
  • the first information indicates that the terminal will send CSI-RS feedback information to the base station by reporting part of the CSI-RS feedback information, it is equivalent to an implicit explanation that the terminal does not support the AI model to compress CSI-RS feedback information or the terminal does not expect to use the AI model. Compress CSI-RS feedback information.
  • the first information may include: a bit specifically indicating whether the terminal supports AI model compressed CSI-RS feedback information.
  • the first information may include a bitmap, and the bitmap may include N bits, N ⁇ 1, each bit corresponding to an AI model; wherein, in the bitmap One bit is used to indicate whether the terminal supports the corresponding AI model compressed CSI-RS feedback information.
  • the first information may include a bitmap, the bits of the bitmap may correspond to 2 N values, and each value corresponds to an AI model.
  • the complexity information indicates at least one of the following:
  • the total number of floating-point operations corresponding to the AI model where the total number of floating-point operations and the maximum allowed value of the CSI calculation duration are jointly used for the terminal to determine whether to support CSI-RS feedback information compression of the corresponding AI model;
  • the first ratio between the complexity of the corresponding AI model and the complexity of the baseline AI model where the first ratio is used for the terminal to combine the terminal AI capabilities with the complexity of the baseline AI model. , determine whether the CSI-RS feedback information compression of the corresponding AI model is supported.
  • the baseline AI model here can be any one of multiple AI models that perform CSI-RS feedback information compression, for example, it can be an AI model specified by the base station or the terminal.
  • the complexity of the baseline AI model can be an AI model known to both the base station and the UE.
  • the baseline AI model may be an AI model with the lowest complexity or the highest complexity among multiple AI models that support CSI-RS feedback information compression. In this way, the value range of the first ratio can be limited to a specific range, thereby reducing the bit overhead for indicating the first ratio.
  • the complexity information can directly indicate the total number of floating point operations of the corresponding AI model.
  • the terminal After receiving it, the terminal can determine itself based on the maximum allowable value of the CSI calculation duration configured by the base station or the maximum allowable value agreed upon by the protocol. Whether to support using the corresponding AI model to compress CSI-RS feedback information.
  • the complexity information is represented by a ratio between the complexity of the corresponding AI model and the baseline AI model.
  • the terminal indicates in advance the total number of floating point operations of the baseline AI model. Therefore, after receiving the first ratio, the terminal can also determine whether it supports the corresponding AI model compressed CSI-RS feedback information.
  • an embodiment of the present disclosure provides an information processing method, which is executed by a network side device.
  • the method includes:
  • S420 Determine whether the terminal supports CSI-RS feedback information compression of at least one AI model according to the AI capability information.
  • S430 Configure CSI-RS according to whether the terminal supports CSI-RS feedback information compression of at least one AI model.
  • network side devices such as base stations receive AI capability information sent by the terminal.
  • Network side devices such as base stations combine the AI capability information of the terminal and the complexity information of each AI model to determine whether the terminal supports AI model compressed CSI. -RS feedback information, or which AI model the terminal supports to compress CSI-RS feedback information.
  • the AI capability information indicates at least one of the following:
  • the performance parameters of the terminal may include at least one of the following parameters:
  • the terminal supports the number of floating point operations per unit time (for example, per second);
  • the second ratio indicates the ratio between the terminal AI capability and the complexity of the baseline AI model.
  • performance parameters can also be characterized by other parameters, which will not be described again here.
  • the AI capability information may at least indicate whether the terminal has AI capability. If the terminal does not have AI capability, the AI capability information of the terminal will not carry the float per unit time (for example, per second) supported by the terminal. The number of point operations, or the total number of floating point operations that can be performed within the maximum allowed value of the CSI calculation time.
  • the second ratio in order to reduce signaling bit overhead, is used to indirectly indicate the AI capability of the terminal.
  • the terminal indicated by the AI capability information is 0 within the maximum allowed value of CSI calculation duration, it means that the terminal does not have AI capability. If the number of floating-point operations supported within the maximum allowed value of the CSI calculation duration reported by the terminal is lower than the total number of floating-point operations of any AI model, it also means that the terminal does not support compression of CSI-RS feedback information by any AI model. If the number of floating-point operations supported within the maximum allowed value of the CSI calculation duration reported by the terminal is greater than or equal to the total number of floating-point operations of at least one AI model, it means that at least one AI model of the terminal has compressed the CSI-RS feedback information.
  • the base station When the base station receives the AI capability information reported by the terminal, it will determine whether the corresponding terminal supports AI model compressed CSI-RS feedback information.
  • configuring CSI-RS according to whether the terminal supports CSI-RS feedback information compression of at least one AI model includes:
  • the terminal When the terminal supports compression of CSI-RS feedback information based on at least one AI model, determine the CSI calculation time when the AI model supported by the terminal is used to compress CSI-RS feedback information, and configure the period of the CSI-RS .
  • the types of CSI-RS feedback information reported here may include at least the following two types:
  • Report compression information that compresses CSI-RS feedback information through the AI model
  • CSI-RS feedback information reported can also be other types, which will not be listed one by one here.
  • the network side device determines that the terminal configures the CSI-RS period using the CSI calculation time required by the corresponding AI model to compress CSI-RS feedback information. .
  • the CSI calculation duration is positively related to the period of the CSI-RS.
  • the network side device may also determine the measurement duration of the CSI-RS based on the CSI calculation duration.
  • the measurement duration may be the duration of measuring the CSI-RS in one CSI-RS period.
  • an embodiment of the present disclosure provides an information processing method, which is executed by a terminal.
  • the method includes:
  • S510 Send second information, where the second information is used for the network side device to determine whether the terminal supports compression of CSI-RS feedback information based on at least one AI model.
  • Embodiments of the present disclosure provide an information processing method, in which the terminal sends second information to the base station.
  • the second information can be used by the network side device to determine whether at least one AI model can be used to compress the CSI-RS feedback information.
  • the network side device can know whether the terminal supports using at least one AI model to compress CSI-RS feedback information, and then the network side device can determine whether the terminal supports the use of AI models or which AI model to use to compress CSI-RS feedback information.
  • Perform CSI-RS configuration so that the CSI-RS configuration generated by the base station is more in line with the terminal's capabilities.
  • an embodiment of the present disclosure provides an information processing method, which is executed by a terminal.
  • the method includes:
  • S610 Send AI capability information indicating the AI capability of the terminal to the network side device.
  • the AI capability information may be one of the aforementioned second information, or may be independent of the second information.
  • the network-side device can determine whether the terminal supports using the AI model to compress CSI-RS feedback information, and then generate CSI-RS configuration for the terminal.
  • the AI capability information indicates at least one of the following:
  • the performance parameters of the terminal may include at least one of the following parameters:
  • the terminal supports the number of floating point operations per unit time (for example, per second);
  • the second ratio indicates the ratio between the terminal AI capability and the complexity of the baseline AI model.
  • performance parameters can also be characterized by other parameters, which will not be described again here.
  • the AI capability information may at least indicate whether the terminal has AI capability. If the terminal does not have AI capability, the AI capability information of the terminal will not carry the float per unit time (for example, per second) supported by the terminal. The number of point operations, or the total number of floating point operations that can be performed within the maximum allowed value of the CSI calculation time.
  • the terminal indicated by the AI capability information is 0 within the maximum allowed value of CSI calculation duration, it means that the terminal does not have AI capability. If the number of floating-point operations supported within the maximum allowed value of the CSI calculation duration reported by the terminal is lower than the total number of floating-point operations of any AI model, it also means that the terminal does not support compression of CSI-RS feedback information by any AI model. If the number of floating-point operations supported within the maximum allowed value of the CSI calculation duration reported by the terminal is greater than or equal to the total number of floating-point operations of at least one AI model, it means that at least one AI model of the terminal has compressed the CSI-RS feedback information.
  • the second ratio in order to reduce signaling bit overhead, is used to indirectly indicate the AI capability of the terminal.
  • the base station When the base station receives the AI capability information reported by the terminal, it will determine whether the corresponding terminal supports AI model compressed CSI-RS feedback information.
  • an embodiment of the present disclosure provides an information processing method, which is executed by a terminal.
  • the method includes:
  • S710A Send first information to the network side device, where the first information is used to indicate whether the terminal supports compression of CSI-RS feedback information based on at least one AI model.
  • the first information may be one of the aforementioned second information.
  • the first information may be generated after the terminal determines whether it supports compression of CSI-RS feedback information based on any AI model based on its own AI capabilities and the complexity of each AI model.
  • the first information may be used by various network side devices (such as base stations) to determine whether the terminal supports AI model compressed CSI-RS feedback information, or the terminal determines which AI model compresses CSI-RS feedback information.
  • various network side devices such as base stations
  • the first information may explicitly or implicitly indicate whether the terminal supports AI model compressed CSI-RS feedback information or which AI model supports compressed CSI-RS feedback information.
  • the first information indicating that the terminal supports CSI-RS feedback information compression based on at least one AI model includes:
  • Model identification indicating the AI model supported by the terminal for CSI-RS feedback information compression
  • the CSI calculation duration indicates the duration required for the terminal to use the AI model identified by the model to compress CSI-RS feedback information.
  • the CSI calculation duration can be either the duration itself or a parameter used to determine the CSI calculation duration; for example, it can be the computing power of the AI model or the computing power of the terminal.
  • the model identifier refers to an AI model supported by the terminal.
  • the model identification may also indicate an AI model that is not supported by the terminal.
  • the CSI calculation time is: the time required for the terminal to use the corresponding AI model to compress CSI-RS feedback information. If the CSI calculation time is longer than the maximum time allowed by the base station, even if the terminal supports the AI model, the base station may not configure it due to timeout.
  • the terminal uses the AI model to compress CSI-RS feedback information.
  • the first information is equivalent to an implicit indication of whether the terminal supports AI model compressed CSI-RS feedback information. If the first information does not carry any model identifier or the terminal does not send the first information, it is equivalent to indicating that the terminal does not support any AI model compressed CSI-RS feedback information. If the first information carries at least one model identifier, it means that the terminal supports at least one AI model to compress CSI-RS feedback information.
  • CSI-RS feedback information adopts a partial reporting method.
  • the first information indicates that the terminal will send CSI-RS feedback information to the base station by reporting part of the CSI-RS feedback information, it is equivalent to an implicit explanation that the terminal does not support the AI model to compress CSI-RS feedback information or the terminal does not expect to use the AI model. Compress CSI-RS feedback information.
  • the first information may include: a bit specifically indicating whether the terminal supports AI model compressed CSI-RS feedback information.
  • the first information may include a bitmap, and the bitmap may include N bits, N ⁇ 1, each bit corresponding to an AI model; wherein one of the bitmaps The bit is used to indicate whether the terminal supports the corresponding AI model compressed CSI-RS feedback information.
  • the first information may include a bitmap, the bits of the bitmap may correspond to 2 N values, and each value corresponds to an AI model.
  • the method further includes:
  • S710B Receive the complexity information of the AI model
  • S720B Send the first information to the base station according to the complexity information of the AI model and the AI capability of the terminal.
  • the first information is determined based on the complexity information of the AI model received from the network side and its own AI capabilities.
  • the complexity information of the AI model may also be determined by the terminal according to protocol agreement or other methods.
  • the first information is not directly the AI capability information of the terminal's AI capability, but is generated and sent to the base station based on the complexity information of the AI model and the terminal's AI capability.
  • sending the first information to the base station according to the complexity information of the AI model and the AI capability of the terminal includes at least one of the following:
  • a message indicating that at least one AI model is supported is sent to the base station.
  • An AI model indicates to the base station the first information that supports CSI-RS feedback information compression of an AI model
  • the terminal sends a message to the base station indicating that the base station supports at least one AI without paying attention to which AI model the terminal specifically supports to compress CSI-RS feedback information.
  • the first information of model CSI-RS feedback information compression may not include the model identifier of the AI model supported by the terminal, but the base station can consider that the terminal supports at least the baseline AI model. Therefore, when configuring the CSI-RS for the terminal, it can at least base on the baseline AI model. Configure CIS-RS according to the complexity or required CSI calculation time.
  • the terminal not only determines whether the terminal supports at least one AI model to compress CSI-RS feedback information based on the complexity information and AI capabilities of each AI model, but also determines which AI model supports CSI-RS feedback information. Compression of feedback information, and carrying the model identifier of the AI model or the reference information of the AI model for which the terminal supports compression of CSI-RS feedback information in the first information and sending it to the terminal, so that the base station not only knows that the terminal supports at least one AI model for Compression of CSI-RS feedback information, and which AI model specifically supports the compression of CSI-RS feedback information.
  • the terminal If the terminal does not have AI capabilities or does not support compressed CSI-RS feedback information for any AI model indicated by the base station, the terminal sends first information indicating that at least one AI model does not support CSI-RS feedback information to the base station.
  • the first information indicating that the terminal supports CSI-RS feedback information compression of at least one AI model includes:
  • Model identification indicating the AI model supported by the terminal for CSI-RS feedback information compression
  • the CSI calculation duration indicates the duration required for the terminal to use the AI model identified by the model to compress CSI-RS feedback information.
  • the first information indication indicating that the terminal does not support CSI-RS information feedback compression based on at least one AI model CSI-RS feedback information adopts a partial reporting method.
  • the complexity information indicates at least one of the following:
  • the total number of floating-point operations corresponding to the AI model where the total number of floating-point operations and the maximum allowed value of the CSI calculation duration are jointly used for the terminal to determine whether to support CSI-RS feedback information compression of the corresponding AI model;
  • the first ratio between the complexity of the corresponding AI model and the complexity of the baseline AI model where the first ratio is used for the terminal to combine the terminal AI capabilities with the complexity of the baseline AI model. , determine whether the CSI-RS feedback information compression of the corresponding AI model is supported.
