Communication method and apparatus

Abstract

The present application discloses a communication method and apparatus. The method comprises: acquiring first data, wherein the first data is data synthesized on the basis of a simulation scenario; and evaluating the first data on the basis of an evaluation rule of the first data, wherein the evaluation rule comprises an evaluation parameter of the first data and a condition required to be satisfied by the evaluation parameter, the evaluation rule is related to a performance requirement of a first model, and the first data is used for training the first model. In the present application, if the quality of the first data satisfies the performance requirement of the first model, the first data may be used to train the first model; alternatively, if the quality of the first data does not satisfy the performance requirement of the first model, the first data may be updated, for example, the quality of the updated first data may satisfy the performance requirement of the first model, and the updated first data may then be used to train the first model. Thus, the probability that synthetic data whose quality does not satisfy the performance requirement of a model is used as training data for the model is reduced, helping to improve the performance of the model.

Description

A communication method and apparatus

[0001] Cross-references to related applications

[0002] This application claims priority to Chinese Patent Application No. 202510126353.4, filed on January 27, 2025, entitled "A Communication Method and Apparatus", the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of communication technology, and in particular to a communication method and apparatus. Background Technology

[0004] Model training in the field of artificial intelligence (AI) has always been a core element driving technological progress. For a long time, real-world data collected in real-world scenarios has been crucial for model training, enabling the model to learn patterns from the real world and thus achieve more accurate predictions and decisions.

[0005] Current data acquisition methods have shifted from data collection to a combination of data collection and data synthesis. This is because, on the one hand, obtaining real-world data is becoming increasingly difficult. High collection costs, time-consuming and laborious annotation processes, and reduced data availability due to strict privacy regulations all make it difficult to meet the ever-increasing performance demands of models by relying solely on real-world data. On the other hand, as model application scenarios continue to expand and diversify, real-world data often cannot fully cover certain specific, rare, but crucial scenarios. Synthesis, on the other hand, is less costly and can generate data for specific scenarios through algorithms, thereby enhancing the model's generalization ability and its ability to handle complex situations.

[0006] However, the quality of synthetic data varies greatly. If poor-quality synthetic data is used as training data for a model, it will affect the model's performance. Summary of the Invention

[0007] This application provides a communication method and apparatus for effectively evaluating the quality of synthetic data, thereby improving the performance of the trained model.

[0008] In a first aspect, embodiments of this application provide a communication method, which can be executed by a first communication device. For example, the first communication device is a first communication equipment, or other equipment including the functions of a first communication equipment, or a chip system (or chip) or other functional module capable of implementing the functions of the first communication equipment. The chip system or functional module is, for example, disposed within the first communication device, wherein the first communication device is a communication device for evaluating synthetic data (e.g., referred to as a synthetic data evaluation function (SDEF)). The method includes: acquiring first data, the first data being data synthesized based on a simulation scenario; evaluating the first data according to evaluation rules for the first data, the evaluation rules including evaluation parameters of the first data and conditions that the evaluation parameters need to satisfy, the evaluation rules being related to the performance requirements of a first model, and the first data being used to train the first model.

[0009] In this embodiment, for first data synthesized based on a simulation scenario, the first communication device can evaluate the quality of the first data using evaluation parameters and the conditions that these parameters must meet. The evaluation parameters and the conditions they must meet can be related to the performance requirements of the first model (e.g., accuracy, generalization ability). For example, if the quality of the first data meets the performance requirements, the first model can be trained using the first data; or, if the quality of the first data does not meet the performance requirements, the first data can be updated. For instance, if the quality of the updated first data meets the performance requirements, the first model can be trained using the updated first data. By evaluating the first data, the probability of using synthesized data whose quality does not meet the model's performance requirements as training data for the model can be reduced, which is beneficial for improving the model's performance.

[0010] In one possible implementation, the evaluation rule further includes one or more of the following: a first method, wherein the first method is a method for obtaining the evaluation parameters; the type of the first data; or, the size of the first data.

[0011] The size of the first data can be related to the time required to evaluate its quality, the type of the first data can be related to the evaluation parameters of the first data and the method for obtaining those parameters, and the method for obtaining those parameters can be related to their accuracy. Therefore, when the first communication device evaluates the first data, in addition to considering the evaluation parameters and the conditions they must meet, it can also consider one or more of the following: the size of the first data, the type of the first data, or the method for obtaining those parameters. This allows the first communication device to comprehensively and accurately evaluate the quality of the first data.

