Communication method and device

Abstract

The embodiment of the invention provides a communication method and a communication device. The method comprises the steps of determining at least two pieces of channel information, such as latest channel information and at least one piece of historical channel information, and then determining input of a first AI model or determining a replaced second AI model according to a time interval between channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model. Therefore, the input of the AI model can be matched with the capability requirement of the AI model, so that the input of the AI model is more suitable, and the performance loss caused by taking improper channel information as the input of the AI model is avoided.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and more specifically, to a communication method and device. Background Art

[0002] With the development of wireless communication technology, in order to support more services and meet higher requirements in terms of system capacity, communication delay and other indicators, the size of antenna arrays is growing, and the supported frequency bands are also increasing, making the communication system more complex. Improving the performance of complex communication systems through artificial intelligence (AI) methods has become an important research direction. For example, the application value of AI in many wireless communication technologies such as channel state information (CSI) feedback, beam management, and pilot design management has been continuously revealed.

[0003] In wireless communications, channels generally change continuously in time, so communication performance can be improved by mining channel time domain correlation. If AI models are applied to wireless communications, then how to effectively utilize channel time domain correlation to improve the performance of AI models is a question worth considering. Summary of the invention

[0004] The present application provides a communication method and device to improve the performance of AI models.

[0005] In a first aspect, a communication method is provided, which can be performed by a communication device. The communication device can be a terminal device, or a chip or circuit for a terminal device, or a network device, or a chip or circuit for a network device, or an AI node, or a chip or circuit for an AI node, and the present application does not limit this.

[0006] The method may include: determining at least two channel information, the at least two channel information including first channel information and second channel information, the first channel information being obtained before the second channel information; determining the input of the first AI model or determining the second AI model according to the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first artificial intelligence AI model.

[0007] Based on the above technical solution, the input of the first AI model or the second AI model is determined according to the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model, so that the input of the first AI model can be matched with the capability requirement of the first AI model, making the input of the first AI model more appropriate, avoiding performance loss caused by using inappropriate channel information as the input of the first AI model. In addition, when it is determined that the first AI model is inappropriate according to the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model, the second AI model can also be determined, and then the second AI model can continue to be used for related operations, avoiding performance loss caused by using an inappropriate AI model to process channel information.

[0008] In combination with the first aspect, in some implementations of the first aspect, determining the input of the first AI model or determining the second AI model according to the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model, includes any one of the following: determining the second AI model when the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, and the input of the second AI model includes at least two channel information; or determining the second AI model when the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model. The input of the second AI model includes the second channel information, and the state information of the second AI model is a default value; or, when it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes M channel information of at least two channel information and X default values, the M channel information does not include the first channel information, M is an integer greater than or equal to 1, and X is an integer greater than 0 or equal to 0; or, when it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes the second channel information, and the state information of the first AI model is the default value.

[0009] Optionally, when it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the first capability requirement of the first AI model, the input of the first AI model is determined, that is, the AI ​​model is not switched. The input of the first AI model may include M channel information and X default values ​​from at least two channel information; or may also include at least two channel information; or may also include the second channel information, and the state information of the first AI model is the default value.

[0010] Optionally, when determining that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the second capability requirement of the first AI model, determining that the input of the first AI model includes at least two channel information, or determining that the input of the first AI model includes the second channel information and the state information of the first AI model. The state information of the first AI model may be determined by the previous output of the first AI model.

[0011] Based on the above technical solution, different operations can be performed according to the different capability requirements of the first AI model. This not only makes it possible to improve the performance of the AI ​​model through a suitable AI model or the input of a suitable AI model, but also to select different rules based on the functions of the AI ​​model.

[0012] In combination with the first aspect, in some implementations of the first aspect, determining at least two pieces of channel information includes: receiving a reference signal, and performing measurement based on the reference signal to obtain at least two pieces of channel information.

[0013] In combination with the first aspect, in certain implementations of the first aspect, the channel acquisition time corresponding to the first channel information is determined based on: the reception time or the transmission time of the reference signal used to measure and obtain the first channel information, or, the reporting time of the first channel information; the channel acquisition time corresponding to the second channel information is determined based on: the reception time or the transmission time of the reference signal used to measure and obtain the second channel information, or, the reporting time of the second channel information.

[0014] Based on the above technical solution, the network side and the terminal side can be aligned with the channel acquisition time corresponding to the channel information, so that when the network device and the terminal device determine the input of the first AI model or determine the second AI model according to the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model, the determined results can be consistent, thereby ensuring the symmetry of the behaviors of the network side and the terminal side.

[0015] In combination with the first aspect, in some implementations of the first aspect, the reference signal is periodic, and the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is determined according to the period of the reference signal.

[0016] The reference signal is periodic, which means that the time domain behavior of the reference signal is periodic, that is, the reference signal is transmitted periodically.

[0017] Optionally, the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is: k*Y, wherein Y represents a period of the reference signal, and k represents the number of channel information between the first channel information and the second channel information.

[0018] In combination with the first aspect, in certain implementations of the first aspect, the reference signal is semi-static, and the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is determined based on the time interval between the activation signaling and the deactivation signaling, wherein the activation signaling is used to activate the reference signal used to measure and obtain the second channel information, and the deactivation signaling is used to deactivate the reference signal used to measure and obtain the first channel information; or, the activation signaling is used to activate the reporting of the second channel information, and the deactivation signaling is used to deactivate the reporting of the first channel information.

[0019] The reference signal is semi-static, which means that the time domain behavior of the reference signal is semi-static. For the semi-static and the activation signaling and deactivation signaling involved in the semi-static, please refer to the description in the following embodiments, which will not be repeated here.

[0020] Optionally, the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is: S+P+Δ. Wherein, S represents the time interval between the reception time or transmission time of the reference signal used to measure the first channel information and this deactivation signaling, P represents the time interval between the reception time or transmission time of the reference signal used to measure the second channel information and this activation signaling, and Δ represents the time interval between the above deactivation signaling and the activation signaling. Alternatively, S represents the time interval between the reporting time of the first channel information and this deactivation signaling, P represents the time interval between the reporting time of the second channel information and this activation signaling, and Δ represents the time interval between the above deactivation signaling and the activation signaling.

[0021] It can be understood that the above is an example and is not limited to this. For example, if the first channel information and the second channel information are in the same activation period, the first channel information and the second channel information can also be determined according to the period of the reference signal in the activation period.

[0022] In combination with the first aspect, in some implementations of the first aspect, determining the input of the first AI model or determining the second AI model according to the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model, includes: determining the second AI model when it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, and the input of the second AI model includes feedback information of at least two channel information; or determining that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, determining the second AI model, and the input of the second AI model includes feedback information of at least two channel information. The input includes feedback information of the second channel information, and the state information of the second AI model is a default value; or, when it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes feedback information of M channel information in at least two channel information and X default values, where the M channel information does not include the first channel information, M is an integer greater than or equal to 1, and X is an integer greater than 0 or equal to 0; or, when it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes feedback information of the second channel information, and the state information of the first AI model is the default value.

[0023] Optionally, the feedback information may be feedback information output by a terminal-side encoder, or may be obtained by processing the output of the encoder, such as coarse-precision channel information restored by a decoder.

[0024] Based on the above technical solution, different operations can be performed according to the different capability requirements of the first AI model. This not only makes it possible to improve the performance of the AI ​​model through a suitable AI model or the input of a suitable AI model, but also to select different rules based on the functions of the AI ​​model.

[0025] In combination with the first aspect, in certain implementations of the first aspect, determining at least two channel information includes: sending a reference signal and receiving feedback information of at least two channel information, obtaining at least two channel information according to the feedback information of at least two channel information, and the channel acquisition time corresponding to the first channel information is determined according to: the sending time of the reference signal used to measure and obtain the first channel information, or, the reporting time of the first channel information; the channel acquisition time corresponding to the second channel information is determined according to: the sending time of the reference signal used to measure and obtain the second channel information, or, the reporting time of the second channel information.

[0026] Based on the above technical solution, the network side and the terminal side can be aligned with the channel acquisition time corresponding to the channel information, so that when the network device and the terminal device determine the input of the first AI model or determine the second AI model according to the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model, the determined results can be consistent, thereby ensuring the symmetry of the behaviors of the network side and the terminal side.

[0027] In combination with the first aspect, in some implementations of the first aspect, the M channel information does not include channel information before the first channel information, or the M channel information does not include channel information before the second channel information.

[0028] Based on the above technical solution, considering that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it can be assumed that the time interval between the channel information obtained before the first channel information (for distinction, recorded as channel information #1) or the channel information obtained before the second channel information (for distinction, recorded as channel information #2) and the channel acquisition time corresponding to the second channel information also does not meet the second capability requirement of the first AI model, thereby reducing the overhead caused by determining whether the time interval between the channel information #1 or the channel information #2 and the channel acquisition time corresponding to the second channel information meets the second capability requirement of the first AI model.

[0029] In combination with the first aspect, in certain implementations of the first aspect, the first capability requirement of the first AI model is that the time interval between channel acquisition times is within a first numerical range, and / or the second capability requirement of the first AI model is that the time interval between channel acquisition times is within a second numerical range.

[0030] In combination with the first aspect, in certain implementations of the first aspect, the first numerical range is less than or equal to T1; and / or the second numerical range is less than or equal to T2; wherein T1 is the maximum value of the time interval between the channel acquisition times corresponding to the first capability requirement of the first AI model, and T2 is the maximum value of the time interval between the channel acquisition times corresponding to the second capability requirement of the first AI model.

[0031] In combination with the first aspect, in some implementations of the first aspect, the value of T1 and / or T2 is determined by any one of the following methods: determined by a network device, determined by a terminal device, or predefined.

[0032] In combination with the first aspect, in some implementations of the first aspect, the method also includes: obtaining first indication information, the first indication information indicating at least one capability requirement of the first AI model.

[0033] Optionally, obtaining the first indication information includes: receiving the first indication information, or determining the first indication information by itself.

[0034] Based on the above technical solution, the communication device can obtain at least one capability requirement of the first AI model. For example, the communication device receives the first indication information sent by other devices. Accordingly, the other devices send the first indication information to the communication device. As an example, the other devices may be network-side devices or chips or circuits for network-side devices. The network-side device may be, for example, an access network device, or may be a core network device, without limitation. For another example, the communication device may determine at least one capability requirement of the first AI model by itself.

[0035] In combination with the first aspect, in certain implementations of the first aspect, the first channel information is any one of the at least two channel information except the second channel information, or the first channel information is separated from the second channel information by k channel information, where k is an integer greater than 1 or equal to 1.

[0036] In combination with the first aspect, in some implementations of the first aspect, the method further includes: obtaining second indication information, where the second indication information indicates k.

[0037] Optionally, obtaining the second indication information includes: receiving the second indication information, or determining the second indication information by itself.

[0038] Based on the above technical solution, the communication device can obtain k. For example, the communication device receives the second indication information sent by the other device, and the second indication information indicates k. Correspondingly, the other device sends the second indication information to the communication device. As an example, the other device can be a network side device or a chip or circuit for a network side device. The network side device can be, for example, an access network device, or can also be a core network device, without limitation. For another example, the communication device can determine k by itself.

[0039] In a second aspect, a communication device is provided, which is used to execute the method provided in the first aspect. Specifically, the device may include a unit and / or module, such as a processing unit and / or a communication unit, for executing the method provided in any one of the above implementations of the first aspect.

[0040] In one implementation, the apparatus is a communication device (such as a terminal device or a network device). When the apparatus is a communication device, the communication unit may be a transceiver or an input / output interface; the processing unit may be at least one processor. Optionally, the transceiver may be a transceiver circuit. Optionally, the input / output interface may be an input / output circuit.

[0041] In another implementation, the device is a chip, a chip system or a circuit used in a communication device. When the device is a chip, a chip system or a circuit used in a terminal device, the communication unit may be an input / output interface, an interface circuit, an output circuit, an input circuit, a pin or a related circuit on the chip, the chip system or the circuit; the processing unit may be at least one processor, a processing circuit or a logic circuit.

[0042] In a third aspect, a communication device is provided, comprising: a memory for storing programs; and at least one processor for executing computer programs or instructions stored in the memory to execute a method provided in any one of the above implementations of the first aspect.

[0043] In one implementation, the apparatus is a communication device (such as a terminal device or a network device).

[0044] In another implementation, the apparatus is a chip, a chip system or a circuit used in a communication device.

[0045] In a fourth aspect, the present application provides a processor for executing the methods provided in the above aspects.

[0046] For the operations such as sending and acquiring / receiving involved in the processor, unless otherwise specified, or unless they conflict with their actual function or internal logic in the relevant description, they can be understood as operations such as processor output and input, or as sending and receiving operations performed by the radio frequency circuit and antenna, and this application does not limit this.

[0047] In a fifth aspect, a computer-readable storage medium is provided, which stores a program code for execution by a device, wherein the program code includes a method for executing any one of the above-mentioned implementations of the first aspect.

[0048] In a sixth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the method provided in any one of the above-mentioned implementations of the first aspect.

[0049] In a seventh aspect, a chip is provided, the chip comprising a processor and a communication interface, the processor reads instructions stored in a memory through the communication interface, and executes the method provided by any one of the above-mentioned implementations of the first aspect.

[0050] Optionally, as an implementation method, the chip also includes a memory, in which a computer program or instructions are stored, and the processor is used to execute the computer program or instructions stored in the memory. When the computer program or instructions are executed, the processor is used to execute the method provided in any one of the above implementation methods of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1A and Figure 1B is a schematic diagram of a wireless communication system applicable to an embodiment of the present application.