  • the total number of floating-point operations of each AI model and the number of floating-point operations performed per unit time (for example, per second) that the terminal can support can be obtained by using the corresponding AI model for the terminal to compress CSI-RS feedback information. If the actual calculation time is less than the CSI calculation time and is less than or greater than the maximum allowed value, it can be determined whether the terminal supports CSI-RS feedback information compression of the corresponding AI model.
  • the terminal If the complexity information is: the first ratio; then after receiving the first ratio, the terminal first determines the number of floating point operations it can perform based on the maximum allowable value of the CSI calculation duration and the number of floating point operations required by the baseline AI model. and then compare the third ratio with the first ratio. If the third ratio is greater than or equal to the first ratio, it can be considered that the terminal supports the corresponding AI model compression CSI-RS feedback information compression. If If the third ratio is less than the first ratio, it can be considered that the terminal does not support the corresponding AI model compression CSI-RS feedback information compression.
  • the terminal can directly report the CSI-RS feedback information indicating that the AI model compression is not supported. first information.
  • Embodiments of the present disclosure provide a CSI calculation duration definition method based on AI-based CSI-RS feedback information compression, aiming to solve the problem of how to estimate the terminal-side CSI calculation duration when an AI model is introduced to compress CSI-RS feedback information.
  • an AI model is introduced to compress CSI-RS feedback information
  • the main factors that affect the CSI calculation time are the AI processing capabilities of the terminal hardware and the complexity of the AI model used.
  • the method provided by the embodiment of the present disclosure comprehensively considers the impact of these two aspects and estimates the CSI calculation time after the introduction of the AI model, thereby providing strong support for determining whether the terminal computing power is sufficient to support the use of the AI model.
  • the base station side can estimate the Configure CSI-RS for the CSI calculation duration.
  • the CSI calculation time may include: the time required to use the AI model to compress CSI-RS feedback information.
  • the CSI calculation duration can be either the duration itself or a parameter used to determine the CSI calculation duration; for example, it can be the computing power of the AI model or the computing power of the terminal, etc.
  • Embodiments of the present disclosure provide a method for defining CSI calculation duration based on AI-based CSI-RS feedback information compression.
  • the method includes:
  • the network side equipment (hereinafter taking the base station as an example) has the complexity information of the available AI model, and the terminal side has its own hardware computing capability information.
  • the base station can decide the CSI calculation of AI-based CSI-RS feedback information compression on the terminal side or the base station side. Estimated duration. Specific methods for defining the CSI calculation duration include a definition method based on terminal processing time and a definition method based on model complexity.
  • the terminal side When using the method based on terminal processing time to estimate the CSI calculation time, the terminal side has its own hardware processing capability information, such as the number of floating point operations that can be performed per second.
  • the AI model complexity and the terminal unit time (for example, per second) processing capability By processing the AI model complexity and the terminal unit time (for example, per second) processing capability, the CSI calculation time under the AI model can be obtained.
  • the terminal side may have the CSI calculation time used to process the baseline AI model or the processing time to store preset AI models for multiple complexities. If the terminal side has a CSI calculation time for processing the baseline AI model, the CSI calculation time under the selected AI model can be obtained based on the complexity of the AI model selected by the base station and the complexity of the baseline AI model; if the terminal side has a specific For the processing time of preset AI models of various complexities, the CSI calculation time under the selected AI model can be obtained by matching the complexity of the AI model selected by the base station with the complexity of the preset model.
  • the base station When estimating the CSI calculation time on the terminal side, the base station first delivers the complexity information of one or a group of available AI models and the maximum calculation time requirements to the terminal.
  • the terminal uses the complexity information of the AI model delivered by the base station. Combined with the own hardware computing power, use terminal processing time or model complexity-based methods to estimate the CSI calculation time under the corresponding model.
  • the terminal compares the estimated CSI calculation time with the maximum calculation time requirement. If the requirements are met, it means that the terminal's computing power can support the use of AI models for CSI-RS feedback information compression, and the terminal reports the adopted model and the estimated CSI calculation time.
  • the base station configures CSI-RS according to the situation reported by the terminal; if the maximum calculation time requirement is not met, it means that the terminal computing power does not support using the AI model to compress CSI-RS feedback information, and the terminal does not report AI-related information and uses traditional Algorithm performs CSI feedback.
  • the terminal When estimating the CSI calculation time on the base station side, the terminal first reports the computing power of its own terminal to the base station.
  • the base station uses the terminal processing time or model complexity based on the terminal computing power and the complexity information of the available AI models. method to estimate the CSI calculation time under the corresponding model.
  • the base station compares the estimated CSI calculation time with the maximum calculation time requirement. If the requirements are met, the base station uses the AI model to compress the CSI-RS feedback information.
  • the base station configures the CSI-RS based on the estimated CSI calculation time and will use the The AI model parameters are delivered to the terminal; if the maximum calculation time requirement is not met, traditional methods are used for CSI feedback.
  • a method for determining the CSI calculation duration of AI-based CSI-RS feedback information compression may include:
  • the network side device determines the definition method of the CSI calculation duration.
  • the setting definition method is defined based on the terminal processing time or the complexity of the AI model based on compressed CSI-RS feedback information. For example, it is determined that the base station or the terminal determines the terminal's CSI calculation duration according to the protocol agreement or its own computing load.
  • the CSI calculation duration is estimated on the base station side or the terminal side. Specifically, it can be divided into defining CSI calculation duration based on terminal processing time on the terminal side, defining CSI calculation duration based on terminal processing time on the base station side, defining CSI calculation duration based on model complexity on the terminal side, and defining CSI calculation duration based on model complexity on the base station side.
  • CSI calculation time e.g., CSI calculation duration based on terminal processing time on the terminal side.
  • the base station side configures the CSI-RS according to the estimated CSI calculation duration.
  • a flow chart for defining the CSI calculation duration based on terminal processing time on the terminal side in the CSI calculation duration definition method for AI-based CSI-RS feedback information compression may include:
  • the network side device (for example, the base station) sets the CSI calculation duration definition method to be defined based on the terminal processing time, and determines to estimate the CSI calculation duration on the terminal side.
  • the network side device then delivers the complexity information of the AI model required for CSI-RS feedback information compression to the terminal, and at the same time delivers the maximum calculation time requirement to the terminal.
  • the complexity information of the AI model can be defined by the floating-point operations per second (FLOPs) of the AI model.
  • the floating point operand is the total number of additions or multiplications required for calculation using the AI model; the maximum calculation time requirement sets the requirements for the CSI calculation time, and the CSI calculation time using the AI model must be within the maximum calculation time requirement.
  • the terminal side estimates the CSI calculation time based on its own hardware computing capabilities and the complexity information of the AI model delivered by the base station.
  • the terminal's own hardware computing capability can be defined by the number of floating point operations FLOPS per unit time (for example, per second).
  • the number of floating-point operations per second is the total number of additions or multiplications that the hardware can perform per second.
  • the base station only sends the complexity information of a single AI model I, and then the terminal can obtain the CSI calculation time T corresponding to the model I.
  • the terminal side determines whether the estimated CSI calculation time meets the maximum calculation time requirement. If it is satisfied, the terminal can use the AI model to compress the CSI-RS feedback information, and report the model information of the AI model used and the estimated CSI calculation time to the base station; if it is not satisfied, the terminal can use the traditional mode to perform CSI-RS feedback information.
  • the terminal uses the traditional algorithm to report CSI-RS feedback information and does not report the estimated CSI calculation time or reports no AI model or reports The AI model is not used to process CSI-RS feedback information.
  • the terminal reports the model information and the corresponding CSI calculation duration of the AI model to the base station.
  • the terminal selects the optimal model based on its own characteristics and hardware resource allocation, and compares the optimal model information with the corresponding The CSI calculation duration is reported to the base station; or the terminal reports two or more models with better performance; or the terminal reports the list and priority of all available models.
  • the base station configures the CSI-RS based on the information reported by the terminal.
  • the terminal reports to the base station that the traditional mode is used for CSI feedback, and the base station side obtains the CSI calculation duration according to the traditional algorithm and configures the CSI-RS.
  • the terminal reports the adopted AI model information and the estimated CSI calculation duration to the base station.
  • the base station After receiving the CSI-RS feedback information compressed using the AI model, the base station can use the corresponding decompression model to obtain the restored CSI-RS feedback information.
  • the CSI-RS can be configured according to the estimated CSI calculation duration.
  • a flow chart is provided for defining the CSI calculation duration based on the terminal processing time on the base station side in the CSI calculation duration definition method for AI-based CSI-RS feedback information compression provided by an embodiment of the present disclosure.
  • the specific process is as follows:
  • the base station side sets the CSI calculation duration definition method to be based on the terminal processing time definition, and determines to estimate the CSI calculation duration on the base station side.
  • the terminal reports its own hardware computing capabilities to the base station.
  • the terminal's own hardware computing capability can be defined by the number of floating point operations per second (FLOPS).
  • FLOPS floating point operations per second
  • the number of floating-point operations per second is the total number of additions or multiplications that the hardware can perform per second.
  • the base station estimates the CSI calculation time based on the terminal hardware computing capabilities and the complexity information of the AI model used.
  • the complexity information of the AI model can be defined by the floating point operands FLOPs of the AI model.
  • the floating point operation number is the total number of additions or multiplications required when calculating using the AI model.
  • only one model I is available, and the base station can obtain the corresponding CSI calculation duration T according to the complexity information of model I.
  • the base station determines whether the estimated CSI calculation time meets the maximum calculation time requirement. If it is satisfied, the AI model can be used at the terminal to compress the CSI-RS feedback information, and the base station will use the corresponding decompression model to decompress it after receiving it; if it is not satisfied, the traditional mode will be used for CSI feedback.
  • the terminal side uses a traditional algorithm to report CSI-RS feedback information.
  • the traditional algorithm here for reporting CS feedback information may at least include: reporting partial CSI-RS feedback information.
  • the base station configures the CSI-RS according to the estimated CSI calculation duration and sends the corresponding AI model information to the terminal.
  • the base station selects the optimal model based on its own characteristics and computing resource allocation, and configures the CSI calculation time according to the optimal model.
  • CSI-RS and delivers the optimal model information to the terminal.
  • a flow chart is provided for defining the CSI calculation duration based on model complexity on the terminal side in the CSI calculation duration definition method for AI-based CSI-RS feedback information compression provided by an embodiment of the present disclosure.
  • the specific process is as follows:
  • the base station side sets the CSI calculation duration definition method to be based on model complexity definition, and determines to estimate the CSI calculation duration on the terminal side.
  • the base station side delivers the complexity information of the AI model required for CSI-RS feedback information compression to the terminal, and at the same time delivers the maximum calculation time requirement to the terminal.
  • the terminal side estimates the CSI calculation time required under the AI model used based on the preset model complexity calculation information.
  • the terminal side is preset with calculation duration information for models of different complexity.
  • the terminal matches the complexity information of the AI model sent by the base station with the preset information to obtain the corresponding CSI calculation time.
  • the terminal side is preset to process baseline CSI calculation duration information corresponding to a certain baseline model.
  • the complexity information of the AI model sent by the terminal and the base station is related to the complexity of the baseline AI model. After scaling the CSI calculation time of the baseline AI model, the CSI calculation time of the AI model used can be obtained.
  • the base station only sends the complexity information of a single AI model I, and then the terminal can obtain the CSI calculation time T corresponding to the model I.
  • the terminal side determines whether the estimated CSI calculation time meets the maximum calculation time requirement. If satisfied, the terminal can use the AI model to compress the CSI-RS feedback information, and report the AI model information and estimated CSI calculation time to the base station; if not satisfied, the terminal uses the traditional mode for CSI feedback.
  • the terminal uses the traditional algorithm to report CSI-RS feedback information and does not report the estimated CSI calculation duration.
  • the terminal reports the model information, that is, the corresponding CSI calculation duration, to the base station.
  • the terminal selects the optimal model based on its own characteristics and hardware resource allocation, and compares the optimal model information with the corresponding The CSI calculation duration is reported to the base station.
  • the base station configures the CSI-RS based on the information reported by the terminal.
  • the terminal reports to the base station that the traditional mode is used for CSI feedback, and the base station side obtains the CSI calculation duration according to the traditional algorithm and configures the CSI-RS.
  • the terminal reports the adopted AI model information and the estimated CSI calculation duration to the base station.
  • the base station After receiving the CSI-RS feedback information compressed using the AI model, the base station can use the corresponding decompression model to obtain the restored CSI-RS feedback information.
  • the CSI-RS can be configured according to the estimated CSI calculation duration.
  • the flow chart of defining the CSI calculation duration based on model complexity on the base station side is as follows:
  • the base station side Based on the preset conditions, the base station side sets the CSI calculation duration definition method to be based on model complexity definition, and determines to estimate the CSI calculation duration on the base station side.
  • the terminal reports the preset model complexity calculation information to the base station.
  • the terminal side presets calculation duration information for models of different complexity.
  • the terminal side presets processing of baseline CSI calculation duration information corresponding to a certain baseline model.
  • the base station estimates the CSI calculation time based on the terminal's preset model complexity calculation information and the complexity information of the adopted AI model.
  • the terminal side is preset with calculation duration information for models of different complexity.
  • the terminal matches the complexity information of the AI model sent by the base station with the preset information to obtain the corresponding CSI calculation time.
  • the terminal side stores baseline CSI calculation duration information corresponding to a certain baseline model.
  • the complexity information of the AI model sent by the terminal and the base station is related to the complexity of the baseline model. After scaling the CSI calculation time of the baseline model, the CSI calculation time of the AI model used can be obtained.
  • only one model I is available, and the base station can obtain the corresponding CSI calculation duration T according to the complexity information of model I.
