Communication method and apparatus

Abstract

The present application discloses a communication method and apparatus. The method comprises: a first network element receives first indication information from a second network element or sends first indication information to the second network element, the first indication information being used for indicating the format of first data; the first network element obtains the first data on the basis of the format of the first data, the first data being used for training a first model. By using the present application, alignment of data values at both ends of an AI model can be achieved, and the CSI recovery performance of the AI model can be improved.

Description

Communication method and device

[0001] This application claims priority to the Chinese patent application with application number 202410412972.5 filed with the State Intellectual Property Office of China on April 3, 2024, and priority to the Chinese patent application with the invention name “A Communication Method and Device”, all contents of which are incorporated by reference into this application. Technical Field

[0002] The present invention relates to the field of communication technology, and in particular to a communication method and device. Background Art

[0003] With the continuous development of wireless communication technology, to support more services and enable spatial multiplexing between users or data streams, base stations (BSs) need to obtain channel state information (CSI) for the downlink channel. Based on this CSI, they then determine precoding matrices and user scheduling. In currently widely used frequency division duplex (FDD)-based communication systems, uplink and downlink channels are not reciprocal. Therefore, the base station requires user equipment (UE) to obtain CSI for the downlink channel through uplink feedback. For example, the base station transmits a downlink reference signal to the UE, which then receives it. Since the UE knows the transmission information of the downlink reference signal, it can estimate (measure) the downlink channel traversed by the received downlink reference signal. The UE then generates CSI based on the measured downlink channel matrix and feeds this CSI back to the base station.

[0004] At present, in order to meet the higher requirements of communication systems in terms of system capacity, communication delay and other indicators, the scale of antenna arrays continues to increase, the number of supported antenna ports increases, and the dimensions of the corresponding channel matrix and precoding matrix increase. In this scenario, the overhead of the base station sending the reference signal increases, and the CSI obtained by the base station is usually compressed to a large extent, and the accuracy of the CSI is low. In order to improve the performance of the communication system, CSI feedback can be implemented based on a two-end artificial intelligence (AI) model. The AI ​​model consists of two sub-models: an encoder and a decoder. The encoder and decoder of the AI ​​model are usually trained together and can be used in combination with each other. For example, the UE can compress and quantize the CSI through the encoder, and the base station can recover the CSI through the decoder. However, if the two ends of the AI ​​model are trained separately, the encoder and decoder may not understand each other, which in turn affects the CSI recovery performance of the decoder. Summary of the Invention

[0005] The embodiments of the present application provide a communication method and device that can achieve data value alignment on both ends of an AI model and improve the CSI recovery performance of the AI ​​model.

[0006] In a first aspect, an embodiment of the present application provides a communication method, which can be performed by a first network element, or by a module (such as a processor, a chip, or a chip system) applied to the first network element, or by a logical node, a logical module, or software that can implement all or part of the functions of the first network element. The method includes:

[0007] Receive first indication information from a second network element or send first indication information to the second network element, where the first indication information is used to indicate a format of first data; obtain the first data based on the format of the first data, where the first data is used to train a first model.

[0008] By transmitting the first indication information, the format of the data used by the first network element and the second network element is made the same, thereby achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0009] In one possible design, a parameter value sequence is received from the second network element, where the parameter value sequence includes M parameter values ​​corresponding to the first input data and N parameter values ​​corresponding to the first output data, where M is an integer greater than 0, and N is an integer greater than 0. Receiving the parameter value sequence facilitates the first network element to obtain the first data for training the first model, thereby achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0010] In another possible design, the parameter value sequence is converted into the first data, where the first data includes P first input data and Q first output data, where P is an integer greater than 0 and less than or equal to M, and Q is an integer greater than 0 and less than or equal to N. By parsing the parameter value sequence, the first network element obtains the first data for training the first model, which facilitates data value alignment on both ends of the AI ​​model, thereby improving the CSI recovery performance of the AI ​​model.

[0011] In another possible design, the format of the first data includes at least one of the following: an arrangement of input data, an arrangement of output data, a parsing format of input data parameter values, input data dimensions, a parsing format of output data parameter values, output data dimensions, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid period corresponding to the first data, a valid time corresponding to the first data, or an invalid time corresponding to the first data. This facilitates data value alignment on both ends of the AI ​​model, thereby improving the CSI recovery performance of the AI ​​model.

[0012] In a second aspect, an embodiment of the present application provides a communication method, which can be performed by a second network element, or by a module (such as a processor, a chip, or a chip system) applied to the second network element, or by a logical node, a logical module, or software that can implement all or part of the functions of the second network element. The method includes:

[0013] First indication information is sent to a first network element or received from the first network element, where the first indication information is used to indicate a format of first data.

[0014] By transmitting the first indication information, the format of the data used by the first network element and the second network element is made the same, which is conducive to achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0015] In one possible design, a parameter value sequence is sent to the first network element, where the parameter value sequence includes M parameter values ​​corresponding to the first input data and N parameter values ​​corresponding to the first output data, where M is an integer greater than 0, and N is an integer greater than 0. Sending the parameter value sequence facilitates the first network element to obtain the first data for training the first model, thereby achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0016] In another possible design, the format of the first data includes at least one of the following: an arrangement of input data, an arrangement of output data, a parsing format of input data parameter values, input data dimensions, a parsing format of output data parameter values, output data dimensions, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid period corresponding to the first data, a valid time corresponding to the first data, or an invalid time corresponding to the first data. This facilitates data value alignment on both ends of the AI ​​model, thereby improving the CSI recovery performance of the AI ​​model.

[0017] In a third aspect, an embodiment of the present application provides a communication device that can implement the method of the first aspect or any possible implementation of the first aspect. The communication device includes corresponding units or modules for executing the above-mentioned method. The units or modules included in the communication device can be implemented through software and / or hardware. The communication device can be, for example, a first network element, or a chip, chip system, or processor that supports the first network element to implement the above-mentioned method. It can also be a logical node, logical module, or software that can implement all or part of the functions of the first network element.

[0018] In a fourth aspect, an embodiment of the present application provides a communication device that can implement the method in the second aspect or any possible implementation of the second aspect. The communication device includes corresponding units or modules for executing the above-mentioned method. The units or modules included in the communication device can be implemented through software and / or hardware. The communication device can be, for example, a second network element, or a chip, chip system, or processor that supports the second network element to implement the above-mentioned method. It can also be a logical node, logical module, or software that can implement all or part of the functions of the second network element.

[0019] In a fifth aspect, the present application provides a communication device, which includes a processor and a memory, wherein the memory is used to store a computer program; the processor is used to execute the computer program stored in the memory, so that the communication device performs the method as described in any one of the first aspects.

[0020] In a sixth aspect, the present application provides a communication device, comprising a processor and a memory, wherein the memory is used to store a computer program; the processor is used to execute the computer program stored in the memory so that the communication device performs a method as described in any one of the second aspects.

[0021] In the seventh aspect, an embodiment of the present application provides a communication system, which includes at least one first network element and at least one second network element, the first network element is used to execute the steps in the above-mentioned first aspect, and the second network element is used to execute the steps in the above-mentioned second aspect.

[0022] In an eighth aspect, an embodiment of the present application provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the methods in the above aspects.

[0023] In a ninth aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the methods in each of the above aspects.

[0024] In the tenth aspect, an embodiment of the present application provides a chip, including a processor and a communication interface, wherein the communication interface is used to communicate with an external device or an internal device, and the processor is used to implement the methods of the above aspects.

[0025] In one possible design, the chip may further include a memory storing a computer program or instructions, and the processor is configured to execute the computer program or instructions stored in the memory, or other programs or instructions. When the computer program or instructions are executed, the processor is configured to implement the aforementioned various aspects of the method.

[0026] In another possible design, the chip can be integrated on the first network element or the second network element. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0028] FIG1a is a schematic diagram of a communication system applicable to an embodiment of the present application;

[0029] FIG1b is a schematic diagram of another communication system applicable to an embodiment of the present application;

[0030] FIG2 is a schematic diagram of a possible application framework in a communication system;

[0031] FIG3 is a schematic diagram of another possible application framework in a communication system;

[0032] FIG4 is a schematic diagram of the structure of a neuron;

[0033] FIG5 is a schematic diagram of the structure of a neural network;

[0034] FIG6 is a schematic diagram of an AI application framework;

[0035] FIG7 is a flow chart of a communication method provided in an embodiment of the present application;

[0036] FIG8 is a flow chart of another communication method provided in an embodiment of the present application;

[0037] FIG9 is a flow chart of another communication method provided in an embodiment of the present application;

[0038] FIG10 is a schematic structural diagram of a communication device provided in an embodiment of the present application;

[0039] FIG11 is a schematic structural diagram of another communication device provided in an embodiment of the present application;

[0040] FIG12 is a schematic structural diagram of a first network element provided in an embodiment of the present application;

[0041] FIG13 is a schematic structural diagram of a second network element provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] Some of the terms used in this application are explained below to facilitate understanding by those skilled in the art.

[0043] (1) AI: Giving machines human intelligence, using computer hardware and software to simulate certain human intelligent behaviors, including machine learning and many other methods.

[0044] (2) Machine learning (ML): Learning models or rules from raw data. There are many different machine learning methods, such as neural networks, decision trees, support vector machines, etc.

[0045] (3) AI model: This refers to a functional model that maps inputs of a certain dimension to outputs of a certain dimension, whose model parameters are obtained through machine learning training. The type of AI model can be a neural network, linear regression model, decision tree model, support vector machine (SVM), Bayesian network, Q learning model, or other machine learning model.

[0046] (4) Neural network (NN): This refers to an artificial neural network, which is a mathematical model that imitates the behavioral characteristics of animal neural networks and performs distributed parallel information processing. It is a special form of AI model.

[0047] (5) Deep neural network (DNN): A neural network with multiple hidden layers.

[0048] (6) Deep learning (DL): machine learning using deep neural networks.

[0049] (7) Auto-encoders (AE) model: also known as bilateral model, collaborative model, dual model or two-side model, etc. A two-side model refers to a model composed of multiple sub-models. The multiple sub-models that constitute the model need to match each other, and the multiple sub-models can be deployed in different nodes.

[0050] (8) CSI: Also known as channel information or channel environment information, it is a type of information that can reflect channel characteristics and channel quality.

[0051] (9) Model training: By selecting a suitable loss function, the model parameters are trained using an optimization algorithm to minimize the loss function value.

[0052] (10) Loss function: used to measure the difference between the model's predicted value and the true value.

[0053] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.

[0054] It should be understood that, in the description of this application, "at least one" means one or more, and "a plurality" means two or more. In addition, unless otherwise specified, the terms "first" and "second" are used only for descriptive purposes and are not to be construed as indicating or implying relative importance or order.