[0052] Figure 2 This is a schematic diagram of processing channel information through the AI ​​model.

[0053] Figure 3 It is a schematic diagram of a communication method 300 provided in an embodiment of the present application.

[0054] Figure 4 It is a schematic flow chart of a communication method 400 provided according to an embodiment of the present application.

[0055] Figure 5 It is a schematic diagram of a semi-static CSI reporting method.

[0056] Figure 6 It is a schematic flow chart of a communication method 600 provided according to another embodiment of the present application.

[0057] Figure 7 It is a schematic flow chart of a communication method 700 provided according to another embodiment of the present application.

[0058] Figure 8 It is a schematic block diagram of a communication device 800 provided in an embodiment of the present application.

[0059] Fig. 9It is a schematic diagram of another communication device 900 provided in an embodiment of the present application.

[0060] Fig.10 It is a schematic diagram of a chip system 1000 provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0062] The technical solution provided in the present application can be applied to various communication systems, such as: the fifth generation (5th generation, 5G) or new radio (new radio, NR) system, long term evolution (long term evolution, LTE) system, LTE frequency division duplex (frequency division duplex, FDD) system, LTE time division duplex (time division duplex, TDD) system, wireless local area network (wireless local area network, WLAN) system, satellite communication system, future communication system, such as the sixth generation (6th generation, 6G) mobile communication system, or a fusion system of multiple systems, etc. The technical solution provided in the present application can also be applied to device to device (D2D) communication, vehicle-to-everything (V2X) communication, machine to machine (M2M) communication, machine type communication (MTC), and Internet of things (IoT) communication system or other communication systems.

[0063] A network element in a communication system can send a signal to another network element or receive a signal from another network element. The signal may include information, signaling or data, etc. The network element may also be replaced by an entity, a network entity, a device, a communication device, a communication module, a node, a communication node, etc. The network element is used as an example for description in the present disclosure. For example, the communication system may include at least one terminal device and at least one network device. The network device may send a downlink signal to the terminal device, and / or the terminal device may send an uplink signal to the network device. It is understandable that the terminal device in the present disclosure may be replaced by a first network element, and the network device may be replaced by a second network element, and the two perform the corresponding communication method in the present disclosure. Alternatively, the corresponding communication method in the present disclosure may be applied between network devices, or between terminal devices, which is not limited here.

[0064] In an embodiment of the present application, the terminal device may also be referred to as user equipment (UE), access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent or user device.

[0065] The terminal device may be a device that provides voice / data, such as a handheld device or a vehicle-mounted device with a wireless connection function. At present, some examples of terminals are: mobile phones, tablet computers, laptop computers, PDAs, mobile internet devices (MID), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self driving, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication functions, computing devices or other processing devices connected to wireless modems, wearable devices, terminal devices in 5G networks or future evolved public land mobile communication networks (public land mobile The embodiments of the present application do not limit this.

[0066] As an example but not limitation, in the embodiments of the present application, the terminal device may also be a wearable device. Wearable devices may also be referred to as wearable smart devices, which are a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothing, and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also powerful functions achieved through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include full-featured, large-sized, and fully or partially independent of smartphones, such as smart watches or smart glasses, as well as devices that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various types of smart bracelets and smart jewelry for vital sign monitoring.

[0067] It should be understood that in some scenarios, the UE can also be used to act as a base station. For example, the UE can act as a scheduling entity that provides sidelink signals between UEs in scenarios such as V2X, D2D or P2P.

[0068] In the embodiment of the present application, the device for realizing the function of the terminal device may be a terminal device, or a device capable of supporting the terminal device to realize the function, such as a chip system or a chip, which may be installed in the terminal device. In the embodiment of the present application, the chip system may be composed of a chip, or may include a chip and other discrete devices. In the embodiment of the present application, only the device for realizing the function of the terminal device is used as an example for explanation, and the scheme of the embodiment of the present application is not limited.

[0069] The network device in the embodiment of the present application may be a device for communicating with a terminal device, and the network device may also be referred to as an access network device or a wireless access network device, such as a base station. The network device in the embodiment of the present application may refer to a wireless access network (RAN) node (or device) that connects a terminal device to a wireless network. Base station can broadly cover various names as follows, or replace with the following names, such as: NodeB, evolved NodeB (eNB), next generation NodeB (gNB), relay station, access point, transmission point (TRP), transmission point (TP), master station, auxiliary station, multi-standard wireless (motor slide retainer, MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), positioning node, etc. The base station can be a macro base station, a micro base station, a relay node, a donor node or the like, or a combination thereof. The base station can also refer to a communication module, a modem or a chip used to be set in the aforementioned device or apparatus. The base station can also be a mobile switching center and a device that performs the base station function in D2D, V2X, and M2M communications, a network-side device in a 6G network, or a device that performs the base station function in a future communication system. The base station can support networks with the same or different access technologies. The embodiments of this application do not limit the specific technology and specific device form used by the network device.

[0070] A base station can be fixed or mobile. For example, a helicopter or drone can be configured to act as a mobile base station, and one or more cells can move according to the location of the mobile base station. In other examples, a helicopter or drone can be configured to act as a device that communicates with another base station.

[0071] In some deployments, the network device mentioned in the embodiments of the present application may be a device including a CU, or a DU, or a device including a CU and a DU, or a control plane CU node (central unit control plane (central unit-control plane, CU-CP)) and a user plane CU node (central unit user plane (central unit-user plane, CU-UP)) and a DU node. For example, the network device may include a gNB-CU-CP, a gNB-CU-UP, and a gNB-DU.

[0072] In some deployments, multiple RAN nodes collaborate to assist the terminal in achieving wireless access, and different RAN nodes implement part of the functions of the base station. For example, the RAN node can be a CU, DU, CU-CP, CU-UP, or RU. The CU and DU can be set separately, or can also be included in the same network element, such as a BBU. The RU can be included in a radio frequency device or a radio frequency unit, such as an RRU, AAU, or RRH.

[0073] The RAN node can support one or more types of fronthaul interfaces, and different fronthaul interfaces correspond to DUs and RUs with different functions. If the fronthaul interface between the DU and the RU is a common public radio interface (CPRI), the DU is configured to implement one or more baseband functions, and the RU is configured to implement one or more radio frequency functions. If the fronthaul interface between the DU and the RU is an enhanced common public radio interface (eCPRI), relative to CPRI, part of the downlink and / or uplink baseband functions are moved from the DU to the RU for implementation. The division method between the DU and the RU is different, corresponding to different types (category, Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, and F.

[0074] Taking eCPRI Cat A as an example, for downlink transmission, based on layer mapping, the DU is configured to implement layer mapping and one or more functions before it (i.e., one or more functions of coding, rate matching, scrambling, modulation, and layer mapping), while other functions after layer mapping (e.g., RE mapping, digital beamforming (BF), or one or more functions of inverse fast Fourier transform (IFFT) / adding cyclic prefix (CP)) are moved to the RU for implementation. For uplink transmission, based on de-RE mapping, the DU is configured to implement de-mapping and one or more functions before it (i.e., one or more functions of decoding, de-rate matching, de-scrambling, demodulation, inverse discrete Fourier transform (IDFT), channel equalization, and de-RE mapping), while other functions after de-mapping (e.g., one or more functions of digital BF or fast Fourier transform (FFT) / removing CP) are moved to the RU for implementation. It can be understood that for the functional description of DU and RU corresponding to various types of eCPRI, reference can be made to the eCPRI protocol and will not be repeated here.

[0075] In one possible design, the processing unit for implementing the baseband function in the BBU is called a baseband high layer (BBH) unit, and the processing unit for implementing the baseband function in the RRU / AAU / RRH is called a baseband low layer (BBL) unit.

[0076] In different systems, CU (or CU-CP and CU-UP), DU or RU may also have different names, but those skilled in the art can understand their meanings. For example, in the ORAN system, CU may also be called O-CU (open CU), DU may also be called O-DU, CU-CP may also be called O-CU-CP, CU-UP may also be called O-CU-UP, and RU may also be called O-RU. Any unit in the CU (or CU-CP, CU-UP), DU and RU in this application may be implemented by a software module, a hardware module, or a combination of a software module and a hardware module.

[0077] In the embodiment of the present application, the device for realizing the function of the network device may be a network device, or a device capable of supporting the network device to realize the function, such as a chip system or a chip, which may be installed in the network device. In the embodiment of the present application, the chip system may be composed of a chip, or may include a chip and other discrete devices. In the embodiment of the present application, only the device for realizing the function of the network device is a network device as an example for explanation, and the solution of the embodiment of the present application is not limited.

[0078] Network devices and terminal devices can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; they can also be deployed on the water; they can also be deployed on aircraft, balloons and satellites in the air. The scenarios in which network devices and terminal devices are located are not limited in the embodiments of the present application. In addition, terminal devices and network devices can be hardware devices, or they can be software functions running on dedicated hardware, software functions running on general-purpose hardware, such as virtualization functions instantiated on a platform (e.g., a cloud platform), or entities including dedicated or general-purpose hardware devices and software functions. The present application does not limit the specific forms of terminal devices and network devices.

[0079] In addition, in order to support AI technology in wireless networks, AI nodes may also be introduced into the network.

[0080] Optionally, the AI ​​node can be deployed in one or more of the following locations in the communication system: access network equipment, terminal equipment, or core network equipment, etc., or the AI ​​node can also be deployed separately, for example, deployed in a location other than any of the above devices, such as a host or cloud server in an over the top (OTT) system. The AI ​​node can communicate with other devices in the communication system, and the other devices can be, for example, one or more of the following: network equipment, terminal equipment, or network elements of the core network, etc.

[0081] It is understood that the present application does not limit the number of AI nodes. For example, when there are multiple AI nodes, the multiple AI nodes can be divided based on functions, such as different AI nodes are responsible for different functions.

[0082] It can also be understood that AI nodes can be independent devices, or they can be integrated into the same device to implement different functions, or they can be network elements in hardware devices, or they can be software functions running on dedicated hardware, or they can be virtualized functions instantiated on a platform (for example, a cloud platform). This application does not limit the specific form of the above-mentioned AI nodes.

[0083] An AI node can be an AI network element or an AI module.

[0084] First, a communication system applicable to an embodiment of the present application is briefly introduced as follows.

[0085] Figure 1A and Figure 1B is a schematic diagram of a wireless communication system applicable to an embodiment of the present application.

[0086] like Figure 1A As shown, the wireless communication system includes a wireless access network 100. The wireless access network 100 may be a next generation (e.g., 6G or higher) wireless access network, or a traditional (e.g., 5G, 4G, 3G, or 2G) wireless access network. One or more terminal devices (120a-120j, collectively referred to as 120) may be connected to each other or to one or more network devices (110a, 110b, collectively referred to as 110) in the wireless access network 100. Network elements in the wireless communication system are connected through interfaces (e.g., NG, Xn) or air interfaces. In addition, one or more AI modules may be provided in each network element in the wireless communication system. The AI ​​modules deployed in different network elements may be the same or different.

[0087] Figure 1A This is just a schematic diagram. The wireless communication system may also include other devices, such as core network devices, wireless relay devices and / or wireless backhaul devices. Figure 1A Not drawn in.

[0088] like Figure 1B As shown, the wireless communication system includes a RAN intelligent controller (RAN intelligentcontroller, RIC). As an example, RIC can be used to implement AI-related functions. As an example, the RIC includes a near-real time RIC (near-real time RIC, near-RT RIC) and a non-real time RIC (non-real time RIC, Non-RTRIC). Among them, the non-real-time RIC mainly processes non-real-time information, such as data that is not sensitive to delay, and the delay of the data can be in seconds. Real-time RIC mainly processes near real-time information, such as data that is relatively sensitive to delay, and the delay of the data is in the order of tens of milliseconds.

[0089] The near real-time RIC is used for model training and reasoning. For example, it is used to train an AI model and use the AI ​​model for reasoning. The near real-time RIC can obtain information on the network side and / or the terminal side from a RAN node (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or a terminal. This information can be used as training data or reasoning data. Optionally, the near real-time RIC can submit the reasoning results to the RAN node and / or the terminal. Optionally, the reasoning results can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the near real-time RIC submits the reasoning results to the DU, and the DU sends it to the RU.

[0090] The non-real-time RIC is also used for model training and reasoning. For example, it is used to train an AI model and use the model for reasoning. The non-real-time RIC can obtain information on the network side and / or the terminal side from a RAN node (such as a CU, CU-CP, CU-UP, DU and / or RU) and / or a terminal. The information can be used as training data or reasoning data, and the reasoning results can be submitted to the RAN node and / or the terminal. Optionally, the reasoning results can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the non-real-time RIC submits the reasoning results to the DU, and the DU sends it to the RU.

[0091] The near real-time RIC and the non-real-time RIC may also be separately set as a network element. Optionally, the near real-time RIC and the non-real-time RIC may also be part of other devices, for example, the near real-time RIC is set in a RAN node (for example, in a CU or DU), and the non-real-time RIC is set in an OAM, a cloud server, a core network device, or other network devices.

[0092] In practical applications, the wireless communication system may include multiple network devices (also called access network devices) at the same time, and may also include multiple terminal devices at the same time, without limitation. A network device may serve one or more terminal devices at the same time. A terminal device may also access one or more network devices at the same time. The embodiment of the present application does not limit the number of terminal devices and network devices included in the wireless communication system.

[0093] In order to facilitate understanding of the embodiments of the present application, the terms involved in the embodiments of the present application are briefly explained below.