  • the base station determines whether the estimated CSI calculation time meets the maximum calculation time requirement. If it is satisfied, the AI model can be used at the terminal to compress the CSI-RS feedback information, and the base station will use the corresponding decompression model to decompress it after receiving it; if it is not satisfied, the traditional mode will be used for CSI feedback.
  • a traditional algorithm is used to report CSI-RS feedback information.
  • the base station configures the CSI-RS according to the estimated CSI calculation duration and sends the corresponding AI model information to the terminal.
  • the base station selects the optimal model based on its own characteristics and computing resource allocation, and configures the CSI-based model according to the CSI calculation time of the optimal model.
  • RS and sends the optimal model information to the terminal.
  • the signaling interaction diagram for calculating the CSI calculation duration on the terminal side is as follows:
  • the base station sends an AI baseline model determination (AI-based CSI Delay Computation) message to the terminal, requiring the terminal to evaluate the CSI calculation time.
  • AI baseline model determination AI-based CSI Delay Computation
  • the terminal sends the CSI calculation duration of the baseline AI model to the base station to confirm completion (AI-based CSI Delay Computation Ready), confirming that the evaluation of the CSI calculation duration has begun.
  • the base station sends an AI Capability Requirement (AI Capability Requirement) to the terminal.
  • AI Capability Requirement includes information such as model complexity and maximum calculation time stored in the base station.
  • the terminal uses a method based on terminal processing time or model complexity to obtain the CSI calculation time, and determines whether to use AI mode based on the maximum calculation time.
  • the terminal sends the complexity information of the baseline AI model (AI-based CSI Delay Computation Information) to the base station, and reports whether to use the AI model for CSI-RS feedback information compression, the AI model information used, and the corresponding CSI calculation time.
  • AI-based CSI Delay Computation Information AI-based CSI Delay Computation Information
  • the base station configures the CSI-RS based on the complexity information of the baseline AI model (AI-based CSI Delay Computation Information).
  • the base station will deliver the AI model parameters used to the terminal.
  • the signaling interaction diagram for calculating the CSI calculation duration on the base station side is as follows:
  • the base station sends a terminal capability inquiry (Capability Inquiry) to the terminal, requiring the terminal to report its own computing capabilities.
  • Capability Inquiry a terminal capability inquiry
  • the terminal sends terminal capability information (Capability Information) to the base station and reports its own computing capability information.
  • Capability Information can at least indicate the AI capability of the terminal.
  • the base station uses methods based on terminal processing time or model complexity to evaluate the AI capabilities of the terminal and determine whether to adopt the AI mode.
  • the base station configures the CSI-RS based on the evaluation results.
  • the base station will deliver the model information of the AI model used to the terminal.
  • the CSI calculation time of AI-based CSI compression is calculated on the terminal side or the base station side.
  • the base station sends the complexity information of the AI model, and the terminal estimates the CSI calculation time based on its own computing power; when calculating on the base station side, the terminal reports its own computing power information, and the base station combines the stored AI model The complexity information estimates the CSI calculation time.
  • CSI calculation time defined based on terminal processing time is proposed.
  • the base station side stores the complexity information of the AI model, and the terminal side stores its own computing power information. By processing the model complexity and the terminal's own computing power, the CSI calculation time can be obtained.
  • CSI calculation time defined based on model complexity is proposed.
  • the terminal side has the CSI calculation time used to process the baseline AI model, according to the relationship between the complexity of the AI model selected by the base station and the complexity of the baseline AI model, the CSI calculation time under the selected AI model can be obtained; when the terminal side has When considering the processing time of preset AI models of various complexities, the CSI calculation time under the selected AI model can be obtained by matching the complexity of the AI model selected by the base station with the complexity of the preset model.
  • an embodiment of the present disclosure provides an information processing device, wherein the method includes:
  • Determining module 110 used to determine whether the terminal supports channel state information-reference signal CSI-RS feedback information compression of at least one AI model
  • the configuration module 120 is configured to configure CSI-RS according to whether the terminal supports CSI-RS feedback information compression of at least one AI model.
  • the information processing device may be included in the terminal.
  • the determination module 110 and the configuration module 120 may be program modules; after the program modules are executed by a processor, the above functions can be implemented.
  • the determination module 110 and the configuration module 120 may be software-hardware combination modules; the software-hardware combination modules include, but are not limited to: programmable arrays; the programmable arrays include, but are not limited to, on-site Programmable arrays and/or complex programmable arrays.
  • the determination module 110 and the configuration module 120 may be pure hardware modules; the pure hardware modules include but are not limited to: application specific integrated circuits.
  • the determination module 110 is configured to send the complexity information of the AI model to the terminal; receive the first information provided by the terminal according to the complexity information; and determine the Whether the terminal supports CSI-RS feedback information compression of at least one AI model.
  • the first information indicating that the terminal supports CSI-RS feedback information compression of at least one AI model includes:
  • Model identification indicating the AI model supported by the terminal for CSI-RS feedback information compression
  • the CSI calculation duration indicates the duration required for the terminal to use the AI model identified by the model to compress CSI-RS feedback information.
  • the first information indicating that the terminal does not support at least one AI model for CSI-RS information feedback compression indicates that the CSI-RS feedback information adopts a partial reporting method.
  • the complexity information indicates at least one of the following:
  • the total number of floating-point operations corresponding to the AI model where the total number of floating-point operations and the maximum allowed value of the CSI calculation duration are jointly used for the terminal to determine whether to support CSI-RS feedback information compression of the corresponding AI model;
  • the first ratio between the complexity of the corresponding AI model and the complexity of the baseline AI model where the first ratio is used for the terminal to combine the terminal AI capabilities with the complexity of the baseline AI model. , determine whether the CSI-RS feedback information compression of the corresponding AI model is supported.
  • the determination module 110 is configured to receive AI capability information sent by the terminal; determine whether the terminal supports CSI-RS feedback information compression of at least one AI model based on the AI capability information. .
  • the AI capability information indicates at least one of the following:
  • the terminal supports a number of floating point operations per second
  • the second ratio indicates the ratio between the terminal AI capability and the complexity of the baseline AI model.
  • the configuration module 120 is configured to configure the type of CSI-RS feedback information reported by the terminal according to whether the terminal supports CSI-RS feedback information compression of at least one AI model; When the terminal supports CSI-RS feedback information compression of at least one AI model, it determines the CSI calculation duration when using the supported AI model to compress CSI-RS feedback information, and configures the CSI-RS period.
  • an embodiment of the present disclosure provides an information processing device, which includes:
  • the sending module 210 is configured to send second information, where the second information is used for the base station to determine whether the terminal supports compression of CSI-RS feedback information by at least one AI model.
  • the information processing device may be included in the terminal.
  • the sending module 210 may be a program module; after the program module is executed by the processor, the base station can determine whether the terminal supports compression of CSI-RS feedback information by the AI model.
  • the sending module 210 may be a combination of soft and hard modules; the combination of soft and hard modules includes but is not limited to: a programmable array.
  • the programmable array includes, but is not limited to: field programmable array and/or complex programmable array.
  • the sending module 210 may be a pure hardware module; the pure hardware module includes but is not limited to: an application specific integrated circuit.
  • the second information includes:
  • AI capability information indicating the AI capability of the terminal.
  • the AI capability information indicates at least one of the following:
  • the terminal supports a number of floating point operations per second
  • the second ratio indicates the ratio between the terminal AI capability and the complexity of the baseline AI model.
  • the second information includes:
  • the first information is used to indicate whether the terminal supports compression of CSI-RS feedback information by at least one AI model.
  • the first information indicating that the terminal supports CSI-RS feedback information compression of at least one AI model includes:
  • Model identification indicating the AI model supported by the terminal for CSI-RS feedback information compression
  • the CSI calculation duration indicates the duration required for the terminal to use the AI model identified by the model to compress CSI-RS feedback information.
  • CSI-RS feedback information adopts a partial reporting method.
  • the device further includes:
  • the receiving module is configured to receive the complexity information of the AI model
  • the sending module 210 is configured to send the first information to the base station according to the complexity information of the AI model and the AI capability of the terminal.
  • the sending module 210 is further configured to perform at least one of the following:
  • a message indicating that at least one AI model is supported is sent to the base station.
  • An AI model indicates to the base station the first information that supports CSI-RS feedback information compression of an AI model
  • the complexity information indicates at least one of the following:
  • the total number of floating-point operations corresponding to the AI model where the total number of floating-point operations and the maximum allowed value of the CSI calculation duration are jointly used for the terminal to determine whether to support CSI-RS feedback information compression of the corresponding AI model;
  • the first ratio between the complexity of the corresponding AI model and the complexity of the baseline AI model where the first ratio is used for the terminal to combine the terminal AI capabilities with the complexity of the baseline AI model. , determine whether the CSI-RS feedback information compression of the corresponding AI model is supported.
  • An embodiment of the present disclosure provides a communication device, including:
  • Memory used to store instructions executable by the processor
  • the processor is configured to execute the information processing method provided by any of the foregoing technical solutions.
  • the processor may include various types of storage media, which are non-transitory computer storage media that can continue to store information stored thereon after the communication device is powered off.
  • the communication device includes: a terminal or a network element, and the network element may be any one of the aforementioned first to fourth network elements.
  • the processor may be connected to the memory through a bus or the like, and be used to read the executable program stored on the memory, for example, at least one of the methods shown in FIGS. 2 to 14 .
  • Figure 17 is a block diagram of a terminal 800 according to an exemplary embodiment.
  • the terminal 800 may be a mobile phone, a computer, a digital broadcast user device, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, or the like.
  • the terminal 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input/output (I/O) interface 812, a sensor component 814, and communications component 816.
  • Processing component 802 generally controls the overall operations of terminal 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations.
  • the processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above method.
  • processing component 802 may include one or more modules that facilitate interaction between processing component 802 and other components.
  • processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.
  • Memory 804 is configured to store various types of data to support operations at terminal 800. Examples of such data include instructions for any application or method operating on the terminal 800, contact data, phonebook data, messages, pictures, videos, etc.
  • Memory 804 may be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EEPROM), Programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EEPROM erasable programmable read-only memory
  • EPROM Programmable read-only memory
  • PROM programmable read-only memory
  • ROM read-only memory
  • magnetic memory flash memory, magnetic or optical disk.
  • Power supply component 806 provides power to various components of terminal 800.
  • Power component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to terminal 800.
  • Multimedia component 808 includes a screen that provides an output interface between the terminal 800 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user.
  • the touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide action.
  • multimedia component 808 includes a front-facing camera and/or a rear-facing camera.
  • the front camera and/or the rear camera may receive external multimedia data.
  • Each front-facing camera and rear-facing camera may be a fixed optical lens system or have focal length and optical zoom capabilities.
  • Audio component 810 is configured to output and/or input audio signals.
  • audio component 810 includes a microphone (MIC) configured to receive external audio signals when terminal 800 is in operating modes, such as call mode, recording mode, and voice recognition mode. The received audio signal may be further stored in memory 804 or sent via communication component 816 .
  • audio component 810 also includes a speaker for outputting audio signals.
  • the I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which may be a keyboard, a click wheel, a button, etc. These buttons may include, but are not limited to: Home button, Volume buttons, Start button, and Lock button.
  • Sensor component 814 includes one or more sensors that provide various aspects of status assessment for terminal 800 .
  • the sensor component 814 can detect the open/closed state of the device 800, the relative positioning of components, such as the display and keypad of the terminal 800, and the sensor component 814 can also detect the position change of the terminal 800 or a component of the terminal 800. , the presence or absence of user contact with the terminal 800 , the orientation or acceleration/deceleration of the terminal 800 and the temperature change of the terminal 800 .
  • Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact.
  • Sensor assembly 814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications.
  • the sensor component 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • the communication component 816 is configured to facilitate wired or wireless communication between the terminal 800 and other devices.
  • the terminal 800 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof.
  • the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communications component 816 also includes a near field communications (NFC) module to facilitate short-range communications.
  • NFC near field communications
  • the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • the terminal 800 may be configured by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable Gate array (FPGA), controller, microcontroller, microprocessor or other electronic components are implemented for executing the above method.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable Gate array
  • controller microcontroller, microprocessor or other electronic components are implemented for executing the above method.
  • a non-transitory computer-readable storage medium including instructions such as a memory 804 including instructions, which can be executed by the processor 820 of the terminal 800 to complete the above method is also provided.
  • the non-transitory computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
  • an embodiment of the present disclosure shows the structure of an access device.
  • the communication device 900 may be provided as a network side device.
  • the communication device may be the aforementioned terminal or base station.
  • the communication device 900 may be the aforementioned base station.
  • the communication device 900 may include a processing component 922, which further includes one or more processors, and memory resources represented by a memory 932 for storing data that may be processed by the processing component 922.
  • the instructions to execute such as an application.
  • the application program stored in memory 932 may include one or more modules, each corresponding to a set of instructions.
  • the processing component 922 is configured to execute instructions to perform any of the foregoing methods applied to the access device, for example, the methods shown in any one of Figures 4 to 9.
  • Communication device 900 may also include a power supply component 926 configured to perform power management of communication device 900, a wired or wireless network interface 950 configured to connect communication device 900 to a network, and an input-output (I/O) interface 958 .
  • the communication device 900 may operate based on an operating system stored in the memory 932, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Databases & Information Systems (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本公开实施例提供一种信息处理方法及装置、通信设备及存储介质。由基站执行的所述信息处理方法可包括确定终端是否支持至少一种人工智能AI模型的信道状态信息-参考信号CSI-RS反馈信息压缩；根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。