[0055] The technical solutions provided in this application can be applied to various communication systems, such as: fifth generation (5G) or new radio (NR) systems, long term evolution (LTE) systems, LTE FDD systems, LTE time division duplex (TDD) systems, wireless local area networks (WLAN) systems, satellite communication systems, future communication systems, such as sixth generation (6G) mobile communication systems, or fusion systems of multiple systems. The technical solutions provided in this application can also be applied to device to device (D2D) communication, vehicle to everything (V2X) communication, machine to machine (M2M) communication, machine type communication (MTC), and Internet of Things (IoT) communication systems or other communication systems.

[0056] A network element in a communication system can send a signal to another network element or receive a signal from another network element. The signal may include information, signaling, or data, etc. The network element can also be replaced by an entity, a network entity, a device, a communication device, a communication module, a node, a communication node, etc. The present disclosure uses the network element as an example for description. For example, the communication system may include at least one terminal device and at least one network device. The network device can send a downlink signal to the terminal device, and / or the terminal device can send an uplink signal to the network device. It is understandable that the terminal device in the present disclosure can be replaced by the first network element, and the network device can be replaced by the second network element, and the two perform the corresponding communication methods in the present disclosure.

[0057] As shown in Figure 1a, Figure 1a is a schematic diagram of a communication system applicable to the communication method of an embodiment of the present application. The communication system may include at least one network device, such as the network device 101 shown in Figure 1a; the communication system may also include at least one terminal device, such as the terminal device 102 and the terminal device 103 shown in Figure 1a. The network device 101 and the terminal device (such as the terminal device 102 and the terminal device 103) can communicate via a wireless link. The communication devices in the communication system, for example, the network device 101 and the terminal device 102, can communicate via multi-antenna technology.

[0058] In an embodiment of the present application, the terminal device may also be referred to as UE, access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent or user equipment.

[0059] The terminal device may be a device that provides voice / data, such as a handheld device or vehicle-mounted device with a wireless connection function. At present, some examples of terminals are: mobile phones, tablet computers, laptop computers, PDAs, mobile internet devices (MIDs), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self-driving, wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, wireless terminals in smart homes, cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (PDAs), handheld devices with wireless communication capabilities, computing devices or other processing devices connected to wireless modems, wearable devices, terminal devices in 5G networks or future evolved public land mobile communication networks (PLMNs). The terminal equipment in the network (PLMN), etc., is not limited to this in the embodiments of the present application.

[0060] As an example and not a limitation, in the embodiments of the present application, the terminal device may also be a wearable device. Wearable devices may also be called wearable smart devices, which are a general term for wearable devices that are intelligently designed and developed using wearable technology for daily wear, such as glasses, gloves, watches, clothes, and shoes. A wearable device is a portable device that is worn directly on the body or integrated into the user's clothes or accessories. Wearable devices are not only hardware devices, but also achieve powerful functions through software support, data interaction, and cloud interaction. Broadly speaking, wearable smart devices include those that are fully functional, large in size, and can achieve complete or partial functions without relying on smartphones, such as smart watches or smart glasses, as well as those that only focus on a certain type of application function and need to be used in conjunction with other devices such as smartphones, such as various smart bracelets and smart jewelry for vital sign monitoring.

[0061] In the embodiment of the present application, the device for realizing the function of the terminal device can be a terminal device, or a device that can support the terminal device to realize the function, such as a chip, a chip system, or a processor, etc. It can also be a logical node, a logical module or software that can realize all or part of the terminal device function. The device can be installed in the terminal device or used in combination with the terminal device. In the embodiment of the present application, the chip system can be composed of chips, or it can include chips and other discrete devices. In the embodiment of the present application, only the device for realizing the function of the terminal device is used as an example for explanation, and the solution of the embodiment of the present application is not limited.

[0062] The network side in the embodiments of the present application may include network equipment, wherein the network equipment includes equipment for communicating with the terminal device, and the network equipment includes access network equipment or radio access network equipment, such as a base station; or operation, administration and maintenance (OAM) equipment or core network (CN) equipment. The access network equipment in the embodiments of the present application may refer to a radio access network (RAN) node (or device) that connects the terminal device to the wireless network. A base station may broadly cover various names as follows, or replace the following names, such as: NodeB, evolved NodeB (eNB), next generation NodeB (gNB), relay station, access point, transmitting and receiving point (TRP), transmitting point (TP), master station, secondary station, multi-standard radio (motor slide retainer, MSR) node, home base station, network controller, access node, wireless node, access point (AP), transmission node, transceiver node, baseband unit (BBU), remote radio unit (RRU), active antenna unit (AAU), remote radio head (RRH), central unit (CU), distributed unit (DU), radio unit (RU), positioning node, etc. A base station may be a macro base station, a micro base station, a relay node, a donor node, or the like, or a combination thereof. The base station may also refer to a communication module, modem or chip used to be set in the aforementioned equipment or device. The base station may also be a mobile switching center and a device that performs the base station function in D2D, V2X, and M2M communications, a network side device in a 6G network, a device that performs the base station function in future communication systems, etc. The base station can support networks with the same or different access technologies. Optionally, the RAN node may also be a server, a wearable device, a vehicle or an on-board device, etc. For example, the access network device in the V2X technology may be a road side unit (RSU). The embodiments of the present application do not limit the specific technology and specific device form adopted by the network equipment.

[0063] Base stations can be fixed or mobile. For example, a helicopter or drone can be configured to act as a mobile base station, and one or more cells can move based on the location of the mobile base station. In other examples, a helicopter or drone can be configured to act as a device that communicates with another base station.

[0064] In one possible scenario, a radio access network device may also be a module or unit that performs some of the functions of a base station, for example, a CU or a DU. In another possible scenario, multiple radio access network devices collaborate to assist a terminal in achieving wireless access, with different radio access network devices each performing some of the functions of a base station. For example, a radio access network device may be a CU, DU, CU-control plane (CP), CU-user plane (UP), or RU. The CU and DU may be configured separately or included in the same network element, such as a BBU. The RU may be included in a radio frequency device or radio frequency unit, such as an RRU, AAU, or RRH.

[0065] In different systems, CU (or CU-CP and CU-UP), DU or RU may also have different names, but those skilled in the art can understand their meanings. For example, in the ORAN system, CU may also be called an open central unit (O-CU), DU may also be called an open distributed unit (O-DU), CU-CP may also be called O-CU-CP, CU-UP may also be called O-CU-UP, and RU may also be called O-RU. For the convenience of description, this application takes DU and RU as examples for description. Any unit in the DU and RU in this application may be implemented by a software module, a hardware module, or a combination of a software module and a hardware module. The embodiments of this application do not limit the specific technology and specific device form adopted by the wireless access network device. For the convenience of description, the following description takes the base station as an example of the wireless access network device.

[0066] The RAN node may support one or more types of fronthaul interfaces, and different fronthaul interfaces correspond to DUs and RUs with different functions. If the fronthaul interface between the DU and the RU is a common public radio interface (CPRI), the DU is configured to implement one or more baseband functions, and the RU is configured to implement one or more radio frequency functions. If the fronthaul interface between the DU and the RU is another interface, relative to the CPRI, it moves part of the downlink and / or uplink baseband functions, such as, for downlink, precoding, digital beamforming (BF), or one or more of inverse fast Fourier transform (IFFT) / adding a cyclic prefix (CP), from the DU to the RU for implementation, and for uplink, digital BF, or one or more of fast Fourier transform (FFT) / removing CP, from the DU to the RU for implementation. In one possible implementation, the interface may be an enhanced common public radio interface (eCPRI). In the eCPRI architecture, the division between DU and RU is different, corresponding to different types (category, Cat) of eCPRI, such as eCPRI Cat A, B, C, D, E, and F.

[0067] Taking eCPRI Cat A as an example, for downlink transmission, based on layer mapping, the DU is configured to implement layer mapping and one or more functions before it (i.e., one or more functions among coding, rate matching, scrambling, modulation, and layer mapping), while other functions after layer mapping (e.g., one or more functions among resource element (RE) mapping, digital BF, or IFFT / CP addition) are moved to the RU for implementation. For uplink transmission, based on RE demapping, the DU is configured to implement demapping and one or more functions before it (i.e., one or more functions among decoding, rate matching, descrambling, demodulation, inverse discrete Fourier transform (IDFT), channel equalization, and RE demapping), while other functions after demapping (e.g., one or more functions among digital BF or FFT / CP removal) are moved to the RU for implementation. It is understood that for the functional description of the DU and RU corresponding to various types of eCPRI, please refer to the eCPRI protocol and will not be repeated here.

[0068] In the embodiments of the present application, the device for realizing the function of the network device can be a network device; it can also be a device that can support the network device to realize the function, such as a chip system, a hardware circuit, a software module, or a hardware circuit plus a software module, or it can be a logical node, a logical module or software that can realize all or part of the network device function. The device can be installed in the network device or used in combination with the network device. In the embodiments of the present application, only the device for realizing the function of the network device is used as an example for description, and the scheme of the embodiments of the present application is not limited.

[0069] The network device and / or terminal device can be deployed on land, including indoors or outdoors, handheld or vehicle-mounted; it can also be deployed on the water surface; it can also be deployed on aircraft, balloons and satellites in the air. The embodiments of this application do not limit the scenarios in which the network device and the terminal device are located. In addition, the terminal device and the network device can be hardware devices, or they can be software functions running on dedicated hardware, software functions running on general-purpose hardware, such as virtualization functions instantiated on a platform (e.g., a cloud platform), or entities including dedicated or general-purpose hardware devices and software functions. This application does not limit the specific forms of the terminal device and the network device.

[0070] In wireless communication networks, such as mobile communication networks, the services supported by the networks are becoming increasingly diverse, and therefore the demands that need to be met are becoming increasingly diverse. For example, the network needs to be able to support ultra-high speeds, ultra-low latency, and / or ultra-large connections. This feature makes network planning, network configuration, and / or resource scheduling increasingly complex. In addition, as network functionality becomes increasingly powerful, such as supporting higher spectrum, supporting high-order multiple input multiple output (MIMO) technology, supporting beamforming, and / or supporting new technologies such as beam management, network energy saving has become a hot research topic. These new demands, new scenarios, and new features have brought unprecedented challenges to network planning, operation and maintenance, and efficient operation. To meet this challenge, artificial intelligence technology can be introduced into wireless communication networks to achieve network intelligence.

[0071] In order to support AI technology in wireless networks, AI nodes may also be introduced into the network.