[0094] 1. Artificial Intelligence: It is to enable machines to learn, accumulate experience, and solve problems that humans can solve through experience, such as natural language understanding, image recognition, and chess. Artificial Intelligence can be understood as the intelligence displayed by machines made by humans. Usually artificial intelligence refers to the technology that presents human intelligence through computer programs. The goals of artificial intelligence include understanding intelligence by building computer programs with symbolic reasoning or inference.

[0095] 2. Machine learning: It is a way to implement artificial intelligence. Machine learning is a method that can give machines the ability to learn, so that machines can complete functions that cannot be completed by direct programming. In a practical sense, machine learning is a method of using data to train a model and then using the model to predict. There are many methods of machine learning, such as neural networks (NN), decision trees, support vector machines, etc. The theory of machine learning is mainly to design and analyze some algorithms that allow computers to learn automatically. Machine learning algorithms are a type of algorithm that automatically analyzes data to obtain rules and uses the rules to predict unknown data.

[0096] 3. Neural network: It is a specific embodiment of machine learning methods. Neural network is a mathematical model that imitates the behavioral characteristics of animal neural networks and processes information. The idea of ​​neural network comes from the neuron structure of brain tissue. Each neuron can perform a weighted sum operation on its input value, and the result of the weighted sum operation is used to generate an output through an activation function.

[0097] A neural network generally includes a multi-layer structure, and each layer may include one or more logic judgment units, which may be called neurons. By increasing the depth and / or width of the neural network, the expressive power of the neural network can be improved, providing more powerful information extraction and abstract modeling capabilities for complex systems. Among them, the depth of the neural network can be understood as the number of layers included in the neural network, and the number of neurons included in each layer can be called the width of the layer. In one possible implementation, the neural network includes an input layer and an output layer. The input layer of the neural network processes the received input through neurons and passes the result to the output layer, and the output layer obtains the output result of the neural network. In another possible implementation, the neural network includes an input layer, a hidden layer, and an output layer. The input layer of the neural network processes the received input through neurons and passes the result to the middle hidden layer, and the hidden layer then passes the calculation result to the output layer or the adjacent hidden layer, and finally the output layer obtains the output result of the neural network. A neural network may include one or more hidden layers connected in sequence, without limitation.

[0098] During the training process of a neural network, a loss function can be defined. The loss function is used to measure the difference between the predicted value of the model and the true value. During the training process of a neural network, the loss function describes the gap or difference between the output value of the neural network and the ideal target value. The training process of a neural network is to adjust the parameters of the neural network so that the value of the loss function is less than the threshold value or meets the target requirements. Among them, the neural network parameters may include at least one of the following: the number of layers of the neural network, the width, the weight of the neuron, and the parameters in the activation function of the neuron.

[0099] 4. Deep neural network: A neural network with multiple hidden layers.

[0100] 5. Deep learning: machine learning using deep neural networks.

[0101] 6. AI model: It is an algorithm or computer program that can realize AI functions. AI model represents the mapping relationship between the input and output of the model, or the AI ​​model is a function model that maps input of a certain dimension to output of a certain dimension. The parameters of the function model can be obtained through machine learning training. For example, f(x) = ax 2 + b is a quadratic function model, which can be regarded as an AI model, a and b are parameters of the AI ​​model, and a and b can be obtained through machine learning training. Exemplarily, the AI ​​models mentioned in the embodiments below of the present application are not limited to neural networks, linear regression models, decision tree models, support vector machines (SVM), Bayesian networks, Q learning models or other machine learning (ML) models.

[0102] AI model design mainly includes data collection links (for example, collecting training data and / or reasoning data), model training links, and model reasoning links. It can also include reasoning result application links. In the aforementioned data collection link, the data source is used to provide training data sets and reasoning data. In the model training link, the AI ​​model is obtained by analyzing or training the training data provided by the data source. Learning the AI ​​model through the model training node is equivalent to using the training data to learn the mapping relationship between the input and output of the AI ​​model. In the model reasoning link, the AI ​​model trained through the model training link is used to reason based on the reasoning data provided by the data source to obtain the reasoning result. This link can also be understood as: inputting the reasoning data into the AI ​​model, and obtaining the output through the AI ​​model, which is the reasoning result. The reasoning result can indicate: the configuration parameters used (executed) by the execution object, and / or the operation performed by the execution object. The reasoning result is published in the reasoning result application link, for example, the reasoning result can be uniformly planned by the execution (actor) entity, for example, the execution entity can send the reasoning result to one or more execution objects (for example, core network equipment, access network equipment, or terminal equipment, etc.) for execution. For example, the execution entity can also provide feedback on the performance of the AI ​​model to the data source, facilitating the subsequent update and training of the AI ​​model.

[0103] It is understood that the implementation of the AI ​​model can be a hardware circuit, or software, or a combination of software and hardware, without limitation. Non-limiting examples of software include: program code, program, subroutine, instruction, instruction set, code, code segment, software module, application, or software application, etc.

[0104] In addition, the encoder can deploy multiple AI models, so that the encoder can encode based on the deployed AI models, such as compressing channel information; the decoder can deploy multiple AI models, so that the decoder can decode based on the deployed AI models, such as recovering compressed information to obtain channel information. For simplicity, the AI ​​model deployed by the encoder is referred to as the encoder's AI model, and the AI ​​model deployed by the decoder is referred to as the decoder's AI model.

[0105] In addition, in an embodiment of the present application, the encoder and the decoder may be a communication device (such as a terminal device, or a network device) or may be arranged in a communication device. The embodiment of the present application is mainly described by taking the encoder and the decoder as communication devices as an example. The encoder is a terminal device and the decoder is a network device; or the encoder is a terminal device and the decoder is another terminal device; or the encoder is a network device and the decoder is another network device; or the encoder is a network device and the decoder is a terminal device, which is not limited. The embodiment of the present application is mainly described by taking the encoder as a terminal device and the decoder as a network device as an example.

[0106] In addition, in the embodiments of the present application, the functions of the AI ​​model are not limited. As an example, the functions of the AI ​​model include at least one of the following: channel status information (CSI) feedback, CSI prediction, positioning, beam management, or beam prediction.

[0107] 7. Model training: The process of selecting a suitable loss function and using an optimization algorithm to train the model parameters so that the value of the loss function is less than the threshold, or the value of the loss function meets the target requirements.

[0108] 8. Model application: Use the trained model to solve practical problems.

[0109] In wireless communications, channels generally change continuously in time, so communication performance can be improved by mining channel time domain correlation. For example, channel prediction can be achieved by mining channel time domain correlation to combat the problem of channel aging. For another example, channel information compression can be achieved by mining channel time domain correlation to reduce the overhead of feedback channel information. For another example, the optimal beam can be predicted by mining channel time domain correlation. Among them, the mining of channel time domain correlation benefits from the slow-changing characteristics of the channel in the time domain. Taking channel measurement as an example, if the channel change speed is slower than the channel measurement cycle, the channel time domain correlation will be stronger, so that performance gain can be achieved by mining channel time domain correlation; if the channel change speed is faster than the channel measurement cycle, the channel time domain correlation will be weakened, so that performance gain may not be achieved by mining channel time domain correlation.

[0110] Figure 2 This is a schematic diagram of processing channel information through the AI ​​model. Figure 2 In the example shown, the encoder may deploy at least one AI model, so that the encoder may perform encoding based on the deployed AI model, such as processing the channel information input to the encoder. t Indicates the latest channel information obtained by the encoder, e t-1 represents the historical channel information obtained by the encoder (also called the state information of the encoder), e t-1 With {H t-1 ,H t-2 ,H t-3 ,…}. The AI ​​model deployed in the encoder or as an input of the encoder includes e t-1 With H t , the output is c t and e t , where e t The updated encoder state information can be used as the next input of the encoder. If the time interval between the channel acquisition time of the latest channel information and the channel acquisition time of the historical channel information is large, the channel time domain correlation is weakened, and the AI ​​model cannot use the historical channel information to obtain a better model output.

[0111] In view of this, the present application proposes that when using an AI model to process multiple channel information (such as the latest channel information and historical channel information), the time interval between the channel acquisition times corresponding to the channel information should be considered, so that appropriate historical channel information can be selected for processing, thereby avoiding performance loss caused by using outdated historical channel information as model input.

[0112] It should be noted that in this application, "indication" may include direct indication, indirect indication, explicit indication, and implicit indication. When describing that a certain indication information is used to indicate A, it can be understood that the indication information carries A, directly indicates A, or indirectly indicates A.

[0113] In the present application, the information indicated by the indication information is referred to as the information to be indicated. In the specific implementation process, there are many ways to indicate the information to be indicated, such as but not limited to, the information to be indicated can be directly indicated, such as the information to be indicated itself or the index of the information to be indicated. The information to be indicated can also be indirectly indicated by indicating other information, wherein there is an association relationship between the other information and the information to be indicated. It is also possible to indicate only a part of the information to be indicated, while the other parts of the information to be indicated are known or agreed in advance. For example, the indication of specific information can also be achieved by means of the arrangement order of each information agreed in advance (for example, specified by the protocol), thereby reducing the indication overhead to a certain extent. In addition, the information to be indicated can be sent as a whole, or it can be divided into multiple sub-information and sent separately, and the sending period and / or sending time of these sub-information can be the same or different.

[0114] The communication method provided by the embodiment of the present application will be described in detail below with reference to the accompanying drawings. The embodiment provided by the present application can be applied to the above Figure 1A or Figure 1B The communication system shown is not limited.

[0115] Figure 3 It is a schematic diagram of a communication method 300 provided in an embodiment of the present application. For the convenience of description, the following is an exemplary description taking the execution subject of method 300 as a communication device as an example. It can be understood that the communication device can be a terminal device or a component of a terminal device (such as a chip or circuit), or the communication device can also be a network device or a component of a network device (such as a chip or circuit), or the communication device can also be an AI node or a component of an AI node (such as a chip or circuit), without limitation. Figure 3 The illustrated method 300 may include the following steps.

[0116] 301. A communication device determines at least two pieces of channel information, where the at least two pieces of channel information include first channel information and second channel information, and the first channel information is obtained before the second channel information.

[0117] The channel information may be obtained by measuring the channel based on a reference signal. Accordingly, at least two channel information may be obtained by measuring the channel based on at least two reference signals, wherein each channel information corresponds to a reference signal measurement. As an example, the reference signal is a channel status information reference signal (CSI-RS).

[0118] The first channel information is obtained before the second channel information, which means that the first channel information is measured earlier than the second channel information. For example, at a first moment, the first channel information is obtained by measuring the channel based on the reference signal; at a second moment, the second channel information is obtained by measuring the channel based on the reference signal, wherein the first moment is earlier than the second moment. In other words, the first channel information and the second channel information reflect the information of the channel at different times, and the first channel information reflects the information of the channel at an earlier time.

[0119] In one possible scenario, the communication device is a terminal device, that is, the executor of method 300 is a terminal device. In this scenario, optionally, in step 301, the communication device determines at least two channel information, including: the terminal device receives a reference signal from a network device, and obtains at least two channel information by measuring based on the reference signal.

[0120] In another possible scenario, the communication device is a network device, that is, the execution subject of method 300 is a network device. In this case, optionally, in step 301, the communication device determines at least two channel information, including: the network device receives feedback information of at least two channel information from the terminal device, and obtains at least two channel information based on the feedback information of the at least two channel information.

[0121] 302. The communication device determines an input of the first AI model or determines a second AI model according to a time interval between a channel acquisition time corresponding to the first channel information and a channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model.

[0122] In an embodiment of the present application, the communication device determines the input of the first AI model or determines the second AI model according to the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model, so that the input of the first AI model can match the capability requirement of the first AI model, making the input of the first AI model more appropriate, avoiding performance loss caused by using inappropriate channel information as the input of the first AI model. In addition, if the first AI model is determined to be inappropriate according to the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model, the second AI model can also be determined, and then the second AI model can continue to be used for related operations, avoiding performance loss caused by using an inappropriate AI model to process channel information.

[0123] The first AI model may represent the original AI model, and the second AI model may represent the switched AI model, or the second AI model is an AI model whose capability requirements are not completely the same as those of the first AI model, that is, the second AI model has at least one capability requirement different from that of the first AI model. It can be understood that the second AI model may have a variety of structures. In order to distinguish, the embodiments of the present application refer to AI models that have different capability requirements from the first AI model as second AI models.

[0124] As an example, the capability requirement of the AI ​​model is a capability requirement related to the time interval between channel acquisition times supported by the AI ​​model. The capability requirement of the AI ​​model can be used by the communication device to determine whether to switch from the first AI model to the second AI model, or can be used by the communication device to determine the input of the first AI model. Among them, the capability requirement of the AI ​​model can also be referred to as the working condition of the AI ​​model, or the state condition of the AI ​​model, and its naming does not limit the scope of protection of the embodiments of the present application.

[0125] In the embodiments of the present application, the functions of the AI ​​model are not limited. As an example, the functions of the AI ​​model include at least one of the following: channel status information (CSI) feedback (or channel feedback), CSI prediction (or channel prediction), positioning, beam management, or beam prediction.

[0126] As an example, the channel acquisition time corresponding to the channel information may be determined according to the reception time or transmission time of the reference signal used to measure the channel information, or may be determined according to the reporting time of the channel information. Taking the first channel information as an example, assuming that the first channel information is obtained by performing channel measurement on a reference signal (for distinction, referred to as the first reference signal), then the channel acquisition time corresponding to the first channel information may be determined according to the reception time or transmission time of the first reference signal, or the channel acquisition time corresponding to the first channel information may be determined according to the reporting time of the first channel information. For example, the reporting time of the first channel information may be the reporting time of the feedback information of the first channel information. The same applies to the other channel information among at least two channel information, which will not be elaborated here.

