Communication method and related apparatus

Abstract

The present application provides a communication method and a related apparatus. The method may be executed by a first apparatus. The method comprises: the first apparatus acquires wireless data quality feedback information of at least one second apparatus, wherein the wireless data quality feedback information is used for indicating usage performance of wireless data, and the wireless data is used by a model. Then, the first apparatus determines quality information of the wireless data on the basis of the wireless data quality feedback information. In the method, the first apparatus determines the quality information of the wireless data on the basis of the usage performance of the wireless data in the model, dynamically reflecting the quality information of the wireless data in a real usage scenario, and enabling real-timeliness and accuracy of the quality information of the wireless data.

Description

A communication method and related apparatus

[0001] This application claims priority to Chinese Patent Application No. 202510085114.9, filed on January 17, 2025, entitled "A Communication Method and Related Device", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of wireless communication, and more particularly to a communication method, apparatus, communication system, and computer storage medium. Background Technology

[0003] Currently, artificial intelligence (AI) technology is being introduced into wireless communication systems. AI technology can be used for wireless channel information compression and reconstruction, beam management, and positioning enhancement. Trained AI models can be used to improve the performance of wireless communication.

[0004] AI-based radio access networks (RANs) require different types of wireless data at each stage of the AI ​​model. For example, model training requires a large amount of diverse data, model inference requires real-time data, and model monitoring requires a large amount of near-real-time data.

[0005] The quality of wireless data is a key factor determining the performance of data usage in model training, monitoring, or inference. However, current wireless data quality metrics fail to accurately reflect the quality of wireless data used in models, thus hindering guidance for models to operate on appropriate wireless data and impacting both model and wireless communication performance. Summary of the Invention

[0006] This application provides a communication method for dynamically updating the quality information of wireless data based on the performance of wireless data in model usage, thereby making the quality information of wireless data more real-time and accurate.

[0007] In a first aspect, this application provides a communication method, which can be executed by a first device, the method comprising:

[0008] The first device acquires wireless data quality feedback information from at least one second device, wherein the wireless data quality feedback information is used to indicate the performance of wireless data used by the model. Then, the first device determines the quality information of the wireless data based on the wireless data quality feedback information.

[0009] In this method, the first device determines the quality information of wireless data based on the performance of wireless data in the model, which dynamically reflects the quality information of wireless data in real usage scenarios, making the quality information of wireless data more real-time and accurate.

[0010] In some possible implementations, wireless data quality feedback information may include wireless data attribute identifiers and the performance of AI models based on wireless data. Wireless data attribute identifiers may include at least one of the following: function type identifier, model identifier, model operation type identifier, or data identifier. The function type identifier refers to the specific function implemented by the AI ​​model in wireless communication; the model operation type may include model training, model inference, etc. The data identifier may be a dataset identifier, data subset identifier, or data sample identifier, or it may be a data acquisition configuration identifier, data processing type identifier, or data source type identifier. The data acquisition configuration identifier and data source type identifier are mainly for the wireless data acquisition phase. The data acquisition configuration identifier indicates the parameter configuration options when acquiring wireless data; the data source type may be wireless air interface, UE-side sensors, network-side environmental data, etc. Furthermore, if the first device is equipped with a data processing node that can perform preprocessing operations on the wireless data, the data identifier may include a data processing type identifier to indicate the type of preprocessing performed on the wireless data. In this way, the attribute identifiers of the wireless data can accurately identify the performance information of each AI model, thereby ensuring the uniqueness of each performance information and facilitating the recording, identification, retrieval, and management of relevant performance information.

[0011] In some possible implementations, the performance of an AI model based on wireless data can be the absolute performance of the model under a specific target attribute. For example, the attribute of the wireless data can be the type of operation performed by the AI ​​model. During model training, performance metrics can be based on the data update gradient, the change in loss value, or the accuracy of prediction; during model inference, performance metrics can be based on the accuracy of prediction. In this way, the quality of wireless data usage in the AI ​​model can be measured using absolute numerical values, and appropriate evaluation metrics can be selected based on different target attributes.

[0012] In other possible implementations, the performance of the AI ​​model based on wireless data can also be relative performance, that is, a relative ranking or relative value obtained by sorting or comparing within a target range. In this case, the performance in the wireless data quality feedback information sent by the second device to the first device can be represented in the form of a relative ranking, or, when the performance information involves hierarchical ranking, the second device can also send the corresponding level to the first device. This allows for a more intuitive display of the differences between different performance information records of the wireless data.

[0013] In some possible implementations, the first device can also acquire the time interval between acquiring wireless data quality feedback information and determining the quality information of the wireless data, and determine the quality information of the wireless data under the target attribute based on this time interval and the performance of the AI ​​model based on the wireless data under the target attribute. For example, the first device can determine the weight of multiple pieces of information based on the size of the time interval, or in other words, the order in which they are acquired, mainly determining the final quality information of the wireless data according to the feedback information that is closer in time. In this way, the real-time performance of the quality information can be improved, making it closer to actual application scenarios.

[0014] In some possible implementations, the communication system may include multiple second devices. The target attribute may include the function type implemented by the AI ​​model, the identifier of the AI ​​model, or the operation type of the AI ​​model. The first device can determine the quality information of the wireless data under the target attribute based on the relationships between the multiple second devices. For example, when the performance of the AI ​​model based on the wireless data is relative performance, the first device can determine the comprehensive ranking result based on the performance ranking information fed back by the multiple second devices and the weight information of each device. In this way, the quality information of the wireless data under the target attribute can be determined more accurately by utilizing the weight relationships between the multiple second devices.

[0015] In some possible implementations, the first device can also comprehensively utilize the time interval between acquiring wireless data quality feedback information and the first device determining the quality information of the wireless data, as well as the weight information among multiple second devices, to determine the quality information of the wireless data under the target attribute. For example, when the performance of the AI ​​model based on the wireless data is an absolute performance, the first device can determine a comprehensive value based on the performance values ​​in each piece of information, the aforementioned time interval, and the weight information among the second devices, as the performance and quality of the wireless data under that target attribute. This makes the updating of wireless data quality information more accurate and real-time.

[0016] In some possible implementations, the first device can send the determined quality information to the second device, for example, in the form of wireless data attribute identifiers and corresponding performance. This allows the second device to obtain timely performance information based on the use of wireless data in the AI ​​model, enabling it to better allocate tasks and utilize wireless data.

[0017] In some possible implementations, the first device may also, in response to a data request based on quality information sent by at least one second device, determine the target wireless data and then send the target wireless data to the at least one second device. In this way, the second device can better perform wireless AI tasks based on the target wireless data that meets the performance requirements.

[0018] In some possible implementations, the first device may also send a wireless data quality feedback instruction to the second device to instruct the second device to generate wireless data quality feedback information. The wireless data quality feedback instruction may be sent by the first device along with the wireless data provided to the second device, or it may be sent separately by the first device to the second device.

[0019] Secondly, this application provides a communication method, which can be executed by a second device, the method comprising:

[0020] The second device generates wireless data quality feedback information based on the performance of the AI ​​model based on wireless data, and sends the wireless data quality feedback information to the first device to determine the quality information of the wireless data.

[0021] The content and specific form of the wireless data quality feedback information are as described in the method of the first aspect of this application. In this method, the second device can provide timely feedback to the first device on the performance of the wireless data as shown in the AI ​​model, which helps the first device determine the quality information of the wireless data based on the wireless data quality feedback information, and then dynamically update the quality of the wireless data.

[0022] In some possible implementations, the second device can also receive quality information sent by the first device, which is determined by the first device based on the wireless data quality feedback information sent by the second device. This allows for timely synchronization of wireless data quality information with the first device, facilitating the second device's subsequent request for qualified wireless data from the first device and thus better completing the wireless communication task.

[0023] In some possible implementations, the second device may also send a data request based on quality information to the first device, wherein the data request is used to determine the target wireless data. The second device can then receive the target wireless data sent by the first device in response and apply it to the operational tasks of the wireless AI model.

[0024] In some possible implementations, the second device may send wireless data quality feedback information as a response to the first device after receiving a wireless data quality feedback indication sent by the first device. The wireless data quality feedback indication may be sent by the first device along with the wireless data it provides to the second device, or it may be sent separately by the first device to the second device.

[0025] Thirdly, this application provides a first apparatus comprising units or modules for implementing the communication method described in the first aspect of this application or any possible implementation thereof.

[0026] Fourthly, this application provides a second apparatus comprising units or modules for implementing the communication method described in any possible implementation of the second or first aspect of this application.

[0027] Fifthly, this application provides a communication system, including a first device and a second device. The first device is used to perform the communication method as described in the first aspect of this application or any possible implementation thereof, and the second device is used to perform the communication method as described in the second aspect of this application or any possible implementation thereof.

[0028] Sixthly, this application provides a computer storage medium for storing a computer program, which, when executed, implements the communication method described in the first aspect or any possible implementation of the first aspect, or the communication method described in the second aspect or any possible implementation of the second aspect.

[0029] In a seventh aspect, this application provides a computer program product containing instructions that, when executed by a computer, cause the computer to implement the communication method described in any possible implementation of the first or second aspect.