Description

信息处理方法及装置、通信设备及存储介质 技术领域
本公开涉及无线通信技术领域但不限于无线通信技术领域，尤其涉及一种信息处理方法及装置、通信设备及存储介质。
背景技术
信道状态信息可对当前信道环境进行描述，在移动通信网络中，基站发射信道状态信息-参考信号(Channel State Information-Reference Signal，CSI-RS)终端对信道状态信息进行评估并将其量化反馈给基站，通过引入信道状态信息(Channel State Information，CSI)反馈信息，基站侧在发送信道状态信息参考信号时可及时进行调整，从而在终端降低误码率，获得最优接收信号。
发明内容
本公开实施例提供一种信息处理方法及装置、通信设备及存储介质。
本公开实施例第一方面提供一种信息处理方法，由基站执行，所述方法包括：
确定终端是否支持至少一种人工智能(Artificial Intelligence，AI)模型的CSI-RS反馈信息压缩；
根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。
本公开实施例第二方面提供一种信息处理方法，由终端执行，所述方法包括：
发送第二信息，其中，所述第二信息，用于供基站确定所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
本公开实施例第三方面提供一种信息处理装置，其中，所述方法包括：
确定模块，用于确定终端是否支持至少一种人工智能AI模型的信道状态信息-参考信号CSI-RS反馈信息压缩；
配置模块，用于根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。
本公开实施例第四方面提供一种信息处理装置，所述装置包括：
发送模块，被配置为发送第二信息，其中，所述第二信息，用于供基站确定所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
本公开实施例第五方面提供一种通信设备，包括处理器、收发器、存储器及存储在存储器上并能够有所述处理器运行的可执行程序，其中，所述处理器运行所述可执行程序时执行如前述第一方面或第二方面提供的信息处理方法。
本公开实施例第六方面提供一种计算机存储介质，所述计算机存储介质存储有可执行程序；所 述可执行程序被处理器执行后，能够实现前述的第一方面或第二方面提供的信息处理方法。
本公开实施例提供的技术方案，基站会根据终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩，进行针对对应终端的CSI-RS配置，如此，生成的CSI-RS配置是与终端的AI能力相适配的，从而减少了不适配的CSI-RS配置导致的UE测量异常现象。应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，并不能限制本公开实施例。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本发明实施例，并与说明书一起用于解释本发明实施例的原理。
图1是根据一示例性实施例示出的一种无线通信系统的结构示意图；
图2是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图3是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图4是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图5是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图6是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图7A是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图7B是根据一示例性实施例示出的一种信息处理方法的流程示意图
图8是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图9是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图10是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图11是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图12是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图13是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图14是根据一示例性实施例示出的一种信息处理方法的流程示意图；
图15是根据一示例性实施例示出的一种信息处理装置的结构示意图；
图16是根据一示例性实施例示出的一种信息处理装置的结构示意图；
图17是根据一示例性实施例示出的一种终端的结构示意图；
图18是根据一示例性实施例示出的一种通信设备的结构示意图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本发明实施例相一致的所有实施方式。相反，它们仅是本发明实施例的一些方面相一致 的装置和方法的例子。
在本公开实施例使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本公开实施例。在本公开所使用的单数形式的“一种”、“”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本文中使用的术语“和/或”是指并包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本公开实施例可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本公开实施例范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所使用的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
请参考图1，其示出了本公开实施例提供的一种无线通信系统的结构示意图。如图1所示，无线通信系统是基于蜂窝移动通信技术的通信系统，该无线通信系统可以包括：若干个终端11以及若干个接入设备12。
其中，终端11可以是指向用户提供语音和/或数据连通性的设备。终端11可以经无线接入网(Radio Access Network，RAN)与一个或多个核心网进行通信，终端11可以是物联网终端，如传感器设备、移动电话(或称为“蜂窝”电话)和具有物联网终端的计算机，例如，可以是固定式、便携式、袖珍式、手持式、计算机内置的或者车载的装置。例如，站(Station，STA)、订户单元(subscriber unit)、订户站(subscriber station)、移动站(mobile station)、移动台(mobile)、远程站(remote station)、接入点、远程终端(remote terminal)、接入终端(access terminal)、用户装置(user terminal)、用户代理(user agent)、用户设备(user device)、或用户终端(user equipment，终端)。或者，终端11也可以是无人飞行器的设备。或者，终端11也可以是车载设备，比如，可以是具有无线通信功能的行车电脑，或者是外接行车电脑的无线通信设备。或者，终端11也可以是路边设备，比如，可以是具有无线通信功能的路灯、信号灯或者其它路边设备等。
接入设备12可以是无线通信系统中的网络侧设备。其中，该无线通信系统可以是第四代移动通信技术(the 4th generation mobile communication，4G)系统，又称长期演进(Long Term Evolution，LTE)系统；或者，该无线通信系统也可以是5G系统，又称新空口(new radio，NR)系统或5G NR系统。或者，该无线通信系统也可以是5G系统的再下一代系统。其中，5G系统中的接入网可以称为NG-RAN(New Generation-Radio Access Network，新一代无线接入网)。或者，MTC系统。
其中，接入设备12可以是4G系统中采用的演进型接入设备(eNB)。或者，接入设备12也可以是5G系统中采用集中分布式架构的接入设备(gNB)。当接入设备12采用集中分布式架构时，通常包括集中单元(central unit，CU)和至少两个分布单元(distributed unit，DU)。集中单元中设置有分组数据汇聚协议(Packet Data Convergence Protocol，PDCP)层、无线链路层控制协议(Radio Link Control，RLC)层、媒体访问控制(Media Access Control，MAC)层的协议栈；分布单元中设置有物理(Physical，PHY)层协议栈，本公开实施例对接入设备12的具体实现方式不加以限定。
接入设备12和终端11之间可以通过无线空口建立无线连接。在不同的实施方式中，该无线空口是基于第四代移动通信网络技术(4G)标准的无线空口；或者，该无线空口是基于第五代移动通信网络技术(5G)标准的无线空口，比如该无线空口是新空口；或者，该无线空口也可以是基于5G的更下一代移动通信网络技术标准的无线空口。
如图2所示，本公开实施例提供一种信息处理方法，由网络侧设备执行，所述方法包括：
S110：确定终端是否支持基于至少一种人工智能AI模型的信道状态信息-参考信号CSI-RS反馈信息处理；
S120：根据所述终端是否支持基于至少一种AI模型的CSI-RS反馈信息处理，为所述终端配置CSI-RS。
在本公开实施例中，网络侧设备可以为基站，该基站可为演进型基站(eNB)和/或下一代基站(gNB)，或是任何一代通信系统的基站。当然，网络侧设备不仅限于基站，还可以为网络中的任何设备，在此不做限定。
基站发射CSI-RS，终端会接收CSI-RS，终端会根据自身对CSI-RS的接收状况，生成CSI-RS反馈信息。即，所述CSI-RS反馈信息可用于网络侧设备确定终端的CSI-RS的接收状况，例如，是否接收到对应传输资源块上的CSI-RS和/或终端对CSI-RS的接收功率等。
在一些实施例中，所述CSI-RS反馈信息的数据量可能比较大，在一些情况下，终端可能会仅仅将部分CSI-RS反馈信息发送给网络侧设备。在一些情况下，终端可以采用AI模型对CSI-RS反馈信息进行处理，终端上报的是经过处理的CSI-RS反馈信息，该经过处理的CSI-RS反馈信息被网络侧设备(例如基站)进行相应处理后就可以获得终端的完整CSI-RS反馈信息。
在一种实现方式中，对采用AI模型对CSI-RS反馈信息进行处理可以为：终端采用AI模型对CSI-RS反馈信息进行压缩。基于此，终端上报经过压缩的CSI-RS反馈信息，该经过压缩的CSI-RS反馈信息被基站进行相应的解压缩后就可以获得终端的完整CSI-RS反馈信息。
但是使用AI模型处理CSI-RS反馈信息，通常至少要求终端具有AI运算能力，例如，终端包含AI芯片，则终端具有AI运算能力。不同AI芯片的AI计算能力不同，不同的AI模型具有不同的计算能力要求，因此，有的终端可能支持通过所有AI模型对CSI-RS反馈信息进行处理，有的终端可能仅仅支持通过部分AI模型对CSI-RS信息进行处理，而有的终端完全不支持采用AI模型对CSI-RS反馈信息进行处理。
在以下的举例说明中，都是以采用AI模型对CSI-RS反馈信息进行压缩来举例说明的，但是本领域内技术人员可以理解，采用AI模型对CSI-RS反馈信息进行处理还可以包括其他操作，在此不再赘述。
在本公开实施例中，基站等网络侧设备会先确定出终端是否支持基于至少一种AI模型的CSI-RS反馈信息处理。根据该确定结果，生成针对该UE的CSI-RS配置，如此，可以确保生成的CSI-RS配置与UE是否支持AI模型对CSI-RS反馈信息的压缩相适配，减少不适配的CSI-RS配置导致的UE测量异常现象。
生成的CSI-RS配置可以通过RRC消息或者MAC CE发送给终端。
本公开实施例提供一种信息处理方法，由网络侧设备执行，所述方法包括：
响应于确定终端支持基于AI模型的CSI-RS反馈信息处理，确定所述终端使用的CSI-RS，其中所述CSI-RS与所述终端支持的AI模型相对应。
如图3所示，本公开实施例提供一种信息处理方法，由网络侧设备执行，所述方法包括：
S310：向所述终端发送AI模型的复杂度信息；
S320：接收所述终端根据复杂度信息提供的第一信息；
S330：根据所述第一信息，确定所述终端是否支持基于至少一种AI模型的CSI-RS反馈信息处理；
S340：根据所述终端是否基于支持至少一种AI模型的CSI-RS反馈信息处理，配置CSI-RS。
在一种实现方式中，对采用AI模型对CSI-RS反馈信息进行处理可以为：终端采用AI模型对CSI-RS反馈信息进行压缩。基于此，终端上报经过压缩的CSI-RS反馈信息，该经过压缩的CSI-RS反馈信息被网络侧设备(例如基站)进行相应的解压缩后就可以获得终端的完整CSI-RS反馈信息。在以下的举例说明中，都是以采用AI模型对CSI-RS反馈信息进行压缩来举例说明的，但是本领域内技术人员可以理解，采用AI模型对CSI-RS反馈信息进行处理还可以包括其他操作，在此不再赘述。
本公开实施例中，网络侧设备会向终端发送AI模型的复杂度信息，该复杂度信息发送给终端之后，终端就可以结合自身的AI能力，确定出是否支持使用对应AI模型对CSI-RS反馈信息压缩。当然，网络侧设备与终端还可以通过通信协议确定的候选AI模型来确定，在此不再赘述。
该复杂度信息可由基站通过RRC消息或者MAC CE发送给终端的。
所述第一信息可为各种用于网络侧设备(例如基站)确定终端是否支持AI模型压缩CSI-RS反馈信息，或者，所述第一信息可为终端确定那种AI模型压缩CSI-RS反馈信息。
在这种方式中，实际上是由终端自行确定是否支持AI模型压缩CSI-RS反馈信息并告知基站。
在一些实施例中，所述第一信息可以显性指示所述终端是否支持AI模型压缩CSI-RS反馈信息或者支持哪种AI模型压缩CSI-RS反馈信息。
在另一些实施例中，所述第一信息还可以隐性指示所述终端是否支持AI模型压缩CSI-RS反馈信息和/或支持那种AI模型压缩CSI-RS反馈信息。
在一些实施例中，表明所述终端支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括以下的至少一个参数：
模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
当然，CSI计算时长既可以是时长本身，也可以是用于确定CSI计算时长的参数；例如，可以是AI模型的计算能力，还可以是终端的计算能力，等。在本公开实施例中，所述模型标识指代的所 述终端支持的AI模型。
在另一些实施例中，所述模型标识还可以指示指代终端不支持的AI模型。
CSI计算时长可为：终端采用对应AI模型进行CSI-RS反馈信息压缩所需的时长，若该CSI计算时长大于基站允许的最大时长，即便终端支持该AI模型，但是由于超时网络侧设备也可能不会配置终端使用该AI模型压缩CSI-RS反馈信息。
在本公开实施例中，所述第一信息相当于隐性指示所述终端是否支持AI模型压缩CSI-RS反馈信息。若第一信息未携带任何一个模型标识或终端未发送第一信息，则相当于说明终端不支持任何AI模型压缩CSI-RS反馈信息。若第一信息至少携带一个模型标识，则说明终端至少支持一种AI模型压缩CSI-RS反馈信息。
在一些实施例中，表明所述终端不支持至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
