Communication method and device

Abstract

The invention provides a communication method and device, and relates to the field of communication, and the method comprises the steps that a network device sends first information to a terminal device, the first information comprises a first configuration used for reasoning, and the first configuration carries a first condition, the first condition is used for indicating a demand for a first model associated with the first configuration and / or a demand for prediction performance of the first model; and under the condition that the first model and / or the prediction performance of the first model meets the first condition, the terminal equipment sends second information to the network equipment, and the second information comprises a reasoning result of the first model or information used for indicating that the first configuration is available. Therefore, under the condition that the terminal equipment needs to activate the configuration sent by the network equipment for the first time for reasoning and report the reasoning result, the first configuration can be activated and the reasoning result can be reported according to the configuration condition of the network equipment, so that the signaling overhead can be reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication, and in particular to a communication method and device. BACKGROUND

[0002] With the development of artificial intelligence (AI), a terminal device can perform beam management, beam selection, mobility management, and channel state information (CSI) feedback enhancement through an AI model, thereby improving device performance.

[0003] Currently, a network device can send one or more CSI report configurations (CSI-ReportConfig) for inference to a terminal device, and the terminal device can determine whether the one or more CSI report configurations for inference are available. The terminal device can also activate an AI model associated with the available configuration to obtain an inference result of the AI model, and report the inference result to the network device.

[0004] However, when the terminal device reports the inference result to the network device, there is a problem of large signaling overhead. SUMMARY

[0005] The present application provides a communication method and device, which can reduce signaling overhead.

[0006] In a first aspect, a communication method is provided. The method can be applied to a terminal side, such as a terminal device or a communication module in the terminal device, or a circuit or chip responsible for communication functions in the terminal device (such as a modem chip, also known as a baseband chip, or a system on chip (SoC) chip or system in package (SIP) chip containing a modem core). Hereinafter, the terminal device is taken as an example for description.

[0007] The method includes receiving first information, the first information including a first configuration for inference, the first configuration carrying a first condition, the first condition being used to indicate a requirement for a first model associated with the first configuration and / or a requirement for a predicted performance of the first model; and in a case where the first model and / or the predicted performance of the first model meets the first condition, sending second information, the second information including an inference result of the first model or information indicating that the first configuration is available.

[0008] Based on the method provided in the application, the terminal device can determine the demand of the network device for the first model and / or the demand of the network device for the predicted performance of the first model according to the first condition. If the first model and / or the demand of the network device for the predicted performance of the first model meets the first condition, it means that the first model meets the performance demand of the network device for the model. Therefore, the terminal device can send the inference result of the first model to the network device or send information indicating that the first configuration is available. The information indicating that the first configuration is available is used for the terminal device to subsequently report the inference result of the first model. In this way, the terminal device determines the first model meeting the performance demand of the network device for the model based on the first condition, and then reports the inference result of the first model, which is beneficial to avoid reporting the inference result of the model not meeting the first condition and is beneficial to reduce signaling overhead.

[0009] With reference to the first aspect, in some implementations of the first aspect, the first model meeting the first condition comprises: a complexity of the first model being less than or equal to a first value, and / or, an occupied memory of the first model being less than or equal to a second value.

[0010] With reference to the first aspect, in some implementations of the first aspect, the predicted performance of the first model meeting the first condition comprises: a loss value of the first model being less than or equal to a third value, and / or, a prediction accuracy of the first model being greater than or equal to a fourth value.

[0011] With reference to the first aspect, in some implementations of the first aspect, the first configuration is used for availability determination. In this way, the first condition is carried in the configuration used for availability determination, which is beneficial to further reduce signaling overhead.

[0012] With reference to the first aspect, in some implementations of the first aspect, the first configuration comprises a first CSI reporting configuration.

[0013] With reference to the first aspect, in some implementations of the first aspect, in a case where the second information comprises the information indicating that the first configuration is available, the method further comprises: receiving information indicating that the first configuration is activated; activating the first configuration to obtain the inference result of the first model; and sending the inference result of the first model. In this way, the terminal device reports the inference result of the first model meeting the first condition, which is beneficial to avoid reporting the inference result of the model not meeting the first condition and is beneficial to reduce signaling overhead.

[0014] In some implementations of the first aspect, the first information further includes a second configuration for inference, the second configuration carrying a second condition, the second condition indicating a requirement for a second model associated with the second configuration and / or a requirement for a prediction performance of the second model; and the method further includes: in a case where the second model and / or the prediction performance of the second model does not satisfy the second condition, sending third information, the third information indicating that the second configuration is unavailable. Since the second configuration is unavailable, the terminal device can not report an inference result of the model associated with the second configuration, which helps to reduce signaling overhead.

[0015] In some implementations of the first aspect, the third information further indicates that the reason why the second configuration is unavailable includes that the second model and / or the prediction performance of the second model does not satisfy the second condition. The terminal device reports the reason why the configuration is unavailable to the network device, so that the network device makes a decision that matches the reason, for example, reduces the requirement for the second model and / or the prediction performance of the second model.

[0016] In some implementations of the first aspect, the method further includes: receiving a third condition, the third condition indicating a requirement for the second model and / or a requirement for a prediction performance of the second model; and in a case where the second model and / or the prediction performance of the second model satisfies the third condition, sending information indicating that the second configuration is available. In this way, the second configuration is activated, and the performance benefit brought by AI is obtained.

[0017] In the second aspect, a communication method is provided, which can be applied to a network side, for example, a network device or a communication module in the network device, or a circuit or chip (such as a modem chip, also known as a baseband chip, or a SoC chip or SIP chip containing a modem core) responsible for communication functions in the network device. Hereinafter, the method is described by taking the case where the method is applied to a network device.

[0018] The method includes: sending first information, the first information including a first configuration for inference, the first configuration carrying a first condition, the first condition indicating a requirement for a first model associated with the first configuration and / or a requirement for a prediction performance of the first model; and in a case where the first model and / or the prediction performance of the first model satisfies the first condition, receiving second information, the second information including an inference result of the first model or information indicating that the first configuration is available.

[0019] The method provided in the application, the first configuration sent by the network device to the terminal device carries the first condition, so that the terminal device determines the performance requirement of the network device on the model, and aligns the requirement of the network device and the terminal device on the inference result of the model, which is beneficial to obtain the inference result of the model meeting the first condition, and is also beneficial to avoid receiving the inference result of the inference model not meeting the first condition, and is beneficial to reduce the signaling overhead.

[0020] Optionally, the first condition and the first configuration can refer to the above description, which will not be repeated here.

[0021] With reference to the second aspect, in some implementations of the second aspect, when the second information includes information used to indicate that the first configuration is available, the method further includes: sending information used to indicate that the first configuration is activated; and receiving the inference result of the first model. In this way, the network device can flexibly determine the timing of activating the first configuration, which is beneficial to improve flexibility.

[0022] With reference to the second aspect, in some implementations of the second aspect, the first information further includes a second configuration used for inference, the second configuration carries a second condition, and the second condition is used to indicate the requirement of a second model associated with the second configuration and / or the requirement of the predicted performance of the second model; and the method further includes: when the performance of the second model and / or the predicted performance of the second model does not meet the second condition, receiving third information, the third information being used to indicate that the second configuration is unavailable.

[0023] With reference to the second aspect, in some implementations of the second aspect, the third information is further used to indicate that the reason why the second configuration is unavailable includes that the second model and / or the predicted performance of the second model does not meet the second condition. In this way, the network device can determine the reason why the second configuration is unavailable, which is beneficial to the network device to make a decision matching the reason.

[0024] With reference to the second aspect, in some implementations of the second aspect, the method further includes: sending a third condition, the third condition being used to indicate the requirement of the second model and / or the requirement of the predicted performance of the second model; and when the performance of the second model and / or the predicted performance of the second model meets the third condition, receiving information used to indicate that the second configuration is available. In this way, the network device can flexibly adjust the activation condition of the configuration to obtain the inference result of the model meeting the third condition.

[0025] The third aspect provides a communication apparatus for executing the method in any possible implementation manner of the above aspects. Specifically, the communication apparatus includes units for executing the method in any possible implementation manner of the above aspects. The units can also be referred to as modules, which will be described below by way of example.

[0026] In an implementation, the communication apparatus can include a module corresponding to each of the methods / operations / steps / actions described in any of the aspects above, which can be implemented in hardware circuit, software, or both.

[0027] In another implementation, the communication apparatus is a communication chip, which can include an input circuit or interface for sending information or data, and an output circuit or interface for receiving information or data.

[0028] In another implementation, the communication apparatus is a terminal device or a network device, which can include a transmitter for sending information or data, and a receiver for receiving information or data.

[0029] In another implementation, the communication apparatus is configured to perform the method in any of the possible implementation manners of the aspects above, and the apparatus can be configured in a terminal device or a network device.

[0030] In a fourth aspect, a communication apparatus is provided, which includes a processor coupled to a memory, and configured to execute instructions in the memory to implement the method in any of the possible implementation manners of the aspects above. Optionally, the communication apparatus further includes the memory. Optionally, the communication apparatus further includes a communication interface, and the processor is coupled to the communication interface.

[0031] In an implementation, the communication interface above can be a transceiver, or an input / output interface.

[0032] In a fifth aspect, a processor is provided, which includes an input circuit, an output circuit, and a processing circuit. The processing circuit is configured to receive a signal through the input circuit, and transmit a signal through the output circuit, so that the processor performs the method in any of the possible implementation manners of the aspects above.

[0033] In a specific implementation process, the processor above can be a chip, the input circuit can be an input pin, the output circuit can be an output pin, and the processing circuit can be a transistor, a gate circuit, a flip-flop, and various logic circuits, etc. The input signal received by the input circuit can be received and input by, for example but not limited to, a receiver, the output signal output by the output circuit can be output to and transmitted by, for example but not limited to, a transmitter, and the input circuit and the output circuit can be the same circuit, which is used as the input circuit and the output circuit at different times respectively. The specific implementation manners of the processor and various circuits are not limited in the embodiments of the present application.

[0034] In a sixth aspect, a communication apparatus is provided, which includes a processor. The processor can receive signals through a receiver, and transmit signals through a transmitter, to perform the method in any possible implementation of the aspects above. The processor in the communication apparatus can be one or more.