[0127] In the first possible implementation manner, the channel acquisition time corresponding to the channel information is equal to the reception time or transmission time of the reference signal used to measure the channel information. By this means, the network device and the terminal device can align the channel acquisition time corresponding to the channel information, so that when the network device and the terminal device determine the input of the first AI model or determine the second AI model according to the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model, the determined results can be consistent, thereby ensuring the symmetry of the behaviors on the network side and the terminal side.

[0128] For example, the terminal device receives the first reference signal from the network device and performs channel measurement based on the first reference signal to obtain the first channel information. The channel acquisition time of the first channel information determined by the terminal device is equal to the time when the terminal device receives the first reference signal. Correspondingly, the channel acquisition time of the first channel information determined by the network device is equal to the time when the network device transmits the first reference signal. Among them, the communication device may be a terminal device or a network device.

[0129] In the above two examples, the transmission time of the reference signal can be ignored.

[0130] Taking the case where the AI model is used for channel prediction as an example, at this time, the input of the AI model is the channel information obtained by channel measurement based on the reference signal. At least two channel information are obtained by at least two reference signal measurements, and each channel information corresponds to one reference signal measurement. Assuming that at least two channel information are {H t-n , H t-n+1 ,..., H t}, where each H m (m = t - n,…, t - 1, t) corresponds to one reference signal measurement. When l < m, the reference signal transmission time corresponding to H m is later than that of H lThe corresponding reference signal sending time, that is, H l In H m Based on the first possible implementation, H m The corresponding channel acquisition time (m=t, ..., tn) is equal to the corresponding reference signal sending time.

[0131] In a second possible implementation, the channel acquisition time corresponding to the channel information is equal to the sum of the receiving time or sending time of the reference signal used to measure the channel information and the first offset. In this way, not only can the network device and the terminal device align the channel acquisition time corresponding to the channel information, but also the transmission time of the reference signal is taken into account. Specifically, considering the transmission time of the signal, the terminal device determines the channel acquisition time corresponding to the channel information, which may be the receiving time of the reference signal used to measure the channel information, and the network device determines the channel acquisition time corresponding to the channel information, which may be the sending time of the reference signal used to measure the channel information plus the first offset. In this way, not only can the network device and the terminal device align the channel acquisition time corresponding to the channel information, but also the transmission time of the reference signal is taken into account. In this way, when the transmission time of the reference signal cannot be ignored, the network device and the terminal device can still align the channel acquisition time corresponding to the channel information by introducing the offset.

[0132] As an example, the value of the first offset may be predefined by a protocol, or pre-stored, or pre-configured. The specific value of the first offset is not limited.

[0133] Take the AI ​​model for channel prediction as an example. At this time, the input of the AI ​​model is the channel information obtained by measuring the channel based on the reference signal. At least two channel information are obtained through at least two reference signal measurements, and each channel information corresponds to one reference signal measurement. Assume that the at least two channel information are {H t-n ,H t-n+1 ,...,H t}, based on the second possible implementation, H m The channel acquisition time corresponding to (m=t, ..., tn) is equal to the sum of the corresponding reference signal sending time and the first offset.

[0134] The second possible implementation method can refer to the description of the first possible implementation method, which is not repeated here.

[0135] It can be understood that the second possible implementation is an example, and the variations belonging to the second possible implementation are also applicable to the protection scope of the present application. For example, the channel acquisition time corresponding to the channel information is equal to the difference between the reception time or the transmission time of the reference signal used to measure the channel information and a certain offset (such as recorded as the second offset). Specifically, the terminal device determines the channel acquisition time corresponding to the channel information, which can be the difference between the reception time of the reference signal used to measure the channel information and the second offset, and the network device determines the channel acquisition time corresponding to the channel information, which can be the transmission time of the reference signal used to measure the channel information. This can also enable the network device and the terminal device to align the channel acquisition time corresponding to the channel information.

[0136] In a third possible implementation, the channel acquisition time corresponding to the channel information is equal to the reporting time of the channel information. Taking the first channel information as an example, the channel acquisition time corresponding to the first channel information is equal to the reporting time of the first channel information. In this way, the network device and the terminal device can align the channel acquisition time corresponding to the channel information, so that when the network device and the terminal device determine the input of the first AI model or determine the second AI model according to the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model, the results can be consistent, thereby ensuring symmetry of the behavior of the network side and the terminal side.

[0137] For example, the terminal device receives a first reference signal from a network device, performs channel measurement based on the first reference signal to obtain first channel information, and reports feedback information of the first channel information to the network device. The channel acquisition time of the first channel information determined by the terminal device is equal to the time when the terminal device sends the feedback information of the first channel information. Accordingly, the channel acquisition time of the first channel information determined by the network device is equal to the time when the network device receives the feedback information of the first channel information. The communication device may be a terminal device, or a network device.

[0138] In the above two examples, the transmission time of the feedback information of the channel information can be ignored.

[0139] Taking the AI ​​model for channel feedback as an example, the input of the AI ​​model includes the latest channel information H t and a related information of the historical channel (t-1) (e (t-1) With {H (t-1) ,H (t-2) ,H (t-3) , …}), where each H m Corresponding to a channel report (the channel report will H mThe information contained in the report is used for network equipment to perform channel recovery). Based on the third possible implementation method, H m The corresponding channel acquisition time is equal to the corresponding channel reporting time.

[0140] In a fourth possible implementation, the channel acquisition time corresponding to the channel information is equal to the sum of the reporting time of the channel information and the third offset. Taking the first channel information as an example, the channel acquisition time corresponding to the first channel information is equal to the sum of the reporting time of the first channel information and the third offset. In this way, not only can the channel acquisition time corresponding to the channel information be aligned with the network device and the terminal device, but also the transmission time of the feedback information is taken into account. Specifically, considering the transmission time of the information, the channel acquisition time corresponding to the channel information determined by the network device is the reception time of the feedback information of the channel information received by the network device, and the channel acquisition time corresponding to the channel information determined by the terminal device is the sending time of the feedback information of the channel information sent by the terminal device plus the third offset. In this way, not only can the channel acquisition time corresponding to the channel information be aligned with the network device and the terminal device, but also the transmission time of the reference signal is taken into account. In this way, when the transmission time of the reference signal cannot be ignored, the network device and the terminal device can still align the channel acquisition time corresponding to the channel information by introducing the offset.

[0141] As an example, the value of the third offset may be predefined by a protocol, or pre-stored, or pre-configured. The specific value of the third offset is not limited.

[0142] Taking the AI ​​model for channel feedback as an example, the input of the AI ​​model includes the latest channel information H t and a related information of the historical channel (t-1) (e (t-1) With {H (t-1) ,H (t-2) ,H (t-3) , …}), where each H m Corresponding to a channel report (the channel report will H m The information contained in the report is used for network equipment to perform channel recovery). Based on the fourth possible implementation method, H m The corresponding channel acquisition time is equal to the sum of the corresponding channel reporting time and the third offset.

[0143] The fourth possible implementation method may refer to the description in the third possible implementation method, which will not be described in detail here.

[0144] It can be understood that the fourth possible implementation is an example, and the variations belonging to the fourth possible implementation are also applicable to the protection scope of the present application. For example, the channel acquisition time corresponding to the channel information is equal to the difference between the reporting time of the channel information and a certain offset (such as recorded as the fourth offset). Specifically, the channel acquisition time corresponding to the channel information determined by the network device is the difference between the reception time of the feedback information of the channel information received by the network device and the fourth offset, and the channel acquisition time corresponding to the channel information determined by the terminal device is the sending time of the feedback information of the channel information sent by the terminal device.

[0145] The above several possible implementations are described by way of example and are not limited thereto. For example, considering that the reference signal may be sent periodically, the channel acquisition time corresponding to the channel information may also be determined based on the period of the reference signal. This will be described in detail later.

[0146] The solution of step 302 is introduced below by taking the communication device as a terminal device or a network device as an example.

[0147] Scenario 1: The communication device is a terminal device.

[0148] In this case, optionally, step 302 includes any one of the following rules.

[0149] Rule #1: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, it determines a second AI model, and the input of the second AI model includes at least two channel information.

[0150] Based on this rule, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, the first AI model can be deactivated and the second AI model can be determined. In this way, when it is determined that the first AI model is inappropriate, the second AI model can be used to perform related operations, thereby avoiding performance loss caused by processing channel information with an inappropriate AI model.

[0151] There is no limitation on the method of determining the second AI model. Several possible implementation methods are introduced below.

[0152] One possible implementation method is to select any one of the AI ​​models deployed in the communication device.

[0153] Another possible implementation manner may be to select an AI model from the AI ​​models deployed by the communication device according to the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information. Specifically, the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the first capability requirement of the AI ​​model.

[0154] Another possible implementation manner is to select an AI model from the AI ​​models deployed by the communication device according to the first capability requirement of the first AI model. Specifically, the first capability requirement of the selected AI model is lower than the first capability requirement of the first AI model.

[0155] The above-mentioned possible implementation methods are described below as examples, and the embodiments of the present application are not limited thereto.

[0156] In the embodiment of the present application, the first capability requirement of the AI ​​model can be understood as: the minimum requirement for using the AI ​​model, that is, if this requirement is not met, the AI ​​model cannot be used.

[0157] Optionally, the first capability requirement of the first AI model is: the time interval between channel acquisition times is within a first numerical range. The time interval between channel acquisition times may be the time interval between channel acquisition times corresponding to any two channel information. The first numerical range may be one numerical range or multiple numerical ranges, without limitation. Furthermore, the higher the first capability requirement of the AI ​​model, the larger the first numerical range. For example, if the first capability requirement of the first AI model is lower than the first capability requirement of the second AI model, it means that the first numerical range corresponding to the second AI model is larger than the first numerical range corresponding to the first AI model.

[0158] In a possible implementation, the first numerical range is a numerical range, and the first numerical range is less than or equal to T1. Based on this, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than T1, the communication device determines the second AI model. It can be understood that the first numerical range can also be less than T1. At this time, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than or equal to T1, the communication device determines the second AI model.

[0159] Among them, T1 is the maximum value of the time interval between channel acquisition times corresponding to the first capability requirement of the first AI model.

[0160] In one example, T1 represents the maximum time interval between channel acquisition times corresponding to any two channel information. In this example, the first channel information and the second channel information may be any two channel information of the at least two channel information.

[0161] In another example, T1 represents the maximum time interval between the channel acquisition time corresponding to the earliest channel information and the channel acquisition time corresponding to the latest channel information. In this example, the first channel information may be the earliest obtained channel information of at least two channel information, and the second channel information may be the latest obtained channel information of at least two channel information.

[0162] In another example, T1 represents the maximum time interval between the channel acquisition times corresponding to two channel information separated by k channel information, and k is an integer greater than 1 or equal to 1. Assume that the channel information in the at least two channel information is arranged in sequence according to the acquisition time. In this example, the first channel information and the second channel information are separated by k channel information. For example, the input of the first AI model is n channel information, denoted as {H t-n ,H t-n+1 ,...,H t-1}, then for example, the first channel information is H t-1 , the second channel information is H t-k-1 ; or the first channel information is H t-m , the second channel information is H t-k-m .

[0163] Optionally, T1 is related to at least one of the following: model structure, training data, training method, model function. For example, in the model verification stage, channel data at different time intervals can be used to test the model, and then the value of T1 can be determined. In this way, after the model training is completed, the T1 corresponding to the model can also be determined, so that when the communication device deploys the model, the value of T1 can also be obtained at the same time.

[0164] In another possible implementation, the first numerical range is a plurality of numerical ranges, such as the first numerical range is: less than or equal to T11, greater than or equal to T12 (T11>T12). Based on this, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than T11 or less than T12, the communication device determines the second AI model. It can be understood that the first numerical range can also be less than T11 and greater than T12. At this time, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than or equal to T11, or, less than or equal to T12, the communication device determines the second AI model.

[0165] Based on rule #1, further optionally, when the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the first capability requirement of the first AI model, the AI ​​model is not switched, that is, the first AI model is determined to be used to process part or all of the at least two channel information. Specifically, the input of the first AI model includes part or all of the at least two channel information. Specifically, what the input of the first AI model is, for example, can be determined according to other rules and is not limited.

[0166] Rule #2: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, it determines a second AI model, and the input of the second AI model includes the second channel information, and the status information of the second AI model is the default value.

[0167] Based on this rule, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, the first AI model can be deactivated and the second AI model can be determined. In this way, when it is determined that the first AI model is inappropriate, the second AI model can be used to perform related operations, thereby avoiding performance loss caused by processing channel information with an inappropriate AI model.

[0168] Regarding the method of determining the second AI model, please refer to the description in Rule #1, which will not be repeated here.

[0169] In an embodiment of the present application, the state information of the AI ​​model is an input of the AI ​​model. For example, the input of the AI ​​model includes channel information and the state information of the AI ​​model, and the output of the AI ​​model is the feedback information of the channel information and the updated state information of the AI ​​model, which serves as the next input of the AI ​​model. Alternatively, the input of the AI ​​model includes the feedback information of the channel information and the state information of the AI ​​model, and the output of the AI ​​model is the restored channel information and the updated state information of the AI ​​model, which serves as the next input of the AI ​​model. Figure 2 For example, the e input to the encoder t-1 It can be understood as the status information of the AI ​​model deployed in the encoder, e t-1 is the previous output of the AI ​​model, so it is related to the channel information input to the AI ​​model last time, so e t-1 The information is related to the historical channel information. When the AI ​​model is enabled and runs for the first time, the status information of the AI ​​model can take the default value.

[0170] The default value may be, for example, 0 or other values, which are not limited. The default value will not be described in detail below.