若第一信息指示终端将采用CSI-RS反馈信息部分上报的方式向基站发送CSI-RS反馈信息，则相当于隐性说明终端不支持AI模型压缩CSI-RS反馈信息或者终端不期望采用AI模型压缩CSI-RS反馈信息。
在另一些实施例中，所述第一信息可包括：一个专门指示终端是否支持AI模型压缩CSI-RS反馈信息的比特。
在另一些实施例中，所述第一信息可包括一个比特位图，该比特位图可以包括N个比特位，N≥1，每一个比特位对应一种AI模型；其中，比特位图中一个比特用于指示终端是否支持对应AI模型压缩CSI-RS反馈信息。
在还有一些实施例中，所述第一信息可包括一个比特位图，该比特位图的比特位可以对应于2 N个取值，每一个取值对应一种AI模型。
在一些实施例中，所述复杂度信息指示以下至少之一：
对应AI模型的总浮点运算次数，其中，所述总浮点运算次数与CSI计算时长的最大允许值，共同用于供所述终端确定是否支持对应AI模型的CSI-RS反馈信息压缩；
对应AI模型的复杂度相对于基线AI模型的复杂度之间的第一比值，其中，所述第一比值，用于供所述终端结合所述终端AI能力与所述基线AI模型的复杂度，确定是否支持对应AI模型的CSI-RS反馈信息压缩。
此处的基线AI模型可为多个进行CSI-RS反馈信息压缩的AI模型中的任意一个，例如，可为基站或者终端指定的一个AI模型。该基线AI模型的复杂度可为基站和UE都知晓的AI模型。
在一些实施例中，该基线AI模型可为多个支持CSI-RS反馈信息压缩的多个AI模型中复杂度最低或者复杂度最高的AI模型。如此第一比值的取值范围就可以限制特定范围内，从而减少指示第一比值的比特开销。
在一些实施例中，复杂度信息可以直接指示对应AI模型的总浮点运算次数，如此终端接收到之后，可以根据基站配置的CSI计算时长的最大允许值或者协议约定的最大允许值，确定自身是否支 持使用对应AI模型压缩CSI-RS反馈信息。
在另一些实施例中，为了减少信令开销，复杂度信息由对应AI模型和基线AI模型的复杂度之间的比值来指代。终端是预先指示基线AI模型的总浮点运算次数的，因此，终端在接收到第一比值之后，同样可以确定出自身是否支持对应的AI模型压缩CSI-RS反馈信息。
如图4所示，本公开实施例提供一种信息处理方法，由等网络侧设备执行，所述方法包括：
S410：接收所述终端发送的AI能力信息；
S420：根据所述AI能力信息，确定所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩。
S430：根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。
在本公开实施例中，基站等网络侧设备接收的是终端发送的AI能力信息，基站等网络侧设备结合终端的AI能力信息和各个AI模型的复杂度信息，确定终端是否支持AI模型压缩CSI-RS反馈信息，或者，终端支持哪种AI模型压缩CSI-RS反馈信息。
在一些实施例中，所述AI能力信息指示以下至少之一：
所述终端是否具有AI能力；
所述终端的性能参数。
其中所述终端的性能参数可以包括以下的至少一种参数：
所述终端的最大AI能力；
所述终端支持单位时间内(例如每秒)的浮点运算次数；
所述终端支持的处理能力；
所述终端在CSI计算时长的最大允许值内支持的浮点运算次数；
第二比值，指示所述终端AI能力与基线AI模型之间复杂度之间的比值。
当然，性能参数还可以通过其他参数来表征，在此不再赘述。
在一些实施例中，所述AI能力信息至少可指示终端是否具有AI能力，若终端不具有AI能力，则终端的AI能力信息将不会携带终端支持的单位时间内(例如每秒)的浮点运算次数，或者自身在CSI计算时长的最大允许值内可以执行的总浮点运算次数。
在一些实施例中，为了减少信令比特开销，会由第二比值来间接指示终端的AI能力。第二比值越大，则说明终端的AI能力越强。若第二比值为0，则可认为终端不具有AI能力。
若AI能力信息指示的终端在CSI计算时长的最大允许值内支持的浮点运算次数为0，则说明终端不具有AI能力。若终端上报的CSI计算时长的最大允许值内支持的浮点运算次数，低于任何一个AI模型的总浮点运算次数，也说明终端不支持任何一种AI模型对CSI-RS反馈信息压缩。若终端上报的CSI计算时长的最大允许值内支持的浮点运算次数大于或等于至少一个AI模型的总浮点运算次数，则说明终端至少一种AI模型对CSI-RS反馈信息的压缩。
基站收到终端上报的AI能力信息，会自行确定对应终端是否在支持AI模型压缩CSI-RS反馈信息。
在一些实施例中，所述根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS，包括：
根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置所述终端上报的CSI-RS反馈信息的类型；
在所述终端支持基于至少一种AI模型的CSI-RS反馈信息压缩时，确定采用终端支持的所述AI模型对CSI-RS反馈信息压缩时的CSI计算时长，配置所述CSI-RS的周期。
此处的上报CSI-RS反馈信息的类型可至少包括以下两种：
上报通过AI模型压缩CSI-RS反馈信息的压缩信息；
上报终端生成的CSI-RS反馈信息。
当然，上报CSI-RS反馈信息的类型还可以为其他类型，在此不再一一列举赘述。
若终端支持AI模型压缩CSI-RS反馈信息，且CSI计算时长小于最大允许值，则网络侧设备确定出终端采用对应AI模型压缩CSI-RS反馈信息所需的CSI计算时长配置CSI-RS的周期。
示例性地，所述CSI计算时长与所述CSI-RS的周期正相关。
在一些实施例中，网络侧设备还可以根据所述CSI计算时长确定CSI-RS的测量时长。该测量时长可为一个CSI-RS的周期测量CSI-RS的持续时长。
如图5所示，本公开实施例提供一种信息处理方法，由终端执行，所述方法包括：
S510：发送第二信息，其中，所述第二信息，用于供网络侧设备确定所述终端是否支持基于至少一种AI模型对CSI-RS反馈信息的压缩。
本公开实施例中提供一种信息处理方法，由终端向基站发送第二信息，该第二信息可以供网络侧设备确定是否可以采用至少一种AI模型对CSI-RS反馈信息进行压缩。如此网络侧设备就能够知晓终端是否支持采用至少一种AI模型进行CSI-RS反馈信息进行压缩，而后方便网络侧设备为终端是否支持使用AI模型或者使用哪种AI模型对CSI-RS反馈信息压缩进行CSI-RS配置，从而使得基站生成的CSI-RS配置更符合终端的能力。
如图6所示，本公开实施例提供一种信息处理方法，由终端执行，所述方法包括：
S610：向网络侧设备发送用于指示所述终端AI能力的AI能力信息。
在一些实施例中，该AI能力信息可为前述第二信息的一种，也可以是独立于所述第二信息的。
通过终端的AI能力的AI能力信息的上报，如此网络侧设备自行判断终端是否支持使用AI模型压缩CSI-RS反馈信息，进而针对终端进行CSI-RS配置的生成。
示例性地，所述AI能力信息指示以下至少之一：
所述终端是否具有AI能力；
所述终端的性能参数。
其中所述终端的性能参数可以包括以下的至少一种参数：
所述终端的最大AI能力；
所述终端支持单位时间内(例如每秒)的浮点运算次数；
所述终端支持的处理能力；
所述终端在CSI计算时长的最大允许值内支持的浮点运算次数；
第二比值，指示所述终端AI能力与基线AI模型之间复杂度之间的比值。
当然，性能参数还可以通过其他参数来表征，在此不再赘述。
在一些实施例中，所述AI能力信息至少可指示终端是否具有AI能力，若终端不具有AI能力，则终端的AI能力信息将不会携带终端支持的单位时间内(例如每秒)的浮点运算次数，或者自身在CSI计算时长的最大允许值内可以执行的总浮点运算次数。
若AI能力信息指示的终端在CSI计算时长的最大允许值内支持的浮点运算次数为0，则说明终端不具有AI能力。若终端上报的CSI计算时长的最大允许值内支持的浮点运算次数，低于任何一个AI模型的总浮点运算次数，也说明终端不支持任何一种AI模型对CSI-RS反馈信息压缩。若终端上报的CSI计算时长的最大允许值内支持的浮点运算次数大于或等于至少一个AI模型的总浮点运算次数，则说明终端至少一种AI模型对CSI-RS反馈信息的压缩。
在一些实施例中，为了减少信令比特开销，会由第二比值来间接指示终端的AI能力。第二比值越大，则说明终端的AI能力越强。若第二比值为0，则可认为终端不具有AI能力。
基站收到终端上报的AI能力信息，会自行确定对应终端是否在支持AI模型压缩CSI-RS反馈信息。
如图7A所示，本公开实施例提供一种信息处理方法，由终端执行，所述方法包括：
S710A：向网络侧设备发送第一信息，其中，第一信息，用于表明所述终端是否支持基于至少一种AI模型对CSI-RS反馈信息的压缩。
该第一信息可为前述第二信息的一种。
所述第一信息可为：终端根据自身AI能力和各个AI模型的复杂度确定出自身是否支持基于任何一种AI模型对CSI-RS反馈信息的压缩之后生成的。
所述第一信息可为各种用于网络侧设备(例如基站)确定终端是否支持AI模型压缩CSI-RS反馈信息，或者，终端确定那种AI模型压缩CSI-RS反馈信息。
在一些实施例中，所述第一信息可以显性或隐性指示所述终端是否支持AI模型压缩CSI-RS反馈信息或者支持哪种AI模型压缩CSI-RS反馈信息。
在一些实施例中，表明所述终端支持基于至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括：
模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
当然，CSI计算时长既可以是时长本身，也可以是用于确定CSI计算时长的参数；例如，可以是AI模型的计算能力，还可以是终端的计算能力等。
在本公开实施例中，所述模型标识指代的所述终端支持的AI模型。
在另一些实施例中，所述模型标识还可以指示指代终端不支持的AI模型。
CSI计算时长为：终端采用对应AI模型进行CSI-RS反馈信息压缩所需的时长，若该CSI计算时长大于基站允许的最大时长，即便终端支持该AI模型，但是由于超时基站也可能不会配置终端使用该AI模型压缩CSI-RS反馈信息。
在本公开实施例中，所述第一信息相当于隐性指示所述终端是否支持AI模型压缩CSI-RS反馈信息。若第一信息未携带任何一个模型标识或终端未发送第一信息，则相当于说明终端不支持任何AI模型压缩CSI-RS反馈信息。若第一信息至少携带一个模型标识，则说明终端至少支持一种AI模型压缩CSI-RS反馈信息。
在一些实施例中，表明所述终端不支持所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
若第一信息指示终端将采用CSI-RS反馈信息部分上报的方式向基站发送CSI-RS反馈信息，则相当于隐性说明终端不支持AI模型压缩CSI-RS反馈信息或者终端不期望采用AI模型压缩CSI-RS反馈信息。
在另一些实施例中，所述第一信息可包括：一个专门指示终端是否支持AI模型压缩CSI-RS反馈信息的比特。在另一些实施例中，所述第一信息可包括一个比特位图，该比特位图可以包括N个比特位，N≥1，每一个比特位对应一种AI模型；其中比特位图中一个比特用于指示终端是否支持对应AI模型压缩CSI-RS反馈信息。或所述第一信息可包括一个比特位图，该比特位图的比特位可以对应于2 N个取值，每一个取值对应一种AI模型。
在一些实施例中，如图7B所示，所述方法还包括：
S710B：接收AI模型的复杂度信息；
S720B：根据所述AI模型的复杂度信息以及所述终端的AI能力，向基站发送第一信息。
在本公开实施例中，所述第一信息是根据从网络侧接收的AI模型的复杂度信息以及自身的AI能力确定的。
在另一些实施例中，AI模型的复杂度信息还可以是终端根据协议约定或者其他方式确定的。
总之，本公开实施例中，第一信息并非直接是终端AI能力的AI能力信息，而是根据AI模型的复杂度信息以及终端的AI能力生成的并向基站发送的。
在一些实施例中，所述根据所述AI模型的复杂度信息以及所述终端的AI能力，向基站发送所述第一信息，包括以下至少之一：
根据所述AI模型的复杂度信息以及所述终端的AI能力，确定所述终端支持一种AI模型对所述CSI-RS反馈信息压缩时，向所述基站发送表明支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息；
根据所述AI模型的复杂度信息以及所述终端的AI能力确定所述终端支持多种AI模型对所述CSI-RS反馈信息压缩时，根据所述终端的资源调度情况选择所述终端支持的一种AI模型向所述基站表明支持一种AI模型的CSI-RS反馈信息压缩的所述第一信息；
根据所述AI模型的复杂度信息以及所述终端的AI能力确定所述终端不支持AI模型对所述 CSI-RS反馈信息压缩时，向所述基站发送表明所述终端不支持至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息。
在一些情况下，终端根据AI模型的复杂度信息以及终端的AI能力，在不关注终端具体支持哪种AI模型压缩CSI-RS反馈信息的情况下，就向基站发送表明基站支持至少一种AI模型的CSI-RS反馈信息压缩的第一信息。在这种情况下，第一信息可能不包含终端支持的AI模型的模型标识，但是基站可以认为终端至少支持基线AI模型，因此，针对终端进行CSI-RS的配置时，可以至少根据基线AI模型的复杂度或者所需的CSI计算时长，进行CIS-RS的配置。
若终端不仅会根据各个AI模型的复杂度信息以及AI能力，会确定出终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩，而且会确定出支持哪一种AI模型对CSI-RS反馈信息的压缩，并将终端支持对CSI-RS反馈信息压缩的AI模型的模型标识或者AI模型的指代信息携带在第一信息发送给终端，如此基站不仅知晓终端支持至少一种AI模型对CSI-RS反馈信息的压缩，而且具体支持哪一种AI模型对CSI-RS反馈信息的压缩。
若终端不具有AI能力或者不支持基站指示的任意一种AI模型压缩CSI-RS反馈信息，则终端向基站发送表明不支持至少一种AI模型对CSI-RS反馈信息的第一信息。示例性地，表明所述终端支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括：
模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
又示例性地，表明所述终端不支持基于至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
在一些实施例中，所述复杂度信息指示以下至少之一：
对应AI模型的总浮点运算次数，其中，所述总浮点运算次数与CSI计算时长的最大允许值，共同用于供所述终端确定是否支持对应AI模型的CSI-RS反馈信息压缩；
对应AI模型的复杂度相对于基线AI模型的复杂度之间的第一比值，其中，所述第一比值，用于供所述终端结合所述终端AI能力与所述基线AI模型的复杂度，确定是否支持对应AI模型的CSI-RS反馈信息压缩。
示例性地，每一个AI模型的总浮点运算次数与终端可支持的单位时间(例如每秒)执行的浮点运算次数，则可以得到终端使用对应AI模型进行CSI-RS反馈信息压缩所需要的实际计算时长，若该实际计算时长小于所述CSI计算时长小于或大于最大允许值，从而可以确定终端是否支持对应的AI模型的CSI-RS反馈信息压缩。
若复杂度信息为：第一比值；则终端在接收到第一比值之后，首先根据自身在CSI计算时长的最大允许值可执行的浮点运算次数，与基线AI模型所需的浮点运算次数之间的第三比值，然后将第三比值与第一比值之间的比较，若第三比值大于或等于第一比值，则可认为终端支持对应的AI模型压缩CSI-RS反馈信息压缩，若第三比值小于第一比值，则可认为终端不支持对应AI模型压缩CSI-RS 反馈信息压缩。