[0030] Based on the implementation methods provided in the above aspects, this application can be further combined to provide more implementation methods. Attached Figure Description

[0031] Figure 1 is a schematic diagram of the application framework of a communication system provided in an embodiment of this application;

[0032] Figure 2 is a schematic diagram of the application framework of another communication system provided in an embodiment of this application;

[0033] Figure 3A is a schematic diagram of a wireless AI framework provided in an embodiment of this application;

[0034] Figure 3B is a schematic diagram of another wireless AI framework provided in an embodiment of this application;

[0035] Figure 4 is a flowchart of a communication method provided in an embodiment of this application;

[0036] Figure 5 is a schematic diagram of wireless data quality feedback information provided in an embodiment of this application;

[0037] Figure 6 is a flowchart of a communication method provided in an embodiment of this application;

[0038] Figure 7 is a schematic diagram of a communication method provided in an embodiment of this application;

[0039] Figure 8 is a schematic diagram of another communication method provided in an embodiment of this application;

[0040] Figure 9 is a schematic diagram of another communication method provided in an embodiment of this application;

[0041] Figure 10 is a schematic diagram of another communication method provided in an embodiment of this application;

[0042] Figure 11 is a schematic diagram of the structure of a first device provided in an embodiment of this application;

[0043] Figure 12 is a schematic diagram of the structure of a second device provided in an embodiment of this application;

[0044] Figure 13 is a schematic diagram of the structure of a terminal device provided in an embodiment of this application;

[0045] Figure 14 is a schematic diagram of the structure of a network device provided in an embodiment of this application. Detailed Implementation

[0046] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0047] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such terms are interchangeable where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0048] It should be understood that the term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone, where A and B can be single or multiple. Additionally, the character " / " in this document generally indicates that the preceding and following related objects are in an "or" relationship. Furthermore, "at least one of the following" or similar expressions in this document are used to represent any combination of the listed items; for example, at least one of A, B, and / or C can represent the following six situations: A alone, B alone, C alone, A and B simultaneously, B and C simultaneously, A and C simultaneously, and A, B, and C simultaneously, where A, B, and C can be single or multiple.

[0049] It is understood that in this application, "instruction" can include direct instruction, indirect instruction, explicit instruction, and implicit instruction. When describing a certain instruction information for the purpose of instructing A, it can be understood that the instruction information carries A, directly instructs A, or indirectly instructs A.

[0050] In this application, the information indicated by the instruction information is called the information to be instructed. In specific implementations, there are many ways to instruct the information to be instructed, such as, but not limited to, directly instructing the information to be instructed, such as the information to be instructed itself or its index; indirectly instructing the information to be instructed by instructing other information, where there is a relationship between the other information and the information to be instructed; or instructing only a part of the information to be instructed, while the other parts are known or pre-agreed upon. For example, the instruction of specific information can be achieved by using a pre-agreed (e.g., protocol-defined) arrangement of various pieces of information, thereby reducing instruction overhead to some extent.

[0051] The information to be instructed can be sent as a whole or divided into multiple sub-information messages, and the sending period and / or timing of these sub-information messages can be the same or different. This application does not limit the specific sending method. The sending period and / or timing of these sub-information messages can be predefined, for example, according to a protocol, or configured by the transmitting device by sending configuration information to the receiving device.

[0052] It is understood that "send" and "receive" in this application refer to the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which can include direct transmission via the air interface or indirect transmission via the air interface from other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which can include direct reception from YY via the air interface or indirect reception from YY via the air interface from other units or modules. "Send" can also be understood as the "output" of the chip interface, and "receive" can also be understood as the "input" of the chip interface.

[0053] In other words, sending and receiving can occur between devices, such as between network devices and terminal devices, or within a device, such as between components, modules, chips, software modules, or hardware modules within the device via buses, wiring, or interfaces.

[0054] It is understandable that information may undergo necessary processing, such as encoding and modulation, between the source and destination, but the destination can understand the valid information from the source. Similar statements in this application can be interpreted in a similar way and will not be elaborated further.

[0055] First, the communication system involved in the embodiments of this application is introduced. The technical solution of this application can be applied to cellular communication systems related to the 3rd Generation Partnership Project (3GPP). For example, fourth-generation (4G) communication systems, 5G communication systems, and communication systems after the fifth generation. For example, future communication systems. For example, the fourth-generation communication system may include the Long Term Evolution (LTE) communication system. The fifth-generation communication system may include the New Radio (NR) communication system. The technical solution of this application can also be applied to wireless fidelity (WiFi) systems, communication systems supporting the convergence of multiple wireless technologies, device-to-device (D2D) systems, or vehicle-to-everything (V2X) communication systems.

[0056] The communication systems to which this application applies include terminal equipment and network equipment. Terminal equipment and network equipment are described below.

[0057] Terminal equipment, also known as user equipment (UE), mobile station (MS), mobile terminal (MT), fixed wireless access (FWA), customer premises equipment (CPE), etc., refers to devices that include wireless communication capabilities (providing voice / data connectivity to users). Examples include handheld devices with wireless connectivity, in-vehicle devices, and machine-type communication (MTC) terminals. Currently, terminal devices can include: mobile phones, tablets, laptops, 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 (e.g., drones, vehicles), wireless terminals in remote medical surgery, wireless terminals in smart grids, wireless terminals in transportation safety, wireless terminals in smart cities, and wireless terminals in smart homes. For example, wireless terminals in self-driving can be drones, helicopters, or airplanes. For example, wireless terminals in vehicle-to-everything (V2X) can be in-vehicle equipment, vehicle-mounted equipment, in-vehicle modules, vehicles, or ships. Wireless terminals in industrial control can be cameras, robots, or robotic arms. Wireless terminals in smart homes can be televisions, air conditioners, robot vacuums, speakers, or set-top boxes. The terminal device can also be a device or module that is connected to the communication system shown above and has corresponding communication functions. The terminal device usually contains a communication module, circuit or chip that performs the corresponding communication function, and the terminal device is also configured with program instructions for performing the corresponding communication function.

[0058] It should be noted that the terminal device can be a device or apparatus with a chip, or a device or apparatus with integrated circuitry, or a chip, chip system, module, or control unit in the device or apparatus shown above; the specific application is not limited to any particular type. It should also be noted that in this application, when referring to a terminal device, it can refer to the terminal device itself, or to the chip, functional module, or integrated circuit within the terminal device that performs the method provided in this application; the specific application is not limited to any particular type.

[0059] A network device is a device deployed in a radio access network to provide wireless communication functions for terminal devices. Network devices may also be referred to as radio access network (RAN) entities, access nodes, network nodes, access network equipment, or communication devices, etc.

[0060] Specifically, the network equipment can be access network equipment for cellular systems related to the 3rd Generation Partnership Project (3GPP). For example, fourth-generation (4G) mobile communication systems, 5G mobile communication systems, or future mobile communication systems. The network equipment can also be access network equipment in open RAN (O-RAN or ORAN) or cloud radio access network (CRAN). Alternatively, the network equipment can also be access network equipment in a communication system resulting from the integration of two or more of the above communication systems.

[0061] Network equipment includes, but is not limited to: evolved Node B (eNB), radio network controller (RNC), Node B (NB), base station controller (BSC), base transceiver station (BTS), home base station (e.g., home evolved Node B, or home Node B, HNB), baseband unit (BBU), access point (AP) in wireless fidelity (WIFI) systems, macro base station, micro base station, wireless relay node, donor node, radio controller in CRAN scenarios, wireless backhaul node, transmission point (TP), or transmission and receiving point (TRP). Network equipment can also be access network equipment in 5G mobile communication systems. For example, a next-generation NodeB (gNB) in a new radio (NR) system, a transmission and reception point (TRP), a TP, or one or more antenna panels (including multiple antenna panels) of a base station in a 5G mobile communication system. Alternatively, network equipment can also be network nodes constituting a gNB or transmission point. Examples include a centralized unit (CU), a distributed unit (DU), a CU-control plane (CP), a CU-user plane (UP), or a radio unit (RU). CUs and DUs can be separate or included in the same network element. For example, a BBU. RUs can be included in radio equipment or radio units. For example, in a remote radio unit (RRU), an active antenna unit (AAU), or a remote radio head (RRH). Alternatively, network equipment can also be a server, wearable device, vehicle, or in-vehicle equipment. For example, in V2X technology, network devices can be roadside units (RSUs).

[0062] To support artificial intelligence (AI) technology in wireless networks, AI nodes may be introduced into the network. AI nodes can be AI network elements or AI modules.

[0063] AI nodes can be deployed in one or more of the following locations within the communication system: access network nodes (RAN nodes), terminal devices, or core network devices. Alternatively, AI nodes can be deployed independently, for example, in a location other than any of the aforementioned devices, such as in the host or cloud server of an over-the-top (OTT) system. AI nodes can communicate with other devices in the communication system, which can be one or more of the following: network devices, terminal devices, or core network elements.

[0064] It is understood that this application does not limit the number of AI nodes. For example, when there are multiple AI nodes, these nodes can be divided based on function, such as different AI nodes being responsible for different functions.

[0065] It can also be understood that AI nodes can be independent devices, or they can be integrated into the same device to achieve different functions. Alternatively, they can be network elements in hardware devices, software functions running on dedicated hardware, or virtualization functions instantiated on a platform (e.g., a cloud platform). This application does not limit the specific form of the aforementioned AI nodes.

[0066] Figure 1 illustrates a possible application framework in a communication system. As shown in Figure 1, network elements in the communication system are connected via interfaces (e.g., NG, Xn) or air interfaces. These network element nodes, such as core network equipment, access network nodes (RAN nodes), terminals, or one or more devices in operations administration and maintenance (OAM), are equipped with one or more AI modules (only one is shown in Figure 1 for clarity). An access network node can be a single RAN node or can include multiple RAN nodes, for example, including CU and DU. The CU and / or DU can also be equipped with one or more AI modules. A CU can also be split into CU-CP and CU-UP, with one or more AI modules configured in the CU-CP and / or CU-UP.

[0067] AI modules are used to implement corresponding AI functions. AI modules deployed in different network elements can be the same or different. The models of AI modules can achieve different functions depending on the parameter configurations. The models of AI modules can be configured based on one or more of the following parameters: structural parameters (e.g., at least one of the following: number of neural network layers, neural network width, inter-layer connections, neuron weights, neuron activation function, or biases in the activation function), input parameters (e.g., the type and / or dimension of the input parameters), or output parameters (e.g., the type and / or dimension of the output parameters). The biases in the activation function can also be referred to as the biases of the neural network.

[0068] In one example, the neural network mentioned above can be a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), or a generative adversarial network (GAN).

[0069] Deep Neural Networks (DNNs) are artificial neural network architectures with multiple layers of nonlinear transformation units stacked in a hierarchical structure to form deep computational models. Compared to shallow neural networks, deep neural networks have more hidden layers, allowing the network model to capture more complex data structures and higher-level abstract features.

[0070] A CNN is a deep neural network with a convolutional structure. A CNN contains a feature extractor consisting of convolutional layers and subsampling layers. This feature extractor can be viewed as a filter, and the convolution process can be seen as performing convolution between a trainable filter and an input image or a convolutional feature map.