[0035] Optionally, the communication apparatus can further include a memory. The processor can be configured to read instructions stored in the memory, and can receive signals through the receiver, and transmit signals through the transmitter, to perform the method in any possible implementation of the aspects above. The memory can be one or more.

[0036] Optionally, the memory can be integrated with the processor, or the memory can be located separately from the processor.

[0037] In a specific implementation process, the memory can be a non-transitory memory, such as a read only memory (ROM), which can be integrated on the same chip as the processor, or can be located separately on different chips. The type of memory and the arrangement of the memory and the processor are not limited in the present application.

[0038] It should be understood that the relevant data interaction process, such as sending indication information, can be a process of outputting indication information from the processor, and receiving capability information can be a process of receiving input capability information by the processor. Specifically, the data output by the processor can be output to the transmitter, and the input data received by the processor can come from the receiver. The transmitter and the receiver can be collectively referred to as a transceiver.

[0039] The communication apparatus in the sixth aspect above can be a chip, and the processor can be implemented by hardware or software. When implemented by hardware, the processor can be a logic circuit, an integrated circuit, etc. When implemented by software, the processor can be a general-purpose processor, which can be implemented by reading software codes stored in the memory. The memory can be integrated in the processor, or can exist independently from the processor.

[0040] In a seventh aspect, a computer program product is provided, which includes a computer program (also referred to as code or instructions), which, when executed by a computer, causes the computer to perform the method in any possible implementation of the aspects above.

[0041] In an eighth aspect, a computer readable storage medium is provided, which stores a computer program (also referred to as code or instructions), which, when executed on a computer, causes the computer to perform the method in any possible implementation of the aspects above.

[0042] In a ninth aspect, a communication system is provided, which can include the terminal device of the first aspect and the network device of the second aspect.

[0043] It should be understood that the third aspect to the ninth aspect of the present application correspond to the technical solutions of the first aspect and the second aspect of the present application, and the beneficial effects obtained by each aspect and the corresponding feasible implementation manners are similar, which will not be repeated. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 is a schematic diagram of an architecture of a communication system provided by an embodiment of the present application;

[0045] Figure 2 is a schematic diagram of an application framework of AI in an NR system;

[0046] Figure 3 is a schematic diagram of spatial domain prediction based on an AI model;

[0047] Figure 4 is a schematic diagram of time domain prediction based on an AI model;

[0048] Figure 5 is a schematic diagram of AI-based CSI feedback;

[0049] Figure 6 is a schematic interaction diagram of LCM of network device decision;

[0050] Figure 7 and Figure 8 is a schematic interaction diagram of a model configuration method;

[0051] Figures 9 to 11 is a schematic interaction diagram of a communication method provided by an embodiment of the present application;

[0052] Figure 12 is a schematic diagram of an architecture of a RAN node provided by an embodiment of the present application;

[0053] Figure 13 is a schematic diagram of a RIC architecture communication system provided by an embodiment of the present application;

[0054] Figure 14 is a schematic flowchart of a communication method applied to an architecture of a RAN node provided by an embodiment of the present application;

[0055] Figure 15 and Figure 16 is a schematic interaction diagram of a communication device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0056] Before introducing the method provided by an embodiment of the present application, the following is explained:

[0057] In the embodiments of the present application, the terms "first", "second", and the like are used to distinguish between similar or identical items or components with substantially the same function and effect. For example, the first model and the second model are merely used to distinguish between different models, and do not limit the order. Those skilled in the art can understand that the terms "first", "second", and the like do not limit the number and execution order, and the terms "first", "second", and the like do not necessarily mean different.

[0058] It should be noted that in the embodiments of the present application, the words "exemplarily" or "for example" are used to represent an example, illustration or description. Any embodiment or design scheme described as "exemplarily" or "for example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the words "exemplarily" or "for example" are intended to present the relevant concept in a specific manner.

[0059] In the embodiments of the present application, "at least one" means one or more, and "multiple" means two or more. The association relationship between the associated objects is described, which means that there can be three relationships, for example, A and / or B, which can represent the following three cases: A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. The character "or" generally represents an "or" relationship between the associated objects. "At least one of the following" or similar expressions means any combination of these items, including any combination of single item or multiple items. For example, at least one of a, b, or c can represent a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0060] In the embodiments of the present application, "sending" and "receiving" represent the direction of signal transmission. For example, "sending second information to a network device" can be understood as that the destination of the second information is the network device, which can include direct transmission through the air interface, or indirect transmission through the air interface by other units or modules. "Receiving second information from a terminal device" can be understood as that the source of the second information is the terminal device, which can include direct reception from the terminal device through the air interface, or indirect reception from the terminal device through the air interface by other units or modules. "Sending" can also be understood as "output" of a chip interface, and "receiving" can also be understood as "input" of a chip interface.

[0061] In other words, sending and receiving can be performed between devices, for example, between a terminal device and a network device; or can be performed within a device, for example, between components, between modules, between chips, between software modules or hardware modules within a device through a bus, a wire or an interface.

[0062] In the embodiments of the present application, "when", "if" and "whether" all refer to the objective situation that the device will make corresponding processing, and are not limited in time, and do not require the device to have a judgment action when implemented, nor mean that there are other limitations. Unless otherwise specified, "if" and "whether" can be replaced, and "when" and "in the case of" can be replaced. "When" and "if" / "whether" can be replaced.

[0063] In the embodiments of the present application, the words such as "exemplarily" or "for example" are used to represent as an example, illustration or description. Any embodiment or design scheme described as "exemplarily" or "for example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the words such as "exemplarily" or "for example" are intended to present the relevant concept in a specific manner.

[0064] Ninthly, in the present application, the schemes in various embodiments can be reasonably combined for use, and the explanation or description of each term appearing in the embodiments, similar operations or steps can be mutually referenced or explained in various embodiments, and this is not limited.

[0065] The technical solutions of the embodiments of the present application can be applied to various communication systems, for example: a fourth generation (4th generation, 4G) communication system (also known as a long term evolution (long term evolution, LTE) communication system), a worldwide interoperability for microwave access (worldwide interoperability for microwave access, WiMAX), a fifth generation (5th generation, 5G) communication system (also known as a new radio (new radio, NR) communication system), or other communication systems that may appear in the future. The technical solutions of the present application can also be applied to a non-terrestrial network (non-terrestrial network, NTN) communication system, or a scenario in which the NTN communication system is integrated with a terrestrial network (terrestrial network, TN). The NTN communication system can be an NTN system integrated with 4G, 5G and any future generation communication system, such as NR NTN, internet of things (internet of things, IoT) NTN, etc. The NTN communication system can be a satellite communication system, or can include unmanned aerial vehicles, high altitude platform stations (high altitude platform station, HAPS), and other aerial access network devices, and the embodiments of the present application do not limit this.

[0066] Exemplarily, Figure 1This is a schematic diagram of the architecture of a communication system 100 provided in an embodiment of this application. Figure 1 As shown, the communication system 100 includes a radio access network (RAN) 10 and a core network 20. Optionally, the communication system 100 also includes an Internet 30. The radio access network 10 may include at least one access network device (such as...). Figure 1 110a and 110b in the above), may also include at least one terminal (such as Figure 1 (Referring to devices 120a-120j). Terminals connect wirelessly to access network equipment, which in turn connects wirelessly or via wired connection to the core network 20. Core network equipment and access network equipment can be independent physical devices, or they can integrate the functions of core network equipment and access network equipment onto the same physical device. Alternatively, a single physical device can integrate some core network equipment functions and some access network equipment functions. Terminals and access network equipment can connect to each other via wired or wireless means. Figure 1 This is just an illustration; the communication system may also include other access network devices, such as wireless repeaters and wireless backhaul devices. Figure 1 Not shown in the image.

[0067] The radio access network 10 can be a cellular system related to the 3rd generation partnership project (3GPP), such as a 5th generation (5G) mobile communication system (also known as an NR system), or it can be applied to future mobile communication systems or other similar communication systems, without specific limitations. The radio access network 10 can also be an open radio access network (O-RAN) or a cloud radio access network (CRAN). The radio access network 10 can also be a non-terrestrial network (NTN), a satellite communication network, a high altitude platform station (HAPS) communication network, an integrated access and backhaul (IAB) communication network, a reconfigurable intelligent surface (RIS) communication network, etc. The radio access network 10 can also be a communication system that integrates two or more of the above systems.

[0068] The nodes in the radio access network 10 can be referred to as RAN nodes. A RAN node, also referred to as a radio access network device, RAN entity, or access node, is configured to help terminals access the communication system wirelessly. The RAN nodes in the communication system 100 can be of the same type or of different types.

[0069] In a possible scenario, the RAN node can be a base station, a transmission reception point (TRP), a next generation NodeB (gNB) in 5G, a base station in a future mobile communication system, an access point (AP) in a satellite, an IAB node, a RAN node in an NTN communication system, i.e., can be deployed in a high altitude platform or a satellite, etc. The RAN node can be a macro base station (e.g., 110a in FIG. 1), a micro base station, or an indoor station (e.g., 110b in FIG. 1), a relay node, or a donor node, or a radio controller in a CRAN scenario. The RAN node can also be a device assuming base station functionalities in device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, drone communication, machine communication, etc. Alternatively, the RAN node can also be a server, a wearable device, a vehicle or a vehicle-mounted device, etc. For example, in V2X technology, the RAN node can be a road side unit (RSU). Figure 1 Figure 1 In a possible scenario, the RAN node can be a base station, a transmission reception point (TRP), a next generation NodeB (gNB) in 5G, a base station in a future mobile communication system, an access point (AP) in a satellite, an IAB node, a RAN node in an NTN communication system, i.e., can be deployed in a high altitude platform or a satellite, etc. The RAN node can be a macro base station (e.g., 110a in FIG. 1), a micro base station, or an indoor station (e.g., 110b in FIG. 1), a relay node, or a donor node, or a radio controller in a CRAN scenario. The RAN node can also be a device assuming base station functionalities in device-to-device (D2D) communication, vehicle-to-everything (V2X) communication, drone communication, machine communication, etc. Alternatively, the RAN node can also be a server, a wearable device, a vehicle or a vehicle-mounted device, etc. For example, in V2X technology, the RAN node can be a road side unit (RSU).