[0012] In one possible implementation, the method further includes: obtaining the evaluation parameters according to a first method, wherein the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to satisfy.

[0013] In this embodiment, several possibilities for determining the first method are provided. One possibility is to determine the first method based on the type of the first data. For example, if the first data is user behavior data, and linear discriminant analysis (LDA) is suitable for analyzing user behavior data because it can consider category information, then the evaluation parameter can be obtained through LDA. Another possibility is to determine the first method based on the conditions that the evaluation parameter of the first data needs to meet. For example, if the evaluation parameter of the first data needs to meet conditions including an accuracy of 95% or higher, and since LDA has higher analytical accuracy than principal component analysis (PCA), then the evaluation parameter can be obtained through LDA. Therefore, the method for obtaining the evaluation parameter can be adaptively selected, thereby improving the accuracy of the obtained evaluation parameter.

[0014] In one possible implementation, the evaluation parameters include one or more of the following: a first parameter, which indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, the real scene being related to the simulation scene, and the third data is sample data used to synthesize the first data; or, a second parameter, which indicates the distribution characteristics of the first data.

[0015] In this implementation, several possible evaluation parameters for the first data are provided. For example, the degree of difference between the first data and real data collected in a real scene can be evaluated, and / or the degree of difference between the first data and the sample data used to synthesize the first data can be evaluated. If the degree of difference between the first data and the real data is small, it indicates that the first data is closer to the real data; if the degree of difference between the first data and the sample data is small, it indicates that the first data is closer to the sample data. Using such first data as training data for the model is beneficial to improving the model's performance. As another example, whether the distribution characteristics of the first data match the distribution characteristics of the model's training data can be evaluated. If they match, it indicates that using the first data as training data for the model is beneficial to improving the model's performance.

[0016] In one possible implementation, the method further includes: receiving a first request, the first request being used to request evaluation of the first data.

[0017] In this embodiment, one method for triggering the evaluation of the first data is provided. For example, the first communication device can trigger the evaluation of the first data upon receiving a first request; if the first request is not received, the first communication device may not perform the evaluation, thereby saving power consumption. In addition, the first communication device can also trigger the evaluation of the first data in other ways, such as triggering the evaluation of the first data upon receiving the evaluation rules for the first data, or the first communication device can actively trigger the evaluation of the first data, etc., without limitation.

[0018] In one possible implementation, the method further includes receiving the evaluation rule.

[0019] The evaluation rules can come from other communication devices, or they can be set by the first communication device itself. Alternatively, if the first communication device evaluates the first data upon receiving a first request, the evaluation rules may or may not be included in the first request.

[0020] In one possible implementation, acquiring the first data includes: synthesizing the first data.

[0021] This can be understood as the generation and evaluation of synthetic data being performed by the same communication device, without the need for additional devices, thus reducing the signaling interaction process.

[0022] In one possible implementation, the evaluation result of the first data is used to indicate that the evaluation parameter does not meet the required condition, and the method further includes: updating the first data according to the evaluation result to obtain fourth data, wherein the evaluation parameter of the fourth data meets the required condition.

[0023] If the evaluation result of the first data indicates that the evaluation parameters of the first data do not meet the required conditions, the first communication device can update the first data based on the evaluation result, for example, the evaluation parameters of the updated first data can meet the required conditions. If the model is trained based on the updated first data, the performance of the model can be improved.

[0024] In one possible implementation, acquiring the first data includes: receiving the first data from a second communication device, the second communication device being used to synthesize the first data.

[0025] This can be understood as the generation and evaluation of synthetic data being performed by different communication devices, with multiple communication devices able to collaborate, thus improving the ability of devices to coordinate.

[0026] In one possible implementation, the evaluation result of the first data is used to indicate that the evaluation parameter does not meet the required condition. After receiving the first data from the second communication device, the method further includes: sending the evaluation result to the second communication device, the evaluation result being used by the second communication device to update the first data.