[0072] Optionally, the AI ​​node can be deployed in one or more of the following locations in the communication system: access network equipment, terminal equipment, or core network equipment. Alternatively, the AI ​​node can be deployed separately, for example, in a location other than any of the above devices, such as a host or cloud server in an over-the-top (OTT) system. The AI ​​node can communicate with other devices in the communication system, such as one or more of the following: network equipment, terminal equipment, or core network elements.

[0073] It is understood that this application does not limit the number of AI nodes. For example, when there are multiple AI nodes, the multiple AI nodes can be divided based on function, such as different AI nodes are responsible for different functions.

[0074] It can also be understood that AI nodes can be independent devices, or they can be integrated into the same device to implement different functions, or they can be network elements in hardware devices, or they can be software functions running on dedicated hardware, or they can be virtualized functions instantiated on a platform (for example, a cloud platform). This application does not limit the specific form of the above-mentioned AI nodes.

[0075] An AI node can be an AI network element or an AI module.

[0076] As shown in Figure 1b, Figure 1b is a schematic diagram of another communication system applicable to the communication method of an embodiment of the present application. Compared to the communication system shown in Figure 1a, the communication system shown in Figure 1b also includes an AI network element 104. AI network element 104 is used to perform AI-related operations, such as building a training dataset or training an AI model.

[0077] In one possible implementation, the network device 101 may send data related to the training of the AI ​​model to the AI ​​network element 104, which constructs a training data set and trains the AI ​​model. For example, the data related to the training of the AI ​​model may include data reported by the terminal device. The AI ​​network element 104 may send the results of the operations related to the AI ​​model to the network device 101, and forward them to the terminal device through the network device 101. For example, the results of the operations related to the AI ​​model may include at least one of the following: an AI model that has completed training, an evaluation result or a test result of the model, etc. Exemplarily, a portion of the trained AI model may be deployed on the network device 101, and another portion may be deployed on the terminal device. Alternatively, the trained AI model may be deployed on the network device 101. Alternatively, the trained AI model may be deployed on the terminal device.

[0078] It should be understood that Figure 1b illustrates only the example of a direct connection between AI network element 104 and network device 101. In other scenarios, AI network element 104 may also be connected to a terminal device. Alternatively, AI network element 104 may be connected to both network device 101 and a terminal device simultaneously. Alternatively, AI network element 104 may be connected to network device 101 through a third-party network element. This embodiment of the present application does not limit the connection relationship between the AI ​​network element and other network elements.

[0079] The AI ​​network element 104 may also be provided as a module in a network device and / or a terminal device, for example, in the network device 101 or the terminal device shown in FIG. 1 a .

[0080] It should be noted that Figures 1a and 1b are simplified schematic diagrams for ease of understanding. For example, the communication system may also include other devices, such as wireless relay devices and / or wireless backhaul devices, which are not shown in Figures 1a and 1b. In actual applications, the communication system may include multiple network devices and multiple terminal devices. The embodiments of the present application do not limit the number of network devices and terminal devices included in the communication system.

[0081] As shown in Figure 2, Figure 2 is a schematic diagram of a possible application framework in a communication system. Network elements in the communication system are connected through interfaces (for example, NG, Xn) or air interfaces. One or more AI modules are provided in one or more devices of these network element nodes, such as core network equipment, access network nodes (RAN nodes), terminals or OAM (for the sake of clarity, only one is shown in Figure 2). The access network node can be a separate RAN node, or it can include multiple RAN nodes, for example, including CU and DU. The CU and / or DU can also be provided with one or more AI modules. Optionally, the CU can also be split into CU-CP and CU-UP. One or more AI models are provided in the CU-CP and / or CU-UP.

[0082] The AI ​​module is used to implement the corresponding AI function. The AI ​​modules deployed in different network elements may be the same or different. The model of the AI ​​module can implement different functions according to different parameter configurations. The model of the AI ​​module can be configured based on one or more of the following parameters: structural parameters (for example, the number of neural network layers, the width of the neural network, the connection relationship between layers, the weight of the neuron, the activation function of the neuron, or at least one of the bias in the activation function), input parameters (for example, the type of input parameters and / or the dimension of the input parameters), or output parameters (for example, the type of output parameters and / or the dimension of the output parameters). Among them, the bias in the activation function can also be called the bias of the neural network.

[0083] An AI module can have one or more models. A model can infer an output, which includes one or more parameters. The learning, training, or inference processes of different models can be deployed on different nodes or devices, or on the same node or device.

[0084] As shown in Figure 3, Figure 3 is a schematic diagram of another possible application framework in a communication system. The communication system includes a RAN intelligent controller (RIC). For example, the RIC can be the AI ​​module shown in Figure 3, which is used to implement AI-related functions. The RIC includes a near-real-time RIC (near-real time RIC, Near-RT RIC) and a non-real-time RIC (non-real time RIC, Non-RT RIC). Among them, the non-real-time RIC mainly processes non-real-time information, such as data that is not sensitive to delay, and the delay of this data can be in the order of seconds. The real-time RIC mainly processes near-real-time information, such as data that is relatively sensitive to delay, and the delay of this data is in the order of tens of milliseconds.

[0085] The near real-time RIC is used for model training and reasoning. For example, it is used to train an AI model and use the AI ​​model for reasoning. The near real-time RIC can obtain network-side and / or terminal-side information from the RAN node (e.g., CU, CU-CP, CU-UP, DU and / or RU) and / or the terminal. This information can be used as training data or reasoning data. Optionally, the near real-time RIC can deliver the reasoning result to the RAN node and / or the terminal. Optionally, the reasoning result can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the near real-time RIC delivers the reasoning result to the DU, and the DU sends it to the RU.

[0086] The non-real-time RIC is also used for model training and reasoning. For example, it is used to train an AI model and use the model for reasoning. The non-real-time RIC can obtain network-side and / or terminal-side information from the RAN node (e.g., CU, CU-CP, CU-UP, DU and / or RU) and / or the terminal. This information can be used as training data or reasoning data, and the reasoning results can be submitted to the RAN node and / or the terminal. Optionally, the reasoning results can be exchanged between the CU and the DU, and / or between the DU and the RU. For example, the non-real-time RIC submits the reasoning results to the DU, and the DU sends it to the RU.

[0087] The near real-time RIC and non-real-time RIC can each be set up as a separate network element. Optionally, the near real-time RIC and non-real-time RIC can also be part of other devices. For example, the near real-time RIC is set up in a RAN node (e.g., a CU or DU), while the non-real-time RIC is set up in an OAM, a cloud server, a core network device, or other network devices.

[0088] The following further illustrates the machine learning in the embodiments of the present application.

[0089] Machine learning is an important technical approach to achieving AI. Machine learning can be divided into supervised learning, unsupervised learning, and reinforcement learning.

[0090] Supervised learning uses a machine learning algorithm to learn the mapping relationship between sample values ​​and sample labels based on collected sample values ​​and sample labels. This learned mapping relationship is then expressed using a machine learning model. The process of training a machine learning model is the process of learning this mapping relationship. For example, in signal detection, a noisy received signal is a sample, and the true constellation point corresponding to this signal is the label. Through training, machine learning aims to learn the mapping relationship between samples and labels, essentially enabling the machine learning model to become a signal detector. During training, the model parameters are optimized by calculating the error between the model's predicted values ​​and the true labels. Once the mapping relationship is learned, it can be used to predict the sample label for each new sample. The mapping relationship learned by supervised learning can include linear and nonlinear mappings. Learning tasks can be categorized into classification and regression based on the type of label.

[0091] Unsupervised learning relies solely on collected sample values, using algorithms to discover inherent patterns within them. One type of unsupervised learning algorithm uses the samples themselves as supervisory signals, meaning the model learns the mapping from one sample to another. This is called self-supervised learning. During training, the model parameters are optimized by calculating the error between the model's predictions and the samples themselves. Self-supervised learning can be used in signal compression and decompression recovery applications. Common algorithms include autoencoders and generative adversarial networks.

[0092] Reinforcement learning, unlike supervised learning, is a type of algorithm that learns problem-solving strategies through interaction with the environment. Unlike supervised and unsupervised learning, reinforcement learning problems lack explicit label data for "correct" actions. Instead, the algorithm must interact with the environment to obtain reward signals from the environment, and then adjust its decision-making actions to maximize the reward signal value. For example, in downlink power control, the reinforcement learning model adjusts the downlink transmit power of each user based on the overall system throughput fed back by the wireless network, hoping to achieve higher system throughput. The goal of reinforcement learning is also to learn the mapping between environmental states and optimal decision-making actions. However, because the labels for "correct actions" are not available in advance, network optimization cannot be achieved by calculating the error between actions and "correct actions." Reinforcement learning training is achieved through iterative interaction with the environment.

[0093] DNNs are a specific implementation of machine learning. According to the universal approximation theorem, neural networks can theoretically approximate any continuous function, enabling them to learn arbitrary mappings. Traditional communication systems require extensive expert knowledge to design communication modules. However, DNN-based deep learning communication systems can automatically discover implicit patterns in massive data sets and establish mapping relationships between data, achieving performance superior to traditional modeling methods.

[0094] The idea of ​​DNN comes from the neuron structure of brain tissue. Each neuron performs a weighted sum operation on its input value, and the weighted sum result is used to generate an output through a nonlinear function, as shown in Figure 4. Figure 4 is a schematic diagram of the structure of a neuron. Specifically, assuming that the input of the neuron is x = [x0, x1, ..., x n ], and the weights corresponding to each input are w=[w0,w1,...,w n ], the bias value of the weighted sum is b, and n is a positive integer. Since the form of nonlinear functions can be diversified, assuming that max{0,x} is the maximum function, the output of the neuron satisfies:

[0095] Among them, y represents the output of the neuron, w i represents the weight, x i represents the input of the neuron, b represents the bias, b can be a decimal, an integer (0, a positive integer or a negative integer) or a complex number, and i is an integer greater than or equal to 0 and less than or equal to n.

[0096] A DNN typically has a multi-layered structure, with each layer containing multiple neurons. The input layer processes the values ​​it receives and then passes them to the intermediate hidden layer. Similarly, the hidden layer passes the calculation results to the final output layer, generating the DNN's final output. Figure 5 shows a schematic diagram of a neural network structure consisting of an input layer, a hidden layer, and an output layer. The input layer has three neurons, the hidden layer has four neurons, and the output layer has two neurons. A neuron may have multiple input connections, each of which calculates an output based on its inputs. For example, each neuron performs a weighted summation operation on its input values ​​and passes the result of this weighted summation through an activation function to generate an output. A neuron may have multiple output connections, with the output of one neuron serving as the input to the next neuron. It should be understood that the input layer only has output connections, and each neuron in the input layer receives a value that is input to the neural network, and each neuron's value serves as the input to all output connections. The output layer only has input connections. The number of neural network layers and the number of neural network elements in each layer shown in Figure 5 are examples only.