[0070] ​In another possible scenario, a terminal is assisted by multiple RAN nodes to implement wireless access, and different RAN nodes respectively implement part of functions of a base station. For example, a RAN node can be a central unit (CU), a distributed unit (DU), or a radio unit (RU). Here, the CU implements functions of a radio resource control (RRC) protocol and a packet data convergence layer protocol of the base station, and can also implement functions of a service data adaptation protocol; the DU implements functions of a radio link control layer and a medium access control (MAC) layer of the base station, and can also implement part of functions or all functions of a physical layer; for specific descriptions of the above protocol layers, refer to relevant technical specifications of the 3GPP. The RU can be used to implement functions of transceiving a radio frequency signal. The CU and the DU can be two independent RAN nodes, or can be integrated in a same RAN node, for example, integrated in a baseband unit. The RU can be included in a radio frequency device, for example, included in a remote radio unit (RRU) or an active antenna unit (AAU). The CU can be further divided into two types of RAN nodes, CU-control plane (CP) and CU-user plane (UP).

[0071] In different systems, a RAN node can have different names. For example, in an O-RAN system, a CU can be referred to as an open CU (O-CU), a DU can be referred to as an open DU (O-DU), and an RU can be referred to as an open RU (O-RU). A RAN node in an embodiment of the present application can be implemented by means of a software module, a hardware module, or a combination of a software module and a hardware module. For example, a RAN node can be a server loaded with a corresponding software module. Embodiments of the present application do not limit specific technologies and specific device forms adopted by a RAN node. For ease of description, a base station is described below as an example of a RAN node.

[0072] A terminal is a device with wireless transceiver function, which can send signals to an access network device or receive signals from an access network device. A terminal can also be referred to as a terminal device, a terminal equipment, a user equipment (UE), a mobile station, a mobile terminal, etc. A terminal device can be widely applied to various scenarios, such as D2D, V2X communication, machine-type communication (MTC), internet of things (IoT), virtual reality, augmented reality, industrial control, automatic driving, remote medical treatment, smart power grid, smart furniture, smart office, smart wear, smart transportation, smart city, etc. A terminal can be a mobile phone, a tablet computer, a computer with wireless transceiver function, a wearable device, a vehicle, an airplane, a ship, a robot, a mechanical arm, a smart home device, etc. Embodiments of the present application do not limit the specific technology and specific device form of the terminal. The device for implementing the function of the terminal can be a terminal; or can be a device capable of supporting the terminal to implement the function, such as a chip system. The device can be installed in the terminal or used in matching with the terminal. In embodiments of the present application, a chip system can be composed of a chip, or can include a chip and other discrete devices. All or part of the functions of the terminal in the present application can also be implemented through software functions running on hardware, or through virtualized functions instantiated on a platform (such as a cloud platform).

[0073] The access network device and the terminal can be fixed in position or mobile. The access network device and the terminal can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; can be deployed on water surface; or can be deployed on an airplane, a balloon and a man-made satellite. Embodiments of the present application do not limit the application scenarios of the access network device and the terminal.

[0074] The roles of the access network device and the terminal can be relative. For example, Figure 1 The helicopter or the unmanned aerial vehicle 120i in FIG. 1 can be configured as a mobile access network device. For those terminals 120j accessing the wireless access network 10 through the 120i, the terminal 120i is an access network device; but for the access network device 110a, the 120i is a terminal, that is, the 110a and the 120i communicate with each other through a wireless air interface protocol. Of course, the 110a and the 120i can also communicate with each other through an interface protocol between access network devices, at this time, the 120i is also an access network device relative to the 110a. Therefore, the access network device and the terminal can be collectively referred to as a communication device, Figure 1 The 110a and the 110b in FIG. 1 can be referred to as a communication device with an access network device function, Figure 1 The 120a-120j in FIG. 1 can be referred to as a communication device with a terminal function.

[0075] The access network device and the terminal, the access network device and the access network device, and the terminal and the terminal can communicate through a licensed spectrum, an unlicensed spectrum, or both. They can also communicate through a spectrum below 6 gigahertz (GHz) or above 6 GHz, or both. The embodiments of the present application do not limit the spectrum resources used for wireless communication.

[0076] In the embodiments of the present application, the functions of the access network device can also be performed by a module (such as a chip) in the access network device, or by a control subsystem containing the functions of the access network device. The control subsystem containing the functions of the access network device herein can be a control center in the above-mentioned application scenarios such as smart grid, industrial control, intelligent transportation, and smart city. The functions of the terminal can also be performed by a module (such as a chip or modem) in the terminal, or by a device containing the functions of the terminal.

[0077] The core network device refers to a device in the core network that provides service support for the terminal. Currently, some examples of core network devices are: access and mobility management function network elements, session management function network elements, user plane function network elements, and the like, which are not listed one by one here.

[0078] The embodiments of the present application can be applied to an AI-based wireless communication scenario, in which the terminal device, the access network device, or the core network device can have AI capability and be configured with an AI model or function for inference. The AI model or function can be trained internally by the device or transmitted to the device by other devices, and the embodiments of the present application do not limit this.

[0079] In the embodiments of the present application, the core network device and the access network device can both be referred to as network devices, and the terminal can be referred to as a UE. For ease of description, the network device and the UE are taken as examples for description below. In order to better understand the embodiments of the present application, first, the application of AI in the wireless communication scenario is introduced.

[0080] 1. The necessity of AI application in the wireless field

[0081] As mobile networks support increasingly diverse services, they need to meet varying demands such as ultra-high speed, ultra-low latency, ultra-high reliability, and massive connectivity. This makes network planning, configuration, and resource scheduling increasingly complex. Furthermore, the increasing use of higher frequencies by mobile networks places greater demands on base station energy efficiency. These new requirements, scenarios, and characteristics present unprecedented challenges to mobile network planning, operation, and efficient management. Relying on manual experience or simple algorithms for network planning, self-optimization of network configuration, and resource scheduling suffers from drawbacks such as high time consumption, high cost, and poor adaptability of self-optimization and scheduling algorithms, making it unsuitable for addressing these new challenges.

[0082] Introducing AI and machine learning into mobile networks can significantly improve the efficiency of network planning, configuration, and resource scheduling, enabling network intelligence. AI can simulate arbitrary nonlinear models, thus effectively adapting to real-world environments and approaching performance limits. AI and machine learning acquire massive amounts of data, using machine learning algorithms to train models and / or make decision inferences, outputting AI models and / or decision results (such as predicting service data volume over a certain future timeframe). To achieve RAN intelligence, it is necessary to research key technologies such as the RAN intelligent wireless network framework, the related functions and protocol processes of AI modules / platforms, etc.

[0083] 2. Application Framework of AI in NR System

[0084] For example, Figure 2 A schematic diagram illustrating an application framework of AI in an NR system is shown. For example... Figure 2 As shown, the data collection entity can store data inputs from sources including the gNB, CUs within the gNB, DUs within the gNB, UEs, or other management entities, serving as a database for AI model training and data analysis inference. The model training entity analyzes the training data provided by the data collection entity to produce the optimal AI model. The model storage entity can store trained or updated models and, under the control of the management entity, transmits the models to the model inference entity. The model inference entity uses the AI ​​model, based on the data provided by the data collection entity, to provide reasonable AI-based predictions about network operation or guide network strategy adjustments. The management entity can also, based on the inference output of the model inference entity, instruct the model training entity to update the model, or provide feedback on model performance to determine whether to update the model. Furthermore, the management entity can, based on the inference output of the model inference entity, select, activate, deactivate, switch, or roll back the inference model to leverage AI for more accurate prediction results.

[0085] 3. AI use cases

[0086] Applications of AI in wireless field can include: beam management, beam selection, mobility management, CSI feedback enhancement (e.g. CSI compression and prediction).

[0087] In wireless communication, CSI refers to known channel properties of a communication link, which describes how a signal propagates from a transmitter to a receiver and represents the combined effects of scattering, fading, and power decay with distance. CSI makes it possible to adapt the transmission to the current channel conditions, which is essential for achieving reliable communication at high data rates in multi-antenna systems. In the process of implementation and application, the base station sends CSI reference signals to the UE for measurement, and the UE calculates various values through measurement and reports them to the base station for CSI acquisition or beam management, or does not report and only uses it for UE to select the receiving beam.

[0088] 1) AI for beam management

[0089] In recent years, AI has played a great role in beam management, especially in reducing the overhead of beam sweeping. Typically, AI takes the received power of wide beams or sparsely scanned narrow beams measured at the UE side as input, and the model inference outputs the candidate better narrow beams, for example, the top-k beams with the best signal quality (i.e. Top-k candidate beams). In one example, the model inference outputs the received signal reference power (RSRP) values or IDs of the beam set. The network side performs scanning based on the Top-k candidate beams to finally determine the better beam. AI models can be deployed at the UE side or the network side, which is not limited in the present application.

[0090] The application of AI in beam management is mainly in two aspects, which are spatial domain prediction and time domain prediction. The following will introduce Figure 3 spatial domain prediction, and Figure 4 time domain prediction.

[0091] Exemplarily, Figure 3 a schematic diagram of spatial domain prediction based on AI model is shown. As Figure 3As shown, the input of the AI model is the beam information of a certain pattern scanned at a certain time, which can be the RSRP value. The set of beam information can be referred to as set B (set B). The complete set A (set A) of beams is predicted by the AI model. The UE can select the top k beams with the best quality from the output set and report the related information to the network.

[0092] Exemplarily, Figure 4 An example of time domain prediction based on an AI model is shown. As shown, Figure 4 The UE can collect the set B beam RSRP from time (t-N+1) to time (t) through a sliding time window T1, as shown in RSRP (t-N+1) to RSRP (t) in the middle. Figure 4 The AI model can output the prediction result of the future time window based on the input information.

[0093] If the AI model is a regression model, the output of the AI model can be set A. If the AI model is a classification model, the output of the AI model can be the top k beams with the best quality. The UE can report the top k beams with the best quality and related information to the network.