在本公开实施例中，若终端不支持AI芯片或者不具有AI能力，则终端在接收到基站发送的AI模型的复杂度信息之后，就可以直接上报表明不支持AI模型压缩CSI-RS反馈信息的第一信息。
本公开实施例提供基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法，旨在解决当引入AI模型对CSI-RS反馈信息进行压缩时，如何对终端侧CSI计算时长进行估计的问题。在引入AI模型进行CSI-RS反馈信息压缩时，影响CSI计算时长的主要因素有终端硬件的AI处理能力及所采用AI模型的复杂度。
本公开实施例提供的方法综合考虑该两方面的影响，对引入AI模型后的CSI计算时长进行预估，从而为判断终端算力是否足以支持使用AI模型提供有力支撑，同时基站侧可根据估计的CSI计算时长配置CSI-RS。
CSI计算时长可包括：使用AI模型压缩CSI-RS反馈信息所需的时长。
当然，CSI计算时长既可以是时长本身，也可以是用于确定CSI计算时长的参数；例如，可以是AI模型的计算能力，还可以是终端的计算能力，等。
本公开实施例提供基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法，所述方法包括：
网络侧设备(以下以基站为例)有可用AI模型的复杂度信息，终端侧有自身硬件计算能力信息，基站可决定在终端侧或基站侧对基于AI的CSI-RS反馈信息压缩的CSI计算时长进行估计。具体的CSI计算时长定义方法有基于终端处理时间的定义方法与基于模型复杂度的定义方法。
当采用基于终端处理时间的方法估计CSI计算时长时，终端侧有自身硬件处理能力信息，如每秒可进行的浮点运算次数等。将AI模型复杂度与终端单位时间(例如每秒)处理能力进行处理，则可得到该AI模型下的CSI计算时长。
当采用基于模型复杂度的方法估计CSI计算时长时，终端侧可具有处理基线AI模型所用CSI计算时长或存储针对多种复杂度的预设AI模型的处理时长。若终端侧有处理基线AI模型所用CSI计算时长，则根据基站所选用AI模型的复杂度与基线AI模型的复杂度关系，可得到在所选AI模型下的CSI计算时长；若终端侧有针对多种复杂度的预设AI模型的处理时长，将基站所选用AI模型复杂度与预设模型的复杂度进行匹配，即可得到在所选AI模型下的CSI计算时长。
当在终端侧对CSI计算时长进行估计时，基站首先将可用的一个或一组AI模型的复杂度信息及最大计算时长要求下发给终端，终端利用基站下发的AI模型的复杂度信息，结合自身硬件算力情况，采用基于终端处理时间或基于模型复杂度的方法，估计出在对应模型下的CSI计算时长。
终端再将估计的CSI计算时长与最大计算时长要求进行对比，若满足要求，则说明终端算力可支持使用AI模型进行CSI-RS反馈信息压缩，终端将采用的模型及估计的CSI计算时长上报基站，基站根据终端上报情况对CSI-RS进行配置；若不满足最大计算时长要求，则说明终端算力不支持使用AI模型进行CSI-RS反馈信息压缩，终端不上报与AI相关信息，采用传统算法进行CSI反馈。
当在基站侧对CSI计算时长进行估计时，终端首先将自身终端的计算能力情况上报给基站，基站根据终端算力，结合可用AI模型的复杂度信息，采用基于终端处理时间或基于模型复杂度的方法， 估计出对应模型下的CSI计算时长。基站再将估计的CSI计算时长与最大计算时长要求进行对比，若满足要求，则使用AI模型进行CSI-RS反馈信息压缩，基站根据估计的CSI计算时长对CSI-RS进行配置，并将选用的AI模型参数下发终端；若不满足最大计算时长要求，则采用传统方法进行CSI反馈。需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。
参考图8所示，为本公开实施例提供的一种基于AI的CSI-RS反馈信息压缩的CSI计算时长确定方法，可包括：
网络侧设备(以下以基站为例)确定CSI计算时长的定义方式，设置定义方式为基于终端处理时间定义或基于压缩CSI-RS反馈信息的AI模型的复杂度定义的。例如，根据协议约定或者自身计算负载等情况，确定由基站或者终端来确定终端的CSI计算时长。
根据设置的CSI计算时长定义方式，在基站侧或终端侧对CSI计算时长进行估计。具体地，可分为在终端侧基于终端处理时间定义CSI计算时长、在基站侧基于终端处理时间定义CSI计算时长、在终端侧基于模型复杂度定义CSI计算时长及在基站侧基于模型复杂度定义CSI计算时长。
基站侧根据估计的CSI计算时长，对CSI-RS进行配置。
需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。需要说明的是，该方法可以结合本公开的其他实施例一起被实施，也可以独立被实施，在此不再赘述。
参考图9，为本公开实施例提供的基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法中在终端侧基于终端处理时间定义CSI计算时长的流程图，可包括：
网络侧设备(例如基站)将CSI计算时长定义方式设置为基于终端处理时间定义，并确定在终端侧对CSI计算时长进行估计。
等网络侧设备将CSI-RS反馈信息压缩所需AI模型的复杂度信息下发终端，同时将最大计算时长要求下发终端。在一种实施例中，AI模型的复杂度信息可用AI模型的浮点运算数(floating-point operations per second，FLOPs)进行定义。浮点运算数即使用该AI模型计算时所需进行的加法或乘法总次数；最大计算时长要求对CSI计算时长做出要求，采用AI模型的CSI计算时长需在该最大计算时长规定内。
终端侧根据自身硬件计算能力及基站下发的AI模型的复杂度信息，对CSI计算时长进行估计。
在一种实施例中，终端自身硬件计算能力可用单位时间(例如每秒)浮点运算次数FLOPS进行定义。每秒浮点运算次数即硬件每秒可进行的加法或乘法总次数。
在一种实施例中，CSI计算时长可用AI模型复杂度与终端硬件算力的比值进行定义，即CSI计算时长＝FLOPs/FLOPS。用模型所需加法或乘法的总运算次数除以硬件每秒可进行加法或乘法运算次数，即可得到该AI模型下所需的计算时长。
在一种实施例中，基站侧仅下发单个AI模型I的复杂度信息，则终端可得到该模型I对应的CSI计算时长T。
在一种实施例中，基站侧下发一组AI模型Ii(i＝1,2,3…)的复杂度信息，则终端可得到该组模型分别对应的CSI计算时长Ti(i＝1,2,3…)。
终端侧判断估计出的CSI计算时长是否满足最大计算时长要求。若满足，则终端可采用AI模型对CSI-RS反馈信息进行压缩，并向基站上报所用AI模型的模型信息及估计的CSI计算时长；若不满足，则终端则采用传统模式进行CSI-RS反馈信息。
在一种实施例中，基站下发的所有模型中，没有模型满足最大计算时长要求，则终端采用传统算法上报CSI-RS反馈信息，且不上报估计的CSI计算时长或上报没有AI模型或上报不使用AI模型处理CSI-RS反馈信息。
在一种实施例中，基站下发的所有AI模型中，仅有一个AI模型满足最大计算时长要求，则终端将该AI模型的模型信息及对应的CSI计算时长上报给基站。
在一种实施例中，基站下发的所有模型中，有多个模型满足最大计算时长要求，则终端根据自身特性及硬件资源分配情况选出最优模型，并将该最优模型信息与对应的CSI计算时长上报基站；或终端上报性能较佳的两个或多个模型；或终端上报所有可用模型列表和优先级。
基站根据终端上报信息，对CSI-RS进行配置。
在一种实施例中，终端向基站上报采用传统模式进行CSI反馈，则基站侧根据传统算法得到CSI计算时长，并对CSI-RS进行配置。
在一种实施例中，终端向基站上报采用的AI模型信息及估计的CSI计算时长，则基站侧接收到使用AI模型压缩的CSI-RS反馈信息后，可采用对应的解压缩模型得到恢复后的CSI-RS反馈信息。在配置CSI-RS时，可根据估计的CSI计算时长对CSI-RS进行配置。
需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。需要说明的是，该方法可以结合本公开的其他实施例一起被实施，也可以独立被实施，在此不再赘述。
参考图10，为本公开实施例提供的基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法中在基站侧基于终端处理时间定义CSI计算时长的流程图，具体流程如下：
基站侧根据预设条件，将CSI计算时长定义方式设置为基于终端处理时间定义，并确定在基站侧对CSI计算时长进行估计。
终端向基站上报自身硬件计算能力。
在一种实施例中，终端自身硬件计算能力可用每秒浮点运算次数FLOPS进行定义。每秒浮点运算次数即硬件每秒可进行的加法或乘法总次数。
基站根据终端硬件计算能力及采用的AI模型的复杂度信息，对CSI计算时长进行估计。
在一种实施例中，AI模型的复杂度信息可用AI模型的浮点运算数FLOPs进行定义。浮点运算数即使用该AI模型计算时所需进行的加法或乘法总次数，CSI计算时长可用AI模型复杂度与终端硬件算力的比值进行定义，即CSI计算时长＝FLOPs/FLOPS。用模型所需加法或乘法的总运算次数除以硬件每秒可进行加法或乘法运算次数，即可得到该AI模型下所需的计算时长。
在一种实施例中，仅有一个模型I可用，则基站根据该模型I的复杂度信息可得到对应的CSI计算时长T。
在一种实施例中，有一组模型Ii(i＝1,2,3…)可用，则基站根据该组模型的复杂度信息可得到该组模型对应的CSI计算时长Ti(i＝1,2,3…)。
基站判断估计出的CSI计算时长是否满足最大计算时长要求。若满足，则可在终端使用AI模型对CSI-RS反馈信息进行压缩，基站收到后用对应的解压缩模型进行解压缩；若不满足，则采用传统模式进行CSI反馈。
在一种实施例中，所有可用模型中，没有AI模型满足最大计算时长要求，则终端侧采用传统算法上报CSI-RS反馈信息。此处的传统算法上报CS反馈信息可至少包括：上报部分CSI-RS反馈信息。
在一种实施例中，所有可用模型中，仅有一个模型满足最大计算时长要求，则基站根据估计的CSI计算时长配置CSI-RS，并将对应的AI模型信息下发给终端。
在一种实施例中，所有可用AI模型中，有多个AI模型满足最大计算时长要求，则基站根据自身特性及计算资源分配情况选出最优模型，根据该最优模型的CSI计算时长配置CSI-RS，并将该最优模型信息下发给终端。
需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。需要说明的是，该方法可以结合本公开的其他实施例一起被实施，也可以独立被实施，在此不再赘述。
参考图11，为本公开实施例提供的基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法中在终端侧基于模型复杂度定义CSI计算时长的流程图，具体流程如下：
基站侧根据预设条件，将CSI计算时长定义方式设置为基于模型复杂度定义，并确定在终端侧对CSI计算时长进行估计。
基站侧将CSI-RS反馈信息压缩所需AI模型的复杂度信息下发终端，同时将最大计算时长要求下发终端。
终端侧根据预设模型复杂度计算信息，估计出所用AI模型下所需的CSI计算时长。
在一种实施例中，终端侧预设有针对不同复杂度模型的计算时长信息。终端根据基站下发的AI模型的复杂度信息与预设信息进行匹配，可得到对应的CSI计算时长。
在一种实施例中，终端侧预设有处理某基线模型对应的基线CSI计算时长信息。终端跟基站下发的AI模型的复杂度信息与该基线AI模型复杂度关系，在对基线AI模型CSI计算时长进行缩放，则可得到所用AI模型的CSI计算时长。
在一种实施例中，基站侧仅下发单个AI模型I的复杂度信息，则终端可得到该模型I对应的CSI计算时长T。
在一种实施例中，基站侧下发一组AI模型Ii(i＝1,2,3…)的复杂度信息，则终端可得到该组模型分别对应的CSI计算时长Ti(i＝1,2,3…)。
终端侧判断估计出的CSI计算时长是否满足最大计算时长要求。若满足，则终端可采用AI模型对CSI-RS反馈信息进行压缩，并向基站上报AI模型信息及估计的CSI计算时长；若不满足，则终端则采用传统模式进行CSI反馈。
在一种实施例中，基站下发的所有模型中，没有模型满足最大计算时长要求，则终端采用传统算法上报CSI-RS反馈信息，且不上报估计的CSI计算时长。
在一种实施例中，基站下发的所有模型中，仅有一个模型满足最大计算时长要求，则终端将该模型信息即对应的CSI计算时长上报给基站。
在一种实施例中，基站下发的所有模型中，有多个模型满足最大计算时长要求，则终端根据自身特性及硬件资源分配情况选出最优模型，并将该最优模型信息与对应的CSI计算时长上报基站。
基站根据终端上报信息，对CSI-RS进行配置。
在一种实施例中，终端向基站上报采用传统模式进行CSI反馈，则基站侧根据传统算法得到CSI计算时长，并对CSI-RS进行配置。
在一种实施例中，终端向基站上报采用的AI模型信息及估计的CSI计算时长，则基站侧接收到使用AI模型压缩的CSI-RS反馈信息后，可采用对应的解压缩模型得到恢复后的CSI-RS反馈信息。在配置CSI-RS时，可根据估计的CSI计算时长对CSI-RS进行配置。
需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。需要说明的是，该方法可以结合本公开的其他实施例一起被实施，也可以独立被实施，在此不再赘述。
参考图12，本公开实施例提供的基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法中在基站侧基于模型复杂度定义CSI计算时长的流程图，具体流程如下：
基站侧根据预设条件，将CSI计算时长定义方式设置为基于模型复杂度定义，并确定在基站侧对CSI计算时长进行估计。
终端向基站上报预设模型复杂度计算信息。
在一种实施例中，终端侧预设针对不同复杂度模型的计算时长信息。
在一种实施例中，终端侧预设处理某基线模型对应的基线CSI计算时长信息。
基站根据终端预设模型复杂度计算信息及采用的AI模型的复杂度信息，对CSI计算时长进行估计。
在一种实施例中，终端侧预设有针对不同复杂度模型的计算时长信息。终端根据基站下发的AI模型的复杂度信息与预设信息进行匹配，可得到对应的CSI计算时长。
在一种实施例中，终端侧存有处理某基线模型对应的基线CSI计算时长信息。终端跟基站下发的AI模型的复杂度信息与该基线模型复杂度关系，在对基线模型CSI计算时长进行缩放，则可得到所用AI模型的CSI计算时长。
在一种实施例中，仅有一个模型I可用，则基站根据该模型I的复杂度信息可得到对应的CSI计算时长T。
在一种实施例中，有一组模型Ii(i＝1,2,3…)可用，则基站根据该组模型的复杂度信息可得到该组模型对应的CSI计算时长Ti(i＝1,2,3…)。
基站判断估计出的CSI计算时长是否满足最大计算时长要求。若满足，则可在终端使用AI模型对CSI-RS反馈信息进行压缩，基站收到后用对应的解压缩模型进行解压缩；若不满足，则采用传统模式进行CSI反馈。
在一种实施例中，所有可用AI模型中没有AI模型满足最大计算时长要求，则采用传统算法上报CSI-RS反馈信息。
在一种实施例中，所有可用模型中，仅有一个模型满足最大计算时长要求，则基站根据估计的CSI计算时长配置CSI-RS，并将对应的AI模型信息下发给终端。
在一种实施例中，所有可用模型中，有多个模型满足最大计算时长要求，则基站根据自身特性及计算资源分配情况选出最优模型，根据该最优模型的CSI计算时长配置CSI-RS，并将该最优模型信息下发给终端。
需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。需要说明的是，该方法可以结合本公开的其他实施例一起被实施，也可以独立被实施，在此不再赘述。
参考图13，本公开实施例提供的基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法中在终端侧计算CSI计算时长的信令交互图，具体如下：
基站向终端发送AI基线模型的确定(AI-based CSI Delay Computation)消息，要求终端对CSI计算时长进行评估。