[0097] DNNs typically have more than one hidden layer, which directly impacts their ability to extract information and fit functions. Increasing the number of hidden layers or increasing the width of each layer can improve the DNN's function-fitting capabilities. The weighted values ​​in each neuron are the parameters of the DNN network model. These parameters are optimized through training, enabling the DNN network to extract data features and express mapping relationships. DNNs typically use supervised or unsupervised learning strategies to optimize model parameters. Based on the network construction method, DNNs can be categorized as feedforward neural networks (FNNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs). Figure 5 shows an FNN network, characterized by fully connected neurons in adjacent layers. This typically requires a large amount of storage space and results in high computational complexity.

[0098] CNN is a neural network specifically designed to process data with a grid-like structure. For example, time series data (discrete sampling along the time axis) and image data (discrete sampling along two dimensions) can both be considered grid-like data. CNNs do not utilize all input information at once for computation. Instead, they use a fixed-size window to intercept a portion of the information for convolution operations, significantly reducing the computational complexity of model parameters. Furthermore, depending on the type of information intercepted by the window (e.g., people and objects in an image represent different types of information), each window can use a different convolution kernel, enabling CNNs to better extract features from the input data.

[0099] RNNs are a type of DNN that utilizes feedback time series information. Their input consists of a new input value at the current moment and their own output value at the previous moment. RNNs are suitable for capturing temporally correlated sequence features and are particularly well-suited for applications such as speech recognition and channel coding.

[0100] The following further illustrates the AI ​​model in the embodiments of the present application.

[0101] The embodiments of the present application relate to an encoder for compressing CSI and a decoder for recovering compressed CSI. The encoder and decoder are used in combination, and it can be understood that the encoder and decoder are matching AI models. An encoder may include one or more AI models, and the decoder matched with the encoder also includes one or more AI models. The number of AI models included in the matching encoder and decoder is the same and corresponds one to one.

[0102] In one possible design, a matching encoder and decoder can be implemented as two components of a single automated automatic translation (AE) model. For example, as shown in Figure 6, which illustrates an AI application framework, the encoder and decoder are deployed on different nodes. The AE model is a typical bilateral model. The encoder and decoder of an AE model are typically trained together and can be used in conjunction with each other. The encoder processes the input V to produce the processed output z, and the decoder decodes the encoder output z into the desired output V'.

[0103] An autoencoder is a type of neural network that performs unsupervised learning. Its characteristic is that it uses input data as labeled data. Therefore, an autoencoder can also be understood as a self-supervised learning neural network. Autoencoders can be used for both data compression and recovery. For example, the encoder in an autoencoder can compress (encode) data A to produce data B; the decoder in the autoencoder can decompress (decode) data B to recover data A. Alternatively, the decoder can be understood as the inverse operation of the encoder.

[0104] For example, the AI ​​model in the embodiments of the present application may include an encoder and a decoder. The encoder and decoder are used in combination, and it can be understood that the encoder and decoder are a matching AI model. The encoder and decoder can be deployed on terminal devices and network devices respectively.

[0105] Alternatively, the AI ​​model in the embodiment of the present application may be a single-end model, which may be deployed on a terminal device or a network device.

[0106] The following further illustrates the CSI and CSI feedback in the embodiments of the present application.

[0107] In communication systems (e.g., LTE or NR), network equipment needs to determine the resources, modulation and coding scheme (MCS), and precoding configurations for downlink data channels of terminal devices based on CSI. It can be understood that CSI is a type of channel information that can reflect channel characteristics and channel quality.

[0108] The channel information may be determined based on a channel measurement result of a reference signal. Alternatively, the channel information may be a channel measurement result of a reference signal. In an embodiment of the present application, the channel measurement result of a reference signal may also be replaced by the channel information.

[0109] CSI measurement refers to the receiver solving the channel information based on the reference signal sent by the transmitter, that is, estimating the channel information using the channel estimation method. Exemplarily, the reference signal may include one or more of the channel state information reference signal (CSI-RS), synchronization signal / physical broadcast channel block (SSB), sounding reference signal (SRS), or demodulation reference signal (DMRS). CSI-RS, SSB, and DMRS can be used to measure downlink CSI. SRS and DMRS can be used to measure uplink CSI.

[0110] In the present application, the meaning of CSI is broader than that of CSI in traditional schemes, and is not limited to channel quality indication (CQI), precoding matrix indicator (PMI), rank indicator (RI), or CSI-RS resource indicator (CRI). It can also be one or more of channel response information (for example, channel response matrix, frequency domain channel response information, time domain channel response information), weight information corresponding to the channel response, reference signal receiving power (RSRP), or signal to interference plus noise ratio (SINR), etc.

[0111] The RI indicates the number of downlink transmission layers recommended by the reference signal receiver (e.g., a terminal device). The CQI indicates the modulation and coding scheme supported by the reference signal receiver (e.g., a terminal device) based on the current channel conditions. The PMI indicates the precoding layer recommended by the reference signal receiver (e.g., a terminal device). The number of precoding layers indicated by the PMI corresponds to the RI.

[0112] As previously described, channel information can be obtained by measuring the reference signal. Feedback information can be obtained by compressing and / or quantizing the channel information. The feedback information can be reported via a channel information report. Channel information can be recovered by decompressing and / or dequantizing the feedback information.

[0113] Feedback information may also be referred to as channel information feedback information, CSI feedback information, CSI feedback information, compressed information, compressed channel information, compressed CSI information, compressed channel information, or compressed CSI, etc. Recovered channel information may also be referred to as CSI recovery information.

[0114] Taking FDD communication scenarios as an example, in FDD communication scenarios, because uplink and downlink channels are not reciprocal or cannot be guaranteed, network equipment typically sends downlink reference signals to terminal devices. The terminal devices perform channel and interference measurements based on the received downlink reference signals to estimate the downlink CSI. The terminal devices generate CSI reports based on a protocol predefined method or a network device configuration method and feed them back to the network device to obtain the downlink CSI.

[0115] In FDD systems, a crucial component of CSI feedback is the PMI, which uses 0-1 bits in the CSI to quantize the channel matrix or precoding matrix. PMI design (also known as codebook design) is a fundamental issue in mobile communication systems. Traditional codebook design methods predefine a series of precoding matrices and their corresponding numbers in the protocol. These precoding matrices are called codewords. The channel matrix or precoding matrix can be approximated using predefined codewords or linear combinations of multiple predefined codewords. Therefore, the terminal device can use the PMI to provide feedback to the network device, including the corresponding codeword number and one or more weighting coefficients, for the network device to recover the channel matrix or precoding matrix.

[0116] However, as the size of antenna arrays continues to increase, the number of supported antenna ports increases, and the dimensions of the corresponding channel matrix and precoding matrix increase. In order to enable terminal devices to measure the downlink channel, the overhead of network equipment sending reference signals increases, and the error of using limited predefined codewords to approximate large-scale channel matrices and precoding matrices will increase. The channel recovery accuracy can be improved by increasing the number of codewords in the codebook, but this will lead to an increase in the overhead of CSI feedback (including the corresponding codeword number and one or more weighting coefficients), thereby reducing the available resources for data transmission and causing system capacity loss.

[0117] In addition, there is a correlation between different elements in the downlink channel matrix between the network device and the terminal device, and there is also a correlation between the downlink channel matrices of different time slots. For example, the existence of correlation between different elements in the channel matrix means that there is a set of bases. When the channel matrix H is projected onto this set of bases, a sparse matrix H' of the equivalent channel can be obtained. In theory, the channel matrix H can be restored by only estimating and feeding back the non-zero elements in H' through the reference signal transmission. Therefore, there is room for compression in the overhead of reference signal transmission and CSI feedback. However, the channel compression space in traditional CSI feedback schemes (such as the codebook-based feedback method mentioned above) is not fully utilized, and the channel compression process may cause significant information loss.

[0118] Because machine learning methods (such as deep learning) have stronger nonlinear feature extraction capabilities, the introduction of AI technology into wireless communication networks has led to the development of a CSI feedback method based on AI models. Terminal devices can use AI models to compress and feedback CSI, and network equipment uses AI models to recover the compressed CSI. AI-based CSI feedback transmits a sequence (such as a bit sequence), which reduces the overhead compared to traditional CSI feedback.

[0119] Taking Figure 6 as an example, the encoder in Figure 6 can be a CSI generator, and the decoder can be a CSI reconstructor. The encoder can be deployed in a terminal device, and the decoder can be deployed in a network device. The terminal device can use the encoder to generate CSI feedback information z from the original CSI information V. The terminal device then reports a CSI report, which can include the CSI feedback information z. The network device can use the decoder to reconstruct the CSI information, thereby obtaining the recovered CSI information V'.

[0120] The CSI original information V may be obtained by the terminal device through CSI measurement. For example, the CSI original information V may include the channel response of the downlink channel or the eigenvector matrix (a matrix composed of eigenvectors) of the downlink channel. The encoder processes the eigenvector matrix of the downlink channel to obtain CSI feedback information z. In other words, the compression and / or quantization operation of the eigenmatrix according to the codebook in the related scheme is replaced by the operation of processing the eigenmatrix by the encoder to obtain CSI feedback information z. The terminal device reports the CSI feedback information z. The network device processes the CSI feedback information z through the decoder to obtain CSI recovery information V'.

[0121] The following further illustrates the training process and reasoning process of the AI ​​model in the embodiments of the present application.

[0122] The training data used to train AI models includes training samples and sample labels. For example, the training samples are channel information determined by the terminal device, and the sample labels are the actual channel information, i.e., the true value CSI. If the encoder and decoder belong to the same autoencoder, the training data can only include the training samples, or the training samples are the sample labels.

[0123] In the field of wireless communications, the true CSI may be high-precision CSI.

[0124] The specific training process is as follows: the model training node uses the encoder to process the channel information, that is, the training sample, to obtain CSI feedback information, and uses the decoder to process the feedback information to obtain the recovered channel information, that is, the CSI recovery information. Then calculate the difference between the CSI recovery information and the corresponding sample label, that is, the value of the loss function, and update the parameters of the encoder and decoder according to the value of the loss function, so that the difference between the recovered channel information and the corresponding sample label is minimized, that is, minimize the loss function. Exemplarily, the loss function can be the minimum mean square error (MSE) or cosine similarity. Repeat the above operations to obtain an encoder and decoder that meet the target requirements. The above model training node can be a terminal device, a network device or other network elements with AI functions in a communication system.

[0125] It should be understood that the above description uses the AI ​​model for CSI compression as an example. The AI ​​model can also be used in other scenarios in CSI feedback. For example, the AI ​​model can be used for CSI prediction, that is, predicting channel information at one or more future moments based on channel information measured at one or more historical moments. The embodiments of this application do not limit the specific use of the AI ​​model in the CSI feedback scenario.