[0094] 2) AI for mobility management

[0095] In the mobility management process, the base station configures the UE to perform measurement reporting and makes handover decisions based on the measurement reporting results. After the introduction of AI, the base station can make predictions based on the limited measurement results of the UE (e.g., configuring a small number of measurement beams, or the UE reducing the measurement of beams and cells) and select the optimal cell for handover decisions. AI can also be used for positioning.

[0096] Currently, AI-based positioning scenarios can include the following two types of five specific use cases.

[0097] The first type: AI model or machine learning (ML) model is directly used for positioning.

[0098] Use case 1: Apply UE-side model, directly use AI model or ML model for positioning. For example, the UE performs signal measurement, inputs the measurement result into the AI model or ML model, and the output of the AI model or ML model is the predicted positioning result.

[0099] Use case 2b: UE-assisted location management function (LMF) side positioning, i.e., the LMF side applies AI model or ML model for positioning based on the assistance information on the UE side.

[0100] Use case 3b: NG-RAN assisted LMF-side positioning, i.e., LMF-side positioning based on the assistance information from NG-RAN side, applying AI model or ML model.

[0101] Second type: AI model or ML model assisted positioning.

[0102] Use case 2a: UE-side model assisted LMF positioning, i.e., UE-side prediction based on UE measurement results, and sending the prediction results to LMF for assisting LMF in positioning.

[0103] Use case 3a: NG-RAN-side model assisted LMF positioning, i.e., NG-RAN-side prediction based on measurement results, and sending the prediction results to LMF for assisting LMF in positioning.

[0104] 3) AI for CSI feedback enhancement

[0105] Currently, 5G system uses codebook as the basic tool for CSI feedback, and multiple schemes such as Type I / II codebook are defined for different feedback accuracy. However, the above codebooks are designed for uniformly arranged antenna arrays, and are not optimized for special antennas such as 3D antennas, and the performance limitations are obvious. AI-based CSI feedback can break this bottleneck and achieve better feedback performance through optimization for specific channel environments.

[0106] The basic principle of AI-based CSI feedback is to regard the high-dimensional channel information feedback task as an end-to-end CSI image compression and recovery task. As shown in Figure 5 , the basic signal flow has a structure similar to an autoencoder: first, the encoder at the encoding end (e.g., the terminal side) uses the encoder to compress the complete channel information H" into a bit stream of information that meets the feedback requirements after feature extraction; then, the information is fed back to the decoding end (e.g., the base station side) through the feedback link; finally, the decoding end uses the decoder to decompress and reconstruct the feature, and finally restores the complete channel information

[0107] The above encoder and decoder are jointly optimized in the end-to-end training process to obtain better CSI reconstruction performance. In actual deployment, the encoder and decoder need to be used in pairs according to the training process, i.e., the compressed CSI output by a certain encoder needs to be recovered by the corresponding decoder.

[0108] In addition, the AI-based channel state information prediction can also include: based on the existing CSI, obtaining the CSI of unknown time-frequency resources without increasing new air interface resource overhead. The CSI of different time or space dimensions is not completely the same, but there is a certain degree of correlation, which makes it possible to predict the CSI.

[0109] The traditional CSI prediction scheme is limited in processing complex data and is difficult to be practically applied due to the prediction accuracy. The AI-based CSI prediction is expected to significantly improve the prediction accuracy, so as to achieve the goal of obtaining unknown CSI with low overhead in the actual system. According to the data correlation category, the AI-based CSI prediction can be divided into four categories: the first category considers the time correlation, that is, the CSI in the next time or the next period of time is predicted according to the CSI in the previous period of time, which is mainly applied to the channel changing with time or the high-speed moving scene. The second category considers the correlation in the frequency angle, for example, the downlink CSI is predicted and reconstructed according to the uplink CSI of FDD. The third category considers the prediction problem in the spatial angle. The fourth category considers the channel correlation between adjacent users.

[0110] 4. Life cycle management (LCM) of AI

[0111] Exemplarily, Figure 6 An exemplary interaction diagram of LCM of network device decision is shown. As Figure 6 shown, the method can include the following steps:

[0112] S601, the UE reports capability information and / or assistance information to the network device, wherein the capability information indicates the features supported by the UE, such as the use cases supported by the UE, and / or the AI functions supported by the UE, etc., and the assistance information indicates the AI model or ML model supported by the UE, and the related information about the function of the AI model or ML model, such as the conditions under which the model (or function) is applicable (or suitable), or whether the model or function is applicable in the current scenario. The use cases supported by the UE can be as shown above, which will not be described here. The model can be replaced by the function, and in this application, the model is taken as an example for description.

[0113] S602, the network device makes a decision based on the capability information and / or assistance information reported by the UE.

[0114] S603, the network device sends a management instruction 1 to the UE, wherein the management instruction 1 is used to activate, deactivate, switch, or fallback the AI model or ML model.

[0115] For example, the information reported by the UE indicates that it has applicable models or functions, and the network device sends a model or function activation instruction to the UE, that is, activates the applicable model or function of the UE.

[0116] S604, if the UE performs measurement by the model or function, the UE reports the performance monitored by the UE or reports actual measurement results to the network device, for assisting the network device in monitoring. In the present application, the actual measurement results can be referred to as actual measurement results.

[0117] S605, the network device performs monitoring based on the performance or actual measurement results reported by the UE.

[0118] S606, the network device sends a management instruction 2 to the UE for subsequent management. The management instruction 2 is used to activate, deactivate, switch, or fallback the AI model or ML model.

[0119] For example, the network device monitors that the model or function performance of the UE is too poor, and sends a deactivation or switching instruction to the UE.

[0120] 5, configuration of UE-side AI model

[0121] The present application provides two configuration methods of UE-side AI model, which are option A and option B. The following describes option A in combination with Figure 7 Option B is described in combination with Figure 8 Option B.

[0122] Exemplarily, Figure 7 A schematic interaction diagram of a model configuration method is shown. As Figure 7 shown, the method can include the following steps:

[0123] S701, the network device sends UE capability query information to the UE, and the UE capability query information is used to query or ask for the capability of the UE.

[0124] S702, based on the UE capability query information, the UE reports UE capability information to the network device, and the UE capability information can include AI capability information supported by the UE.

[0125] For example, the UE capability information includes one or more of the following: support for AI-based beam management, support for AI-based positioning, support for AI-based mobility management, or support for AI-based CSI feedback enhancement.

[0126] In some other possible cases, the UE capability information includes information about support for AI or ML function.

[0127] S703、network equipment sends an RRC reconfiguration message to the UE based on the capability information of the UE and network requirements, wherein the RRC reconfiguration message includes multiple CSI report configurations for inference, so as to confirm whether there is a usable model under the configuration required by the network.

[0128] The CSI report configuration can include complete measurement resource configuration, reporting content, and reporting method configuration, etc. Different CSI report configurations have different identities. The requirements of different CSI report configurations can be different. For example, according to scene division, CSI report configuration 1 is used for beam management prediction, and CSI report configuration 2 is used for mobility management prediction configuration. Or, in the same scene, different configurations, CSI report configuration 3 is used for beam management, and is 4-beam prediction 8-beam. CSI report configuration 2 is used for beam management, and is 8-beam prediction 16-beam.

[0129] In the scene of AI-based beam management, the CSI report configuration can include the beam number configuration of set A and set B, and the measurement resource configuration corresponding to set A and set B. In some examples, the beam number configuration of set A is 16 (i.e. SETA = 16, SETB = 8), and the beam number configuration of set B is 8 (i.e. SETB = 8). The network equipment additionally configures conditions corresponding to set A and set B, respectively, and the mapping relationship can be established by an associated ID method to achieve logical association of configuration items. The reporting type can be periodic, semi-periodic, or one-time. The reporting content can be the top K values (i.e. topK values).

[0130] If in the time-domain prediction scene under AI-based beam management, the CSI report configuration can also include configurations such as the number of time slots that need to be predicted and / or the number of measurement time slots. It can be understood that if the UE also supports one or more of AI-based mobility management, AI-based CSI feedback enhancement, or AI-based positioning, the CSI report configuration can also include configuration parameters required by these scenes.

[0131] S704, based on the RRC reconfiguration message, the UE sends an RRC reconfiguration complete message to the network equipment, and the RRC reconfiguration complete message can include an applicability report.

[0132] Based on the RRC reconfiguration message, the UE confirms whether there is a usable model under the configuration required by the network, and sends the usable model and the unusable model to the network equipment in the form of an applicability report. In some examples, the applicability report can also be called application report or availability report, which is not limited in the present application.

[0133] The RRC reconfiguration message includes multiple CSI report configurations for inference. When an AI model included by the UE matches a CSI report configuration and is available, the UE can mark the CSI report configuration as available. When an AI model included by the UE matches a CSI report configuration but is unavailable, or when an AI model included by the UE does not match a CSI report configuration, the UE can mark the CSI report configuration as unavailable. The UE can send information about which CSI report configurations are available and which CSI report configurations are unavailable to the network device.

[0134] If the UE periodically reports the applicability report, the UE can actively activate an available CSI report configuration, use the available CSI report configuration for inference to obtain an inference result of an AI model associated with the CSI report configuration, and report the inference result to the network device, that is, S706 and S707 are performed. As shown in Figure 7 , the inference result can be carried in an L1 report signal. In some examples, the L1 report signal can be carried in uplink control information (UCI).

[0135] If the UE semi-statically or statically reports the applicability report, the UE can not actively activate an available CSI report configuration and needs to be instructed by the network device. The network device can instruct the UE to activate the available CSI report configuration. Based on the instruction, the UE uses the available CSI report configuration for inference to obtain an inference result of an AI model associated with the CSI report configuration, and reports the inference result to the network device, as shown in S705 to S707 in Figure 7 . As shown in Figure 7 , the network device can instruct the UE to activate the available CSI report configuration through an L1 trigger signal. In some examples, the L1 trigger signal can be carried in downlink control information (DCI) or a medium access control control element (MAC CE).

[0136] As can be known from the above description, S705 is optional.

[0137] Exemplarily, Figure 8 another model configuration method is shown. As shown in Figure 8 , the method can include the following steps:

[0138] S801 and S802 can refer to the description of S701 and S702 above, which will not be repeated here.