[0071] RNN is a type of recursive neural network that takes sequence data as input, recursively moves along the direction of sequence evolution, and connects all nodes (recurrent units) in a chain-like manner.

[0072] GAN is a deep learning model. It consists of a generator and a discriminator, and is trained through adversarial learning. Its purpose is to estimate the potential distribution of data samples and generate new data samples.

[0073] 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 they can be deployed on the same node or device.

[0074] Figure 2 illustrates a possible application framework in a communication system. As shown in Figure 2, the communication system includes a RAN intelligent controller (RIC). For example, the RIC can be the AI ​​modules 117 and 118 shown in Figure 1, used to implement AI-related functions. RICs include near-real-time RICs (near-RT RICs) and non-real-time RICs (non-RT RICs). Non-real-time RICs primarily process non-real-time information, such as data that is not sensitive to latency, with latency in the order of seconds. Real-time RICs primarily process near-real-time information, such as data that is relatively sensitive to latency, with latency in the order of tens of milliseconds.

[0075] Near real-time (NRT) RICs are used for model training and inference. For example, they are used to train AI models and then use those models for inference. NRT RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, compute nodes, and / or RUs) and / or terminals. This information can be used as training data or inference data. NRT RICs can deliver inference results to RAN nodes and / or terminals. Inference results can be exchanged between CUs and DUs, and / or between DUs and RUs. For example, a NRT RIC delivers an inference result to a DU, which then forwards it to an RU.

[0076] Non-real-time RICs are also used for model training and inference. For example, they are used to train AI models and then use those models for inference. Non-real-time RICs can obtain network-side and / or terminal-side information from RAN nodes (e.g., CUs, CU-CPs, CU-UPs, DUs, compute nodes, and / or RUs) and / or terminals. This information can be used as training data or inference data, and the inference results can be delivered to RAN nodes and / or terminals. Inference results can be exchanged between CUs and DUs, and / or between DUs and RUs; for example, a non-real-time RIC delivers inference results to a DU, which then forwards them to an RU.

[0077] Near real-time RICs and non-real-time RICs can also be configured as separate network elements. Near real-time RICs and non-real-time RICs can also be part of other devices. For example, near real-time RICs can be set in RAN nodes (e.g., CU, DU, compute nodes), while non-real-time RICs can be set in OAM, cloud servers, core network devices, or other network devices.

[0078] AI can endow machines with human-like intelligence, for example, allowing them to use computer hardware and software to simulate certain intelligent human behaviors. To achieve artificial intelligence, machine learning methods can be employed. In machine learning, machines learn (or train) a model using training data. This model represents the mapping between inputs and outputs. The learned model can be used for reasoning (or prediction), that is, it can be used to predict the output corresponding to a given input. This output can also be called the reasoning result (or prediction result).

[0079] This document explains some basic concepts in the field of AI, which does not limit the scope of protection of the embodiments of this application.

[0080] (1) Machine learning (ML):

[0081] Machine learning is a crucial technological approach to achieving AI. AI endows machines with human-like intelligence, using computer hardware and software to simulate certain intelligent human behaviors, including machine learning and other methods. Machine learning refers to learning models or rules from raw data, such as neural networks, decision trees, and support vector machines. Machine learning can be categorized into supervised learning, unsupervised learning, and reinforcement learning.

[0082] Supervised learning, based on collected sample values ​​and labels, uses machine learning algorithms to learn the mapping relationship between sample values ​​and labels, and expresses this learned mapping relationship using a machine learning model. The process of training the machine learning model is the process of learning this mapping relationship. For example, in signal detection, the noisy received signal is the sample, and the corresponding real constellation point is the label. Machine learning aims to learn the mapping relationship between samples and labels through training, that is, to enable the machine learning model to learn a signal detector. During training, the model parameters are optimized by calculating the error between the model's predicted values ​​and the real labels. Once the mapping relationship is learned, it can be used to predict the sample label of each new sample. The mapping relationship learned in supervised learning can include linear mappings and nonlinear mappings. Based on the type of label, the learning task can be divided into classification tasks and regression tasks.

[0083] Unsupervised learning relies solely on collected sample values, using algorithms to discover inherent patterns within the samples. One type of unsupervised learning algorithm uses the samples themselves as supervisory signals; that is, the model learns the mapping relationship from sample to sample, which is called self-supervised learning. During training, model parameters are optimized by calculating the error between the model's predictions and the samples themselves. Self-supervised learning can be used for signal compression and decompression recovery applications; common algorithms include autoencoders and generative adversarial networks.

[0084] 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 do not have explicit "correct" action labels. The algorithm needs to interact with the environment to obtain reward signals from the environment, and then adjust its decision actions to obtain a larger reward signal value. For example, in downlink power control, the reinforcement learning model adjusts the downlink transmission power of each terminal device based on the total system throughput feedback from the wireless network, aiming to achieve a higher system throughput. The goal of reinforcement learning is also to learn the mapping relationship between the environment state and the optimal decision action. However, because the label of the "correct action" cannot be obtained in advance, the network cannot be optimized by calculating the error between the action and the "correct action." Reinforcement learning training is achieved through iterative interaction with the environment.

[0085] Deep neural networks (DNNs) are a specific implementation of machine learning. According to the general approximation theorem, neural networks can theoretically approximate any continuous function, thus enabling them to learn arbitrary mappings. Traditional communication systems rely on extensive expert knowledge to design communication modules, while DNN-based deep learning communication systems can automatically discover hidden pattern structures from large datasets, establish mapping relationships between data, and achieve performance superior to traditional modeling methods.

[0086] Based on their construction method, DNNs can be divided into feedforward neural networks (FNNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs). FNNs can be neural networks where neurons in adjacent layers are completely connected pairwise, which makes FNNs typically require a large amount of storage space and have high computational complexity.

[0087] CNNs are neural networks specifically designed to process data with a grid-like structure. For example, time-series data (discrete sampling along the time axis) and image data (two-dimensional discrete sampling) can both be considered grid-like data. CNNs do not use all the input information at once for computation; instead, they use a fixed-size window to extract a portion of the information for convolution operations, which significantly reduces the computational cost of model parameters. Furthermore, depending on the type of information extracted by the window (such as people and objects in an image representing different types of information), each window can use different convolution kernels, allowing CNNs to better extract features from the input data.

[0088] Recurrent Neural Networks (RNNs) are a type of distributed neural network (DNN) that utilizes feedback time-series information. Their input includes the current input value and their own output value from the previous time step. RNNs are well-suited for acquiring temporally correlated sequence features, and are particularly applicable to applications such as speech recognition and channel coding / decoding.

[0089] AI models refer to function models that map a certain-dimensional input to a certain-dimensional output, and their parameters can be obtained through machine learning training. For example, f(X) = aX² + b is a quadratic function model, which can be viewed as an AI model. a and b correspond to the model's parameters and can be obtained through machine learning training. Data used for model training, validation, and / or testing in machine learning can form datasets or training datasets. The quantity and / or quality of data in these datasets or training datasets will affect the effectiveness of machine learning. Model training involves selecting an appropriate loss function (which measures the difference between the model's predictions and the true values) and using optimization algorithms to train the model parameters to minimize the loss function value. Model testing involves evaluating the model's performance using test data after training. Model application involves using the trained model to solve real-world problems.

[0090] A neural network, or artificial neural network, is a mathematical model that mimics the behavioral characteristics of animal neural networks to perform distributed parallel information processing. It is a special form of AI model.

[0091] (2) Model training:

[0092] Model training involves selecting an appropriate function (such as a loss function) and using optimization algorithms to train the model parameters so that the difference between the model's predicted values ​​and the ground truth (or target values, labels) tends to be minimized.

[0093] For example, model training methods include, but are not limited to, supervised learning, self-supervised learning, and knowledge distillation.

[0094] (3) Model file and model parameters:

[0095] Model files and / or model parameters can be used to determine the model. Optionally, the model in this application may refer to the model itself, or it may refer to the model files and / or model parameters used to determine the model.

[0096] The model file can be used to indicate the model structure, which may include, but is not limited to, FNN, CNN, or RNN. The model file can have a fixed format, such as a standard predefined format, or a format pre-negotiated by both ends of the interface. Model parameters can refer to parameters in the neural network model, such as, but not limited to, the number of layers in the neural network, the type and weights of neurons in each layer, etc. This application does not limit the method of distributing model parameters.

[0097] Take DNN as an example. The idea behind DNN comes from the neuronal structure of the brain. Each neuron can perform a weighted summation operation on its inputs and output the result of the weighted summation through a nonlinear function. For example, the input of a neuron is x = [x_0, x_1, ..., x_(N-1)], the corresponding weights are w = [w_0, w_1, ..., w_(N-1)], the bias of the weighted summation is b, and the nonlinear function f() can take various forms; for example, the nonlinear function f() can be the maximum value function max{0, x}. Then the effect of one neuron's execution is... Where N is a positive integer, and n is a positive integer greater than or equal to 0 and less than or equal to (N-1). The weights of the weighted summation operation of neurons in a neural network and the nonlinear function are called the parameters of the neural network. The parameters of all neurons in a neural network constitute the parameters of the neural network.

[0098] A DNN typically has multiple neural network layers, including an input layer, one or more hidden layers, and an output layer. Generally, the first layer is the input layer, the last layer is the output layer, and the layers in between are hidden layers. Each layer contains multiple neurons. Layers are fully connected; that is, any neuron in the i-th layer is connected to any neuron in the (i+1)-th layer. The input layer processes the received values ​​(i.e., the DNN's input) through neurons and then passes them to the hidden layers. Similarly, the hidden layers pass the computation results to the final output layer, producing the DNN's output. This application does not limit the structure and parameters used in the AI ​​model.

[0099] One of the model structure or model parameters can be predefined, while the other can be sent by the sender (e.g., the network side). Alternatively, both the model structure and model parameters can be sent by the sender (e.g., the network side). This application does not impose any restrictions on this.