[0126] In the existing technology, the encoder and decoder of an AI model can be trained together or separately. If the two ends of the AI ​​model are trained by different manufacturers without knowing the situation of the other side, the encoder and decoder may not understand each other, which will affect the CSI recovery performance of the decoder. For the above-mentioned independent training situation, there are several possible solutions: (1) achieve docking by aligning the model; (2) achieve docking by aligning the data.

[0127] Among them, the method corresponding to the existing technical solution (2) is to achieve docking by aligning data. However, how to align the data is not known. Therefore, how to achieve data alignment on both ends is an urgent problem to be solved.

[0128] In view of this, the present application provides a communication method and device, which, by standardizing a series of data set formats, is conducive to achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0129] The technical solution provided in this application will be described in detail below with reference to more drawings.

[0130] The technical solutions provided in this application can be described in detail through multiple embodiments, with specific reference to the description of each embodiment below. It should be understood that the technical solutions described in the various embodiments of this application can be combined in any way to form new embodiments, and that the same or similar parts of the concepts or solutions involved can be referenced or combined with each other. Each embodiment is described in detail below.

[0131] The deployment of the AI ​​model of the present application can be implemented on a chip inside the device. That is to say, the first network element in the present application can be a terminal device (such as UE) or a network device (such as a base station); correspondingly, the second network element can be a network device (such as a base station) or a terminal device (such as UE). It can be understood that the first network element and the second network element are for the two communicating parties. In a communication process, one party of the communication is the first network element and the other party of the communication is the second network element. For example, in a data transmission process, the first network element is the UE and the second network element is the base station; in another data transmission process, the first network element is the base station and the second network element is the UE.

[0132] As shown in Figure 7, Figure 7 is a flow chart of a communication method provided in an embodiment of the present application. The communication method includes but is not limited to the following steps:

[0133] S701: The second network element sends first indication information to the first network element.

[0134] Among them, the first indication information is used to indicate the format of the first data; the format of the first data is also called the attribute of the first data, and the format of the first data includes at least one of the following: the arrangement of input data, the arrangement of output data, the parsing format of input data parameter values, the input data dimension, the parsing format of output data parameter values, the output data dimension, the software update version corresponding to the first data, the hardware update version corresponding to the first data, the valid time period corresponding to the first data, the effective time corresponding to the first data or the expiration time corresponding to the first data; the first indication information can be carried and transmitted through signaling, and the first indication information can also be predefined by the protocol. The signaling here can be physical layer signaling or high-level signaling, which is not limited in this application.

[0135] In one possible embodiment, the second network element sends a first indication message to the first network element, and the first indication message may include at least one of the following items of the format of the first data: the arrangement of the input data, the arrangement of the output data, the parsing format of the input data parameter value, the input data dimension, the parsing format of the output data parameter value, the output data dimension, the software update version corresponding to the first data, the hardware update version corresponding to the first data, the valid time period corresponding to the first data, the effective time corresponding to the first data, or the expiration time corresponding to the first data.

[0136] In another possible embodiment, a first network element and a second network element obtain a mapping relationship between formats and indexes, where the mapping relationship includes multiple formats, with different formats corresponding to different indexes. The second network element sends first indication information to the first network element, where the first indication information may include an index of the format of the first data. After receiving the index of the format of the first data, the first network element determines, based on the mapping relationship between the formats and indexes, the arrangement of input data, the arrangement of output data, the parsing format of input data parameter values, the input data dimensions, the parsing format of output data parameter values, the output data dimensions, the software update version corresponding to the first data, the hardware update version corresponding to the first data, the validity period corresponding to the first data, the effective time corresponding to the first data, or the expiration time corresponding to the first data. The mapping relationship between formats and indexes may be predefined, prestored, pre-burned, or preconfigured. Predefinition may include predefinition, such as protocol definition. Alternatively, the mapping relationship may be configured or preconfigured. Preconfiguration may be implemented by pre-saving corresponding code, tables, or other methods that can be used to indicate relevant information in the device. This application does not limit the specific implementation method.

[0137] Optionally, the parameter value parsing format is also called the parameter value quantization method, and the parameter value parsing format may include at least one of the following: integer quantization, floating-point quantization, fixed-point, complex, rational, boolean, string, binary, date and time, logarithmic quantization, piecewise linear quantization, non-linear quantization, adaptive quantization, or layered quantization. Specific examples are as follows:

[0138] (1) Integer quantization: including 8-bit integer (abbreviated as INT8), 16-bit integer (abbreviated as INT16) and 32-bit integer (abbreviated as INT32). Among them, INT8 uses 8-bit binary storage; INT16 uses 16-bit binary storage; INT32 uses 32-bit binary storage.

[0139] (2) Floating-point quantization: including single-precision floating-point numbers (abbreviated as Float32), double-precision floating-point numbers (abbreviated as Float64) and half-precision floating-point numbers (abbreviated as Float16). Among them, Float32 uses 32-bit binary storage to provide an accuracy of about 7 significant digits; Float64 uses 64-bit binary storage to provide an accuracy of about 15 significant digits; Float16 uses 16-bit binary storage to provide an accuracy of about 3-4 significant digits.

[0140] (3) Fixed-point number: represents the decimal point at a fixed position, usually used in situations where fixed precision and number of decimal places are required.

[0141] (4) Complex number: It consists of a real part and an imaginary part and can represent a point on the complex plane.

[0142] (5) Fraction: It is expressed as the ratio of the numerator and denominator and can accurately represent rational numbers.

[0143] (6) Logical value: Usually has only two possible states, true or false.

[0144] (7) String: It is a sequence of characters that can represent text data.

[0145] (8) Binary: uses 0 and 1 to represent numerical values ​​and is the basic form of data processing within a computer system.

[0146] (9) Time and date: Time is usually expressed in hours, minutes and seconds, and date is expressed in years, months and days.

[0147] (10) Logarithmic quantization: The quantization step size increases as the input value increases, making the quantization error more uniform across the entire dynamic range.

[0148] (11) Piecewise linear quantization: The quantization function consists of multiple linear segments, each with a different slope and intercept to adapt to different regions of the data.

[0149] (12) Nonlinear quantization: The quantization function is nonlinear and can be designed based on the statistical characteristics of the data to minimize the quantization error or distortion.

[0150] (13) Adaptive quantization: The quantization step size is dynamically adjusted based on the local characteristics of the signal, for example, based on the texture complexity of the image region or the motion estimation results of the video frame.

[0151] (14) Hierarchical quantization: The data is divided into multiple levels, and each level uses a different quantization step size to preserve different levels of detail in the signal.

[0152] Optionally, the second network element can select any format from the mapping relationship between format and index as the format of the first data, or can select the format of the first data from the mapping relationship between format and index according to actual application requirements. This application does not limit this.

[0153] For example, the mapping relationship between formats and indexes may be as shown in Table 1. The mapping relationship includes multiple formats, for example, format 1, format 2, ... Specifically, the index of format 1 is "1", the input data corresponding to format 1 is arranged in a manner that prioritizes input data, and the output data is arranged in a manner that prioritizes input data and then output data. The parsing format of the input data parameter value is INT8, and the input data dimensions include the number of streams rank of the precoding codebook corresponding to the measured CSI-RS, the number of transmitting antennas or transmitting ports (tx) corresponding to the measured CSI-RS, the number of subbands sb of the frequency domain resources corresponding to the measured CSI-RS, and the number of time domain resources slot of the measured CSI-RS. Alternatively, the input data dimensions include the number of receiving antennas or receiving ports (rx) corresponding to the measured CSI-RS, tx, sb, and slot. The parsing format of the output data parameter value is INT8, and the output data dimensions include rank, tx, sb, and slot, or the output data dimensions include rx, tx, sb, and slot. Among them, prioritizing input data means loading the parameter values ​​of the input data into a signaling first, and arranging the output data after the input data means loading the parameter values ​​of the output data into the same signaling after the parameter values ​​of the input data are loaded. The parsing format of the parameter values ​​refers to the quantization method. INT8 is an integer value format that can indicate a storage range of 8 binary bits (i.e., 1 byte).

[0154] Format 2 has an index of "2." The corresponding input data for Format 2 is arranged in one signaling, and the output data is arranged in another signaling. The parsing format for the input data parameter values ​​is Float16, and the input data dimensions include tx, sb, and slot. The parsing format for the output data parameter values ​​is Float16, and the output data dimensions include tx, sb, and slot. The signaling that carries the input data and the signaling that carries the output data are two independent signalings. Float16 is a numerical format used to represent real numbers, using 16 bits (i.e., 2 bytes) of storage space to encode a floating-point number.

[0155] Format 3 has an index of "3." The corresponding input data is arranged in one signaling format, and the output data is arranged in another signaling format. Format 4 has an index of "4." The corresponding input data parameter values ​​are parsed in Float16 format, and the corresponding output data parameter values ​​are parsed in Float16 format. Format 5 has an index of "5." The corresponding input data dimensions include tx, sb, and slot, and the corresponding output data dimensions include tx, sb, and slot. Format 6 has an index of "6." The first data corresponding to Format 6 corresponds to the software update version first released in 2024. Format 7 has an index of "7." The first data corresponding to Format 7 corresponds to the hardware update version first released in 2024. Format 8 has an index of "8." The first data corresponding to Format 8 has a validity period from January 2024 to August 2025. Format 9 has an index of "9." The first data corresponding to Format 9 has an effective date of January 2024. The index of format 10 is "10," and the expiration date of the first data corresponding to format 10 is January 2030. The index of format 11 is "11," and the output data parameter value corresponding to format 11 is parsed in INT8 format. The output data dimensions include rank, tx, sb, and slot, or the output data dimensions include rx, tx, sb, and slot. Other similarities are not repeated here.

[0156] Table 1

[0157] For example, the mapping relationship between the format and the index can also be shown in Table 2 and Table 3. The mapping relationship includes multiple formats, for example, format A, format B, ... Specifically, as can be seen from Table 2, the index of format A is "A", the arrangement of the input data corresponding to format A is to arrange the input data first, the arrangement of the output data is to arrange the output data after the input data, the parsing format of the input data parameter value is Float16, the input data dimension includes rank and tx, or the input data dimension includes rx and tx, the parsing format of the output data parameter value is Float16, the output data dimension includes rank and tx, or the output data dimension includes rx and tx. The index of format B is "B", the arrangement of the input data corresponding to format B is to load the input data in one signaling, the arrangement of the output data is to load the output data in another signaling, the parsing format of the input data parameter value is INT8, the input data dimension includes tx, the parsing format of the output data parameter value is INT8, and the output data dimension includes tx. Format C has an index of "C." The corresponding input data parameter values ​​are parsed in Float16 format, with the input data dimension including tx. The output data parameter values ​​are parsed in Float16 format, with the output data dimension including tx. Format D has an index of "D." The first data corresponding to format D corresponds to the software update version of the second version of 2024, the hardware update version of the second version of 2024, and the validity period of the first data is from January 2024 to January 2026. Other similarities are omitted here.