[0139] S803, the network device sends an RRC reconfiguration message 1 to the UE based on the capability information of the UE and the network requirement, the RRC reconfiguration message 1 including one or more sets of inference-related configuration parameters for the UE to confirm whether there is a usable model under the configuration required by the network. Each set of inference-related configuration parameters in the one or more sets of inference-related configuration parameters can be part of the CSI report configuration, or can be a model-related parameter, which is not limited in the present application.

[0140] S804, based on the RRC reconfiguration message, the UE sends an RRC reconfiguration complete message to the network device, which can include a suitability report.

[0141] The suitability report includes the availability of each set of inference-related configuration parameters. If a set of inference-related configuration parameters is available, and the set of inference-related configuration parameters is part of a certain CSI report configuration, the network device can issue this CSI report configuration to the UE in the subsequent step. If a set of inference-related configuration parameters is available, and the set of inference-related configuration parameters is a model-related parameter, the network device can determine a CSI report configuration based on the correspondence between the set of inference-related configuration parameters and the CSI report configuration, and issue this CSI report configuration to the UE in the subsequent step.

[0142] If a set of inference-related configuration parameters is not available, the network device does not perform the subsequent step.

[0143] S805, based on the RRC reconfiguration complete message, the network device sends an RRC reconfiguration message 2 to the UE, the RRC reconfiguration message 2 including a usable CSI report configuration for inference, for the UE to activate the usable CSI report configuration for inference. Wherein, the usable CSI report configuration for inference is used to represent the CSI report configuration corresponding to the available inference-related configuration parameters.

[0144] S806, based on the RRC reconfiguration message 2, the UE activates the usable CSI report configuration to obtain the inference result of the AI model associated with the usable CSI report configuration.

[0145] S807, the UE sends an L1 report signal to the network device, the L1 report signal including the inference result of the AI model associated with the usable CSI report configuration.

[0146] From the above Figure 7 and Figure 8The method shown can know that the UE can activate the available CSI report configuration, use the available CSI report configuration for reasoning to obtain the reasoning result of the AI model associated with the CSI report configuration, and report the reasoning result to the network device. In the scenario where the UE uses the AI model associated with the CSI report configuration for reasoning for the first time, the network device is not sure about the performance of the AI model, so the network device instructs the UE to activate all available CSI report configurations, so that the UE reports the reasoning result of the AI model associated with all available CSI report configurations, or the UE actively reports all reasoning results. However, not all reasoning results can meet the performance requirements of the network device on the model. In this case, reporting all reasoning results will cause unnecessary signaling overhead.

[0147] Therefore, the embodiments of the present application provide a communication method. When the network device sends a configuration (for example, a CSI report configuration) to the UE, the configuration can carry a condition, which is used to indicate the requirement for the model associated with the configuration and / or the requirement for the predicted performance of the model. In the case where the model associated with the configuration and / or the predicted performance meets the condition, the UE sends information indicating that the configuration is available to the network device, or the reasoning result of the model associated with the configuration. In this way, in the case where the model associated with the configuration meets the condition, the UE reports the reasoning result of the model, which is beneficial to avoid reporting the reasoning result of the model that does not meet the first condition, beneficial to reduce signaling overhead, and beneficial to make the reported reasoning result meet the performance requirement of the network device on the model, and align the requirement of the network device and the UE for the reasoning result of the model.

[0148] In order to better understand the embodiments of the present application, the following will be described in combination with Figures 9 to 11 The method provided by the embodiments of the present application will be described in detail. The embodiments shown by the embodiments of the present application show the method provided by the embodiments of the present application from the perspective of device interaction. The specific form and quantity of each device shown are only examples, and should not constitute any limitation on the implementation of the method provided by the embodiments of the present application.

[0149] In the following, the method of the embodiments of the present application will be described in detail taking the network device and the terminal device as the execution subject.

[0150] It should be understood that the network device can be the network device itself, or a chip, chip system or processor supporting the network device to implement the method provided by the embodiments of the present application, or a logic module or software capable of implementing all or part of the network device; the terminal device can be the terminal device itself, or a chip, chip system or processor supporting the terminal device to implement the method provided by the embodiments of the present application, or a logic module or software capable of implementing all or part of the terminal device, which is not limited by the embodiments of the present application.

[0151] Exemplarily, Figure 9A schematic interaction diagram of a communication method provided by an embodiment of the present application is shown. The method can be applied to the communication system shown in Figure 1 but the embodiments of the present application are not limited thereto. In some examples, the network device can be the radio access network 10 in Figure 1 and the terminal device can be 120j, 120a or 120e in Figure 1 .

[0152] As shown in Figure 9 , the method can include the following steps:

[0153] S901, the network device sends first information to the terminal device, the first information including a first configuration for inference, the first configuration carrying a first condition, the first condition being used to indicate a requirement for a first model associated with the first configuration and / or a requirement for a predicted performance of the first model.

[0154] Optionally, the first configuration is used for availability determination. In this way, the first condition is carried in the configuration for availability confirmation, which is conducive to further reducing signaling overhead.

[0155] Optionally, the first configuration can include a CSI report configuration. When the first configuration includes the CSI report configuration, the first information can be carried in the RRC reconfiguration message shown in Figure 7 or the first information can be carried in the RRC reconfiguration message 2 shown in Figure 8 .

[0156] The first model is a model that uses information (or parameters) in the first configuration for inference. The first model associated with the first configuration can also be referred to as the first model corresponding to the first configuration, and the embodiments of the present application are not limited thereto. In different scenarios, the first model predicts different indicators.

[0157] For example, in a beam prediction scenario, the first model can predict a first layer reference signal received power (L1-RSRP) or top-k downlink transmission beams (Top-k DL Tx beam). The L1-RSRP is an important indicator for measuring the strength of the received reference signal in a wireless communication system, and is used to evaluate the quality of the radio signal.

[0158] For another example, in a mobility management scenario, the first model can predict a layer 3 cell-reference signal received power (L3-cell-RSRP). The L3-cell-RSRP refers to a cell-reference signal received power after a third layer filtering. The third layer filtering can consider a signal condition in a longer time range, for high layer radio resource management, such as cell selection, handover decision, etc.

[0159] For another example, in a CSI prediction scenario, the first model can predict a user perceived throughput (UPT).

[0160] In an example, the first condition is used to indicate a requirement for the first model, which can be understood as a requirement for the first model itself. The requirement can also be referred to as a demand, which is not limited in the embodiments of the present application.

[0161] For example, the first condition is used to indicate a requirement for a complexity of the first model and / or a requirement for a memory occupied by the first model. The complexity of the first model can refer to a complexity of a model structure (e.g., a number of parameters, a depth of layers). Alternatively, the first condition is used to indicate a requirement for an algorithm (e.g., K nearest neighbors (KNN) or Transformer) of the first model.

[0162] In an example, the first condition is used to indicate a requirement for a prediction performance of the first model. The prediction performance can also be referred to as an inference performance, which is not limited in the embodiments of the present application.

[0163] For example, the first condition is used to indicate a requirement for a prediction accuracy of the first model and / or a requirement for a loss value of the first model. For example, in a beam prediction scenario, the first model is used to predict a L1-RSRP, and the first condition can be used to indicate a requirement for an accuracy of the first model in predicting the L1-RSRP. In a CSI prediction scenario, the first model is used to predict a UPT, and the first condition can be used to indicate a requirement for an accuracy of the first model in predicting the UPT. The prediction accuracy of the first model can be measured by a normalized mean squared error (NMSE) or a squared generalized cosine similarity (SGCS), for example.

[0164] In an example, the first condition is used to indicate a requirement for the first model associated with the first configuration and a requirement for a prediction performance of the first model.

[0165] For example, the first condition is used to indicate a requirement on a memory occupied by the first model and a prediction accuracy of the first model. For another example, the first condition is used to indicate a requirement on a memory occupied by the first model and a loss value of the first model.

[0166] The first condition is used to determine whether the first configuration needs to be activated, that is, in a case where the first model and / or a prediction performance of the first model meets the first condition, the terminal device can activate the first configuration, and therefore the first condition can also be referred to as an activation condition. Since the embodiments of the present application are applicable to a scenario in which the terminal device uses an AI model associated with a CSI report configuration for inference for the first time, the first condition can also be referred to as a first activation condition. The first activation condition is used to represent a condition for determining whether to activate the CSI report configuration for the first time to obtain an inference result of the AI model associated with the CSI report configuration. The first activation condition is used when determining whether to activate the CSI report configuration for the first time. If the terminal device is not activating the CSI report configuration for the first time, the terminal device can ignore the condition, that is, the condition is not used as a condition for determining whether to activate the CSI report configuration.

[0167] The first information can include one or more configurations for inference, and each configuration carries a condition, which is used to indicate a requirement on a model associated with the configuration and / or a prediction performance of the model. In the embodiments of the present application, the first configuration carrying the first condition is taken as an example for description.

[0168] The first configuration carried by the network device and sent to the terminal device carries the first condition, so that the terminal device determines a performance requirement of the network device on the model.

[0169] S902, in a case where the first model and / or a prediction performance of the first model meets the first condition, the terminal device sends, to the network device, second information, the second information including an inference result of the first model or information indicating that the first configuration is available.

[0170] If the first condition is used to indicate a requirement on the first model, for example, a requirement on a complexity of the first model and / or a memory occupied by the first model, in an example, the first condition includes a first value and / or a second value, the first value corresponding to the complexity of the first model, and the second value corresponding to the memory occupied by the first model, so that the first model meeting the first condition can include that the complexity of the first model is less than or equal to the first value, and / or the memory occupied by the first model is less than or equal to the second value.

[0171] If the first condition is used to indicate a requirement on the prediction performance of the first model, such as a requirement on the prediction accuracy of the first model and / or a requirement on the loss value of the first model, in an example, the first condition comprises a third value and / or a fourth value, the third value corresponds to the loss value of the first model, and the fourth value corresponds to the prediction accuracy of the first model. In this way, the first model satisfying the first condition can include that the loss value of the first model is less than or equal to the third value, and / or the prediction accuracy of the first model is greater than or equal to the fourth value.