[0100] The transmitting model can refer to sending model files and / or model parameters, while the receiving model can refer to receiving model files and / or model parameters. Currently, AI technology is being introduced into wireless communication systems. AI technology can be used for wireless channel information compression and reconstruction, beam management, and positioning enhancement, improving wireless communication performance based on trained AI models. Figure 3A is a schematic diagram of a possible wireless AI framework. As shown in Figure 3A, the wireless AI framework can include multiple modules such as data collection, model training, model management, model inference, and model storage.

[0101] AI-based radio access networks (RANs) require different types of wireless data at each stage of the AI ​​model development process. For example, model training requires a large amount of diverse data, model inference requires real-time data, and model monitoring requires near-real-time large amounts of data. Based on this, as shown in Figure 3B, a wireless AI framework can be divided from a data node perspective into data acquisition nodes, data processing nodes, data management nodes, data storage nodes, and data usage nodes. Data acquisition nodes acquire wireless data and transmit it to data processing or data storage nodes. Data processing nodes preprocess the acquired wireless data to better suit the AI ​​model's usage requirements, such as through data cleaning, filtering, normalization, or compression. Data storage nodes store, update, or delete wireless data or its identifiers. Data management nodes receive data usage requests from data usage nodes and establish data transmission between different nodes and configure data transmission parameters based on the different AI-enabled features in the requests. The data usage nodes are equipped with AI models to perform multiple functions, including AI model training, AI model validation, AI model monitoring, AI model inference, and data analysis. It should be noted that these data nodes can be deployed in separate modules or within the same module. For example, a device can simultaneously deploy data acquisition nodes, data processing nodes, data management nodes, and data storage nodes to realize the corresponding functions of each node.

[0102] The quality of wireless data is a key factor determining the performance of data usage in model training, monitoring, and inference. However, current wireless data quality information is primarily defined by static metrics such as data distribution (e.g., statistical characteristics), signal-to-noise ratio (SNR) at acquisition time, and label accuracy. These metrics fail to accurately reflect the quality of data during model usage. Therefore, selecting wireless data for model operations based on this static quality information before use cannot fully predict the model's performance, thus impacting both model and wireless communication performance.

[0103] In view of this, this application provides a communication method for dynamically updating the quality information of wireless data based on the performance of wireless data in model usage, thereby making the quality information of wireless data more real-time and accurate. Specifically, the method can be executed by a first device, and the method includes: the first device acquiring wireless data quality feedback information from at least one second device, wherein the wireless data quality feedback information is used to indicate the performance of wireless data usage, and the wireless data is used by the model. Then, the first device determines the quality information of the wireless data based on the wireless data quality feedback information.

[0104] In this method, the first device determines the quality information of wireless data based on the performance of wireless data in the model, dynamically reflecting the quality information of wireless data in real-world usage scenarios. This makes the quality information of wireless data more real-time and accurate, thereby enabling the model to operate according to appropriate wireless data and improving the performance of the model and wireless communication.

[0105] It should be noted that the first device refers to a communication device deployed with data management nodes or data storage nodes, capable of updating data quality. When the first device deploys data storage nodes, it can store and update the quality information of wireless data in a database. When the first device deploys data management nodes, it can store and update the quality information of wireless data, instruct data storage nodes in the same device to store and update the quality information of wireless data in a database, and instruct data storage nodes in other devices to store and update the quality information of wireless data, while simultaneously obtaining the updated quality information of wireless data from the aforementioned other devices. Of course, the first device may also include data acquisition nodes and data processing nodes, and can perform corresponding functions. The second device refers to a communication device deployed with an AI model, capable of using wireless data based on the AI ​​model and feeding back quality information to the first device. In some possible implementations, the second device may also deploy functional nodes such as data acquisition nodes and data processing nodes. In other possible implementations, the second device may be the same device as the first device.

[0106] To make the technical solution of this application clearer and easier to understand, the communication method of this application will first be described from the perspective of the first device. In the embodiments of this application, the first device refers to a communication device that is equipped with a data management node or a data storage node and is capable of updating data quality.

[0107] Referring to the flowchart of a communication method shown in Figure 4, the method includes the following steps:

[0108] S402: The first device acquires wireless data quality feedback information from at least one second device.

[0109] The wireless data quality feedback information is determined by the second device based on the performance of the AI ​​model when using wireless data. Specifically, the wireless data quality feedback information may include attribute identifiers of the wireless data and performance information of the AI ​​model based on the wireless data.

[0110] The attribute identifiers of wireless data can be identifiers of the function type implemented by the AI ​​model, identifiers of the AI ​​model itself, identifiers of the operation type of the AI ​​model, or data identifiers corresponding to the wireless data used by the AI ​​model, etc. In this way, the attribute identifiers of wireless data can accurately identify the performance information of each AI model, thereby ensuring the uniqueness of each piece of performance information and facilitating the recording, identification, retrieval, and management of relevant performance information.

[0111] The functional type implemented by the AI ​​model depends on its specific application in wireless communication. For example, the AI ​​model can be used for sending or receiving information. In sending information, the AI ​​model can be used for Channel State Information (CSI) compression, channel coding, symbol modulation, resource mapping, waveform modulation, and radio frequency (RF) processing, etc. Correspondingly, in receiving, the AI ​​model can be used for CSI reconstruction, channel decoding, symbol demodulation, de-resource mapping, waveform demodulation, or RF processing, etc. Of course, the AI ​​model may also be used in other processes in wireless communication to achieve other functions, and this application embodiment does not limit this in any way.

[0112] Furthermore, the second device can deploy multiple AI models to implement different functions. For a given function, multiple AI models can serve as alternatives, or multiple AI models may need to collaborate to complete a single possible function. Therefore, the attribute identifier of the wireless data can also include the identifier of the AI ​​model, used to pinpoint the specific model.

[0113] Furthermore, the performance of an AI model using wireless data at different stages of model operation may not be the same. For example, when an AI model uses wireless data during the training and inference phases, the channel state may change, causing the same wireless data to exhibit different performance in AI models at different operational stages. Simultaneously, the performance metrics for the data at different stages of the AI ​​model may also differ. For instance, during model training, performance metrics may be based on the data update gradient, the change in loss value, or prediction accuracy; during model inference, performance metrics may be based on prediction accuracy. Therefore, the attribute identifiers of wireless data can also include identifiers of the AI ​​model's operation type.

[0114] In some possible implementations, the attribute identifier of the wireless data may include the data identifier corresponding to the wireless data. The data identifier can be a dataset identifier, a data subset identifier, or a data sample identifier; it can also be a data acquisition configuration identifier, a data processing type identifier, or a data source type identifier. Among these, the data acquisition configuration identifier and the data source type identifier primarily pertain to the wireless data acquisition phase. The data acquisition configuration identifier indicates the parameter configuration options when acquiring wireless data, such as SNR, bandwidth, reference signal configuration, etc., during data acquisition; the data source type can be the wireless air interface, UE-side sensors, network-side environmental data, etc. Furthermore, if the first device is equipped with a data processing node capable of performing preprocessing operations on the wireless data, the data identifier may include a data processing type identifier to indicate the type of preprocessing performed on the wireless data. For example, preprocessing could include data cleaning, data filtering, data normalization, or data compression of the wireless data.

[0115] In some possible implementations, the performance of an AI model based on wireless data can be the absolute performance of the model under a specific target attribute. For example, the attribute of the wireless data can be the operation type of the AI ​​model. During model training, performance metrics can be based on data update gradients, changes in loss values, or prediction accuracy; during model inference, performance metrics can be based on prediction accuracy. Furthermore, the performance indicators can be normalized, allowing for better comparison of data usage within the AI ​​model. For instance, for different operations of the same model performing the same function, performance values ​​obtained when each operation uses data from different subsets can be normalized. This allows for a more accurate assessment of the performance information of different wireless data for that function, leading to a comparison and identification of the most suitable wireless data.

[0116] In other possible implementations, the performance of the AI ​​model based on wireless data can also be relative performance, that is, a relative ranking or relative value obtained by sorting or comparing within the target range. For example, Model A, used to implement channel estimation, used 10 data samples during training, correspondingly obtaining 10 performance information pieces. In this case, the performance in the wireless data quality feedback information sent by the second device to the first device can be represented in the form of a relative ranking, such as ranking the best to worst performance from 1 to 10. Alternatively, when the performance information involves hierarchical ranking, the second device can also send the corresponding level to the first device. It should be noted that the hierarchical rules can be formulated by the second device after obtaining the relevant performance information and sent to the first device, or they can be agreed upon in advance by the two devices. This application embodiment does not impose any limitations on this.

[0117] To make the above explanation clearer and easier to understand, this application provides a schematic diagram of wireless data quality feedback information. Referring to Figure 5, each row in the table represents a possible composition of data quality feedback information. For example, the wireless data quality feedback information corresponding to dataset ID-1 indicates the absolute quality of a specific dataset under a specific function, model, and operation. Specifically, it indicates the training performance of AI model 1 using dataset ID-1 during CSI compression / reconstruction. The wireless data quality feedback information corresponding to dataset ID-2 indicates the absolute quality of a specific dataset under a specific function and operation, and is not limited to a specific model. For example, it could be the value obtained by the second device by integrating various models. The wireless data quality feedback information corresponding to data subset ID-3 indicates the absolute quality of a data subset under a specific function and operation. The wireless data quality feedback information corresponding to data subset ID-4 indicates the quality ranking of a specific data subset under a specific operation. As shown in Figure 5, this information indicates that when implementing the CSI compression / reconstruction function, data subset ID-4 ranks 8th among the 10 data sets participating in the ranking during the model training phase. The wireless data quality feedback information corresponding to data sample IDs-5 and 6 indicates the quality level of a specific data sample under a specific function, model, and operation. The wireless data quality feedback information corresponding to data sample ID-7 indicates the absolute quality of a specific data sample under a specific function, model, and operation. The wireless data quality feedback information corresponding to data acquisition configuration ID-8 indicates the absolute quality of data corresponding to a specific acquisition configuration under a specific function, model, and operation. The information corresponding to dataset ID-9 indicates the static quality of a specific dataset, such as label accuracy, and can be used as the default or initial quality of the data. It is stored in the first device, and when the second device sends the wireless data quality feedback information corresponding to dataset ID-9, the first device can update the information accordingly.