[0158] As can be seen from Table 3, the index of format E is "E", and the arrangement of input data corresponding to format E is to arrange input data first, and the arrangement of output data is to arrange output data after input data. The parsing format of input data parameter values ​​is INT8, and the input data dimensions include sb and slot. The parsing format of output data parameter values ​​is INT8, and the output data dimensions include sb and slot. The index of format F is "F", and the arrangement of input data corresponding to format F is to load input data in one signaling, and the arrangement of output data is to load output data in another signaling. The parsing format of input data parameter values ​​is INT16, and the input data dimensions include slot. The parsing format of output data parameter values ​​is INT16, and the output data dimensions include slot. The index of format G is "G", and the parsing format of output data parameter values ​​corresponding to format G is Float16, and the output data dimensions include rank or rx. Other similarities will not be repeated here.

[0159] Table 2

[0160] Table 3

[0161] Optionally, the second network element may also send a parameter value sequence to the first network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

[0162] For example, if the format of the first data is format 1 in Table 1, the second network element will prioritize loading the parameter values ​​corresponding to the M first input data into signaling A according to the arrangement of the input data and the arrangement of the output data corresponding to format 1. After the parameter values ​​corresponding to the M first input data are loaded, the parameter values ​​corresponding to the N first output data are loaded into signaling A, and then the second network element sends signaling A to the first network element.

[0163] For example, if the format of the first data is format 2 in Table 1, the second network element loads the parameter values ​​corresponding to the M first input data into signaling B and loads the parameter values ​​corresponding to the N first output data into signaling C according to the arrangement of the input data and the arrangement of the output data corresponding to format 2. Signaling B and signaling C are two independent signalings, and then the second network element sends signaling B and signaling C to the first network element.

[0164] Optionally, the first indication information may further include X bits, where a value corresponding to the X bits is used to indicate a format of the first data in multiple formats, and X is an integer greater than 0.

[0165] For example, multiple formats include format a, format b, format c and format d, format a is the first format among the multiple formats, format b is the second format among the multiple formats, format c is the third format among the multiple formats, and format d is the fourth format among the multiple formats. The first indication information includes 2 bits. If the value of the 2 bits is "00", the format of the first data is format a, that is, the first indication information is used to indicate format a; if the value of the 2 bits is "01", the format of the first data is format b, that is, the first indication information is used to indicate format b; if the value of the 2 bits is "10", the format of the first data is format c, that is, the first indication information is used to indicate format c; if the value of the 2 bits is "11", the format of the first data is format d, that is, the first indication information is used to indicate format d.

[0166] Optionally, the first indication information may further include a bitmap, where the bitmap includes at least one bit, and the j-th bit in the bitmap represents the j-th format among multiple formats, where j is an integer greater than 0.

[0167] For example, the multiple formats include format a, format b, format c and format d, format a is the first format among the multiple formats, format b is the second format among the multiple formats, format c is the third format among the multiple formats, and format d is the fourth format among the multiple formats. The bit map includes 4 bits. If the bit map is "1000", the first indication information is used to indicate format a; if the bit map is "0100", the first indication information is used to indicate format b; if the bit map is "0010", the first indication information is used to indicate format c; if the bit map is "0001", the first indication information is used to indicate format d.

[0168] Optionally, after the second network element sends the first indication information to the first network element, the second network element may further receive confirmation information returned by the first network element, where the confirmation information is used to indicate confirmation that the first indication information has been received.

[0169] Optionally, after the second network element sends the parameter value sequence to the first network element, the second network element may further receive confirmation information returned by the first network element, where the confirmation information is used to indicate confirmation that the parameter value sequence has been received.

[0170] It should be noted that if the second network element sends the first indication information to the first network element, then in this scenario, the second network element may first send the first indication information and then send the parameter value sequence, or the second network element may first send the parameter value sequence and then send the first indication information; or, the second network element may also send the first indication information and the parameter value sequence at the same time. The specific sending order is not limited in this application.

[0171] S702: The first network element obtains first data based on the first indication information, where the first data is used to train a first model.

[0172] Specifically, the first network element can receive a parameter value sequence sent by the second network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data. Then, the first network element converts the parameter value sequence into first data according to the format of the first data, where the first data includes P first input data and Q first output data, where P is an integer greater than 0 and less than or equal to M, and Q is an integer greater than 0 and less than or equal to N.

[0173] Among them, the first network element can parse all parameter values ​​in the parameter value sequence, or can parse part of the parameter values ​​in the parameter value sequence, which is not limited in this application.

[0174] Furthermore, the first network element parses parameter values ​​corresponding to P first input data among the parameter values ​​corresponding to M first input data according to the arrangement of the input data in the format of the first data, the parsing format of the input data parameter values, and the input data dimension, to obtain multiple groups of input data, each group of input data including P first input data in the corresponding parsing format and corresponding data dimension; and the first network element parses parameter values ​​corresponding to Q first output data among the parameter values ​​corresponding to N first output data according to the arrangement of the output data in the format of the first data, the parsing format of the output data parameter values, and the output data dimension, to obtain multiple groups of output data, each group of output data including Q first output data in the corresponding parsing format and corresponding data dimension.

[0175] For example, if the format of the first data is format 1 in Table 1, the first network element receives signaling A from the second network element, and signaling A is used to carry parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data. Then, the first network element parses the parameter values ​​corresponding to P first input data among the parameter values ​​corresponding to the M first input data carried by signaling A according to the arrangement of the input data corresponding to format 1, the parsing format of the input data parameter values, and the input data dimension, to obtain four groups of input data, wherein the first group of input data includes P first input data with a parsing format of INT8 and a data dimension of rank, or the first group of input data includes P first input data with a parsing format of INT8 and a data dimension of rx, the second group of input data includes P first input data with a parsing format of INT8 and a data dimension of tx, and the third group of input data includes P first input data with a parsing format of INT8 and a data dimension of sb. , the fourth group of input data includes P first input data with a parsing format of INT8 and a data dimension of slot; and the first network element parses the parameter values ​​corresponding to Q first output data among the parameter values ​​corresponding to the N first output data carried by signaling A according to the arrangement method of the output data corresponding to format 1, the parsing format of the output data parameter values ​​and the output data dimension, to obtain four groups of output data, wherein the first group of output data includes Q first output data with a parsing format of INT8 and a data dimension of rank, or the first group of output data includes Q first output data with a parsing format of INT8 and a data dimension of rx, the second group of output data includes Q first output data with a parsing format of INT8 and a data dimension of tx, the third group of output data includes Q first output data with a parsing format of INT8 and a data dimension of sb, and the fourth group of output data includes Q first output data with a parsing format of INT8 and a data dimension of slot.

[0176] For example, if the format of the first data is format 2 in Table 1, the first network element receives signaling B and signaling C from the second network element, signaling B is used to carry parameter values ​​corresponding to M first input data, and signaling C is used to carry parameter values ​​corresponding to N first output data. The first network element parses the parameter values ​​corresponding to the M first input data carried by signaling B according to the arrangement of the input data corresponding to format 2, the parsing format of the input data parameter values, and the input data dimension, and obtains three groups of input data, wherein the first group of input data includes P first input data with a parsing format of Float16 and a data dimension of tx, and the second group of input data includes P first input data with a parsing format of Float16 and a data dimension of sb. According to the present invention, the third group of input data includes P first input data with a parsing format of Float16 and a data dimension of slot; and the first network element parses the parameter values ​​corresponding to the Q first output data among the parameter values ​​corresponding to the N first output data carried by the signaling C according to the arrangement method of the output data corresponding to format 2, the parsing format of the output data parameter values ​​and the output data dimension, and obtains three groups of output data, wherein the first group of output data includes Q first output data with a parsing format of Float16 and a data dimension of tx, the second group of output data includes Q first output data with a parsing format of Float16 and a data dimension of sb, and the third group of output data includes Q first output data with a parsing format of Float16 and a data dimension of slot.

[0177] Optionally, the first network element may also perform model training on the first model based on the first data.

[0178] The first model and the second model in this application are used in conjunction with each other. The first model can be used for data compression and quantization, and correspondingly, the second model can be used for compressed data recovery. For example, the first model can be an encoder for compressing CSI, and the second model can be a decoder for recovering compressed CSI. This application does not limit this.

[0179] Optionally, the first model in this application may be a module or chip in the first network element, such as an AI module or AI chip; the second model in this application may be a module or chip in the second network element, such as an AI module or AI chip.

[0180] The first data may include inputs to the first model, or target outputs of the first model, or both inputs and target outputs of the first model. Specifically, the first data may include one or more training data, and the training data may include training samples input to the first model, or may include target outputs of the first model.

[0181] In the process of training the first model, the first model is first initialized, that is, the parameters of each layer in the AI ​​model are pre-configured. Then, the first model is initially trained with training data. In order to make the output of the first model as close as possible to the value you really want to predict, you can compare the current network's predicted value with the really desired target value, and then update the weight vector of each layer of the AI ​​model according to the difference between the two. For example, if the network's predicted value is too high, you can adjust the weight vector to lower the predicted value. After continuous adjustment, until the first model can predict the really desired target value or a value very close to the really desired target value.

[0182] Alternatively, a loss function or objective function can be used to measure the difference between the predicted value and the target value. For example, a higher loss function output (loss) indicates a greater difference. Therefore, training the first model becomes a process of minimizing this loss as much as possible, making the loss function value less than a threshold, or making the loss function value meet the target requirement.

[0183] Optionally, the prediction accuracy of the first model can be improved by adjusting model parameters. For example, if the first model is a neural network, the prediction accuracy of the first model can be improved by adjusting the model parameters of the neural network. Adjusting the model parameters of the neural network includes adjusting at least one of the following parameters: the number of layers or width of the neural network, the weights of neurons, or parameters in the neuron activation function.

[0184] S703: The first network element sends first information to the second network element.

[0185] The first information includes h second output data, where h is an integer greater than 0 and less than or equal to P.

[0186] Optionally, the first network element may generate the first information based on the first data.

[0187] Specifically, the first network element selects h first input data from P first input data, and then inputs the h first input data into the first model for compression and quantization to obtain h second output data. Then, the first network element generates first information based on the h second output data.