终端向基站发送基线AI模型的CSI计算时长确定完成(AI-based CSI Delay Computation Ready)，确认开始对CSI计算时长进行评估。
基站向终端发送AI能力请求(AI Capability Requirement)，该AI能力请求包含基站存储的模型复杂度及最大计算时长等信息。
终端采用基于终端处理时间或基于模型复杂度的方法，得到CSI计算时长，并根据最大计算时长判断是否采用AI模式。
终端向基站发送基线AI模型的复杂度信息(AI-based CSI Delay Computation Information)，上报是否采用AI模型进行CSI-RS反馈信息压缩、采用的AI模型信息及对应的CSI计算时长。
基站根据基线AI模型的复杂度信息(AI-based CSI Delay Computation Information)对CSI-RS进行配置。
若采用AI模式，基站将所用AI模型参数下发给终端。
参考图14，本公开实施例提供的基于AI的CSI-RS反馈信息压缩的CSI计算时长定义方法中在基站侧计算CSI计算时长的信令交互图，具体如下：
基站向终端发送终端能力问询(Capability Enquiry)，要求终端对自身计算能力进行上报。
终端向基站发送终端能力信息(Capability Information)，上报自身计算能力信息。能力信息至 少可以指示终端的AI能力。
基站根据存储的AI模型的复杂度信息，采用基于终端处理时间或基于模型复杂度的方法，对终端的AI能力进行评估，并判断是否采用AI模式。
基站根据评估结果对CSI-RS进行配置。
若采用AI模式，基站将所用AI模型的模型信息下发给终端。
在终端侧或基站侧计算基于AI的CSI压缩的CSI计算时间。在终端侧进行计算时，基站下发AI模型的复杂度信息，终端结合自身算力情况对CSI计算时间进行估计；在基站侧进行计算时，终端上报自身算力信息，基站结合存储的AI模型的复杂度信息对CSI计算时间进行估计。
在一些实施例中，提出了基于终端处理时间定义的CSI计算时间。基站侧存有AI模型的复杂度信息，终端侧存有自身算力信息，将模型复杂度与终端自身算力进行处理，可得到CSI计算时间。
在另一些实施例中，提出了基于模型复杂度定义的CSI计算时间。当终端侧存有处理基线AI模型所用CSI计算时长，根据基站所选用AI模型的复杂度与基线AI模型的复杂度关系，可得到在所选AI模型下的CSI计算时长；当终端侧存有针对多种复杂度的预设AI模型的处理时长时，将基站所选用AI模型复杂度与预设模型的复杂度进行匹配，即可得到在所选AI模型下的CSI计算时长。
需要说明的是，前述实施例是以基站为例进行说明的。当然，本领域内技术人员都理解，本公开的所有实施例中的网络侧设备可以为基站或任何其他设备，在此不再赘述。需要说明的是，该方法可以结合本公开的其他实施例一起被实施，也可以独立被实施，在此不再赘述。
如图15所示，本公开实施例提供一种信息处理装置，其中，所述方法包括：
确定模块110，用于确定终端是否支持至少一种AI模型的信道状态信息-参考信号CSI-RS反馈信息压缩；
配置模块120，用于根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。
该信息处理装置可包含在终端中。
在一些实施例中，所述确定模块110以及所述配置模块120可为程序模块；所述程序模块被处理器执行之后，能够实现上述功能。
在另一些实施例中，所述确定模块110和所述配置模块120可为软硬结合模块；所述软硬结合模块包括但不限于：可编程阵列；所述可编程阵列包括但不限于现场可编程阵列和/或复杂可编程阵列。
在还有一些实施例中，所述确定模块110和所述配置模块120可为纯硬件模块；所述纯硬件模块包括但不限于：专用集成电路。
在一些实施例中，所述确定模块110，被配置为向所述终端发送AI模型的复杂度信息；接收所述终端根据复杂度信息提供的第一信息；根据所述第一信息，确定所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩。
在一些实施例中，表明所述终端支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括：
模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
在一些实施例中，表明所述终端不支持至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
在一些实施例中，所述复杂度信息指示以下至少之一：
对应AI模型的总浮点运算次数，其中，所述总浮点运算次数与CSI计算时长的最大允许值，共同用于供所述终端确定是否支持对应AI模型的CSI-RS反馈信息压缩；
对应AI模型的复杂度相对于基线AI模型的复杂度之间的第一比值，其中，所述第一比值，用于供所述终端结合所述终端AI能力与所述基线AI模型的复杂度，确定是否支持对应AI模型的CSI-RS反馈信息压缩。
在一些实施例中，所述确定模块110，被配置为接收所述终端发送的AI能力信息；根据所述AI能力信息，确定所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩。
在一些实施例中，所述AI能力信息指示以下至少之一：
所述终端是否具有AI能力；
所述终端支持每秒的浮点运算次数；
所述终端在CSI计算时长的最大允许值内支持的浮点运算次数；
第二比值，指示所述终端AI能力与基线AI模型之间复杂度之间的比值。
在一些实施例中，所述配置模块120，被配置为根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置所述终端上报的CSI-RS反馈信息的类型；在所述终端支持至少一种AI模型的CSI-RS反馈信息压缩时，确定采用支持的所述AI模型对CSI-RS反馈信息压缩时的CSI计算时长，配置所述CSI-RS的周期。
如图16所示，本公开实施例提供一种信息处理装置，所述装置包括：
发送模块210，被配置为发送第二信息，其中，所述第二信息，用于供基站确定所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
该信息处理装置可包含在终端中。
在一些实施例中，该发送模块210可为程序模块；所述程序模块被处理器执行之后，能够实现供基站确定终端是否支持AI模型对CSI-RS反馈信息的压缩。
在另一些实施例中，该发送模块210可为软硬结合模块；所述软硬结合模块包括但不限于：可编程阵列。所述可编程阵列包括但不限于：现场可编程阵列和/或复杂可编程阵列。
在还有一些实施例中，该发送模块210可为纯硬件模块；所述纯硬件模块包括但不限于：专用集成电路。
在一些实施例中，所述第二信息包括：
指示所述终端AI能力的AI能力信息。
在一些实施例中，所述AI能力信息指示以下至少之一：
所述终端是否具有AI能力；
所述终端支持每秒的浮点运算次数；
所述终端在CSI计算时长的最大允许值内支持的浮点运算次数；
第二比值，指示所述终端AI能力与基线AI模型之间复杂度之间的比值。
在一些实施例中，所述第二信息包括：
第一信息，用于表明所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
在一些实施例中，表明所述终端支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括：
模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
在一些实施例中，表明所述终端不支持所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
在一些实施例中，所述装置还包括：
接收模块，被配置为接收AI模型的复杂度信息；
所述发送模块210，被配置为根据所述AI模型的复杂度信息以及所述终端的AI能力，向基站发送所述第一信息。
在一些实施例中，所述发送模块210，还被配置为执行以下至少之一：
根据所述AI模型的复杂度信息以及所述终端的AI能力，确定所述终端支持一种AI模型对所述CSI-RS反馈信息压缩时，向所述基站发送表明支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息；
根据所述AI模型的复杂度信息以及所述终端的AI能力确定所述终端支持多种AI模型对所述CSI-RS反馈信息压缩时，根据所述终端的资源调度情况选择所述终端支持的一种AI模型向所述基站表明支持一种AI模型的CSI-RS反馈信息压缩的所述第一信息；
根据所述AI模型的复杂度信息以及所述终端的AI能力确定所述终端不支持AI模型对所述CSI-RS反馈信息压缩时，向所述基站发送表明所述终端不支持至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息。
在一些实施例中，所述复杂度信息指示以下至少之一：
对应AI模型的总浮点运算次数，其中，所述总浮点运算次数与CSI计算时长的最大允许值，共同用于供所述终端确定是否支持对应AI模型的CSI-RS反馈信息压缩；
对应AI模型的复杂度相对于基线AI模型的复杂度之间的第一比值，其中，所述第一比值，用于供所述终端结合所述终端AI能力与所述基线AI模型的复杂度，确定是否支持对应AI模型的 CSI-RS反馈信息压缩。
本公开实施例提供一种通信设备，包括：
用于存储处理器可执行指令的存储器；
处理器，分别存储器连接；
其中，处理器被配置为执行前述任意技术方案提供的信息处理方法。
处理器可包括各种类型的存储介质，该存储介质为非临时性计算机存储介质，在通信设备掉电之后能够继续记忆存储其上的信息。
这里，所述通信设备包括：终端或者网元，该网元可为前述第一网元至第四网元中的任意一个。
所述处理器可以通过总线等与存储器连接，用于读取存储器上存储的可执行程序，例如，如图2至图14所示的方法的至少其中之一。
图17是根据一示例性实施例示出的一种终端800的框图。例如，终端800可以是移动电话，计算机，数字广播用户设备，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等。
参照图17，终端800可以包括以下一个或多个组件：处理组件802，存储器804，电源组件806，多媒体组件808，音频组件810，输入/输出(I/O)的接口812，传感器组件814，以及通信组件816。
处理组件802通常控制终端800的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件802可以包括一个或多个处理器820来执行指令，以完成上述的方法的全部或部分步骤。此外，处理组件802可以包括一个或多个模块，便于处理组件802和其他组件之间的交互。例如，处理组件802可以包括多媒体模块，以方便多媒体组件808和处理组件802之间的交互。
存储器804被配置为存储各种类型的数据以支持在终端800的操作。这些数据的示例包括用于在终端800上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器804可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电源组件806为终端800的各种组件提供电力。电源组件806可以包括电源管理系统，一个或多个电源，及其他与为终端800生成、管理和分配电力相关联的组件。
多媒体组件808包括在所述终端800和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件808包括一个前置摄像头和/或后置摄像头。当终端800处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光 学变焦能力。
音频组件810被配置为输出和/或输入音频信号。例如，音频组件810包括一个麦克风(MIC)，当终端800处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器804或经由通信组件816发送。在一些实施例中，音频组件810还包括一个扬声器，用于输出音频信号。
I/O接口812为处理组件802和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件814包括一个或多个传感器，用于为终端800提供各个方面的状态评估。例如，传感器组件814可以检测到设备800的打开/关闭状态，组件的相对定位，例如所述组件为终端800的显示器和小键盘，传感器组件814还可以检测终端800或终端800一个组件的位置改变，用户与终端800接触的存在或不存在，终端800方位或加速/减速和终端800的温度变化。传感器组件814可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件814还可以包括光传感器，如CMOS或CCD图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件814还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件816被配置为便于终端800和其他设备之间有线或无线方式的通信。终端800可以接入基于通信标准的无线网络，如WiFi，2G或3G，或它们的组合。在一个示例性实施例中，通信组件816经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件816还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，终端800可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述方法。
在示例性实施例中，还提供了一种包括指令的非临时性计算机可读存储介质，例如包括指令的存储器804，上述指令可由终端800的处理器820执行以完成上述方法。例如，所述非临时性计算机可读存储介质可以是ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
如图18所示，本公开一实施例示出一种接入设备的结构。例如，通信设备900可以被提供为一网络侧设备。该通信设备可为前述终端或者基站。
参照图18，通信设备900可为前述基站，该通信设备900可包括处理组件922，其进一步包括一个或多个处理器，以及由存储器932所代表的存储器资源，用于存储可由处理组件922的执行的指令，例如应用程序。存储器932中存储的应用程序可以包括一个或一个以上的每一个对应于一组指令的模块。此外，处理组件922被配置为执行指令，以执行上述方法前述应用在所述接入设备的任意方法，例如，如图4至图9任意一个所示方法。
通信设备900还可以包括一个电源组件926被配置为执行通信设备900的电源管理，一个有线或无线网络接口950被配置为将通信设备900连接到网络，和一个输入输出(I/O)接口958。通信设备900可以操作基于存储在存储器932的操作系统，例如Windows Server TM，Mac OS XTM，UnixTM，LinuxTM，FreeBSDTM或类似。
本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本发明的其它实施方案。本公开旨在涵盖本发明的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本发明的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本发明的真正范围和精神由下面的权利要求指出。
应当理解的是，本发明并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本发明的范围仅由所附的权利要求来限制。