[0118] It should be noted that the specific form of wireless data quality feedback information is not limited to the table shown in Figure 5; Figure 5 merely provides one possible example. That is, when the first device acquires and stores wireless data quality feedback information, it may use a relational database to store the information in a relational data format, or it may use a non-relational database to store the message in a non-relational data format.

[0119] In some possible implementations, the first device may also send a wireless data quality feedback instruction to the second device to instruct the second device to generate wireless data quality feedback information. The wireless data quality feedback instruction may be sent by the first device along with the wireless data provided to the second device, or it may be sent separately by the first device to the second device.

[0120] S404: The first device determines the quality information of the wireless data based on the wireless data quality feedback information.

[0121] The first device can determine the quality information of the wireless data used in the AI ​​model based on the wireless data quality feedback information sent by the second device. Specifically, the first device can classify, aggregate, or sort the acquired information based on the attribute identifiers of the wireless data.

[0122] In some possible implementations, the first device can determine the target attribute based on the attribute identifier in the wireless data quality feedback information, filter information including the target attribute, and then determine the final quality information based on the performance of the AI ​​model based on the wireless data carried in this information. For example, when the performance of the AI ​​model based on the wireless data is absolute performance, that is, the performance is expressed in the form of absolute values, the first device can determine a comprehensive performance value based on the performance values ​​in each of the filtered information, as the performance and quality of the wireless data under the target attribute. Specifically, the first device can use an average calculation method, selecting all or part of the performance values ​​under the target attribute for calculation. When calculating the average, the weight of each selected value can be the same or different.

[0123] Furthermore, to improve the real-time performance of wireless data quality information, the second device can periodically and proactively send wireless data quality feedback information or send it upon request from the first device. The first device can determine the quality information of the wireless data based on the different times at which the wireless data quality feedback information is acquired. That is, when the second device sends wireless data quality feedback information multiple times, for data with the same target attribute, the first device can determine the weight of multiple pieces of information according to the order of acquisition time, and reflect this in the average value of the computational performance. Specifically, the first device can acquire the time interval between each acquisition of wireless data quality feedback information and the determination of the wireless data quality information, and then determine the weight of the information according to the size of the time interval, assigning a larger weight to information with a shorter time interval and a smaller weight to information with a longer time interval. In other words, information that is closer to the time of determining the wireless data quality information will be assigned a larger weight, and information that is farther away from the time of determining the wireless data quality information will be assigned a smaller weight. In this way, determining the final wireless data quality information mainly based on feedback information that is closer in time can improve the real-time performance of the quality information and is closer to the actual application scenario. For example, the RAN can more effectively utilize AI technology to optimize network performance in areas such as beam management, positioning services, and compression and reconstruction of radio channel information based on time-sensitive quality information, thereby improving the network's intelligence level and service quality. In some possible implementations, the radio data quality feedback information may also include the time when the second device uses radio data. In this case, the first device can also obtain the time interval between the time when the second device uses radio data and the time when the first device determines the quality information of the radio data, using this as timing information to comprehensively determine the quality information of the radio data.

[0124] In some possible implementations, the communication system may include multiple second devices. The target attribute may include the function type implemented by the AI ​​model, the identifier of the AI ​​model, or the operation type of the AI ​​model. The first device can determine the quality information of the wireless data under the target attribute based on the relationships between the multiple second devices. For example, when the second devices are deployed on terminal devices, each second device may be assigned different weights based on factors such as geographical location, Quality of Service (QoS) requirements, and user density. The first device needs to process the wireless data quality feedback information fed back by the second devices according to their weight information. In this way, the quality information of the wireless data under the target attribute can be determined more accurately by utilizing the weight relationships between multiple second devices.

[0125] Furthermore, the first device can comprehensively determine the quality information of wireless data under the target attribute by combining the relationships between multiple second devices and the temporal relationships of multiple feedback quality information. Specifically, it can be determined using the following formula:

[0126] In the formula, i indicates a second device, i is a positive integer greater than 1, d indicates wireless data, and w i (d) indicates the weight information of device i, acci(d) indicates the performance of the AI ​​model of device i based on wireless data under the target attributes, t i The time for the first device to acquire wireless data quality feedback information sent by device i or the time for device i to use wireless data is specified, t indicates the time for the first device to determine the quality information of the wireless data, and f is a decreasing function. In this way, the weighting relationships between multiple second devices and the timing information when quality information is fed back can be used to more accurately and in real-time determine the quality information of wireless data under the target attribute.

[0127] Here, the decreasing function f represents the quality information of the final wireless data determined by the first device mainly based on feedback information that is closer in time. For example, the decreasing function f can be in the following form:

[0128] f(tt i )=exp-α(tt i ), where α is a positive number.

[0129] At this point, the quality information of the wireless data under the target attribute can be determined as follows:

[0130] In some possible implementations, the performance of AI models based on wireless data can also be relative. In this case, the performance in the wireless data quality feedback information sent by the second device to the first device can be represented in the form of a relative ranking, such as ranking the best to worst performance from 1 to 10. The first device can determine the comprehensive ranking result based on the performance ranking information fed back by multiple second devices and the weight information of each of the multiple second devices. For example, for a dataset D, each second device can rank the performance of the data in dataset D in the model according to one or more attributes such as the implemented function, model identifier, or model operation type, and send the ranking information to the first device. Then, the first device obtains the performance ranking information of multiple second devices based on dataset D, and performs a unified ranking of all performance ranking information related to the data in dataset D according to the weight information of each second device to obtain the final ranking result of dataset D.

[0131] It should be noted that the second device needs to obtain numerical performance information before sorting. If the numerical performance information has not undergone uniform normalization, for example, if a performance value of 0.8 indicates different meanings or quality evaluation granularities for data used in training or inference operations, then these two pieces of performance information cannot be compared or ranked. For example, the performance information for implementing CSI compression / reconstruction using model A and performing training operations in dataset D, and the performance information for implementing channel estimation using model B and performing inference operations, cannot be compared and ranked. In this case, the performance information participating in the ranking needs to have the same attributes.

[0132] Similarly, if the performance information based on numerical values ​​has undergone uniform normalization, the second device can sort the normalized performance information. Furthermore, if all performance information related to dataset D has been normalized, the second device can perform an overall sort of all performance information related to dataset D and feed it back to the first device. The first device can then perform an overall sort of all acquired performance information according to the weights of each second device, determining the final sorting result. In this way, all performance information can be sorted at once, facilitating the subsequent selection of wireless data by the communication system.

[0133] Furthermore, the sorting information can also be time-sensitive. That is, when historical data is available, the weight of the most recently acquired data is increased, making the messages more real-time. Specifically, the second device can sort the data based on time weights when feeding back performance sorting information, or the first device can further weight the information sent by each node based on time weights when performing comprehensive sorting.

[0134] In some possible implementations, the first device can send the determined quality information to the second device, for example, in the form of wireless data attribute identifiers and corresponding performance. This allows the second device to obtain timely performance information based on the use of wireless data in the AI ​​model, enabling it to better allocate tasks and utilize wireless data.

[0135] In some possible implementations, the first device may also respond to a data request based on quality information sent by at least one second device, determine target wireless data, and then send the target wireless data to the at least one second device. In one possible implementation, the second device may select target wireless data based on wireless data quality information, allowing the first device to query a database and send the specified data; that is, the data request includes the specifically specified target wireless data. In another possible implementation, the first device may select target wireless data from a database and send it based on a data request sent by the second device; that is, the data request sent by the second device includes a range or conditions for wireless data quality requirements, and the first device determines and sends the target wireless data according to the required range or conditions. In this way, the second device can better implement the functions of wireless AI based on target wireless data with performance requirements.

[0136] Based on the above description, in the communication method of this application, a first device acquires wireless data quality feedback information from at least one second device. This feedback information indicates the performance of wireless data usage, and the wireless data is used by the model. Then, the first device determines the quality information of the wireless data based on the feedback information. Thus, the first device determines the wireless data quality information based on the performance of the wireless data in the model, dynamically reflecting the quality information of the wireless data in real-world usage scenarios. This makes the quality information of the wireless data more real-time and accurate, thereby facilitating guidance for the model to operate based on appropriate wireless data and improving the performance of both the model and wireless communication.

[0137] The communication method of this application will now be described from the perspective of the second device. In this embodiment, the second device refers to a communication device that is equipped with an AI model, can use wireless data based on the AI ​​model, and can feed back quality information to the first device.

[0138] Referring to the flowchart of a communication method shown in Figure 6, the method includes the following steps:

[0139] S602: The second device generates wireless data quality feedback information based on the performance of the AI ​​model based on wireless data.

[0140] The wireless data quality feedback information is determined by the second device based on the performance of the AI ​​model when using wireless data. Specifically, the wireless data quality feedback information may include attribute identifiers of the wireless data and performance information of the AI ​​model based on the wireless data.

[0141] The attribute identifiers of wireless data can be identifiers of the function type implemented by the AI ​​model, identifiers of the AI ​​model itself, identifiers of the operation type of the AI ​​model, or data identifiers corresponding to the wireless data used by the AI ​​model, etc. In this way, the attribute identifiers of wireless data can accurately identify the performance information of each AI model, thereby ensuring the uniqueness of each piece of performance information and facilitating the recording, identification, retrieval, and management of relevant performance information.

[0142] The performance of an AI model based on wireless data can be the absolute performance of the model under a certain target attribute, or a relative ranking or relative value obtained by sorting or comparing within the target range. When the performance information involves hierarchical ranking, the second device can also send the corresponding level to the first device. It should be noted that the hierarchical rules can be formulated by the second device after obtaining relevant performance information and then sent to the first device, or they can be agreed upon in advance by the two devices. This application embodiment does not impose any limitations on this.

[0143] In some possible implementations, before the second device uses the wireless data in the AI ​​model, it can also send a data request to the first device to obtain wireless data. If the first device has not yet updated the wireless data quality information, it can select wireless data and send it to the second device based on the original data quality information. For example, the first device can select wireless data and send it to the second device based on static indicators such as the distribution of the wireless data (e.g., the statistical characteristics of the data), the signal-to-noise ratio (SNR) at the time of acquisition, and tag accuracy.