[0027] If the generation and evaluation of the synthesized data can be performed by different communication devices, and the evaluation result of the first data indicates that the evaluation parameters of the first data do not meet the required conditions, then the first communication device can send the evaluation result to a second communication device used to synthesize the first data. This allows the second communication device to update the first data, for example, so that the evaluation parameters of the updated first data meet the required conditions. If the model is trained based on the updated first data, the model's performance can be improved.

[0028] Secondly, embodiments of this application also provide a communication method, which can be executed by a second communication device. For example, the second communication device is a second communication equipment, or other equipment including the functions of a second communication equipment, or a chip system (or chip) or other functional module capable of implementing the functions of the second communication equipment. The chip system or functional module is, for example, disposed within the second communication equipment, wherein the second communication equipment is a communication device for generating synthetic data (e.g., referred to as a synthetic data generation function (SDGF)). The method includes: receiving an evaluation result of first data, the evaluation result of which is obtained by evaluating the first data according to evaluation rules for the first data, the evaluation rules including evaluation parameters of the first data and conditions that the evaluation parameters need to satisfy, the evaluation rules being related to the performance requirements of a first model, the first data being used to train the first model, and the first data being data synthesized based on a simulation scenario; and updating the first data according to the evaluation result when the evaluation result indicates that the evaluation parameters do not meet the required conditions, to obtain fourth data, wherein the evaluation parameters of the fourth data meet the required conditions.

[0029] In one possible implementation, the evaluation rule further includes one or more of the following: a first method, wherein the first method is a method for obtaining the evaluation parameters; the type of the first data; or, the size of the first data.

[0030] In one possible implementation, the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to satisfy.

[0031] In one possible implementation, the evaluation parameters include one or more of the following: a first parameter, which indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, the real scene being related to the simulation scene, and the third data is sample data used to synthesize the first data; or, a second parameter, which indicates the distribution characteristics of the first data.

[0032] In one possible implementation, the method further includes: synthesizing the first data and sending the first data.

[0033] In one possible implementation, the method further includes: receiving a second request for requesting the synthesis of the first data, the second request including a second method for invoking a first application programming interface (API) for evaluating the first data.

[0034] In this embodiment, one method for triggering the synthesis of first data is provided. For example, the second communication device can trigger the synthesis of first data upon receiving a second request. This method can be understood as directly requesting the synthesis of first data. Besides this, other methods can also be used to trigger the synthesis of first data. If no second request is received, the second communication device may not perform the synthesis, thereby saving power consumption. Furthermore, the second communication device can also trigger the synthesis of first data in other ways, such as actively triggering the synthesis of first data, etc., without limitation. One method for triggering the evaluation of first data is also provided. For example, the second request may include a method of calling a first API. The first API may be the API of a third-party synthetic data evaluation platform. The second communication device can evaluate the first data by calling the first API upon receiving the second request, that is, by evaluating the first data through a third-party synthetic data evaluation platform.

[0035] In one possible implementation, before sending the first data, the method further includes encrypting the first data.

[0036] In this embodiment, if the second communication device evaluates the first data through a third-party synthetic data evaluation platform, the second communication device can encrypt the first data before sending it, thereby ensuring the security of the first data as much as possible.

[0037] In one possible implementation, the method further includes sending the evaluation rule.

[0038] In this embodiment, if the second communication device evaluates the first data through a third-party synthetic data evaluation platform, the second communication device can send evaluation rules to the third-party synthetic data evaluation platform so that the third-party synthetic data evaluation platform can evaluate the first data according to the evaluation rules, thereby effectively evaluating the quality of the first data.

[0039] Thirdly, embodiments of this application also provide a communication method, which can be executed by a third communication device. For example, the third communication device is a third communication equipment, or other equipment including the functions of a third communication equipment, or a chip system (or chip) or other functional module capable of implementing the functions of the third communication equipment. The chip system or functional module is, for example, disposed within the third communication equipment, which is a communication device for orchestrating data (e.g., called a data orchestrator (DO) / data controller (DC)). The method includes: sending a first request, the first request being used to request the evaluation of first data, the first data being data synthesized based on a simulation scenario.

[0040] In one possible implementation, the method further includes: sending an evaluation rule for the first data, the evaluation rule including evaluation parameters of the first data and conditions that the evaluation parameters need to satisfy, the evaluation rule being related to the performance requirements of the first model, and the first data being used to train the first model.