[0171] Based on rule #2, further optionally, when the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the first capability requirement of the first AI model, it is determined that the input of the first AI model includes the second channel information and the status information of the first AI model, and the status information of the first AI model is determined by the previous output of the first AI model.

[0172] Rule #3: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes M channel information and X default values ​​from at least two channel information, and the M channel information does not include the first channel information.

[0173] Based on this rule, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, the first AI model can continue to be used to process the channel information, but the processed channel information at least does not include the first channel information, that is, the input of the first AI model at least does not include the first channel information, so that inappropriate channel information can be removed, avoiding performance loss caused by using inappropriate channel information as the input of the first AI model.

[0174] Wherein, M is an integer greater than 1 or equal to 1, and X is an integer greater than 0 or equal to 0.

[0175] Case #1, X=0. At this time, based on rule #3, when the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, the input of the first AI model includes M channel information. For example, assuming that at least two channel information are n channel information, denoted as {H t-n ,H t-n+1 ,...,H t-1}, the first channel information is H t-k , the second channel information is H t-1 , if H t-k The corresponding channel acquisition time and H t-1 If the time interval between the corresponding channel acquisition times does not meet the second capability requirement of the first AI model, the input of the first AI model does not include H t-k , that is, the input of the first AI model is {H t-n ,...,H t-k-1 ,Ht-k+1 ,...,H t-1}. It is understandable that H t-k The previous channel information and the second channel information are judged. If it is judged that H t-k If the time interval between the previous channel information and the second channel information does not meet the second capability requirement of the first AI model, then the input of the first AI model does not include H t-k Previous channel information.

[0176] In case #2, X>0, based on rule #3, when the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, the input of the first AI model includes M channel information and at least one default value. The number of default values ​​is not limited, and two possible implementations are described below.

[0177] In a first possible implementation, the number of default values ​​depends on how many channel information the time interval between channel acquisition times does not meet the second capability requirement of the first AI model. For example, assuming that at least two channel information are n channel information, denoted as {H t-n ,H t-n+1 ,...,H t-1}, the first channel information is H t-n , the second channel information is H t-1 , the third channel information is H t-n+1 , if H t-n The corresponding channel acquisition time and H t-1 The time interval between the corresponding channel acquisition times does not meet the second capability requirement of the first AI model, and H t-n+1 The corresponding channel acquisition time and H t-1 The time interval between the corresponding channel acquisition times also does not meet the second capability requirement of the first AI model, then the input of the first AI model includes two default values ​​(for distinction, the two default values ​​are recorded as ), and does not include H t-n and H t-n+1 , that is, the input of the first AI model is

[0178] In a second possible implementation, the number of default values ​​is always 1. For example, assuming that at least two channel information is n channel information, denoted as {H t-n ,H t-n+1 ,...,H t-1}, the first channel information is H t-n , the second channel information is H t-1 , if H t-n The corresponding channel acquisition time and Ht-1 If the time interval between the corresponding channel acquisition times does not meet the second capability requirement of the first AI model, the input of the first AI model includes a default value (for distinction, the default value is recorded as ), and does not include H t-n , that is, the input of the first AI model is For another example, assuming that the first channel information is H t-n , the second channel information is H t-1 , the third channel information is H t-n+1 , if H t-n The corresponding channel acquisition time and H t-1 The time interval between the corresponding channel acquisition times does not meet the second capability requirement of the first AI model, and H t-n+1 The corresponding channel acquisition time and H t-1 The time interval between the corresponding channel acquisition times also does not meet the second capability requirement of the first AI model, then the input of the first AI model includes a default value And does not include H t-n and H t-n+1 , that is, the input of the first AI model is

[0179] In an embodiment of the present application, the second capability requirement of the AI ​​model can be used to determine the input of the AI ​​model. For example, if the time interval between the channel acquisition times corresponding to two channel information meets the second capability requirement of the AI ​​model, the input of the AI ​​model includes the two channel information; if the time interval between the channel acquisition times corresponding to the two channel information does not meet the second capability requirement of the AI ​​model, the input of the AI ​​model does not include the channel information obtained earlier of the two channel information.

[0180] Optionally, the second capability requirement of the first AI model is: the time interval between channel acquisition times is within a second numerical range. The time interval between channel acquisition times may be the time interval between channel acquisition times corresponding to any two channel information. The second numerical range may be one numerical range or multiple numerical ranges, without limitation.

[0181] In a possible implementation, the second numerical range is a numerical range, such as the second numerical range is less than or equal to T2. Based on this, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than T2, the communication device determines that the input of the first AI model includes M channel information and X default values ​​in at least two channel information. It can be understood that the second numerical range can also be less than T2. ​​In this case, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than or equal to T2, the communication device determines that the input of the first AI model includes M channel information and X default values ​​in at least two channel information.

[0182] Wherein, T2 is the maximum value of the time interval between the channel acquisition times corresponding to the second capability requirement of the first AI model. For T2, reference can be made to the relevant description of T1, which will not be repeated here.

[0183] In another possible implementation, the second numerical range is a plurality of numerical ranges, such as the second numerical range is: less than or equal to T21, greater than or equal to T22 (T21>T22). Based on this, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than T21 or less than T22, the communication device determines that the input of the first AI model includes M channel information and X default values ​​in at least two channel information. It can be understood that the second numerical range can also be less than T21 and greater than T22. At this time, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is greater than or equal to T21, or, less than or equal to T22, the communication device determines that the input of the first AI model includes M channel information and X default values ​​in at least two channel information.

[0184] Further optionally, the M channel information does not include the channel information obtained before the first channel information. Considering that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, if the time interval between the channel acquisition time corresponding to the first channel information and the second channel information is greater than T2, then the time interval between the channel information obtained before the first channel information (for distinction, recorded as channel information #1) and the channel acquisition time corresponding to the second channel information is also greater than T2, so it can be directly determined that the time interval between the channel acquisition time corresponding to the channel information #1 and the second channel information is also not within the second numerical range, and the overhead caused by determining whether the time interval between the channel acquisition time corresponding to the channel information #1 and the second channel information is greater than T2 can be subtracted.

[0185] For example, assuming that at least two channel information are n channel information, denoted as {H t-n ,H t-n+1 ,...,H t-1}, the first channel information is H t-1-k , the second channel information is H t-1 , if H t-1-k The corresponding channel acquisition time and H t-1 The time interval between the corresponding channel acquisition times is greater than T2, then H t-1-k Previous channel information ({H t-n ,H t-n+1 ,...,H t-k-2}) The corresponding channel acquisition time and H t-1 The time interval between the corresponding channel acquisition times is also greater than T2, so the input of the first AI model does not include {H t-n ,H t-n+1 ,...,H t-k-2 ,H t-1-k Taking the above situation #1 as an example, the input of the first AI model is {H t-k ,H t-k+1 ,...,H t-1}; or, taking the first possible implementation in the above scenario #2 as an example, the input of the first AI model is Alternatively, taking the second possible implementation in the above scenario #2 as an example, the input of the first AI model is

[0186] Further optionally, the M channel information does not include the channel information obtained before the second channel information. In this way, considering that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, such as the time interval between the channel acquisition time corresponding to the first channel information and the second channel information is greater than T2, it can be determined that the input of the first AI model does not include the channel information obtained before the second channel information. In this way, for the second channel information, if there is channel information whose time interval between the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it can be considered that there is little necessity to use the channel information before the second channel information, or it is difficult to achieve performance gain by mining the channel time domain correlation between the second channel information and the channel information before the second channel information, so the input of the first AI model may not include the channel information obtained before the second channel information.

[0187] Based on rule #3, further optionally, when the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the second capability requirement of the first AI model, it is determined that the input of the first AI model includes the at least two channel information.

[0188] Rule #4: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it determines that the input of the first AI model includes the second channel information, and the status information of the first AI model is the default value.

[0189] Based on this rule, if the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, the first AI model can continue to be used to process the second channel information, and the state information of the first AI model is a default value instead of being determined by the previous output of the first AI model. In this way, the state information of the first AI model can be avoided from being related to historical channel information, so as to avoid the impact of state information related to inappropriate channel information and improve the output performance of the first AI model.

[0190] Based on Rule #4, further optionally, when the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the second capability requirement of the first AI model, it is determined that the input of the first AI model includes the second channel information and the status information of the first AI model, and the status information of the first AI model is determined by the previous output of the first AI model.

[0191] The above rules #1 to #4 are for illustration only, and any variation of the above rules are applicable to the embodiments of the present application.

[0192] The above rules #1 to #4 can be used alone or in combination without limitation. For example, the above rules #1 and #3 are used in combination. If the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information neither meets the first capability requirement of the first AI model nor the second capability requirement of the first AI model, then one possible implementation method may be to use a rule with a higher priority according to the priority, or one possible method is to use all rules at the same time, or one possible implementation method is to randomly select a rule to use. For example, assuming that the priority of rule #1 is higher than the priority of rule #2, then if the time interval between the channel acquisition times neither meets the first capability requirement of the first AI model nor the second capability requirement of the first AI model, then it is processed in accordance with rule #1, that is, the second AI model is determined, and the input of the second AI model includes at least two channel information.

[0193] In addition, optionally, different rules can be used for AI models with different functions or different structures. In other words, different rules can be divided based on the function or structure of the AI ​​model. As an example, there are two types of rules. For the purpose of distinction, these two types of rules are recorded as first-type rules and second-type rules, where the first-type rules are related to the state information of the AI ​​model, and the second-type rules are related to the input of the AI ​​model. In one possible implementation, the AI ​​model used for channel feedback corresponds to the first-type rules, and the AI ​​model used for channel prediction corresponds to the second-type rules. Taking the above-mentioned rules #1-rule #4 as an example, rules #2 and rule #4 belong to the first-type rules, and rules #1 and rule #3 belong to the second-type rules.

[0194] Scenario 2: The communication device is a network device.

[0195] In this case, optionally, step 302 includes any of the following solutions.

[0196] Rule #5: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, it determines a second AI model, and the input of the second AI model includes feedback information of at least two channel information.

[0197] In an embodiment of the present application, the feedback information of the channel information may be feedback information output by an encoder on the terminal device side, or may be information obtained by processing the output of the encoder, such as a coarse-precision channel restored by a decoder.

[0198] Rule #5 can refer to Rule #1 above and will not be repeated here.

[0199] Rule #6: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, it determines a second AI model, and the input of the second AI model includes feedback information of the second channel information, and the status information of the second AI model is the default value.

[0200] Rule #6 can refer to Rule #2 above and will not be repeated here.

[0201] Rule #7: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it determines that the input of the first AI model includes feedback information of M channel information of at least two channel information and X default values, where the M channel information does not include the first channel information, M is an integer greater than or equal to 1, and X is an integer greater than or equal to 0.

[0202] Rule #7 can refer to Rule #3 above and will not be repeated here.

[0203] Rule #8: When the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it determines that the input of the first AI model includes feedback information of the second channel information, and the status information of the first AI model is the default value.

[0204] Rule #8 can refer to Rule #4 above and will not be repeated here.

[0205] The scheme of step 302 is introduced above in combination with the two situations that the communication device is a terminal device and a network device. It can be understood that the above is an example description, and any variation of the above rules is applicable to the embodiments of the present application. For example, taking the communication device as a terminal device as an example, when the communication device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, the second AI model is determined, and the input of the second AI model includes M channel information and X default values ​​from at least two channel information, the M channel information does not include the first channel information, M is an integer greater than or equal to 1, and X is an integer greater than or equal to 0. When the communication device is a network device, it is similar and will not be repeated here.

[0206] The following describes how to obtain the rules.

[0207] In a possible implementation, the rules are predefined by the protocol, or pre-stored, or pre-configured.

[0208] Based on this implementation, the communication device can directly use the predefined rules for judgment. For example, assuming that the predefined rule #1 is used for judgment, the communication device can directly determine the input of the second AI model or the first AI model based on the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and the predefined first capability requirement of the first AI model.

[0209] In another possible implementation, the communication device receives indication information (recorded as indication information #1 for distinction), where the indication information #1 indicates usage rules of the first AI model.

[0210] For example, the communication device is a terminal device. In this case, the communication device receiving indication information #1 may include: the terminal device receiving indication information #1 from the network device; or the terminal device receiving indication information #1 from other devices (such as AI nodes, or other terminal devices). For another example, the communication device is a network device. In this case, the communication device receiving indication information #1 may include: the network device receiving indication information #1 from the terminal device; or the network device receiving indication information #1 from other devices (such as AI nodes, or other network devices).

[0211] In one example, the indication information #1 directly indicates the usage rules of the first AI model. For example, the indication information #1 is the usage rules of the first AI model. In another example, the indication information #1 is the identifier or number of the usage rules of the first AI model.

[0212] In another example, indication information #1 indirectly indicates the usage rules of the first AI model. For example, indication information #1 is other information (referred to as information #A for distinction), and the information #A has an association relationship with the rules. In this way, the usage rules of the first AI model can be determined based on the information #A indicated by indication information #1 and the association relationship between information #A and the rules. The association relationship between information #A and the rules can be predefined by the protocol, or pre-stored, or pre-configured, or indicated by other communication devices.

[0213] In a possible implementation, information #A is the value of a parameter involved in the rule.

[0214] For example, assuming that the first capability requirement of the first AI model is that the time interval between the channel acquisition times corresponding to any two channel information is less than or equal to T1, the value of T1 is associated with the rule. For example, the value of T1 can be T11 and T12. If the information #A indicated by the indication information #1 is T11, it means that the usage rule of the first AI model is rule #1; if the information #A indicated by the indication information #1 is T12, it means that the usage rule of the first AI model is rule #2.