[0144] S604: The second device sends wireless data quality feedback information to the first device to determine the quality information of the wireless data.

[0145] The second device can actively and periodically send wireless data quality feedback information to the first device, or it can send wireless data quality feedback information as a response to the first device after receiving a wireless data quality feedback instruction from the first device. The specific form of the wireless data quality feedback information can be seen in an example shown in Figure 5. The wireless data quality feedback instruction can be sent by the first device along with the wireless data it provides to the second device, or it can be sent separately by the first device to the second device.

[0146] In some possible implementations, the second device can also receive quality information sent by the first device, which is determined by the first device based on the wireless data quality feedback information sent by the second device. This allows for timely synchronization of wireless data quality information with the first device, facilitating the second device's subsequent request for qualified wireless data from the first device and better realizing the functions in wireless communication.

[0147] In some possible implementations, the second device may also send a data request based on quality information to the first device, wherein the data request is used to determine target wireless data. In one possible implementation, the second device may select target wireless data based on wireless data quality information, causing the first device to query a database and send the specified data; that is, the data request includes the specifically specified target wireless data. In another possible implementation, the data request sent by the second device includes a range or conditions for wireless data quality requirements, and the first device determines and sends the target wireless data according to the required range or conditions. Then, the second device may receive the target wireless data sent by the first device in response and apply it to the operational tasks of the wireless AI model.

[0148] Based on the above description, embodiments of this application provide a communication method in which a second device generates wireless data quality feedback information based on the performance of an AI model based on wireless data, and sends the wireless data quality feedback information to a first device to determine the quality information of the wireless data. In this way, the second device can promptly provide feedback to the first device on the performance of the wireless data as demonstrated in the AI ​​model, which facilitates the first device in determining the quality information of the wireless data based on the wireless data quality feedback information, and thus dynamically updating the quality of the wireless data.

[0149] The following section will introduce the communication method of this application by showing the specific deployment of the first and second devices in specific scenarios, as illustrated in Figures 7 to 10.

[0150] It should be noted that the following scenario embodiments are merely illustrative examples, and the first and second devices may also be other devices with the above-described functions. In some possible implementations, the first and second devices may also be the same device, such as a base station. The base station can update the quality information of wireless data in real time based on its deployed AI model and related data function modules or nodes, thereby selecting more suitable wireless data for the AI ​​model and improving the quality of wireless network services.

[0151] Refer to Figure 7 for a schematic diagram of a communication method scenario. As shown in Figure 7, the first device is a network device, specifically a base station; the second device is a terminal, which is equipped with an AI model. The method provided in this application embodiment includes:

[0152] S702: The terminal sends a wireless data request to the network device.

[0153] S704: Network devices send wireless data to terminals.

[0154] In some possible implementations, the network device can also send a wireless data quality feedback indication to the terminal. This wireless data quality feedback indication can be sent by the network device along with the wireless data it provides to the terminal, or it can be sent separately by the network device to the terminal.

[0155] S706: The model uses wireless data.

[0156] S708: The terminal sends wireless data quality feedback information to the network device.

[0157] S710: Network devices determine the quality information of wireless data.

[0158] In some possible implementations, the network device can also send specific quality information to the terminal, enabling the terminal to send a wireless data request based on the quality information. The network device can determine the target wireless data based on the wireless data request and the quality information, and then send it to the terminal. The specific details of the above steps are similar to the methods corresponding to Figures 4 and 6, and will not be repeated here.

[0159] Refer to Figure 8 for a schematic diagram of a communication method scenario. As shown in Figure 8, the first device is a terminal; the second device is a network device, specifically a base station (BS), which deploys an AI model. The method provided in this application embodiment includes:

[0160] S802: The network device sends a wireless data request to the terminal.

[0161] S804: The terminal sends wireless data to the network device.

[0162] In some possible implementations, the terminal may also send a wireless data quality feedback indication to the network device. This wireless data quality feedback indication may be sent by the terminal along with the wireless data it provides to the network device, or it may be sent separately by the terminal to the network device.

[0163] S806: The model uses wireless data.

[0164] S808: Network devices send wireless data quality feedback information to terminals.

[0165] S810: The terminal determines the quality information of the wireless data.

[0166] In some possible implementations, the terminal can also send specific quality information to the network device, enabling the network device to send a wireless data request based on the quality information. The terminal can determine the target wireless data based on the wireless data request and the quality information, and then send it to the network device. The specific details of the above steps are similar to the methods corresponding to Figures 4 and 6, and will not be repeated here.

[0167] Refer to Figure 9 for a schematic diagram of a communication method scenario. As shown in Figure 9, the first device is a network device, specifically a base station, core network (CN), or operations administration and maintenance (OAM); the second device is a terminal and / or an over-the-top (OTT) server, on which an AI model is deployed. The OTT server is an internet-based server. The method of this application embodiment demonstrates the process by which a communication system determines the quality information of wireless data when an AI model is deployed in a terminal and / or an OTT server. The method provided in this application embodiment includes:

[0168] S902: The network device sends a downlink reference signal to the terminal.

[0169] S904: Terminal collects data, and the terminal sends the collected dataset.

[0170] Reference signals are predefined signals used by network devices and terminals during communication to provide a reference for communication services in different scenarios. Terminals can measure and acquire wireless data based on downlink reference signals, such as Layer 1 real-time wireless data, CSI, beam received power, etc.; Layer 3 wireless quality data, cell received signal quality, etc.; and minimized drive test data, terminal wireless trajectory data, received signal quality, etc. The specific acquisition method follows the existing communication protocol.

[0171] Then, the network device can send the collected data, or dataset, to the network device or OTT server for storage. In Figure 9, the dashed line represents the case where the dataset is stored in the OTT server and the quality information is determined on the OTT server side, while the solid line represents the case where the dataset is stored in the network device and the quality information is determined on the network device side.

[0172] S906a: The OTT server sends a data request to the network device.

[0173] S906b: The OTT server sends a data request to the terminal.

[0174] When an OTT server has deployed an AI model and does not store a dataset, the OTT server can send data requests directly to the network device or send data requests to the network device through a terminal. When sending data requests to the network device through a terminal, the OTT server can also send instruction information to the terminal, instructing the terminal to send data requests to the network device.

[0175] S906b / c: The terminal sends a data request to the network device.

[0176] When a terminal deploys an AI model, or receives an instruction from an OTT server to send a data request to a network device through the terminal, the terminal can send a data request to the network device.

[0177] S906d: The terminal sends a data request to the OTT server.

[0178] When an AI model is deployed on a terminal and the dataset is stored on an OTT server, the terminal can send a data request to the OTT server.

[0179] S908a: Network devices can send wireless data to OTT servers.

[0180] Corresponding to S906a, the network device sends wireless data to the OTT server. In some possible implementations, the network device may also send wireless data quality feedback instructions at the same time.

[0181] S908b / c: Network devices send wireless data to terminals.

[0182] Corresponding to S906b / c, network devices can send wireless data to terminals for use by AI models within the terminals, or to enable terminals to send wireless data to OTT servers. In some possible implementations, the network device may also send wireless data quality feedback instructions.

[0183] S908b: The terminal sends wireless data to the OTT server.

[0184] Corresponding to S906b and S908b / c, after receiving wireless data sent by the network device, the terminal sends wireless data to the OTT server. In some possible implementations, the terminal can also send a wireless data quality feedback indication at the same time.

[0185] S908d: The OTT server sends wireless data to the terminal.

[0186] Corresponding to S906d, the OTT server can send wireless data to the terminal. In some possible implementations, the OTT server can also send wireless data quality feedback instructions.

[0187] S910: The model uses wireless data.

[0188] When a terminal and / or OTT server is equipped with an AI model, it can utilize the received wireless data to perform wireless AI functions, such as wireless channel information compression and reconstruction, beam management, and positioning enhancement, thereby improving the quality and efficiency of wireless transmission.

[0189] S912a: The OTT server sends wireless data quality feedback information to network devices.

[0190] Corresponding to S906a and S908a, the OTT server can send wireless data quality feedback information to network devices.

[0191] S912b: The OTT server sends wireless data quality feedback information to the terminal.

[0192] Corresponding to S906b and S908b, the OTT server can send wireless data quality feedback information to the terminal, so that the terminal can forward the wireless data quality feedback information to the network device.

[0193] S912b / c: The terminal sends wireless data quality feedback information to the network device.

[0194] Corresponding to S906b / c, S908b / c and S912b, the terminal can send the wireless data quality feedback information it generates to the network device, or it can forward the wireless data quality feedback information generated by the OTT server to the network device.

[0195] S912d: The terminal sends wireless data quality feedback information to the OTT server.

[0196] For S906d and S908d, the terminal can send wireless data quality feedback information to the OTT server.

[0197] S914: Network devices determine the quality information of wireless data, and when an OTT server stores a dataset, the OTT server determines the quality information of the wireless data.

[0198] In some possible implementations, the network device can also send specific quality information to the terminal and / or OTT server to send a wireless data request based on the quality information. The network device can determine the target wireless data based on the wireless data request and the quality information, and send it to the terminal and / or the OTT server. Alternatively, when the OTT server stores a dataset, it can also send specific quality information to the terminal to send a wireless data request based on the quality information. Then, the OTT server can determine the target wireless data based on the wireless data request and the quality information, and send it to the terminal. The specific details of each step are similar to the methods corresponding to Figures 4 and 6, and will not be repeated here.

[0199] Refer to Figure 10 for a scenario diagram of a communication method. As shown in Figure 10, the first device is a network device, specifically a core network (CN), operations administration and maintenance (OAM), etc., with the core network being used as an example for the following description; the second device is a base station. The method of this application embodiment demonstrates the process by which the communication system determines the quality information of wireless data on the network side when an AI model is deployed in a base station. The method provided in this application embodiment includes:

[0200] S1002a: Core network or OAM activated base station collects wireless data.

[0201] S1002b: Base station configuration for collecting wireless data.