[0041] In one possible implementation, the evaluation rule further includes one or more of the following: a first method, wherein the first method is a method for obtaining the evaluation parameters; the type of the first data; or, the size of the first data.

[0042] In one possible implementation, the method further includes: obtaining the evaluation parameters according to a first method, wherein the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to satisfy.

[0043] In one possible implementation, the evaluation parameters include one or more of the following: a first parameter, which indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, the real scene being related to the simulation scene, and the third data is sample data used to synthesize the first data; or, a second parameter, which indicates the distribution characteristics of the first data.

[0044] In one possible implementation, the method further includes: sending a second request for requesting the synthesis of the first data, the second request including a second method for invoking a first API for evaluating the first data.

[0045] Fourthly, embodiments of this application also provide a communication device. The communication device can be the first communication device described in the first aspect above. The communication device possesses the functions of the first communication device described above. The communication device is, for example, a first communication equipment, or other equipment including the functions of a first communication equipment, or a chip system (or chip) or other functional module, which can implement the functions of the first communication equipment, and is, for example, disposed in the first communication equipment. In one optional implementation, the communication device includes a baseband device and a radio frequency device. In another optional implementation, the communication device includes a processing unit (sometimes also called a processing module) and a transceiver unit (sometimes also called a transceiver module). The transceiver unit can implement both transmitting and receiving functions. When the transceiver unit implements the transmitting function, it can be called a transmitting unit (sometimes also called a transmitting module), and when the transceiver unit implements the receiving function, it can be called a receiving unit (sometimes also called a receiving module). The transmitting unit and the receiving unit can be the same functional module, which is called the transceiver unit. This functional module can realize the transmitting and receiving functions; or, the transmitting unit and the receiving unit can be different functional modules, and the transceiver unit is a general term for these functional modules.

[0046] In one possible implementation, the processing unit is configured to acquire first data, which is data synthesized based on a simulation scenario; evaluate the first data according to evaluation rules, the evaluation rules including evaluation parameters of the first data and conditions that the evaluation parameters need to meet, the evaluation rules being related to the performance requirements of a first model, and the first data being used to train the first model.

[0047] Fifthly, embodiments of this application also provide a communication device. The communication device can be the second communication device described in the second aspect above. The communication device possesses the functions of the second communication device described above. The communication device is, for example, a second communication equipment, or other equipment including the functions of a second communication equipment, or a chip system (or chip) or other functional module, which can implement the functions of the second communication equipment, and is, for example, disposed in the second communication equipment. In one optional implementation, the communication device includes a baseband device and a radio frequency device. In another optional implementation, the communication device includes a processing unit (sometimes also called a processing module) and a transceiver unit (sometimes also called a transceiver module). The transceiver unit can implement both transmitting and receiving functions. When the transceiver unit implements the transmitting function, it can be called a transmitting unit (sometimes also called a transmitting module), and when the transceiver unit implements the receiving function, it can be called a receiving unit (sometimes also called a receiving module). The transmitting unit and the receiving unit can be the same functional module, which is called the transceiver unit. This functional module can realize the transmitting and receiving functions; or, the transmitting unit and the receiving unit can be different functional modules, and the transceiver unit is a general term for these functional modules.

[0048] In one embodiment, the transceiver unit is configured to receive the evaluation result of the first data, which is obtained by evaluating the first data according to the evaluation rules of the first data. The evaluation rules include the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet. The evaluation rules are related to the performance requirements of the first model. The first data is used to train the first model and is data synthesized based on the simulation scenario.

[0049] In one embodiment, the processing unit is configured to update the first data according to the evaluation result when the evaluation result indicates that the evaluation parameter does not meet the required condition, thereby obtaining fourth data, wherein the evaluation parameter of the fourth data meets the required condition.

[0050] Sixthly, embodiments of this application also provide a communication device. The communication device can be the third communication device described in the third aspect above. The communication device possesses the functions of the third communication device described above. This communication device is, for example, a third communication equipment, or other equipment including the functions of a third communication equipment, or a chip system (or chip) or other functional module, which can implement the functions of a second communication device, and is, for example, disposed in a third communication device. In one optional implementation, the communication device includes a baseband device and a radio frequency device. In another optional implementation, the communication device includes a processing unit (sometimes also called a processing module) and a transceiver unit (sometimes also called a transceiver module). The transceiver unit can implement both transmitting and receiving functions. When the transceiver unit implements the transmitting function, it can be called a transmitting unit (sometimes also called a transmitting module), and when the transceiver unit implements the receiving function, it can be called a receiving unit (sometimes also called a receiving module). The transmitting unit and the receiving unit can be the same functional module, which is called the transceiver unit. This functional module can realize the transmitting and receiving functions; or, the transmitting unit and the receiving unit can be different functional modules, and the transceiver unit is a general term for these functional modules.