[0215] Another possible implementation method is that information #A is the function of the AI ​​model.

[0216] For example, assuming that channel prediction corresponds to rule #1 and channel feedback corresponds to rule #2. For example, if information #A indicates that the function of the AI ​​model is channel feedback, the use rule of the AI ​​model is rule #2; if information #A indicates that the function of the AI ​​model is channel prediction, the use rule of the AI ​​model is rule #1.

[0217] For another example, assume that channel feedback corresponds to the first type of rules (the first type of rules include, for example, rules #2 and rules #4), and channel prediction corresponds to the second type of rules (the second type of rules include, for example, rules #1 and rules #3). For example, if information #A indicates that the function of the AI ​​model is channel feedback, then the usage rules of the AI ​​model are the first type of rules; if information #A indicates that the function of the AI ​​model is channel prediction, then the usage rules of the AI ​​model are the second type of rules. Among them, the usage rules of the AI ​​model can be one of the first type of rules or the second type of rules, or it can also be multiple rules of the first type of rules or the second type of rules. Two examples are listed below.

[0218] As an example, the usage rule of the AI ​​model may be a rule in the first category of rules or a rule in the second category of rules. In one possible implementation, information #A' may be used to indicate which specific usage rule of the AI ​​model is. For example, information #A indicates that the function of the AI ​​model is channel feedback, and information #A' indicates that the usage rule of the AI ​​model is rule #2 or rule #4. Based on information #A and information #A', it can be determined that the usage rule of the AI ​​model is rule #2 or rule #4 in the first category of rules; information #A indicates that the function of the AI ​​model is channel prediction, and information #A' indicates that the usage rule of the AI ​​model is rule #1 or rule #3. Based on information #A and information #A', it can be determined that the usage rule of the AI ​​model is rule #1 or rule #3 in the second category of rules. In another possible implementation, a rule with a specific number may be used by default or predefined. For example, if information #A indicates that the function of the AI ​​model is channel feedback, then the usage rules of the AI ​​model are the rules with specific numbers in the first category of rules (such as the rule with the smallest number or the rule with the largest number); if information #A indicates that the function of the AI ​​model is channel prediction, then the usage rules of the AI ​​model are the rules with specific numbers in the second category of rules (such as the rule with the smallest number or the rule with the largest number).

[0219] In another example, the usage rules of the AI ​​model can be multiple rules (such as all rules) in the first category of rules or the second category of rules. In this example, when judging based on multiple rules, if multiple rules are satisfied at the same time (such as simultaneously satisfying rule #2 and rule #4 in the first category of rules), you can choose any rule, or choose a rule with a higher priority based on the priority. For details, please refer to the previous description, which will not be repeated here.

[0220] It can be understood that the functions of the above-mentioned AI model can also be replaced by the structure of the AI ​​model.

[0221] It can be understood that the above is an example and is not limited to this. For example, indication information #1 is implemented by at least one bit, and the usage rules of the first AI model are determined by the value of indication information #1. For example, suppose there are two rules, such as rule #1 and rule #3 described above, which are indicated by 1 bit: the usage rules of the first AI model. If the bit is set to "0", it means that the usage rules of the first AI model are rule #1; if the bit is set to "1", it means that the usage rules of the first AI model are rule #3. It should be understood that the above is only an exemplary description and is not limiting.

[0222] Optionally, parameters involved in the rules, such as the capability requirements of the AI ​​model and the numerical range corresponding to the capability requirements of the AI ​​model, can be obtained in any of the following ways.

[0223] In a possible implementation manner, the parameters involved in the rule are predefined by the protocol, or pre-stored, or pre-configured.

[0224] Based on this implementation, after the rule is known, the parameters involved in the rule can also be determined. Regarding the way in which the communication device obtains the rule, reference can be made to the above related description, which will not be repeated here.

[0225] Another possible implementation manner is to receive indication information (recorded as indication information #2 for distinction), where the indication information #2 indicates parameters involved in the rule.

[0226] For example, the communication device is a terminal device. In this case, the communication device receiving indication information #2 may include: the terminal device receiving indication information #2 from the network device; or the terminal device receiving indication information #2 from other devices (such as AI nodes, or other terminal devices). For another example, the communication device is a network device. In this case, the communication device receiving indication information #2 may include: the network device receiving indication information #2 from the terminal device; or the network device receiving indication information #2 from other devices (such as AI nodes, or other network devices).

[0227] In one example, the indication information #2 directly indicates the parameters involved in the rule. For example, the indication information #2 is the parameters involved in the rule.

[0228] In another example, indication information #2 indirectly indicates the parameters involved in the rule. For example, indication information #2 is other information (referred to as information #B for distinction), and the information #B has an association relationship with the parameters involved in the rule. In this way, the parameters involved in the rule can be determined based on the information #B indicated by indication information #2 and the association relationship between information #B and the parameters involved in the rule. The association relationship between information #B and the parameters involved in the rule can be predefined by the protocol, or pre-stored, or pre-configured, or indicated by other communication devices.

[0229] When the above rule is used for judgment, the channel information used may be the latest channel information and any historical channel information, or the latest channel information and channel information separated by k channel information, or any two channel information, or any two channel information separated by k channel information, where k is an integer greater than 1 or equal to 1.

[0230] In a first possible implementation manner, the first channel information is any one of the at least two channel information except the second channel information.

[0231] For example, the second channel information is the latest channel information obtained from the at least two channel information, so that the latest channel information can be compared with the channel acquisition time corresponding to each historical channel information.

[0232] For example, at least two channel information are {H t-n ,H t-n+1 ,...,H t}, the second channel information is the latest channel information H t , then we can t With {H t-n ,H t-n+1 ,...,H t-1 The time interval between each channel information in} is determined according to the above rules.

[0233] For another example, suppose the original input of the first AI model is H t With e t-1 , where e t-1 With {H t-1 ,H t-2 ,H t-3 ,...}, the second channel information is the latest channel information H t , then we can t With {Ht-n ,H t-n+1 ,...,H t-1 The time interval between the channel acquisition times corresponding to each channel information in} is determined according to the above rules.

[0234] In a second possible implementation manner, the first channel information and the second channel information are separated by k channel information.

[0235] For example, at least two channel information are {H t-n ,H t-n+1 ,...,H t}, the second channel information is the latest channel information H t , then we can t With H t-k The time interval between the corresponding channel acquisition times is determined according to the above rules.

[0236] Among them, the method for obtaining k is not limited. For example, k is predefined by the protocol, or pre-stored, or pre-configured. For another example, the communication device receives indication information (for distinction, recorded as indication information #3), and the indication information #3 indicates k. For example, the communication device is a terminal device. In this case, the communication device receives indication information #3, which may include: the terminal device receives indication information #3 from the network device; or, the terminal device receives indication information #3 from other devices (such as AI nodes, or other terminal devices). For another example, the communication device is a network device. In this case, the communication device receives indication information #3, which may include: the network device receives indication information #3 from the terminal device; or, the network device receives indication information #3 from other devices (such as AI nodes, or other network devices).

[0237] The above two implementations are examples, and variations of the above implementations are also applicable to the embodiments of the present application. For example, the time interval between the channel acquisition time corresponding to a certain channel information and any other channel information can be judged according to the above rules; for another example, the time interval between the channel acquisition times corresponding to specific two of at least two channel information (such as the earliest obtained channel information and the latest obtained channel information) can be judged according to the above rules.

[0238] Generally speaking, the time domain behavior of the reference signal can be periodic, semi-static, or non-periodic. The time domain behavior of the reference signal is periodic, indicating that the network device sends the reference signal to the terminal device periodically. The time domain behavior of the reference signal is non-periodic, indicating that the network device sends the reference signal to the terminal device non-periodic. The time domain behavior of the reference signal is semi-static, indicating that the network device sends the reference signal to the terminal device semi-statically. Specifically, after the network device configures the information of the semi-static reference signal, it will not send the reference signal immediately. The network device will first send an activation signaling to notify the terminal device. After the activation signaling takes effect, the network device will periodically send the reference signal. Until the network device sends a deactivation signaling, it stops sending the reference signal. The following mainly takes the communication device as a terminal device, combined with the time domain behavior of the reference signal as periodic or semi-static and the different functions of the AI ​​model as an example for exemplary explanation.

[0239] Scenario 1: The AI ​​model is used for channel prediction.

[0240] In this scenario, the input of the AI ​​model is the channel information obtained by channel detection based on the reference signal. As described above, in this scenario, the channel acquisition time corresponding to the channel information is equal to the sending time of the reference signal corresponding to the channel information, or the channel acquisition time corresponding to the channel information is equal to the sum of the sending time of the reference signal corresponding to the channel information and the first offset.

[0241] 1. The time domain behavior of the reference signal is periodic

[0242] Assume that the period of the reference signal is Y, and the first channel information and the second channel information are separated by k channel information.

[0243] As an example, the above rule #1 can be replaced by: when the communication device determines that k*Y does not meet the first capability requirement of the first AI model, such as k*Y is greater than T1, a second AI model is determined, and the input of the second AI model includes at least two channel information.

[0244] As an example, the above rule #3 can be replaced by: when the communication device determines that k*Y does not meet the second capability requirement of the first AI model, such as k*Y is greater than T2, it is determined that the input of the first AI model includes M channel information and X default values ​​of at least two channel information, and the M channel information does not include the first channel information.

[0245] As an example, the above rule #5 can be replaced by: when the communication device determines that k*Y does not meet the first capability requirement of the first AI model, such as k*Y is greater than T1, a second AI model is determined, and the input of the second AI model includes feedback information of at least two channel information.

[0246] As an example, the above rule #7 can be replaced by: when the communication device determines that k*Y does not meet the second capability requirement of the first AI model, such as k*Y is greater than T2, it is determined that the input of the first AI model includes feedback information of M channel information out of at least two channel information and X default values, and the M channel information does not include the first channel information.

[0247] 2. The time domain behavior of the reference signal is semi-static

[0248] As an example, the above rule #1 can be replaced by: when the communication device determines that the time interval between the last deactivation and this activation does not meet the first capability requirement of the first AI model, such as the time interval between the last deactivation and this activation is greater than T1, a second AI model is determined, and the input of the second AI model includes at least two channel information.

[0249] As an example, the above rule #3 can be replaced as follows: when the communication device determines that the time interval between the last deactivation and this activation does not meet the second capability requirement of the first AI model, such as the time interval between the last deactivation and this activation is greater than T2, it is determined that the input of the first AI model includes M channel information and X default values ​​of at least two channel information, and the M channel information does not include the first channel information.

[0250] As an example, the above rule #5 can be replaced by: when the communication device determines that the time interval between the last deactivation and the current activation does not meet the first capability requirement of the first AI model, such as the time interval between the last deactivation and the current activation is greater than T1, a second AI model is determined, and the input of the second AI model includes feedback information of at least two channel information.

[0251] As an example, the above rule #7 can be replaced by: when the communication device determines that the time interval between the last deactivation and the current activation does not meet the second capability requirement of the first AI model, such as the time interval between the last deactivation and the current activation is greater than T2, it is determined that the input of the first AI model includes feedback information of M channel information out of at least two channel information and X default values, and the M channel information does not include the first channel information.

[0252] Scenario 2: The AI ​​model is used for channel feedback.

[0253] In this scenario, the input of the AI ​​model includes channel information obtained by channel detection based on the reference signal, and the status information of the AI ​​model. Specifically, at least two channel information are obtained through at least two reference signal measurements, and each channel information corresponds to one reference signal measurement. As described above, in this scenario, the channel acquisition time corresponding to the channel information is equal to the corresponding channel reporting time, or the channel acquisition time corresponding to the channel information is equal to the sum of the corresponding channel reporting time and the third bias, or the channel acquisition time corresponding to the channel information is equal to the sending time of the reference signal corresponding to the channel information, or the channel acquisition time corresponding to the channel information is equal to the sum of the sending time of the reference signal corresponding to the channel information and the first bias.

[0254] 1. The time domain behavior of the reference signal is periodic

[0255] Assume that the period of the reference signal is Y, and the first channel information and the second channel information are separated by k channel information.

[0256] As an example, the above rule #2 can be replaced as follows: when the communication device determines that k*Y does not meet the first capability requirement of the first AI model, such as k*Y is greater than T1, a second AI model is determined, and the input of the second AI model includes second channel information, and the status information of the second AI model is a default value.

[0257] As an example, the above rule #4 can be replaced by: when the communication device determines that k*Y does not meet the second capability requirement of the first AI model, such as k*Y is greater than T2, it is determined that the input of the first AI model includes the second channel information, and the state information of the first AI model is the default value.

[0258] As an example, the above rule #6 can be replaced as follows: when the communication device determines that k*Y does not meet the first capability requirement of the first AI model, such as k*Y is greater than T1, a second AI model is determined, and the input of the second AI model includes feedback information of the second channel information, and the status information of the second AI model is a default value.

[0259] As an example, the above rule #8 can be replaced as follows: when the communication device determines that k*Y does not meet the second capability requirement of the first AI model, such as k*Y is greater than T2, it is determined that the input of the first AI model includes feedback information of the second channel information, and the state information of the first AI model is the default value.

[0260] 2. The time domain behavior of the reference signal is semi-static

[0261] As an example, the above rule #2 can be replaced as follows: when the communication device determines that the time interval between the last deactivation and this activation does not meet the first capability requirement of the first AI model, such as the time interval between the last deactivation and this activation is greater than T1, a second AI model is determined, and the input of the second AI model includes second channel information, and the status information of the second AI model is the default value.