[0202] S1004a: The base station sends a downlink reference signal to the terminal.

[0203] S1006a: Terminal data collection, and terminal reporting of collected data.

[0204] A reference signal is a predefined signal used by network devices and terminals during communication to provide a reference for communication services in different scenarios. Terminals can measure and acquire wireless data based on downlink reference signals.

[0205] S1004b: The base station configures the terminal to send an uplink reference signal.

[0206] S1006b: Base station data collection.

[0207] Compared to S1004a and S1006a, in some other possible implementations, the base station can also be configured to send uplink reference signals to enable the base station to perform tasks such as wireless data collection and measurement.

[0208] S1008: The base station sends a dataset to the core network or OAM.

[0209] S1010: The core network or OAM sends wireless data to the base station.

[0210] In some possible implementations, the core network can also send radio data quality feedback instructions together.

[0211] S1012: The model uses wireless data.

[0212] S1014: The base station sends wireless data quality feedback information to the core network or OAM.

[0213] S1016: The core network or OAM determines the quality information of the radio data.

[0214] In some possible implementations, the core network or OAM can also send specific quality information to the base station, enabling the base station to send a radio data request based on the quality information. The core network or OAM can determine the target radio data based on the radio data request and the quality information, and then send it to the base station. The specific details of the above steps are similar to the methods corresponding to Figures 4 and 6, and will not be repeated here.

[0215] The first apparatus provided in the embodiments of this application will now be described. Please refer to FIG11, which is a schematic structural diagram of the first apparatus in the embodiments of this application. The first apparatus 1100 can be used to perform the above-described method embodiments. The first apparatus 1100 includes a transceiver module 1101 and a processing module 1102.

[0216] The processing module 1102 is used for data processing. The transceiver module 1101 can implement the corresponding communication functions. The transceiver module 1101 can also be called a communication interface or a communication module.

[0217] In some possible implementations, the first device 1100 may further include a storage module, which can be used to store program code, program instructions and / or data, and the processing module 1102 can read the instructions and / or data in the storage module so that the first device 1100 can implement the aforementioned method embodiments.

[0218] The first device 1100 can be used to perform the actions performed by the first device in the above method embodiments. The first device 1100 can be a terminal device, a network device, or a component configurable on a terminal device or a network device. The processing module 1102 is used to perform processing-related operations on the first device side in the above method embodiments. The transceiver module 1101 is used to perform receiving-related operations on the first device side in the above method embodiments.

[0219] In some possible implementations, the transceiver module 1101 may include a sending module and a receiving module. The sending module is used to perform the sending operation in the above method embodiments. The receiving module is used to perform the receiving operation in the above method embodiments.

[0220] It should be noted that the first device 1100 may include a transmitting module but not a receiving module. Alternatively, the first device 1100 may include a receiving module but not a transmitting module. Specifically, it depends on whether the above-described scheme executed by the first device 1100 includes both transmitting and receiving actions.

[0221] For example, the first device 1100 is used to perform the following scheme:

[0222] The transceiver module 1101 is used to acquire wireless data quality feedback information sent by at least one second device, wherein the wireless data quality feedback information is used to indicate the performance of wireless data usage, and the wireless data is used by the model.

[0223] The processing module 1102 is used to determine the quality information of wireless data based on the wireless data quality feedback information.

[0224] In some possible implementations, the processing module 1102 is further configured to obtain the time interval between the first device acquiring wireless data quality feedback information and the first device determining the quality information of the wireless data. In this case, the processing module 1102 can determine the quality information of the wireless data under the target attribute based on the time interval and the performance of the AI ​​model based on the wireless data under the target attribute.

[0225] In some possible implementations, the processing module 1102 is also used to obtain weight information for each of the multiple second devices, and then determine the quality information of the wireless data based on the weight information.

[0226] In some possible implementations, transceiver module 1101 is also used to send quality information to at least one second device.

[0227] In some possible implementations, the processing module 1102 is further configured to determine the target wireless data in response to a data request based on quality information sent by at least one second device. The transceiver module 1101 can then be configured to send the target wireless data to at least one second device.

[0228] In some possible implementations, the transceiver module 1101 is further configured to transmit wireless data to at least one second device and to transmit a wireless data quality feedback indication to at least one second device, the wireless data quality feedback indication being used to instruct at least one second device to generate wireless data quality feedback information.

[0229] It should be understood that the specific procedures for each module to perform the above-mentioned corresponding processes have been described in detail in the above method embodiments, and will not be repeated here for the sake of brevity.

[0230] The processing module 1102 in the above embodiments can be implemented by at least one processor or processor-related circuitry. The transceiver module 1101 can be implemented by a transceiver or transceiver-related circuitry. The transceiver module 1101 can also be referred to as a communication module or communication interface. The storage module can be implemented by at least one memory.

[0231] The following is a schematic diagram of the structure of a second device according to an embodiment of this application. Referring to FIG12, the second device 1200 can be used to perform the above-described method embodiment. The second device 1200 includes a transceiver module 1201 and a processing module 1202.

[0232] The processing module 1202 is used for data processing. The transceiver module 1201 can implement the corresponding communication functions. The transceiver module 1201 can also be called a communication interface or a communication module.

[0233] In some possible implementations, the second device 1200 may further include a storage module, which can be used to store program code, program instructions and / or data. The processing module 1202 can read the instructions and / or data in the storage module so that the second device 1200 can implement the aforementioned method embodiments.

[0234] The second device 1200 can be used to perform the actions performed by the second device in the above method embodiments. The second device 1200 can be a terminal device, a network device, or a component configurable on a terminal device or a network device. The processing module 1202 is used to perform processing-related operations on the second device side in the above method embodiments. The transceiver module 1201 is used to perform receiving-related operations on the second device side in the above method embodiments.

[0235] In some possible implementations, the transceiver module 1201 may include a sending module and a receiving module. The sending module is used to perform the sending operation in the above method embodiments. The receiving module is used to perform the receiving operation in the above method embodiments.

[0236] It should be noted that the second device 1200 may include a transmitting module but not a receiving module. Alternatively, the second device 1200 may include a receiving module but not a transmitting module. Specifically, it depends on whether the above-described scheme executed by the second device 1200 includes both transmitting and receiving actions.

[0237] For example, the second device 1200 is used to execute the following scheme:

[0238] The processing module 1202 is used to generate wireless data quality feedback information based on the performance of the artificial intelligence (AI) model based on wireless data.

[0239] The transceiver module 1201 is used to send wireless data quality feedback information to the first device for determining the quality information of wireless data.

[0240] In some possible implementations, the transceiver module 1201 is also used to receive quality information sent by the first device.

[0241] In some possible implementations, transceiver module 1201 is further configured to send a data request based on quality information to the first device, the data request being used to determine target wireless data. Then, transceiver module 1201 is further configured to receive the target wireless data sent by the first device in response.

[0242] In some possible implementations, the transceiver module 1201 is also configured to receive wireless data transmitted by the first device, and to receive a wireless data quality feedback indication transmitted by the first device, the wireless data quality feedback indication being used to instruct at least one second device to generate wireless data quality feedback information.

[0243] It should be understood that the specific procedures for each module to perform the above-mentioned corresponding processes have been described in detail in the above method embodiments, and will not be repeated here for the sake of brevity.

[0244] The processing module 1202 in the above embodiments can be implemented by at least one processor or processor-related circuitry. The transceiver module 1201 can be implemented by a transceiver or transceiver-related circuitry. The transceiver module 1201 can also be referred to as a communication module or communication interface. The storage module can be implemented by at least one memory.

[0245] This application also provides an apparatus 1300, which may be a terminal device, a processor in the terminal device, or a chip. The apparatus 1300 can be used to perform the operations performed by the first or second apparatus in the above method embodiments.

[0246] When device 1300 is a terminal device, Figure 13 shows a simplified structural diagram of the terminal device. As shown in Figure 13, the terminal device includes a processor, a memory, and a transceiver. The memory can store computer program code, and the transceiver includes a transmitter 1331, a receiver 1332, radio frequency circuitry (not shown in the figure), an antenna 1333, and input / output devices (not shown in the figure).

[0247] The processor is mainly used to process communication protocols and communication data; control terminal devices; execute software programs; and process data from software programs.

[0248] Memory is mainly used to store software programs and data.

[0249] Radio frequency (RF) circuits are mainly used for the conversion between baseband signals and RF signals, as well as for the processing of RF signals.

[0250] Antennas are primarily used for transmitting and receiving radio frequency signals in the form of electromagnetic waves.

[0251] Input / output devices can include touchscreens, displays, or keyboards. They are primarily used to receive user input and output data to the user. It should be noted that some types of terminal devices may not have input / output devices.

[0252] When data needs to be transmitted, the processor performs baseband processing on the data to be transmitted and outputs a baseband signal to the radio frequency (RF) circuit. The RF circuit then processes the baseband signal and transmits it outwards via an antenna as electromagnetic waves. When data is sent to the terminal device, the RF circuit receives the RF signal through the antenna. The RF circuit converts the RF signal back into a baseband signal and outputs it to the processor. The processor converts the baseband signal back into data and processes the data. For ease of explanation, Figure 13 only shows one memory, processor, and transceiver. In actual terminal device products, there may be one or more processors and one or more memories. Memory can also be called storage medium or storage device, etc. Memory can be independent of the processor or integrated with the processor; this embodiment does not limit this.

[0253] In this embodiment, the antenna and radio frequency circuit with transceiver function can be regarded as the transceiver module of the terminal device, and the processor with processing function can be regarded as the processing module of the terminal device.

[0254] As shown in Figure 13, the terminal device includes a processor 1310, a memory 1320, and a transceiver 1330. The processor 1310 may also be referred to as a processing unit, processing board, processing module, or processing device, etc. The transceiver 1330 may also be referred to as a transceiver unit, transceiver, or transceiver device, etc.

[0255] In some possible implementations, the devices in transceiver 1330 used for receiving can be considered as receiving modules, and the devices in transceiver 1330 used for transmitting can be considered as transmitting modules. That is, transceiver 1330 includes a receiver and a transmitter. A transceiver is sometimes also called a transceiver unit, transceiver module, or transceiver circuit. A receiver is sometimes also called a receiver unit, receiver module, or receiver circuit. A transmitter is sometimes also called a transmitter, transmitter module, or transmitter circuit.