Claims (21)

  1. 一种信息处理方法，其中，由基站执行，所述方法包括：
    确定终端是否支持至少一种人工智能AI模型的信道状态信息-参考信号CSI-RS反馈信息压缩；
    根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。
  2. 根据权利要求1所述的方法，其中，所述确定终端是否支持至少一种人工智能AI模型的CSI-RS反馈信息压缩，包括：
    向所述终端发送AI模型的复杂度信息；
    接收所述终端根据复杂度信息提供的第一信息；
    根据所述第一信息，确定所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩。
  3. 根据权利要求2所述的方法，其中，表明所述终端支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括：
    模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
    CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
  4. 根据权利要求2所述的方法，其中，表明所述终端不支持至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
  5. 根据权利要求2所述的方法，其中，所述复杂度信息指示以下至少之一：
    对应AI模型的总浮点运算次数，其中，所述总浮点运算次数与CSI计算时长的最大允许值，共同用于供所述终端确定是否支持对应AI模型的CSI-RS反馈信息压缩；
    对应AI模型的复杂度相对于基线AI模型的复杂度之间的第一比值，其中，所述第一比值，用于供所述终端结合所述终端AI能力与所述基线AI模型的复杂度，确定是否支持对应AI模型的CSI-RS反馈信息压缩。
  6. 根据权利要求1所述的方法，其中，所述确定终端是否支持至少一种人工智能AI模型的CSI-RS反馈信息压缩，包括：
    接收所述终端发送的AI能力信息；
    根据所述AI能力信息，确定所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩。
  7. 根据权利要求6所述的方法，其中，所述AI能力信息指示以下至少之一：
    所述终端是否具有AI能力；
    所述终端支持每秒的浮点运算次数；
    所述终端在CSI计算时长的最大允许值内支持的浮点运算次数；
    第二比值，指示所述终端AI能力与基线AI模型之间复杂度之间的比值。
  8. 根据权利要求1至7任一项所述的方法，其中，所述根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS，包括：
    根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置所述终端上报的CSI-RS反馈信息的类型；
    在所述终端支持至少一种AI模型的CSI-RS反馈信息压缩时，确定采用支持的所述AI模型对CSI-RS反馈信息压缩时的CSI计算时长，配置所述CSI-RS的周期。
  9. 一种信息处理方法，由终端执行，所述方法包括：
    发送第二信息，其中，所述第二信息，用于供基站确定所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
  10. 根据权利要求9所述的方法，其中，所述第二信息包括：
    指示所述终端AI能力的AI能力信息。
  11. 根据权利要求10所述的方法，其中，所述AI能力信息指示以下至少之一：
    所述终端是否具有AI能力；
    所述终端支持每秒的浮点运算次数；
    所述终端在CSI计算时长的最大允许值内支持的浮点运算次数；
    第二比值，指示所述终端AI能力与基线AI模型之间复杂度之间的比值。
  12. 根据权利要求9所述的方法，其中，所述第二信息包括：
    第一信息，用于表明所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
  13. 根据权利要求12所述的方法，其中，表明所述终端支持至少一种AI模型的CSI-RS反馈信息压缩的所述第一信息包括：
    模型标识，指示所述终端支持的用于CSI-RS反馈信息压缩的所述AI模型；
    CSI计算时长，指示所述终端采用所述模型标识的AI模型进行CSI-RS反馈信息压缩所需的时长。
  14. 根据权利要求12所述的方法，其中，表明所述终端不支持所述第一信息指示：CSI-RS反馈信息采用部分上报方式。
  15. 根据权利要求12所述的方法，其中，所述方法还包括：
    接收AI模型的复杂度信息；
    所述发送第二信息，包括：
    根据所述AI模型的复杂度信息以及所述终端的AI能力，向基站发送所述第一信息。
  16. 根据权利要求15所述的方法，其中，所述根据所述AI模型的复杂度信息以及所述终端的AI能力，向基站发送所述第一信息，包括以下至少之一：
    根据所述AI模型的复杂度信息以及所述终端的AI能力，确定所述终端支持一种AI模型对所述CSI-RS反馈信息压缩时，向所述基站发送表明支持至少一种AI模型的CSI-RS反馈信息压缩的所述 第一信息；
    根据所述AI模型的复杂度信息以及所述终端的AI能力确定所述终端支持多种AI模型对所述CSI-RS反馈信息压缩时，根据所述终端的资源调度情况选择所述终端支持的一种AI模型向所述基站表明支持一种AI模型的CSI-RS反馈信息压缩的所述第一信息；
    根据所述AI模型的复杂度信息以及所述终端的AI能力确定所述终端不支持AI模型对所述CSI-RS反馈信息压缩时，向所述基站发送表明所述终端不支持至少一种AI模型对CSI-RS信息反馈压缩的所述第一信息。
  17. 根据权利要求15或16所述的方法，其中，所述复杂度信息指示以下至少之一：
    对应AI模型的总浮点运算次数，其中，所述总浮点运算次数与CSI计算时长的最大允许值，共同用于供所述终端确定是否支持对应AI模型的CSI-RS反馈信息压缩；
    对应AI模型的复杂度相对于基线AI模型的复杂度之间的第一比值，其中，所述第一比值，用于供所述终端结合所述终端AI能力与所述基线AI模型的复杂度，确定是否支持对应AI模型的CSI-RS反馈信息压缩。
  18. 一种信息处理装置，其中，所述方法包括：
    确定模块，用于确定终端是否支持至少一种人工智能AI模型的信道状态信息-参考信号CSI-RS反馈信息压缩；
    配置模块，用于根据所述终端是否支持至少一种AI模型的CSI-RS反馈信息压缩，配置CSI-RS。
  19. 一种信息处理装置，所述装置包括：
    发送模块，被配置为发送第二信息，其中，所述第二信息，用于供基站确定所述终端是否支持至少一种AI模型对CSI-RS反馈信息的压缩。
  20. 一种通信设备，包括处理器、收发器、存储器及存储在存储器上并能够有所述处理器运行的可执行程序，其中，所述处理器运行所述可执行程序时执行如权利要求1至8或9至17任一项提供的方法。
  21. 一种计算机存储介质，所述计算机存储介质存储有可执行程序；所述可执行程序被处理器执行后，能够实现如权利要求1至8或9至17任一项提供的方法。
PCT/CN2022/080785 2022-03-14 2022-03-14 信息处理方法及装置、通信设备及存储介质 WO2023173262A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2022/080785 WO2023173262A1 (zh) 2022-03-14 2022-03-14 信息处理方法及装置、通信设备及存储介质
CN202280000750.1A CN114788317A (zh) 2022-03-14 2022-03-14 信息处理方法及装置、通信设备及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/080785 WO2023173262A1 (zh) 2022-03-14 2022-03-14 信息处理方法及装置、通信设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023173262A1 true WO2023173262A1 (zh) 2023-09-21

Family

ID=82422393

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/080785 WO2023173262A1 (zh) 2022-03-14 2022-03-14 信息处理方法及装置、通信设备及存储介质

Country Status (2)

Country Link
CN (1) CN114788317A (zh)
WO (1) WO2023173262A1 (zh)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117280723A (zh) * 2022-03-31 2023-12-22 北京小米移动软件有限公司 基于ai的csi处理能力确定方法、装置、介质、产品及芯片
CN117882426A (zh) * 2022-08-12 2024-04-12 北京小米移动软件有限公司 Csi上报方法、装置、设备及系统
WO2024097594A1 (en) * 2022-11-03 2024-05-10 Google Llc Channel state information reporting based on machine learning techniques and on non learning machine techniques
CN117997770A (zh) * 2022-11-04 2024-05-07 大唐移动通信设备有限公司 管理模型的方法、装置及设备
WO2024138375A1 (zh) * 2022-12-27 2024-07-04 北京小米移动软件有限公司 一种通信方法、装置、设备及存储介质
CN116017543A (zh) * 2022-12-27 2023-04-25 京信网络系统股份有限公司 信道状态信息反馈增强方法、装置、系统和存储介质
CN118282452A (zh) * 2022-12-30 2024-07-02 华为技术有限公司 一种通信方法及装置
WO2024207502A1 (zh) * 2023-04-07 2024-10-10 北京小米移动软件有限公司 一种通信方法、装置及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3454491A1 (en) * 2017-09-12 2019-03-13 Samsung Electronics Co., Ltd. Method and apparatus for mapping uplink control information for channel state information feedback
CN111819872A (zh) * 2020-06-03 2020-10-23 北京小米移动软件有限公司 信息传输方法、装置、通信设备及存储介质
CN111954206A (zh) * 2019-05-17 2020-11-17 株式会社Ntt都科摩 终端和基站
CN112350788A (zh) * 2020-08-07 2021-02-09 中兴通讯股份有限公司 信道状态信息反馈方法、装置、设备和存储介质
CN113922936A (zh) * 2021-08-31 2022-01-11 中国信息通信研究院 一种ai技术信道状态信息反馈方法和设备

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111954309A (zh) * 2019-05-17 2020-11-17 株式会社Ntt都科摩 终端和基站
CN114128164A (zh) * 2019-07-26 2022-03-01 Oppo广东移动通信有限公司 一种信息处理方法、网络设备、用户设备
US11696119B2 (en) * 2019-12-16 2023-07-04 Qualcomm Incorporated Neural network configuration for wireless communication system assistance
US12101142B2 (en) * 2020-01-14 2024-09-24 Nokia Technologies Oy Method, device and computer readable medium of communication
US11387880B2 (en) * 2020-02-28 2022-07-12 Qualcomm Incorporated Channel state information feedback using channel compression and reconstruction

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3454491A1 (en) * 2017-09-12 2019-03-13 Samsung Electronics Co., Ltd. Method and apparatus for mapping uplink control information for channel state information feedback
CN111954206A (zh) * 2019-05-17 2020-11-17 株式会社Ntt都科摩 终端和基站
CN111819872A (zh) * 2020-06-03 2020-10-23 北京小米移动软件有限公司 信息传输方法、装置、通信设备及存储介质
CN112350788A (zh) * 2020-08-07 2021-02-09 中兴通讯股份有限公司 信道状态信息反馈方法、装置、设备和存储介质
CN113922936A (zh) * 2021-08-31 2022-01-11 中国信息通信研究院 一种ai技术信道状态信息反馈方法和设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
ZTE, SANECHIPS: "Applications of Artificial Intelligence in MIMO Networks", 3GPP DRAFT; RP-201771, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), MOBILE COMPETENCE CENTRE ; 650, ROUTE DES LUCIOLES ; F-06921 SOPHIA-ANTIPOLIS CEDEX ; FRANCE, vol. TSG RAN, no. Electronic Meeting; 20200914 - 20200918, 7 September 2020 (2020-09-07), Mobile Competence Centre ; 650, route des Lucioles ; F-06921 Sophia-Antipolis Cedex ; France , XP051931570 *

Also Published As

Publication number Publication date
CN114788317A (zh) 2022-07-22

Similar Documents

Publication Publication Date Title
WO2023173262A1 (zh) 信息处理方法及装置、通信设备及存储介质
US20230098973A1 (en) Direct link data transmission method and apparatus, and storage medium
WO2020019216A1 (zh) 传输配置方法及装置
WO2023240573A1 (zh) 信道状态信息的处理方法、装置及通信设备
WO2023245576A1 (zh) Ai模型确定方法、装置、通信设备及存储介质
WO2021030974A1 (zh) 寻呼配置方法、装置、通信设备及存储介质
WO2021196214A1 (zh) 传输方法、装置及计算机存储介质
JP2022533071A (ja) モニタリング方法、シグナリング下り送信方法及び装置、通信機器及び記憶媒体
WO2024000523A1 (zh) 信道状态信息的处理方法及装置、通信设备及存储介质
WO2023240572A1 (zh) 一种信息传输方法、装置、通信设备及存储介质
US20230269047A1 (en) Positioning reference signaling configuration method and apparatus, user equipment, and storage medium
WO2022205385A1 (zh) 测量间隔处理方法、装置、通信设备及存储介质
WO2022151436A1 (zh) 信息配置方法及装置、通信设备和存储介质
WO2022104605A1 (zh) 调制与编码策略mcs的配置方法、装置及通信设备
WO2023201660A1 (zh) Rsrp门限参数确定方法、装置、通信设备及存储介质
WO2022147730A1 (zh) 省电信号的处理方法及装置、通信设备及存储介质
WO2022016450A1 (zh) 逻辑信道复用方法及装置、通信设备及存储介质
WO2022036610A1 (zh) 一种通信方法、通信装置及存储介质
US20220408469A1 (en) Downlink control information configuration method and apparatus, and communication device and storage medium
WO2023155111A1 (zh) 信息处理方法、装置、通信设备及存储介质
WO2023206504A1 (zh) 系统消息处理方法及装置、通信设备及存储介质
WO2023173260A1 (zh) 信息处理方法及装置、通信设备及存储介质
WO2023230969A1 (zh) 人工智能模型的确定方法及装置、通信设备及存储介质
WO2024060027A1 (zh) 信息处理方法、装置、通信设备及存储介质
WO2024168489A1 (zh) 信息传输方法及装置、通信设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22931302

Country of ref document: EP

Kind code of ref document: A1