[0262] As an example, the above rule #4 can be replaced as follows: when the communication device determines that the time interval between the last deactivation and this activation does not meet the second capability requirement of the first AI model, such as the time interval between the last deactivation and this activation is greater than T2, it is determined that the input of the first AI model includes the second channel information, and the status information of the first AI model is the default value.

[0263] As an example, the above rule #6 can be replaced as follows: when the communication device determines that the time interval between the last deactivation and this activation does not meet the first capability requirement of the first AI model, such as the time interval between the last deactivation and this activation is greater than T1, a second AI model is determined, and the input of the second AI model includes feedback information of the second channel information, and the status information of the second AI model is the default value.

[0264] As an example, the above rule #8 can be replaced as follows: when the communication device determines that the time interval between the last deactivation and this activation does not meet the second capability requirement of the first AI model, such as the time interval between the last deactivation and this activation is greater than T2, it is determined that the input of the first AI model includes feedback information of the second channel information, and the status information of the first AI model is the default value.

[0265] The above description is for illustrative purposes only and is not intended to be limiting.

[0266] In addition, the embodiment of the present application also proposes that, for a semi-static reference signal, the state information of the first AI model may be reset to a default value in response to activation by default, or other AI models (such as a second AI model) different from the first AI model may be used by default to perform related operations. For example, for a semi-static reference signal, each time a semi-static report is triggered, the network device may instruct the terminal device to reset the state information of the first AI model to the default value before the first report in this time period. (t-1) The network device instructs the terminal device to select another AI model different from the first AI model in this time period, or the terminal device defaults to the first AI model status information e before the first report in this time period. (t-1) The default value is set, or the terminal device selects another AI model different from the first AI model by default in this time period. In this case, as an example, it is no longer necessary to determine the time interval between the channel acquisition times corresponding to the channel information.

[0267] For ease of understanding, the following takes the interaction between the terminal device and the network device as an example, assuming that the encoder is deployed on the terminal device side and the decoder is deployed on the network device side, and an exemplary description is given in combination with different scenarios. Contents not described in detail below can refer to the description in method 300, and will not be repeated below.

[0268] Figure 4 4 is a schematic flow chart of a communication method 400 provided according to an embodiment of the present application. The AI ​​model in the method 400 can be used for channel feedback. Figure 4 The illustrated method 400 may include the following steps.

[0269] Optionally, method 400 includes step 401: the network device sends usage rules of the first AI model to the terminal device.

[0270] In a possible implementation, the network device sends indication information #1 to the terminal device, where the indication information #1 indicates the usage rules of the first AI model. For the implementation of indication information #1, reference may be made to the relevant description in method 300, which will not be described in detail here.

[0271] It can be understood that the usage rules of the network device sending the first AI model to the terminal device are exemplary descriptions, such as Figure 3 As described in the embodiment, the usage rules of the first AI model may also be predefined by the protocol, or pre-stored, or pre-configured, so that the terminal device may also determine the usage rules of the first AI model by itself.

[0272] Among them, the usage rules of the first AI model can be at least one of the above rules #1 to rule #8.

[0273] Optionally, the method 400 further includes step 402: the network device sends configuration information of a reference signal to the terminal device.

[0274] As mentioned above, the time domain behavior of the reference signal may be periodic, semi-static, or aperiodic. The network device may configure the time domain behavior of the reference signal and notify the terminal device.

[0275] Optionally, the method 400 further includes step 403: the network device sends configuration information of a channel state information (CSI report) report to the terminal device.

[0276] CSI reporting can also be periodic, semi-static, or non-periodic. The network device can configure the CSI reporting method and notify the terminal device. For example, if the CSI reporting is periodic, optionally, the configuration information of the CSI reporting also includes the period of CSI reporting. For another example, if the CSI reporting is semi-static, optionally, the configuration information of the CSI reporting also includes the period of CSI reporting in each CSI reporting time period.

[0277] 404. The network device sends a reference signal to the terminal device.

[0278] Specifically, the network device may send a reference signal to the terminal device according to the time domain behavior of the reference signal, wherein the reference signal may be used by the terminal device to perform channel measurement.

[0279] 405. The terminal device performs channel measurement according to the reference signal to obtain channel information.

[0280] It is understandable that the network device can send a reference signal to the terminal device according to the configuration of the reference signal, so that the terminal device can perform channel measurement based on the reference signal to obtain channel information. And the terminal device can report according to the configuration information of CSI reporting.

[0281] 406. The terminal device selects the second AI model to perform CSI compression according to the usage rule of the first AI model.

[0282] Assuming that the usage rule of the first AI model is Rule #2, and the terminal device determines, according to Rule #2, that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, then the terminal device determines the second AI model. The terminal device can use the second AI model for CSI compression, that is, the input of the second AI model includes the second channel information, and the status information of the second AI model is the default value. Among them, the first channel information can be obtained by the terminal device performing channel measurement based on the reference signal before obtaining the second channel information. The channel acquisition time corresponding to the first channel information and the second channel information can be determined according to the reporting time of the channel information.

[0283] Assume that the first channel information is H t-k , the second channel information is H t , H t-k The corresponding channel acquisition time can be based on H t-k The reporting time is determined, H t The corresponding channel acquisition time can be based on H t The reporting time is determined.

[0284] When the CSI reporting mode configured by the network device is periodic, the configuration information of the CSI reporting in step 403 includes the period of CSI reporting, so that the terminal device can determine H according to the period of CSI reporting. t The corresponding channel reporting time and H t-k The time interval between the corresponding channel reporting times can then determine the second AI model according to the time interval and the first capability requirement of the first AI model.

[0285] When the CSI reporting mode configured by the network device is semi-static, the CSI reporting configuration information in step 403 includes the CSI reporting period in each CSI reporting time period, and the time of the deactivation signaling of the previous CSI reporting time period and the activation signaling of this CSI reporting can determine the time interval between the previous deactivation and this activation, so that the terminal device can determine the H t The corresponding channel reporting time and H t-k The time interval between the corresponding channel reporting times, and then the second AI model can be determined according to the time interval and the first capability requirement of the first AI model. t The corresponding channel reporting time and H t-k The corresponding channel reporting times are all within the same CSI reporting time period, then H t The corresponding channel reporting time and H t-k The time interval between the corresponding channel reporting times is k times the period; if H t The corresponding channel reporting time and H t-k The corresponding channel reporting time is not in the same CSI reporting time period, such as Figure 5 As shown, H t The corresponding channel reporting time and H t-k The time interval between the corresponding channel reporting times can be expressed as: S+P+Δ, where S represents the H in a CSI reporting time period. t-k The time interval between the corresponding channel reporting time and the deactivation signaling, P represents the activation signaling and H in a CSI reporting time period. t The time interval between the corresponding channel reporting times, Δ represents H t-k The deactivation signaling in the CSI reporting time period of the corresponding channel reporting time is the same as H t The time interval between activation signaling within the CSI reporting time period where the corresponding channel reporting time is located.

[0286] Understandably, Figure 4This mainly introduces the situation where the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, and is not limited to this. For example, if the terminal device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the first capability requirement of the first AI model, the terminal device determines not to switch the AI ​​model, and optionally, the terminal device determines that the input of the first AI model includes the second channel information and the state information of the first AI model, and the state information of the first AI model is determined by the previous output of the first AI model.

[0287] Regarding the method of determining the second AI model, please refer to the description in method 300 and will not be repeated here.

[0288] 407. The terminal device sends information of the second AI model to the network device.

[0289] The information of the second AI model may be information that can identify the second AI model. As an example, the information of the second AI model includes an identifier of the second AI model. Based on the information of the second AI model sent by the terminal device, the network device may select an AI model that matches the second AI model and decode the output of the second AI model.

[0290] 408. The terminal device sends a CSI report to the network device.

[0291] Among them, the CSI report includes feedback information output by the second AI model.

[0292] 409 , the network device selects an AI model that matches the second AI model to perform CSI recovery.

[0293] Based on the above technical solution, the terminal device can determine to switch from the first AI model to the second AI model according to the time interval between the channel acquisition times corresponding to the channel information and the usage rules of the first AI model. Therefore, when it is determined that the first AI model is inappropriate, the second AI model can also be determined, and then the second AI model can continue to be used for related operations, thereby avoiding performance loss caused by using an inappropriate AI model to process channel information.

[0294] Figure 6 6 is a schematic flow chart of a communication method 600 provided according to another embodiment of the present application. The AI ​​model in the method 600 can be used for channel feedback. Figure 6 The illustrated method 600 may include the following steps.

[0295] Optionally, method 600 includes step 601: the network device sends usage rules of the first AI model to the terminal device.

[0296] Optionally, the method 600 further includes step 602: the network device sends configuration information of a reference signal to the terminal device.

[0297] Optionally, the method 600 further includes step 603: the network device sends configuration information of the CSI report to the terminal device.

[0298] 604. The network device sends a reference signal to the terminal device.

[0299] 605. The terminal device performs channel measurement according to the reference signal to obtain channel information.

[0300] For steps 601-605, reference may be made to the description of steps 401-405, which will not be repeated here.

[0301] 606. The terminal device determines the input of the first AI model according to the usage rule of the first AI model, and uses the first AI model to perform CSI compression.

[0302] Step 606 is similar to step 406, except that in step 606, it is assumed that the usage rule of the first AI model is rule #4.

[0303] For example, based on rule #4, if the terminal device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, the terminal device determines to use the first AI model for CSI compression, and the input of the first AI model includes the second channel information, and the status information of the first AI model is the default value.

[0304] For another example, based on rule #4, if the terminal device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information meets the second capability requirement of the first AI model, the terminal device determines to use the first AI model for CSI compression, and the input of the first AI model includes the second channel information and the status information of the first AI model, and the status information of the first AI model is determined by the previous output of the first AI model, and the output of the first AI model includes feedback information corresponding to the second channel information, and the updated status information of the first AI model, which is used for the next input of the first AI model.

[0305] 607, the terminal device sends a CSI report to the network device.

[0306] Among them, the CSI report includes feedback information output by the first AI model.

[0307] 608. The network device uses an AI model that matches the first AI model to perform CSI recovery.

[0308] Based on the above technical solution, the terminal device can determine the input of the first AI model according to the time interval between the channel acquisition times corresponding to the channel information and the usage rules of the first AI model, so that the input of the first AI model can match the capability requirements of the first AI model, making the input of the first AI model more appropriate, and avoiding performance loss caused by using inappropriate channel information as the input of the first AI model.

[0309] Figure 7 700 is a schematic flow chart of a communication method 700 provided according to an embodiment of the present application. The AI ​​model in the method 700 can be used for channel prediction. Figure 7 The illustrated method 700 may include the following steps.

[0310] Optionally, method 700 includes step 701: the network device sends usage rules of the first AI model to the terminal device.

[0311] Optionally, the method 700 further includes step 702: the network device sends configuration information of a reference signal to the terminal device.

[0312] For steps 701-702, reference may be made to the description of steps 401-402, which will not be repeated here.

[0313] 703. The network device sends a reference signal to the terminal device.

[0314] 704. The terminal device performs channel measurement according to the reference signal to obtain channel information.

[0315] For steps 703-704, reference may be made to the description of steps 403-404, which will not be repeated here.

[0316] 705. The terminal device selects the second AI model for prediction according to the usage rules of the first AI model, or determines the input of the first AI model and makes predictions based on the first AI model.

[0317] In one possible scenario, the terminal device determines the second AI model according to the usage rules of the first AI model, and makes predictions based on the second AI model. For example, assuming that the usage rules of the first AI model are rule #1, the terminal device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, so the terminal device determines the second AI model, and the input of the second AI model includes at least two channel information. Among them, the at least two channel information can be obtained by the terminal device performing channel measurement based on at least two reference signals, and each channel information corresponds to a channel measurement.

[0318] In another possible scenario, the terminal device determines the input of the first AI model according to the usage rules of the first AI model, and makes predictions based on the first AI model. For example, assuming that the usage rules of the first AI model are rule #3, the terminal device determines that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model. Therefore, the terminal device determines that the input of the first AI model includes M channel information and X default values ​​from at least two channel information, and the M channel information does not include the first channel information. Among them, at least two channel information can be obtained by the terminal device performing channel measurement based on at least two reference signals, and each channel information corresponds to a channel measurement.

[0319] Based on the above technical solution, the terminal device can determine to use a suitable AI model for prediction, or determine a suitable input for the AI ​​model for prediction according to the time interval between the channel acquisition times corresponding to the channel information and the usage rules of the first AI model, thereby avoiding performance loss caused by using inappropriate channel information as the input of the first AI model and improving prediction performance.

[0320] It can be understood that the above description is mainly combined with the scenarios of channel feedback and channel prediction, and is not limited to this. As mentioned above, the AI ​​model can also be used for positioning, beam management, or beam prediction. These scenarios can refer to the scenarios of channel feedback or channel prediction.

[0321] It can be understood that some optional features in the embodiments of the present application may not depend on other features in some scenarios, and may also be combined with other features in some scenarios, without limitation.

[0322] It can also be understood that in some of the above embodiments, sending information is mentioned multiple times. Taking A sending information to B as an example, A sending information to B may include A sending information directly to B, or may include A sending information to B through other devices or network elements, and there is no limitation on this.

[0323] It can also be understood that in some of the above embodiments, terminal devices and network devices are mainly used as examples for illustrative description, and this is not limited. For example, the terminal device can also be replaced by a component of the terminal device (such as a chip or circuit), and the network device can also be replaced by a component of the network device (such as a chip or circuit).

[0324] It can also be understood that the solutions in the various embodiments of the present application can be used in reasonable combination, and the explanations or descriptions of the various terms appearing in the embodiments can be mutually referenced or explained in the various embodiments, without limitation.