[0256] The processor 1310 is used to perform processing operations on the first or second device side in the above embodiments. The transceiver 1330 is used to perform transmission and reception operations on the first or second device side in the above embodiments.

[0257] It should be understood that Figure 13 is merely an example and not a limitation, and the terminal device described above, including the transceiver module and the processing module, may not depend on the structure shown in Figures 11, 12, or 13.

[0258] When device 1300 is a chip, the chip includes a processor, a memory, and a transceiver. The transceiver can be an input / output circuit or a communication interface. The processor can be a processing module integrated on the chip, a microprocessor, or an integrated circuit. In the above method embodiments, the transmitting operation of the first or second device can be understood as the chip's output, and the receiving operation of the first or second device in the above method embodiments can be understood as the chip's input.

[0259] This application also provides an apparatus 1400, which can be a network device or a chip. The apparatus 1400 can be used to perform the operations performed by the first or second apparatus in the above embodiments.

[0260] When device 1400 is a network device, such as a base station, Figure 14 shows a simplified schematic diagram of a base station structure. The base station includes parts 1410, 1420, and 1430.

[0261] Part 1410 is mainly used for baseband processing and controlling the base station; Part 1410 is usually the control center of the base station, which can be called a processor, and is used to control the base station to perform the processing operations of the first or second device side in the above method embodiments.

[0262] Section 1420 is primarily used to store computer program code and data.

[0263] Section 1430 is primarily used for transmitting and receiving radio frequency (RF) signals, as well as converting RF signals to baseband signals. Section 1430 is commonly referred to as a transceiver module, transceiver, transceiver circuit, or transceiver unit. The transceiver module of section 1430, also called a transceiver or transceiver unit, includes antenna 1433 and RF circuitry (not shown in the figure), where the RF circuitry is mainly used for RF processing. In some possible implementations, the device in section 1430 that performs the receiving function can be considered a receiver, and the device that performs the transmitting function can be considered a transmitter; that is, section 1430 includes receiver 1432 and transmitter 1431. The receiver can also be called a receiving module, receiver circuit, or receiving circuit, and the transmitter can be called a transmitting module, transmitter, or transmitting circuit.

[0264] Sections 1410 and 1420 may include one or more circuit boards, each of which may include one or more processors and one or more memories. The processors are used to read and execute programs in the memories to implement baseband processing functions and control the base station. If multiple circuit boards exist, they can be interconnected to enhance processing capabilities. As an alternative implementation, multiple circuit boards may share one or more processors, multiple circuit boards may share one or more memories, or multiple circuit boards may simultaneously share one or more processors.

[0265] For example, in some possible implementations, the transceiver module of section 1430 is used to execute the transceiver-related processes performed by the first or second device in the above embodiments. The processor of section 1410 is used to execute the processing-related processes performed by the first or second device in the above embodiments.

[0266] It should be understood that Figure 14 is merely an example and not a limitation, and the network device described above, including the processor, memory, and transceiver, may not depend on the structure shown in Figures 11, 12, or 14.

[0267] When device 1400 is a chip, the chip includes a transceiver, a memory, and a processor. The transceiver can be an input / output circuit or a communication interface; the processor can be a processor integrated on the chip, a microprocessor, or an integrated circuit. In the above method embodiments, the transmitting operation of the first or second device can be understood as the chip's output, and the receiving operation of the first or second device in the above method embodiments can be understood as the chip's input.

[0268] This application also provides a computer-readable storage medium having stored thereon computer instructions for implementing the methods executed by the first or second device in the above method embodiments.

[0269] For example, when the computer program is executed by a computer, it enables the computer to implement the method performed by the first device or the second device in the above method embodiments.

[0270] This application also provides a computer program product containing instructions that, when executed by a computer, cause the computer to perform the method described in the above method embodiments by the first device or the second device.

[0271] This application also provides a communication system, which includes a first device and a second device. The first device is used to perform some or all of the operations performed by the first device in the above embodiments, and the second device is used to perform some or all of the operations performed by the second device in the above embodiments.

[0272] This application also provides a chip device, including a processor, for calling computer programs or computer instructions stored in the memory to cause the processor to execute the method provided in the above embodiments.

[0273] In some possible implementations, the input of the chip device corresponds to the receive operation in any of the above embodiments, and the output of the chip device corresponds to the send operation in any of the above embodiments.

[0274] In some possible implementations, the processor is coupled to the memory via an interface.

[0275] In some possible implementations, the chip device also includes a memory that stores computer programs or computer instructions.

[0276] In the embodiments of this application, the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), graphics processing units (GPUs), neural network processing units (NPUs), artificial intelligence processors, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor, any conventional processor, or one or more integrated circuits used to control the execution of a program for controlling the method provided in any of the above embodiments. The memory mentioned above can be read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, random access memory (RAM), etc. Some or all steps of the communication method in the embodiments of this application can be implemented by a GPU or NPU, or by a GPU or NPU in conjunction with other processors.

[0277] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the explanations and beneficial effects of the relevant contents in any of the above-mentioned devices can be referred to the corresponding method embodiments provided above, and will not be repeated here.

[0278] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0279] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0280] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0281] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the essential contribution of the technical solution of this application, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0282] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A communication method, characterized in that, Applied to a first device, the method includes: Obtain wireless data quality feedback information from at least one second device, the wireless data quality feedback information being used to indicate the performance of an artificial intelligence (AI) model based on wireless data; The quality information of the wireless data is determined based on the wireless data quality feedback information.

2. The method according to claim 1, characterized in that, The wireless data quality feedback information includes wireless data attribute identifiers and the performance of the AI ​​model based on wireless data. The wireless data attribute identifiers include at least one of function type identifiers, model identifiers, model operation type identifiers, or data identifiers.

3. The method according to claim 2, characterized in that, The data identifier includes at least one of a dataset identifier, a data subset identifier, a data sample identifier, a data acquisition configuration identifier, a data processing type identifier, or a data source type identifier. The data acquisition configuration identifier indicates the parameter configuration options when acquiring the wireless data, and the data processing type identifier indicates the type of preprocessing performed on the wireless data.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: obtaining the time interval between the first device acquiring the wireless data quality feedback information and the first device determining the quality information of the wireless data; Determining the quality information of the wireless data based on the wireless data quality feedback information includes: Based on the time interval and the performance of the AI ​​model based on the wireless data under the target attribute, the quality information of the wireless data under the target attribute is determined.

5. The method according to claim 4, characterized in that, The at least one second device includes a plurality of second devices, and the quality information of the wireless data under the target attribute is determined according to the following formula: Wherein, i indicates a second device, i is a positive integer greater than 1, d indicates the wireless data, and w i (d) Weight information of indicator device i, the acc i (d) Indicates the performance of the AI ​​model of device i based on the wireless data under the target attribute, wherein t i The time indicated by t is the time when the first device acquires the wireless data quality feedback information sent by device i, the time indicated by t is the time when the first device determines the quality information of the wireless data, and f is a decreasing function.

6. The method according to any one of claims 1 to 3, characterized in that, The wireless data quality feedback information also includes the performance ranking information of the AI ​​model based on the wireless data. The at least one second device includes multiple second devices. Determining the quality information of the wireless data based on the wireless data quality feedback information includes: Based on the performance ranking information fed back by the plurality of second devices and the weight information of each of the plurality of second devices, a comprehensive ranking result is determined.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The quality information is sent to the at least one second device.

8. The method according to claim 7, characterized in that, The method further includes: In response to a data request based on the quality information sent by the at least one second device, target wireless data is determined; The target wireless data is transmitted to the at least one second device.

9. The method according to any one of claims 1 to 8, characterized in that, The method further includes: The system transmits wireless data to the at least one second device and transmits a wireless data quality feedback indication to the at least one second device, the wireless data quality feedback indication being used to instruct the at least one second device to generate the wireless data quality feedback information.

10. A data processing method, characterized in that, Applied to a second device, the method includes: Based on the performance of the AI ​​model based on wireless data, generate wireless data quality feedback information; The wireless data quality feedback information is sent to the first device to determine the quality information of the wireless data.

11. The method according to claim 10, characterized in that, The wireless data quality feedback information includes wireless data attribute identifiers and the performance of the AI ​​model based on wireless data. The wireless data attribute identifiers include at least one of function type identifiers, model identifiers, model operation type identifiers, or data identifiers.

12. The method according to claim 11, characterized in that, The data identifier includes at least one of a dataset identifier, a data subset identifier, a data sample identifier, a data acquisition configuration identifier, a data processing type identifier, or a data source type identifier. The data acquisition configuration identifier indicates the parameter configuration options when acquiring the wireless data, and the data processing type identifier indicates the type of preprocessing performed on the wireless data.

13. The method according to any one of claims 10 to 12, characterized in that, The wireless data quality feedback information also includes the performance ranking information of the AI ​​model based on the wireless data.

14. The method according to any one of claims 10 to 13, characterized in that, The method further includes: Receive the quality information sent by the first device.

15. The method according to claim 14, characterized in that, The method further includes: Send a data request based on the quality information to the first device, the data request being used to determine the target wireless data; Receive the target wireless data sent by the first device in response.

16. The method according to any one of claims 10 to 15, characterized in that, The method further includes: The device receives the wireless data sent by the first device and receives a wireless data quality feedback indication sent by the first device, wherein the wireless data quality feedback indication is used to instruct the at least one second device to generate the wireless data quality feedback information.

17. A first device, characterized in that, The first device includes: A unit for performing the method as described in any one of claims 1 to 9.

18. A second device, characterized in that, The second device includes: A unit for performing the method as described in any one of claims 10 to 16.

19. A communication system, characterized in that, The system includes a first device and a second device, the first device being configured to perform the method as described in any one of claims 1 to 9, and the second device being configured to perform the method as described in any one of claims 10 to 16.

20. A computer storage medium, characterized in that, The computer storage medium is used to store a computer program, which, when executed, is used to implement the method of any one of claims 1 to 16.

Images