[0188] For example, if the first data includes P first input data with a parsing format of Float16 and a data dimension of tx, the first network element can select h first input data with a parsing format of Float16 and a data dimension of tx from the P first input data with a parsing format of Float16 and a data dimension of tx, and input them into the first model for compression and quantization to obtain h second output data with a parsing format of Float16 and a data dimension of tx; then, the first network element generates first information based on the h second output data with a parsing format of Float16 and a data dimension of tx.

[0189] Optionally, the first information may be carried and transmitted via signaling, where the signaling may be physical layer signaling or high-layer signaling, which is not limited in this application.

[0190] Optionally, after the first network element sends the first information to the second network element, the first network element may further receive confirmation information returned by the second network element, where the confirmation information is used to confirm that the first information has been received.

[0191] S704: The second network element inputs the first information into the second model to obtain an inference result.

[0192] Specifically, h second output data in the first information are input into the second model for inference recovery to obtain h third output data.

[0193] Among them, the h third output data are the true outputs of the second model, and the above h first input data are the target outputs of the second model.

[0194] In this embodiment, the format of the first data is sent to the first network element through the second network element, so that the format of the data used by the first network element to train the first model is the same as the format of the data used by the second network element, which is conducive to achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0195] As shown in Figure 8, Figure 8 is a flow chart of another communication method provided in an embodiment of the present application. The communication method includes but is not limited to the following steps:

[0196] S801: A first network element sends first indication information to a second network element.

[0197] The first network element in step S801 is used to execute the various processes involving the second network element in step S701, and the second network element in step S801 is used to execute the various processes involving the first network element in step S701. The specific implementation method can refer to step S701 in the previous embodiment and will not be repeated here.

[0198] Optionally, before the first network element sends the first indication information to the second network element, the first network element may also send docking information to the second network element, where the docking information is used to request the format of data used for docking the first model and the second model.

[0199] Optionally, the docking information can be carried and transmitted through signaling, where the signaling can be physical layer signaling or high-layer signaling, which is not limited in this application.

[0200] Optionally, after the first network element sends the docking information to the second network element, the first network element may further receive confirmation information returned by the second network element, where the confirmation information is used to indicate confirmation that the docking information has been received.

[0201] It should be noted that if the first network element sends the first indication information to the second network element, then in this scenario, the second network element can send the parameter value sequence to the first network element only after receiving the first indication information.

[0202] S802: The first network element obtains first data based on the first indication information, where the first data is used to train a first model.

[0203] S803: The first network element sends first information to the second network element.

[0204] S804: The second network element inputs the first information into the second model to obtain an inference result.

[0205] The specific implementation of steps S802 to S804 is the same as that of steps S702 to S704 in the previous embodiment, and reference may be made to steps S702 to S704, which will not be repeated here.

[0206] In this embodiment, the format of the first data is sent from the first network element to the second network element, so that the format of the data used by the second network element is the same as the format of the data used by the first network element to train the first model, which is conducive to achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0207] In addition, the deployment of the AI ​​model of the present application can also be implemented at a location outside the device. That is to say, the first network element in the present application can also be a location outside the terminal device (such as an OTT system) or a location outside the network device (such as an intelligent network element); correspondingly, the second network element can be a location outside the network device (such as an intelligent network element) or a location outside the terminal device (such as an OTT system). It can be understood that the first network element and the second network element are for the two communicating parties. In a communication process, one party of the communication is the first network element and the other party of the communication is the second network element. For example, in a data transmission process, the first network element is an OTT system and the second network element is an intelligent network element; in another data transmission process, the first network element is an intelligent network element and the second network element is an OTT system. The intelligent network element here can be a near real-time RIC or a non-real-time RIC.

[0208] As shown in Figure 9, Figure 9 is a flow chart of another communication method provided by an embodiment of the present application. In this scenario, the first network element refers to a location outside the terminal device, and the second network element refers to a location outside the network device. The communication method includes but is not limited to the following steps:

[0209] S901: The UE sends first indication information to the base station.

[0210] The UE in step S901 is used to execute the various processes involving the first network element in step S801, and the base station in step S901 is used to execute the various processes involving the second network element in step S801. The specific implementation method can refer to step S801 in the previous embodiment and will not be repeated here.

[0211] Optionally, the UE may also receive a downlink reference signal sent by the base station. Further, the UE receives the downlink reference signal, measures the downlink reference signal, obtains a measurement result of the downlink reference signal, and then sends the measurement result of the downlink reference signal to the first network element.

[0212] S902: The UE obtains first data based on the first indication information.

[0213] The UE in step S902 is used to execute the various processes of obtaining the first data involving the first network element in step S702, and the base station in step S902 is used to execute the various processes of obtaining the first data involving the second network element in step S702. The specific implementation method can refer to step S702 in the above embodiment and will not be repeated here.

[0214] S903: The UE sends first data to the first network element.

[0215] Optionally, the first data may be carried and transmitted via signaling, where the signaling may be physical layer signaling or high-layer signaling, which is not limited in this application.

[0216] S904: The first network element trains the first model based on the first data.

[0217] The specific implementation method of step S904 is the same as the specific implementation method of training the first model in step S702. Please refer to step S702 in the above embodiment and will not be repeated here.

[0218] S905: The first network element sends first information to the second network element.

[0219] S906: The second network element inputs the first information into the second model to obtain an inference result.

[0220] The specific implementation of steps S905 to S906 is the same as the specific implementation of steps S703 to S704 in the above embodiment, and can refer to steps S703 to S704, which will not be repeated here.

[0221] Optionally, the second network element may further receive first information from the UE. The first information may contain the same content as the first information sent by the first network element, and reference may be made to the first information sent by the first network element, which will not be described in detail here.

[0222] By adopting the embodiment of the present application, by transmitting the first indication information, the format of the data used by the first network element and the second network element is made the same, which is conducive to achieving data value alignment on both ends of the AI ​​model and improving the CSI recovery performance of the AI ​​model.

[0223] The above describes in detail the method of the embodiment of the present application. The following is a description of the device provided in the embodiment of the present application.

[0224] As shown in Figure 10, Figure 10 is a schematic diagram of the structure of a communication device provided in an embodiment of the present application. The communication device can be a first network element, or a chip, chip system, or processor that supports the first network element to implement the above-mentioned method. It can also be a logical node, logical module, or software that can implement all or part of the functions of the first network element. The device can be used to implement any method and function involving the first network element in any of the aforementioned embodiments. The device may include a communication module 1001 and a processing module 1002. A detailed description of each module is as follows.

[0225] The communication module 1001 is configured to receive first indication information from a second network element or send first indication information to the second network element, where the first indication information is used to indicate a format of first data.

[0226] The processing module 1002 is used to obtain first data based on the format of the first data, where the first data is used to train a first model.

[0227] Optionally, the communication module 1001 is further used to receive a parameter value sequence from the second network element, the parameter value sequence including parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

[0228] Optionally, the processing module 1002 is further used to convert the parameter value sequence into first data, where the first data includes P first input data and Q first output data, where P is an integer greater than 0 and less than or equal to M, and Q is an integer greater than 0 and less than or equal to N.

[0229] Optionally, the format of the first data includes at least one of the following: an arrangement method of input data, an arrangement method of output data, a parsing format of input data parameter values, an input data dimension, a parsing format of output data parameter values, an output data dimension, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid time period corresponding to the first data, an effective time corresponding to the first data, or an expiration time corresponding to the first data.

[0230] Optionally, the communication module 1001 is further configured to send first information to the second network element, where the first information includes h second output data, where h is an integer greater than 0 and less than or equal to P.

[0231] It should be noted that the implementation of each module may also correspond to the corresponding description of the method embodiment shown in Figures 7 to 9, and execute the method and functions executed by the first network element in the above embodiment.

[0232] As shown in Figure 11, Figure 11 is a schematic diagram of the structure of another communication device provided in an embodiment of the present application. The communication device can be a second network element, or a chip, chip system, or processor that supports the second network element to implement the above method. It can also be a logical node, logical module, or software that can implement all or part of the functions of the second network element. The device can be used to implement any method and function involving the second network element in any of the aforementioned embodiments. The device may include a communication module 1101 and a processing module 1102. A detailed description of each module is as follows.

[0233] The communication module 1101 is configured to send first indication information to a first network element or receive first indication information from the first network element, where the first indication information is used to indicate a format of first data.

[0234] Optionally, the communication module 1101 is further used to send a parameter value sequence to the first network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

[0235] Optionally, the format of the first data includes at least one of the following: an arrangement method of input data, an arrangement method of output data, a parsing format of input data parameter values, an input data dimension, a parsing format of output data parameter values, an output data dimension, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid time period corresponding to the first data, an effective time corresponding to the first data, or an expiration time corresponding to the first data.

[0236] Optionally, the communication module 1101 is further configured to receive first information sent by the first network element, where the first information includes h second output data, where h is an integer greater than 0 and less than or equal to P.

[0237] Optionally, the processing module 1102 is used to input the first information into the second model to obtain an inference result.

[0238] It should be noted that the implementation of each module may also correspond to the corresponding description of the method embodiment shown in Figures 7 to 9, and execute the method and functions executed by the second network element in the above embodiment.

[0239] Figure 12 is a schematic diagram of the structure of a first network element provided in an embodiment of the present application. The first network element is used to perform the functions of the first network element in the above method embodiment, or to implement the steps or processes performed by the first network element in the above method embodiment.

[0240] As shown in Figure 12, the first network element includes a processor 1201 and a transceiver 1202. Optionally, the first network element also includes a memory 1203. The processor 1201, transceiver 1202, and memory 1203 can communicate with each other via an internal connection path to transmit control and / or data signals. The memory 1203 is used to store computer programs, and the processor 1201 is used to call and execute the computer programs from the memory 1203 to control the transceiver 1202 to transmit and receive signals. Optionally, the first network element may also include an antenna for transmitting uplink data or uplink control signaling output by the transceiver 1202 via wireless signals.

[0241] The processor 1201 and the memory 1203 may be combined into a processing device, and the processor 1201 is configured to execute program code stored in the memory 1203 to implement the aforementioned functions. In a specific implementation, the memory 1203 may also be integrated into the processor 1201 or independent of the processor 1201. The processor 1201 may correspond to the processing module in FIG10 .

[0242] The transceiver 1202 may correspond to the communication module in FIG10 and may also be referred to as a transceiver unit or transceiver module. The transceiver 1202 may include a receiver (or receiver, receiving circuit) and a transmitter (or transmitter, transmitting circuit). The receiver is used to receive signals, and the transmitter is used to transmit signals.

[0243] It should be understood that the first network element shown in Figure 12 is capable of implementing the various processes involving the first network element in the method embodiments shown in Figures 7-9. The operations and / or functions of the various modules in the first network element are respectively for implementing the corresponding processes in the above method embodiments. For details, please refer to the description of the above method embodiments. To avoid repetition, detailed description is omitted here.