[0172] In an example, in a case where the first model and / or the prediction performance of the first model satisfies the first condition, the terminal device can activate the first configuration to obtain the inference result of the first model, and send the inference result of the first model to the network device. In this way, the terminal device sends the inference result of the first model satisfying the first condition to the network device, which is beneficial to avoid reporting the inference result of the model not satisfying the first condition, and is beneficial to reduce signaling overhead.

[0173] Optionally, the inference result of the first model can be carried in the L1 report signal shown in the above Figure 8 . Alternatively, the second information can be carried in the L1 report signal shown in the above Figure 8 .

[0174] In an example, in a case where the first model and / or the prediction performance of the first model satisfies the first condition, the terminal device can determine that the first configuration is available, and send information indicating that the first configuration is available to the network device. For example, the information indicating that the first configuration is available can be carried in the RRC reconfiguration complete message shown in the above Figure 7 . Alternatively, the second information can be carried in the RRC reconfiguration complete message shown in the above Figure 7 .

[0175] Optionally, in a case where the first model and / or the prediction performance of the first model satisfies the first condition, the terminal device can actively activate the first configuration, or activate the first configuration based on the indication of the network device, to obtain the inference result of the first model associated with the first configuration, and send the inference result of the first model to the network device.

[0176] In this way, the first model associated with the available first configuration reported by the terminal device to the network device satisfies the first condition, which is beneficial to subsequently report the inference result of the first model satisfying the first condition, avoid reporting the inference result of the model not satisfying the first condition, and reduce signaling overhead.

[0177] In some examples, in a case where the second information includes information used for indicating that the first configuration is available, according to the second information, the network device sends, to the terminal device, information used for indicating to activate the first configuration; and according to the information used for indicating to activate the first configuration, the terminal device activates the first configuration to obtain an inference result of the first model, and sends the inference result of the first model to the network device. In this way, the terminal device reports the inference result of the first model satisfying the first condition, which is beneficial to avoid reporting the inference result of a model not satisfying the first condition, and is beneficial to reduce signaling overhead. In addition, the network device can flexibly determine the timing of activating the first configuration, which is beneficial to improve flexibility.

[0178] For example, the information used for indicating to activate the first configuration can be carried in the LI trigger signal in the above Figure 7 . The inference result of the first model can be carried in the LI report signal in the above Figure 7 .

[0179] Based on the method provided in the embodiments of the present application, the first configuration sent by the network device to the terminal device carries the first condition, so that the terminal device determines the performance requirement of the network device on the model. If the first model associated with the first configuration satisfies the first condition, it means that the first model satisfies the performance requirement of the network device on the model. Therefore, the terminal device can send the inference result of the first model to the network device or send information used for indicating that the first configuration is available, wherein the information used for indicating that the first configuration is available is used for subsequently reporting the inference result of the first model satisfying the first condition, which is beneficial to avoid reporting the inference result of a model not satisfying the first condition, and is beneficial to reduce signaling overhead.

[0180] Optionally, the above first information further includes a second configuration used for inference, the second configuration carries a second condition, and the second condition is used for indicating a requirement on a second model associated with the second configuration and / or a requirement on a predicted performance of the second model; and the method shown in the above Figure 9 may further include: in a case where the second model and / or the predicted performance of the second model does not satisfy the second condition, the terminal device sends, to the network device, third information used for indicating that the second configuration is unavailable.

[0181] In some examples, the third information can be carried in the RRC reconfiguration complete message shown in the above Figure 7 .

[0182] The second condition is a condition for judging whether the second configuration needs to be activated. Since the terminal device can activate an available configuration, the second condition can be set as a condition for judging whether a configuration is available. When the second configuration is available, it means that the second configuration can be activated. When the second configuration is unavailable, it means that the second configuration cannot be activated.

[0183] The second configuration is different from the first configuration, the second condition can be different from the first condition, or can be the same, and embodiments of the present application do not limit this.

[0184] For example, the second condition is different from the first condition. The first condition is used to indicate that the memory occupied by the first model is less than or equal to a second value. The second condition is used to indicate that the prediction accuracy of the second model is greater than or equal to a preset threshold, which can be the same as the fourth value mentioned above, or can be different, and embodiments of the present application do not limit this.

[0185] For another example, the second condition is the same as the first condition. The first condition is used to indicate that the memory occupied by the first model is less than or equal to a second value. The first condition is used to indicate that the memory occupied by the second model is less than or equal to the second value.

[0186] In the case that the second model and / or the prediction performance of the second model does not meet the second condition, it is indicated that the second configuration is not available, that is, the second configuration cannot be activated. The terminal device can send third information to the network device to inform the network device that the second configuration is not available, and the terminal device can not report the inference result of the model associated with the second configuration. In this way, it is beneficial to reduce the signaling overhead.

[0187] Optionally, the third information is also used to indicate that the reason why the second configuration is not available includes that the second model and / or the prediction performance of the second model does not meet the second condition. In this way, the network device can determine the reason why the second configuration is not available, which is beneficial to the network device to make a decision matching the reason.

[0188] Optionally, after the terminal device sends the third information to the network device, the method further includes: the network device sends a third condition to the terminal device, the third condition is used to indicate the demand for the second model and / or the demand for the prediction performance of the second model; and in the case that the second model and / or the prediction performance of the second model meets the third condition, the terminal device sends information indicating that the second configuration is available to the network device.

[0189] The second configuration is not available, which can be due to that the second model associated with the second configuration and / or the prediction performance of the second model does not meet the second condition. The network device can send a third condition to the terminal device, and the demand for the second model and / or the demand for the prediction performance of the second model in the third condition can be less than (or lower than) the second condition. The terminal device can judge whether the second model and / or the prediction performance of the second model meets the third condition. If the third condition is met, the terminal device sends information indicating that the second configuration is available to the network device, that is, updates the availability of the second configuration.

[0190] The second configuration is updated from unavailable to available, which can be understood as the second configuration is updated from not needing to be activated to being able to be activated, and the terminal device can activate the second configuration to obtain an inference result of a second model associated with the second configuration and send the inference result of the second model to the network device.

[0191] In this way, the network device can flexibly adjust the activation condition of the configuration to obtain the inference result of the model that meets the condition.

[0192] To better understand the embodiments of the present application, the following describes the scenarios shown in the method provided by the embodiments of the present application. Figure 7 and Figure 8 application.

[0193] Exemplarily, Figure 10 application. As shown in Figure 10 application. As shown in

[0194] S1001, the network device sends an RRC reconfiguration message to the terminal device, and the RRC reconfiguration message includes a plurality of CSI report configurations for inference, and each CSI report configuration in the plurality of CSI report configurations carries a condition, and the role of the condition can be referred to the description above, which will not be described here.

[0195] S1002, based on the RRC reconfiguration message, the terminal device sends an RRC reconfiguration completion message to the network device, and the RRC reconfiguration completion message includes the availability of each CSI report configuration and the reason why the CSI report configuration is unavailable.

[0196] It can be understood that the RRC reconfiguration completion message includes information about whether each CSI report configuration is available, and if the CSI report configuration is unavailable, the RRC reconfiguration completion message further includes the reason why the CSI report configuration is unavailable.

[0197] The reason why the CSI report configuration is unavailable can include that a model or a prediction performance of the model associated with the CSI report configuration does not meet the condition carried in the CSI report configuration.

[0198] If the CSI report configuration is unavailable, the network device can perform any one of the following processing on the CSI report configuration:

[0199] 1) The network device starts a timer, and during the running of the timer, if no indication information that the CSI report configuration is available is received, the network device releases the CSI report configuration after the timer expires. In this way, the CSI report configuration is retained for a period of time, so that it can be used for inference when it is available, which is beneficial to obtain the performance benefits brought by AI.

[0200] 2) The network device releases the CSI report configuration, and starts a timer, and after the timer expires, the network device sends the CSI report configuration to the terminal device again. In this way, the CSI report configuration that is not available is released, which is beneficial to avoid excessive resource overhead.

[0201] 3) The network device sends the CSI report configuration that is not available to the terminal device, and the CSI report configuration that is not available carries another condition (for example, the third condition involved in the above embodiment), which is different from the last condition (for example, the second condition involved in the above embodiment). If the model associated with the CSI report configuration and / or the prediction performance of the model meets the condition, the terminal device reports to the network device that the CSI report configuration is available.

[0202] S1003, if the terminal device periodically reports the availability of each CSI report configuration, the CSI report configuration is actively activated if the model associated with the CSI report configuration meets the condition carried by the CSI report configuration, and the CSI report configuration is used for inference to obtain the inference result of the AI model associated with the CSI report configuration.

[0203] S1004, the terminal device sends an L1 report signal to the network device, and the L1 report signal includes the inference result of the AI model associated with the activated CSI report configuration.

[0204] The method provided by the embodiments of the present application can set conditions for the CSI report configuration, and when the terminal device needs to activate the configuration sent by the network device for the first time for inference and report the inference result, the configuration can be activated and the inference result can be reported on demand according to the condition configured by the network device, which is beneficial to reduce signaling overhead and reduce power consumption, and is also beneficial to align the performance requirements of the network device and the terminal device for the model.

[0205] It should be noted that, in the examples shown in Figure 10 In the examples shown in

[0206] In the examples shown in Figure 10 In some other examples, if different CSI report configurations correspond to the same condition, one of the different CSI report configurations (for example, CSI report configuration 1) can carry the condition, and the other CSI report configurations can carry an indication information, which is used to indicate that the condition is the same as the condition carried by the CSI report configuration 1. In this way, the same condition can not be repeatedly configured, which is beneficial to reduce complexity.

[0207] In Figure 10 In the example shown, one condition is carried in the CSI report configuration, in some other examples, the network device can send the CSI report configuration and the condition to the terminal device at the same time. For example, the CSI report configuration and the condition can be carried in the RRC reconfiguration message shown above. Figure 7 In another example, the CSI report configuration and the condition can be carried in the RRC reconfiguration message 2 shown above. In this way, the CSI report configuration does not carry the condition, and is more flexible. Figure 8

[0208] Exemplarily, Figure 11 An exemplary interaction diagram of a communication method provided by an embodiment of the present application is shown. As Figure 11 shown, the method can include S801-S804 described above. The method further includes the following steps:

[0209] S1101, the network device sends an RRC reconfiguration message 2 to the terminal device, the RRC reconfiguration message 2 including one or more CSI report configurations available for inference, each of the one or more CSI report configurations carrying a condition, which can refer to the description above and will not be repeated here.