[0325] Above, combined Figures 3 to 7 The method provided by the embodiment of the present application is described in detail. Figures 8 to 10 The device provided in the embodiment of the present application is described in detail. It should be understood that the description of the device embodiment corresponds to the description of the method embodiment, so the contents not described in detail can be referred to the method embodiment above, and for the sake of brevity, they are not described here.

[0326] Figure 8 8 is a schematic diagram of a communication device 800 provided in an embodiment of the present application. The device 800 includes a processing unit 820. The processing unit 820 can be used to perform processing, such as determining the input of the first AI model or determining the second AI model. Optionally, the device 800 also includes a transceiver unit 810. The transceiver unit 810 can be used to implement corresponding communication functions. The transceiver unit 810 can also be called a communication interface or a communication unit.

[0327] Optionally, the device 800 may further include a storage unit, which may be used to store instructions and / or data, and the processing unit 820 may read the instructions and / or data in the storage unit so that the device implements the aforementioned method embodiment.

[0328] The device 800 may be the communication device in the aforementioned embodiment (eg Figure 3 The communication device in Figure 4 , Figure 6 , Figure 7 The terminal device in Figure 4 , Figure 6 , Figure 7 The device 800 may implement the steps or processes performed by the communication device in the above method embodiment. The transceiver unit 810 may be used to perform the transceiver-related operations of the communication device in the above method embodiment (such as the operations of sending and / or receiving data or messages), and the processing unit 820 may be used to perform the processing-related operations of the communication device in the above method embodiment, or operations other than transceiver (such as operations other than sending and / or receiving data or messages).

[0329] Optionally, the processing unit 820 is used to determine at least two channel information, where the at least two channel information include first channel information and second channel information, and the first channel information is obtained before the second channel information; the processing unit 820 is also used to determine the input of the first AI model or determine the second AI model based on the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model.

[0330] It should be understood that the specific process of each unit executing the above corresponding steps has been described in detail in the above method embodiment, and for the sake of brevity, it will not be repeated here.

[0331] It should also be understood that the device 800 here is embodied in the form of a functional unit. The term "unit" here may refer to an application specific integrated circuit (ASIC), an electronic circuit, a processor (such as a shared processor, a proprietary processor or a group processor, etc.) and a memory for executing one or more software or firmware programs, a merged logic circuit and / or other suitable components that support the described functions. In an optional example, those skilled in the art can understand that the device 800 can be specifically the communication device in the above-mentioned embodiment, and can be used to execute the various processes and / or steps corresponding to the communication device in the above-mentioned method embodiments. To avoid repetition, it will not be repeated here.

[0332] The device 800 of each of the above schemes has the function of implementing the corresponding steps performed by the communication device in the above method. The function can be implemented by hardware, or by hardware executing the corresponding software implementation. The hardware or software includes one or more modules corresponding to the above functions; for example, the transceiver unit can be replaced by a transceiver (for example, the sending unit in the transceiver unit can be replaced by a transmitter, and the receiving unit in the transceiver unit can be replaced by a receiver), and other units, such as the processing unit, can be replaced by a processor to respectively perform the transceiver operations and related processing operations in each method embodiment.

[0333] In addition, the transceiver unit 810 may also be a transceiver circuit (for example, may include a receiving circuit and a sending circuit), and the processing unit may be a processing circuit.

[0334] It should be pointed out that Figure 8 The device in the embodiment may be a communication device in the foregoing embodiment, or may be a chip or a chip system, such as a system on chip (SoC). The transceiver unit may be an input / output circuit or a communication interface; the processing unit may be a processor or a microprocessor or an integrated circuit integrated on the chip. This is not limited here.

[0335] Fig. 9 1 is a schematic diagram of another communication device 900 provided in an embodiment of the present application. The device 900 includes a processor 910, the processor 910 is coupled to a memory 920, the memory 920 is used to store computer programs or instructions and / or data, and the processor 910 is used to execute the computer programs or instructions stored in the memory 920, or read the data stored in the memory 920, so as to execute the methods in the above method embodiments.

[0336] Optionally, there are one or more processors 910 .

[0337] Optionally, the memory 920 is one or more.

[0338] Optionally, the memory 920 is integrated with the processor 910 or provided separately.

[0339] Alternatively, if Fig. 9 As shown, the device 900 further includes a transceiver 930, which is used to receive and / or send signals. For example, the processor 910 is used to control the transceiver 930 to receive and / or send signals.

[0340] As an example, the processor 910 may have Figure 8 The processing unit 820 shown in FIG. 8 may have the function of a storage unit, the memory 920 may have the function of a storage unit, and the transceiver 930 may have the function of a Figure 8 The functions of the transceiver unit 810 are shown in FIG.

[0341] As a solution, the device 900 is used to implement the operations performed by the communication device in each of the above method embodiments.

[0342] For example, the processor 910 is configured to execute computer programs or instructions stored in the memory 920 to implement relevant operations of the communication device in the above various method embodiments.

[0343] It should be understood that the processor mentioned in the embodiments of the present application may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0344] It should also be understood that the memory mentioned in the embodiments of the present application may be a volatile memory and / or a non-volatile memory. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM). For example, RAM can be used as an external cache. By way of example and not limitation, RAM includes the following forms: static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0345] It should be noted that when the processor is a general-purpose processor, DSP, ASIC, FPGA or other programmable logic device, discrete gate or transistor logic device, discrete hardware component, the memory (storage module) can be integrated into the processor.

[0346] It should also be noted that the memory described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0347] Fig.10 1 is a schematic diagram of a chip system 1000 provided in an embodiment of the present application. The chip system 1000 (or also referred to as a processing system) includes a logic circuit 1010 and an input / output interface 1020 .

[0348] Among them, the logic circuit 1010 can be a processing circuit in the chip system 1000. The logic circuit 1010 can be coupled to a storage unit and call instructions in the storage unit so that the chip system 1000 can implement the methods and functions of each embodiment of the present application. The input / output interface 1020 can be an input / output circuit in the chip system 1000, outputting information processed by the chip system 1000, or inputting data or signaling information to be processed into the chip system 1000 for processing.

[0349] Specifically, for example, if the communication device is installed with the chip system 1000, the logic circuit 1010 is coupled to the input / output interface 1020, and the input / output interface 1020 can input the reference signal to the logic circuit 1010 for processing, such as performing channel measurement on the reference signal to obtain channel information, or determining the input of the first AI model or determining the second AI model based on the time interval between the channel acquisition times corresponding to the channel information and at least one capability requirement of the first AI model.

[0350] As a solution, the chip system 1000 is used to implement the operations performed by the communication device in the above various method embodiments.

[0351] For example, the logic circuit 1010 is used to implement the processing-related operations performed by the communication device in the above method embodiment; the input / output interface 1020 is used to implement the sending and / or receiving-related operations performed by the communication device in the above method embodiment.

[0352] An embodiment of the present application also provides a computer-readable storage medium on which computer instructions for implementing the methods executed by the communication device in the above-mentioned method embodiments are stored.

[0353] For example, when the computer program is executed by a computer, the computer can implement the method performed by the communication device in each embodiment of the above method.

[0354] An embodiment of the present application also provides a computer program product, comprising instructions, which, when executed by a computer, implement the methods performed by the communication device in the above-mentioned method embodiments.

[0355] The present application also provides a communication system, which includes the terminal device and / or network device in the above embodiments. For example, the system includes Figure 4 , Figure 6 , Figure 7 The terminal device and / or network device in the illustrated embodiment.

[0356] The explanation of the relevant contents and beneficial effects of any of the above-mentioned devices can be referred to the corresponding method embodiments provided above, which will not be repeated here.

[0357] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0358] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. For example, the computer can be a personal computer, a server, or a network device, etc. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media integrations. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state disk (SSD)). For example, the aforementioned available medium includes, but is not limited to, various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0359] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A communication method, characterized in that: include: Determine at least two pieces of channel information, where the at least two pieces of channel information include first channel information and second channel information, where the first channel information is obtained before the second channel information; Determine the input of the first AI model or determine the second AI model based on the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first artificial intelligence AI model.

2. The method according to claim 1, characterized in that Determining an input of the first AI model or determining a second AI model according to a time interval between a channel acquisition time corresponding to the first channel information and a channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model, includes any of the following: When determining that a time interval between a channel acquisition time corresponding to the first channel information and a channel acquisition time corresponding to the second channel information does not meet a first capability requirement of the first AI model, determining the second AI model, wherein an input of the second AI model includes the at least two channel information; or When it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, determining the second AI model, wherein the input of the second AI model includes the second channel information, and the state information of the second AI model is a default value; or, When it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes M channel information of the at least two channel information and X default values, the M channel information does not include the first channel information, M is an integer greater than or equal to 1, and X is an integer greater than or equal to 0; or, When it is determined that a time interval between a channel acquisition time corresponding to the first channel information and a channel acquisition time corresponding to the second channel information does not meet a second capability requirement of the first AI model, it is determined that the input of the first AI model includes the second channel information, and the state information of the first AI model is a default value.

3. The method according to claim 2, characterized in that The determining of at least two pieces of channel information includes: Receiving a reference signal, and performing measurement based on the reference signal to obtain the at least two channel information; The channel acquisition time corresponding to the first channel information is determined according to: the reception time or the transmission time of the reference signal used to measure and obtain the first channel information, or, the reporting time of the first channel information; the channel acquisition time corresponding to the second channel information is determined according to: the reception time or the transmission time of the reference signal used to measure and obtain the second channel information, or, the reporting time of the second channel information.

4. The method according to claim 3, characterized in that The reference signal is periodic, and the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is determined according to the period of the reference signal; or, The reference signal is semi-static, and the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information is determined according to the time interval between the activation signaling and the deactivation signaling, The activation signaling is used to activate the reference signal used to measure and obtain the second channel information, and the deactivation signaling is used to deactivate the reference signal used to measure and obtain the first channel information; or, the activation signaling is used to activate reporting of the second channel information, and the deactivation signaling is used to deactivate the signaling for reporting the first channel information.

5. The method according to claim 1, characterized in that Determining an input of the first AI model or determining a second AI model according to a time interval between a channel acquisition time corresponding to the first channel information and a channel acquisition time corresponding to the second channel information, and at least one capability requirement of the first AI model, includes: When it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, determining the second AI model, and the input of the second AI model includes feedback information of the at least two channel information; or When it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the first capability requirement of the first AI model, determining the second AI model, wherein the input of the second AI model includes feedback information of the second channel information, and the state information of the second AI model is a default value; or, When it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes feedback information of M channel information of the at least two channel information and X default values, the M channel information does not include the first channel information, M is an integer greater than or equal to 1, and X is an integer greater than or equal to 0; or When it is determined that the time interval between the channel acquisition time corresponding to the first channel information and the channel acquisition time corresponding to the second channel information does not meet the second capability requirement of the first AI model, it is determined that the input of the first AI model includes feedback information of the second channel information, and the state information of the first AI model is a default value.

6. The method according to claim 5, characterized in that The determining of at least two pieces of channel information includes: Send a reference signal and receive feedback information of the at least two channel information, obtain the at least two channel information according to the feedback information of the at least two channel information, the channel acquisition time corresponding to the first channel information is determined according to: the sending time of the reference signal used to measure and obtain the first channel information, or, the reporting time of the first channel information; the channel acquisition time corresponding to the second channel information is determined according to: the sending time of the reference signal used to measure and obtain the second channel information, or, the reporting time of the second channel information.

7. The method according to any one of claims 2 to 6, characterized in that The M channel information do not include channel information before the first channel information, or the M channel information do not include channel information before the second channel information.

8. The method according to any one of claims 2 to 7, characterized in that The first capability requirement of the first AI model is: the time interval between channel acquisition times is within a first value range, and / or, The second capability requirement of the first AI model is that the time interval between channel acquisition times is within a second numerical range.

9. The method according to claim 8, characterized in that The first numerical range is less than or equal to T1; and / or, The second numerical range is less than or equal to T2; Among them, T1 is the maximum time interval between channel acquisition times corresponding to the first capability requirement of the first AI model, and T2 is the maximum time interval between channel acquisition times corresponding to the second capability requirement of the first AI model.

10. The method according to claim 9, characterized in that The value of T1 and / or T2 is determined by any one of the following methods: determined by a network device, determined by a terminal device, or predefined.

11. The method according to any one of claims 1 to 10, characterized in that The method further comprises: Obtain first indication information, where the first indication information indicates at least one capability requirement of the first AI model.

12. The method according to any one of claims 1 to 11, characterized in that The first channel information is any one of the at least two channel information except the second channel information, or the first channel information is separated from the second channel information by k channel information, where k is an integer greater than 1 or equal to 1.

13. The method according to claim 12, characterized in that The method further comprises: Second indication information is obtained, where the second indication information indicates the k.

14. A communication device, characterized in that: The method comprises a module or a unit for executing the method according to any one of claims 1 to 13.

15. A communication device, characterized in that: The device comprises a processor, wherein the processor is configured to execute a computer program or instruction stored in a memory, so that the device performs the method according to any one of claims 1 to 13.

16. The device according to claim 15, characterized in that The device further comprises the memory and / or the communication interface, wherein the communication interface is coupled to the processor. The communication interface is used to input and / or output information.

17. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program or instruction. When the computer program or instruction is executed on a communication device, the communication device is caused to execute the method according to any one of claims 1 to 13.

18. A computer program product, characterized in that The computer program product comprises a computer program or instructions for performing the method according to any one of claims 1 to 13.

19. A chip, characterized in that: The chip is coupled to a memory and is used to read and execute program instructions stored in the memory to implement the method according to any one of claims 1 to 13.

Images