[0244] The processor 1201 can be used to execute the actions implemented within the first network element described in the previous method embodiment, and the transceiver 1202 can be used to execute the actions of the first network element sending to or receiving from the second network element described in the previous method embodiment. For details, please refer to the description of the previous method embodiment, which will not be repeated here.

[0245] The processor 1201 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field-programmable gate array (FPGA), or other programmable logic device, a transistor logic device, a hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor 1201 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, and the like. The communication bus 1204 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, for example. The bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, FIG12 is represented by only one thick line, but this does not mean that there is only one bus or one type of bus. The communication bus 1204 is used to implement connection and communication between these components. In the embodiment of the present application, the transceiver 1202 is used to communicate signaling or data with other node devices. The memory 1203 may include volatile memory, such as nonvolatile dynamic random access memory (NVRAM), phase change random access memory (PRAM), magnetoresistive random access memory (MRAM), etc. It may also include non-volatile memory, such as at least one disk storage device, electrically erasable programmable read-only memory (EEPROM), flash memory devices, such as NOR flash memory or NAND flash memory, semiconductor devices, such as solid state disks (SSDs), etc. The memory 1203 may also be at least one storage device located remote from the aforementioned processor 1201. The memory 1203 may also store a set of computer program code or configuration information. The processor 1201 may also execute the program stored in the memory 1203. The processor may cooperate with the memory and the transceiver to perform any of the methods and functions of the first network element in the above-mentioned application embodiments.

[0246] Figure 13 is a schematic diagram of the structure of a second network element provided in an embodiment of the present application. The second network element is used to perform the functions of the second network element in the above method embodiment, or to implement the steps or processes performed by the second network element in the above method embodiment.

[0247] As shown in Figure 13, the second network element includes a processor 1301 and a transceiver 1302. Optionally, the second network element also includes a memory 1303. The processor 1301, transceiver 1302, and memory 1303 can communicate with each other via an internal connection path to transmit control and / or data signals. The memory 1303 is used to store computer programs, and the processor 1301 is used to retrieve and execute the computer programs from the memory 1303 to control the transceiver 1302 to transmit and receive signals. Optionally, the second network element may also include an antenna for transmitting uplink data or uplink control signaling output by the transceiver 1302 via wireless signals.

[0248] The processor 1301 and the memory 1303 may be combined into a processing device, and the processor 1301 is configured to execute program code stored in the memory 1303 to implement the aforementioned functions. In a specific implementation, the memory 1303 may also be integrated into the processor 1301 or independent of the processor 1301. The processor 1301 may correspond to the processing module in FIG11 .

[0249] The transceiver 1302 may correspond to the communication module in FIG11 and may also be referred to as a transceiver unit or transceiver module. The transceiver 1302 may include a receiver (or receiver, receiving circuit) and a transmitter (or transmitter, transmitting circuit). The receiver is used to receive signals, and the transmitter is used to transmit signals.

[0250] It should be understood that the second network element shown in Figure 13 is capable of implementing the various processes involving the second network element in the method embodiments shown in Figures 7-9. The operations and / or functions of the various modules in the second network element are respectively for implementing the corresponding processes in the above method embodiments. For details, please refer to the description of the above method embodiments. To avoid repetition, detailed description is omitted here.

[0251] The processor 1301 may be configured to execute the actions implemented within the second network element as described in the previous method embodiments, and the transceiver 1302 may be configured to execute the actions described in the previous method embodiments in which the second network element sends to or receives from the first network element. For details, please refer to the description in the previous method embodiments and will not be repeated here.

[0252] The processor 1301 can be any of the aforementioned types of processors. The communication bus 1304 can be a PCI bus or an EISA bus, for example. Buses can be classified as address buses, data buses, and control buses. For ease of illustration, Figure 13 shows only one thick line, but this does not imply that there is only one bus or only one type of bus. The communication bus 1304 is used to enable communication between these components. In the embodiment of the present application, the transceiver 1302 of the second network element is used to communicate signaling or data with other devices. The memory 1303 can be any of the aforementioned types of memory. The memory 1303 can also be at least one storage device located remotely from the processor 1301. The memory 1303 stores a set of computer program code or configuration information, and the processor 1301 executes the program in the memory 1303. The processor can cooperate with the memory and transceiver to perform any of the methods and functions of the second network element in the aforementioned embodiment of the application.

[0253] An embodiment of the present application also provides a chip, including a processor and a communication interface, wherein the communication interface is used to communicate with an external device or an internal device, and the processor is used to implement the methods in each of the above aspects.

[0254] In one possible design, the chip may further include a memory storing a computer program or instructions, and the processor is configured to execute the computer program or instructions stored in the memory, or other programs or instructions. When the computer program or instructions are executed, the processor is configured to implement the aforementioned various aspects of the method.

[0255] In another possible design, the chip can be integrated on the first network element or the second network element.

[0256] An embodiment of the present application further provides a processor, which is coupled to a memory and is used to execute any method and function involving the first network element or the second network element in any of the above embodiments.

[0257] An embodiment of the present application also provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute any method and function involving the first network element or the second network element in any of the above embodiments.

[0258] An embodiment of the present application also provides a device for executing any method and function involving the first network element or the second network element in any of the above embodiments.

[0259] An embodiment of the present application also provides a communication system, which includes at least one first network element and at least one second network element involved in any of the above embodiments.

[0260] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the communication device, units or modules within the device described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0261] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, an SSD).

[0262] It should be understood that the "and / or" appearing in the embodiments of the present application is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B can represent three situations: A exists alone, A and B exist at the same time, and B exists alone.

[0263] It should be understood that in the embodiments of the present application, "B corresponding to A" means that B is associated with A and B can be determined based on A. However, it should also be understood that determining B based on A does not mean determining B based solely on A, but B can also be determined based on A and / or other information.

[0264] It should be understood that the symbol " / " in the embodiments of this application can indicate that the preceding and following objects are in an "or" relationship. Furthermore, the symbol " / " can also represent a division sign, i.e., performing a division operation. For example, A / B can mean A divided by B.

[0265] It is understood that in the embodiments of the present application, the first network element and / or the second network element may perform some or all of the steps in the embodiments of the present application. These steps or operations are merely examples. In the embodiments of the present application, other operations or variations of various operations may also be performed. In addition, the various steps may be performed in a different order than those presented in the embodiments of the present application, and it is possible that not all operations in the embodiments of the present application need to be performed.

[0266] The above-described specific implementation methods further illustrate the purpose, technical solutions and beneficial effects of this application. Any modifications, equivalent replacements, improvements, etc. made within the principles of this application shall be included in the scope of protection of this application.

Claims

1. A communication method, characterized in that: Applied to a first network element, the method includes: receiving first indication information from a second network element or sending first indication information to the second network element, where the first indication information is used to indicate a format of the first data; Based on a format of the first data, the first data is obtained, and the first data is used to train a first model.

2. The method according to claim 1, wherein The method further comprises: Receive a parameter value sequence from the second network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

3. The method according to claim 2, wherein The obtaining the first data based on the format of the first data includes: The parameter value sequence is converted into the first data, where the first data includes P first input data and Q first output data, where P is an integer greater than 0 and less than or equal to M, and Q is an integer greater than 0 and less than or equal to N.

4. The method according to any one of claims 1 to 3, wherein The format of the first data includes at least one of the following: an arrangement method of input data, an arrangement method of output data, a parsing format of input data parameter values, an input data dimension, a parsing format of output data parameter values, an output data dimension, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid time period corresponding to the first data, an effective time corresponding to the first data, or an expiration time corresponding to the first data.

5. A communication method, characterized in that: Applied to a second network element, the method includes: First indication information is sent to a first network element or received from the first network element, where the first indication information is used to indicate a format of first data.

6. The method according to claim 5, wherein The method further comprises: A parameter value sequence is sent to the first network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

7. The method according to claim 5 or 6, wherein: The format of the first data includes at least one of the following: an arrangement method of input data, an arrangement method of output data, a parsing format of input data parameter values, an input data dimension, a parsing format of output data parameter values, an output data dimension, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid time period corresponding to the first data, an effective time corresponding to the first data, or an expiration time corresponding to the first data.

8. A communication device, characterized in that: Applied to a first network element, the apparatus includes: a communication module, configured to receive first indication information from a second network element or send first indication information to the second network element, where the first indication information is used to indicate a format of the first data; A processing module is used to obtain the first data based on the format of the first data, where the first data is used to train a first model.

9. The device according to claim 8, wherein The communication module is further used to receive a parameter value sequence from the second network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

10. The device according to claim 9, wherein The processing module is also used to convert the parameter value sequence into the first data, where the first data includes P first input data and Q first output data, where P is an integer greater than 0 and less than or equal to M, and Q is an integer greater than 0 and less than or equal to N.

11. The device according to any one of claims 8 to 10, characterized in that The format of the first data includes at least one of the following: an arrangement method of input data, an arrangement method of output data, a parsing format of input data parameter values, an input data dimension, a parsing format of output data parameter values, an output data dimension, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid time period corresponding to the first data, an effective time corresponding to the first data, or an expiration time corresponding to the first data.

12. A communication device, characterized in that: Applied to a second network element, the device includes: The communication module is used to send first indication information to a first network element or receive first indication information from the first network element, where the first indication information is used to indicate a format of first data.

13. The device according to claim 12, wherein The communication module is further used to send a parameter value sequence to the first network element, where the parameter value sequence includes parameter values ​​corresponding to M first input data and parameter values ​​corresponding to N first output data, where M is an integer greater than 0 and N is an integer greater than 0.

14. The device according to claim 12 or 13, characterized in that The format of the first data includes at least one of the following: an arrangement method of input data, an arrangement method of output data, a parsing format of input data parameter values, an input data dimension, a parsing format of output data parameter values, an output data dimension, a software update version corresponding to the first data, a hardware update version corresponding to the first data, a valid time period corresponding to the first data, an effective time corresponding to the first data, or an expiration time corresponding to the first data.

15. A communication device, characterized in that: The communication device comprises a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the communication device to perform the method according to any one of claims 1 to 4.

16. A communication device, characterized in that: The communication device comprises a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the communication device to perform the method according to any one of claims 5 to 7.

17. A communication system, characterized in that: The method comprises a first network element and a second network element, wherein the first network element is used to execute the method according to any one of claims 1 to 4, and the second network element is used to execute the method according to any one of claims 5 to 7.

18. A computer-readable storage medium, characterized in that The computer-readable storage medium comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 4 or any one of claims 5 to 7 is implemented.

19. A chip, characterized in that: The chip includes a processor and a communication interface, where the communication interface is used to communicate with an external device or an internal device, and the processor is used to implement the method according to any one of claims 1 to 4 or any one of claims 5 to 7.

Images