[0210] S1102, in the case where each CSI report configuration associated model meets the condition carried therein, the CSI report configuration is activated, and inference is performed using the CSI report configuration to obtain an inference result of the AI model associated with the CSI report configuration.

[0211] S1103, the terminal device sends an L1 report signal to the network device, the L1 report signal including the inference result of the AI model associated with the activated CSI report configuration.

[0212] For the CSI report configuration that is not activated, the network device can perform any of the following:

[0213] 1) The network device starts a timer, and during the running of the timer, if the inference result of the AI model associated with the CSI report configuration is not received, the network device releases the CSI report configuration after the timer expires. In this way, the network device retains the CSI report configuration for a period of time, so that the terminal device can use it for inference when it can be activated, which is conducive to obtaining the performance benefits brought by AI.

[0214] 2) The network device releases the CSI report configuration that is not activated, and starts a timer, and after the timer expires, the CSI report configuration is sent to the terminal device again. In this way, the CSI report configuration that is not activated is released, which is conducive to avoiding excessive resource overhead.

[0215] ​The method provided by the embodiments of the present application can be used to send performance conditions for available CSI report configurations by the network device according to the needs after the terminal device reports the applicability report, to judge whether the performance requirements of the network device on the model are met before the terminal device activates the configuration for the first time, and to report the inference result of the model meeting the performance requirements by the terminal device, which is beneficial to reduce the signaling overhead of the terminal device and beneficial to align the performance requirements of the network device and the terminal device on the model.

[0216] The network device in the embodiments of the present application can be a RAN node. In an example, the architecture of the RAN node is a CU and DU separation architecture.

[0217] Exemplarily, Figure 12 A schematic diagram showing the architecture of a RAN node is shown. As Figure 12 shown, the core network device is connected with the RAN node through an NG interface, and the RAN node is connected with the user equipment through an air interface Uu. One or more AI modules can be arranged in the core network device, the RAN node and the user equipment. For the sake of clarity, Figure 12 The RAN node can include a CU and a DU. One or more AI modules can also be arranged in the CU and / or the DU.

[0218] Optionally, the CU can also be split into a central unit control plane (CU-CP) and a central unit-user plane (CU-UP). One or more AI models are arranged in the CU-CP and / or the CU-UP. The AI module is used to implement the corresponding AI function. The AI modules deployed in different network elements can be the same or different. The model of the AI module can implement different functions according to different parameter configurations. One AI module can have one or more models. One model can infer an output including one parameter or multiple parameters. The learning process, training process or inference process of different models can be deployed in different nodes or devices, or can be deployed in the same node or device.

[0219] Exemplarily, Figure 13 A schematic diagram showing a RAN intelligent controller (RIC) architecture communication system is shown. As Figure 13As shown, the RIC architecture communication system includes near-real-time RIC (near-RT RIC) and non-real-time RIC (non-RT RIC). Near-real-time RIC is used for model training and inference. For example, it is used to train an AI model and then use that AI model for inference. Near-real-time RIC can obtain network-side and / or terminal-side information from RAN nodes (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or terminal devices. This information can be used as training data or inference data. Optionally, near-real-time RIC can deliver inference results to RAN nodes and / or terminal devices. Optionally, inference results can be exchanged between CU and DU, and / or between DU and RU. For example, near-real-time RIC delivers inference results to DU, and DU sends them to RU.

[0220] Non-real-time RICs are used for model training and inference. For example, they are used to train AI models and then use those models for inference. Non-real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CU, CU-CP, CU-UP, DU, and / or RU) and / or terminal devices. This information can be used as training data or inference data, and the inference results can be delivered to RAN nodes and / or terminal devices. Optionally, inference results can be exchanged between CUs and DUs, and / or between DUs and RUs; for example, a non-real-time RIC delivers the inference result to a DU, which then forwards it to an RU. Near-real-time and non-real-time RICs can also be configured as separate network elements. Optionally, near-real-time and non-real-time RICs can also be part of other devices; for example, a near-real-time RIC can be located in a RAN node (e.g., in a CU or DU), while a non-real-time RIC can be located in operations, administration, and maintenance (OAM) systems, cloud servers, core network devices, or other network devices.

[0221] The method provided in this application can also be applied to a CU and DU separate architecture. In a CU and DU separate architecture, the conditions in the first configuration can be set by the CU or by the DU, and this application does not limit this.

[0222] In one example, such as Figure 14 As shown in a, the DU can set the conditions in the first configuration. Specifically, the CU sends the first configuration for inference to the DU. The first configuration does not carry the first condition. The DU determines the first condition and sends first information to the terminal device. The first information includes the first configuration for inference. The first configuration carries the first condition. The DU also receives second information from the terminal device, that is, it executes the steps of S901 and S902 described above.

[0223] In an example, as shown by b in Figure 14 , the CU can set the condition in the first configuration. Specifically, the CU sends the first configuration for inference to the DU, the first configuration carries the first condition, the DU sends the first information to the terminal device, the first information includes the first configuration for inference, the first configuration carries the first condition, and receives the second information from the terminal device, that is, the steps of S901 and S902 are executed.

[0224] It can be understood that various numerical numbers involved in the embodiments of the present application are only for differentiation for convenience of description, and do not limit the scope of the embodiments of the present application. The size of the serial number of the above processes does not mean the execution order, and the execution order of the processes should be determined according to its function and inherent logic.

[0225] In various embodiments of the present application, the terms and / or descriptions of different embodiments are consistent and can be mutually referred to if there is no special description and logical conflict, and the technical features in different embodiments can be combined to form new embodiments according to their inherent logical relationship.

[0226] It can be understood that, in order to realize the functions in the above embodiments, the terminal device or the network device includes the corresponding hardware structure and / or software module for executing each function. Those skilled in the art should easily realize that, in combination with the units and method steps of the examples described in the embodiments disclosed in the present application, the present application can be realized in the form of hardware or a combination of hardware and computer software. Whether a certain function is realized in the form of hardware or computer software driving hardware depends on the specific application scene and design constraints of the technical solution.

[0227] Figure 15 and Figure 16 The structural schematic diagram of the possible communication device provided by the embodiments of the present application is shown. These communication devices can be used to realize the functions of the terminal device or the network device in the above method embodiments, and thus can also realize the beneficial effects possessed by the above method embodiments.

[0228] As shown in Figure 15 , the communication device 1500 includes a processing unit 1510 and a transceiver unit 1520. The communication device 1500 is used to realize the functions of the terminal device or the network device in the method embodiments shown in Figure 9 .

[0229] In a possible implementation, the device 1500 is used to realize the steps corresponding to the terminal device in the method embodiments shown in Figure 8 .

[0230] The transceiver 1520 is configured to receive first information, the first information comprising a first configuration for inference, the first configuration carrying a first condition indicating a requirement for a first model associated with the first configuration and / or a requirement for a predicted performance of the first model; and the processing unit 1510 is configured to determine second information in a case where the first model and / or the predicted performance of the first model satisfies the first condition; and the transceiver 1520 is configured to transmit the second information, the second information comprising an inference result of the first model or information indicating that the first configuration is available.

[0231] Optionally, the first model satisfying the first condition comprises that a complexity of the first model is less than or equal to a first value and / or a memory occupied by the first model is less than or equal to a second value.

[0232] Optionally, the predicted performance of the first model satisfying the first condition comprises that a loss value of the first model is less than or equal to a third value and / or a prediction accuracy of the first model is greater than or equal to a fourth value.

[0233] Optionally, the first configuration is used for availability confirmation.

[0234] Optionally, the first configuration comprises a first CSI report configuration.

[0235] Optionally, in a case where the second information comprises the information indicating that the first configuration is available, the transceiver 1520 is configured to receive information indicating that the first configuration is activated; the processing unit 1510 is configured to activate the first configuration to obtain the inference result of the first model; and the transceiver 1520 is configured to transmit the inference result of the first model.

[0236] Optionally, the first information further comprises a second configuration for inference, the second configuration carrying a second condition indicating a requirement for a second model associated with the second configuration and / or a requirement for a predicted performance of the second model; and the transceiver 1520 is configured to transmit third information indicating that the second configuration is unavailable in a case where the second model and / or the predicted performance of the second model does not satisfy the second condition.

[0237] Optionally, the third information further indicates a reason why the second configuration is unavailable, the reason comprising that the second model and / or the predicted performance of the second model does not satisfy the second condition.

[0238] Optionally, the transceiver 1520 is configured to receive a third condition indicating a requirement for the second model and / or a requirement for the predicted performance of the second model; and transmit information indicating that the second configuration is available in a case where the second model and / or the predicted performance of the second model satisfies the third condition.

[0239] In another possible implementation, the apparatus 1500 is configured to implement the above describedFigure 9 The network device corresponds to the steps in the method shown.

[0240] The processing unit 1510 is configured to determine first information, and the transceiver 1520 is configured to send the first information, the first information including a first configuration for inference, the first configuration carrying a first condition, the first condition being used to indicate a requirement for a first model associated with the first configuration and / or a requirement for a predicted performance of the first model; and receive second information in a case where the first model and / or the predicted performance of the first model meets the first condition, the second information including an inference result of the first model or information indicating that the first configuration is available.

[0241] The first condition and the first configuration can refer to the description above, which will not be described here again.

[0242] Optionally, in a case where the second information includes the information indicating that the first configuration is available, the transceiver 1520 is configured to send information indicating that the first configuration is activated, and receive the inference result of the first model.

[0243] Optionally, the first information further includes a second configuration for inference, the second configuration carrying a second condition, the second condition being used to indicate a requirement for a second model associated with the second configuration and / or a requirement for a predicted performance of the second model; and the transceiver 1520 is configured to receive third information in a case where a performance of the second model and / or the predicted performance of the second model does not meet the second condition, the third information being used to indicate that the second configuration is unavailable.

[0244] Optionally, the third information is further used to indicate that a reason why the second configuration is unavailable includes that the second model and / or the predicted performance of the second model does not meet the second condition.

[0245] Optionally, the transceiver 1520 is configured to send a third condition, the third condition being used to indicate the requirement for the second model and / or the requirement for the predicted performance of the second model; and receive information indicating that the second configuration is available in a case where the performance of the second model and / or the predicted performance of the second model meets the third condition.

[0246] It should be understood that the communication apparatus 1500 is embodied in the form of functional units here. The term "unit" here can refer to an application specific integrated circuit (ASIC), an electronic circuit, a processor (for example, a shared processor, a dedicated processor, or a group processor and the like) and a memory for executing one or more software or firmware programs, a combination 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 communication apparatus 1500 can be embodied in the terminal device or the network device in the above embodiments, and the communication apparatus 1500 can be used to execute the respective processes and / or steps corresponding to the terminal device or the network device in the above method embodiments. To avoid repetition, details are not described here.

[0247] The communication apparatus 1500 described above has the functions of implementing the respective steps performed by the terminal device or the network device in the above method; the above functions can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above functions. In the embodiments of the present application, Figure 15 The communication apparatus 1500 in the above embodiments can also be a chip, for example: SOC.

[0248] As shown in Figure 16 , the communication apparatus 1600 can include a processor 1610, a transceiver 1620, and a memory 1630. Among them, the processor 1610, the transceiver 1620 and the memory 1630 communicate with each other through an internal connection path, the memory 1630 is used to store instructions, and the processor 1610 is used to execute the instructions stored in the memory 1630 to control the transceiver 1620 to send and / or receive signals.

[0249] It should be understood that the communication device 1600 can be embodied as the terminal device or the network device in the above-described embodiments, and can be used to perform the steps and / or processes corresponding to the terminal device or the network device in the above-described method embodiments. Optionally, the memory 1630 can include read-only memory and random access memory, and provide instructions and data for the processor. A part of the memory can also include non-volatile random access memory. For example, the memory can also store device type information. The processor 1610 can be used to execute the instructions stored in the memory, and when the processor 1610 executes the instructions stored in the memory, the processor 1610 is used to perform the steps and / or processes of the above-described method embodiments. The transceiver 1620 can include a transmitter and a receiver, and the transmitter can be used to implement the steps and / or processes corresponding to the transmitter for performing the sending actions in the above-described transceiver. For example, the transmitter can be used to send information to another device through the antenna. The receiver can be used to implement the steps and / or processes corresponding to the receiver for performing the receiving actions in the above-described transceiver. For example, the receiver can be used to receive information from another device through the antenna.

[0250] It should be understood that in the embodiments of the present application, the processor can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0251] In the implementation process, the steps of the above-described method can be completed by the integrated logic circuits of hardware in the processor or the instructions in the form of software. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as the execution completed by the hardware processor, or executed by the combination of hardware and software modules in the processor. The software modules can be located in the storage media in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, register, etc. The storage medium is located in the memory, and the processor executes the instructions in the memory, and combines the hardware to complete the steps of the above-described method. To avoid repetition, it will not be described in detail here.

[0252] The embodiments of the present application also provide a processor. The processor can execute the processes and / or steps corresponding to the terminal device or the network device in the above-described method embodiments, and to avoid repetition, it will not be described here.

[0253] The embodiments of the present application further provide a chip or a chip system. The processor can execute each process and / or step corresponding to the terminal device or the network device in the above-mentioned method embodiments, and details are not repeated here to avoid repetition.

[0254] The embodiments of the present application further provide a communication system, which can include the terminal device and the network device.

[0255] The embodiments of the present application further provide a computer readable storage medium for storing a computer program for implementing the method shown in the above-mentioned method embodiments.

[0256] The embodiments of the present application further provide a computer program product, which includes a computer program (also referred to as code or instructions), and when the computer program runs on a computer, the computer can execute the method shown in the above-mentioned method embodiments.

[0257] Those skilled in the art can understand that the modules and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0258] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-mentioned system, device and module can refer to the corresponding processes in the above-mentioned method embodiments, and details are not repeated here.

[0259] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be realized by other ways. For example, the above-mentioned device embodiments are only schematic, for example, the division of modules is only a logical function division, and actual implementation can have another division manner, for example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0260] The modules described as separate components can or can not be physically separated, and the components shown as modules can or can not be physical modules, i.e. they can be located in one place or distributed on multiple network modules. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment.

[0261] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can be physically present alone, or two or more modules can be integrated in one module.

[0262] If the functions are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the present application that essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a terminal device or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0263] The above is only a specific implementation of the present application, but the protection scope of the embodiments of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the embodiments of the present application, which should be covered in the protection scope of the embodiments of the present application. Therefore, the protection scope of the embodiments of the present application should be subject to the protection scope of the claims.

Claims

1. A communication method characterized by comprising: Comprising: receiving first information, the first information comprising a first configuration for inference, the first configuration carrying a first condition, the first condition being used to indicate a requirement for a first model associated with the first configuration and / or a requirement for a predicted performance of the first model; in a case where the first model and / or the predicted performance of the first model satisfies the first condition, sending second information, the second information comprising an inference result of the first model or information indicating that the first configuration is available.

2. The method of claim 1, wherein, The first model satisfies the first condition, comprising: a complexity of the first model being less than or equal to a first value, and / or, a memory occupied by the first model being less than or equal to a second value.

3. The method according to claim 1 or 2, characterized in that, The predicted performance of the first model satisfies the first condition, comprising: a loss value of the first model being less than or equal to a third value, and / or, a prediction accuracy of the first model being greater than or equal to a fourth value.

4. The method according to any one of claims 1 to 3, characterized in that, The first configuration is used for availability confirmation.

5. The method according to any one of claims 1 to 4, characterized in that, The first configuration comprises a first CSI report configuration.

6. The method according to any one of claims 1 to 5, characterized in that, In a case where the second information comprises information indicating that the first configuration is available, the method further comprises: receiving information indicating that the first configuration is activated; activating the first configuration to obtain the inference result of the first model; sending the inference result of the first model.

7. The method according to any one of claims 1 to 6, characterized in that, The first information further comprises a second configuration for inference, the second configuration carrying a second condition, the second condition being used to indicate a requirement for a second model associated with the second configuration and / or a requirement for a predicted performance of the second model; The method further comprises: in a case where the second model and / or the predicted performance of the second model does not satisfy the second condition, sending third information, the third information being used to indicate that the second configuration is unavailable.

8. The method of claim 7, wherein, The third information is further used to indicate a reason why the second configuration is unavailable, comprising: the second model and / or the predicted performance of the second model does not satisfy the second condition.

9. The method of claim 8, wherein, The method further comprises: receiving a third condition, the third condition being used to indicate a requirement for the second model and / or a requirement for the predicted performance of the second model; in a case where the second model and / or the predicted performance of the second model satisfies the third condition, sending information indicating that the second configuration is available.

10. A communication method characterized by comprising: Comprising: sending first information, the first information comprising a first configuration for inference, the first configuration carrying a first condition, the first condition being used to indicate a requirement for a first model associated with the first configuration and / or a requirement for a predicted performance of the first model; in a case where the first model and / or the predicted performance of the first model satisfies the first condition, receiving second information, the second information comprising an inference result of the first model or information indicating that the first configuration is available.

11. The method of claim 10, wherein, The first model satisfies the first condition, comprising: a complexity of the first model being less than or equal to a first value, and / or, a memory occupied by the first model being less than or equal to a second value.

12. The method according to claim 10 or 11, characterized in that, The prediction performance of the first model satisfies the first condition, including: a loss value of the first model is less than or equal to a third value, and / or a prediction accuracy of the first model is greater than or equal to a fourth value.

13. The method according to any one of claims 10 to 12, characterized in that, The first configuration is for availability confirmation.

14. The method according to any one of claims 10 to 13, characterized in that, The first configuration includes a first CSI reporting configuration.

15. The method according to any one of claims 10 to 14, characterized in that, In a case where the second information includes information for indicating that the first configuration is available, the method further includes: sending information for indicating activation of the first configuration; receiving an inference result of the first model.

16. The method according to any one of claims 10 to 15, characterized in that, The first information further includes a second configuration for inference, the second configuration carrying a second condition for indicating a requirement for a second model associated with the second configuration and / or a requirement for a prediction performance of the second model; The method further includes: In a case where the performance of the second model and / or the prediction performance of the second model does not satisfy the second condition, receiving third information for indicating that the second configuration is unavailable.

17. The method of claim 16, wherein, The third information is further for indicating a reason that the second configuration is unavailable, including: the second model and / or the prediction performance of the second model does not satisfy the second condition.

18. The method of claim 17, wherein, The method further includes: sending a third condition for indicating a requirement for the second model and / or a requirement for a prediction performance of the second model; In a case where the performance of the second model and / or the prediction performance of the second model satisfies the third condition, receiving information for indicating that the second configuration is available.

19. A communications device, characterized by comprising means for performing the method of any one of claims 1 to 9, or means for performing the method of any one of claims 10 to 18.

20. A communications device, characterized by comprising: a processor coupled with a memory, the memory for storing a computer program, when the processor invokes the computer program, causing the communication device to perform the method of any one of claims 1 to 9, or causing the communication device to perform the method of any one of claims 10 to 18.

21. A chip, characterized by comprising: a processor for reading instructions stored in a memory, when the processor executes the instructions, causing the chip to implement the method of any one of the above claims 1 to 9, or causing the chip to implement the method of any one of the above claims 10 to 18.

22. A communication system, characterized by comprising: a device for implementing the method of any one of the above claims 1 to 9 and a device for implementing the method of any one of the above claims 10 to 18.

23. A computer-readable storage medium, characterized in that, The computer readable storage medium has stored thereon a computer program, when the computer program is run on a computer, causing the method of any one of claims 1 to 9 to be performed, or causing the method of any one of claims 10 to 18 to be performed.

24. A computer program product, characterised in that, The computer program product includes instructions, when the instructions are executed, causing the method of any one of claims 1 to 9 to be performed, or causing the method of any one of claims 10 to 18 to be performed.

Images