[0051] In one implementation, the transceiver unit is configured to send a first request, the first request being for requesting evaluation of first data, the first data being data synthesized based on a simulation scenario.

[0052] A seventh aspect provides a communication device, which can be the first communication device described in the first aspect above. The communication device possesses the functions of the first communication device described above. The communication device is, for example, a first communication equipment, or other equipment including the functions of a first communication equipment, or a system-on-a-chip (or chip) or other functional module capable of implementing the functions of the first communication equipment, and the system-on-a-chip or functional module is, for example, disposed in the first communication equipment. The communication device includes a processor for executing the functions of the first communication device described in the first or fourth aspect above. Optionally, the communication device further includes a memory. The memory stores a computer program, and the processor is coupled to the memory. When the processor reads the computer program or instructions, it causes the communication device to execute the methods executed by the first communication device in the above aspects. Optionally, the memory and the processor are integrated together.

[0053] Eighthly, a communication device is provided, which can be the second communication device described in the second aspect above. The communication device possesses the functions of the second communication device described above. The communication device is, for example, a second communication equipment, or other equipment including the functions of a second communication equipment, or a system-on-a-chip (or chip) or other functional module capable of implementing the functions of the second communication equipment, and the system-on-a-chip or functional module is, for example, disposed within the second communication equipment. The communication device includes a processor for executing the functions of the second communication device described in the second or fifth aspect above. Optionally, the communication device further includes a memory. The memory stores a computer program, and the processor is coupled to the memory. When the processor reads the computer program or instructions, it causes the communication device to execute the methods performed by the second communication device in the above aspects. Optionally, the memory and the processor are integrated together.

[0054] Ninthly, a communication device is provided, which can be the third communication device described in the third aspect above. The communication device possesses the functions of the third communication device described above. The communication device is, for example, a third communication equipment, or other equipment including the functions of a third communication equipment, or a system-on-a-chip (or chip) or other functional module capable of implementing the functions of the third communication equipment, and the system-on-a-chip or functional module is, for example, disposed within the third communication equipment. The communication device includes a processor for executing the functions of the third communication device described in the third or sixth aspect above. Optionally, the communication device further includes a memory. The memory stores a computer program, and the processor is coupled to the memory. When the processor reads the computer program or instructions, it causes the communication device to execute the methods performed by the third communication device in the above aspects. Optionally, the memory and the processor are integrated together.

[0055] A tenth aspect provides a communication system comprising a first communication device, a second communication device, and a third communication device. The first communication device is used to perform the method described in the first aspect. For example, the first communication device can be implemented using the communication device described in the fourth or seventh aspect. The second communication device is used to perform the method described in the second aspect. For example, the second communication device can be implemented using the communication device described in the fifth or eighth aspect. The third communication device is used to perform the method described in the third aspect. For example, the third communication device can be implemented using the communication device described in the sixth or ninth aspect.

[0056] Eleventhly, a computer-readable storage medium is provided for storing a computer program or instructions that, when executed, cause the method performed by the first, second, or third communication device described above to be implemented.

[0057] In a twelfth aspect, a computer program product containing instructions is provided, which, when the computer program or instructions are run on a computer, causes the methods described in the above aspects to be implemented.

[0058] In a thirteenth aspect, a chip system is provided, including a processor and an interface, the processor being configured to call and execute instructions from the interface to enable the chip system to implement the methods described above.

[0059] The beneficial effects of the second to thirteenth aspects and their embodiments described above can be referred to the beneficial effects of the first aspect and any of its embodiments, and will not be repeated here. Attached Figure Description

[0060] Figure 1 is a schematic diagram of the architecture of a communication network provided in an embodiment of this application;

[0061] Figure 2 is a schematic diagram of a data plane architecture for a future communication network provided in an embodiment of this application;

[0062] Figure 3A is a schematic diagram of another communication network architecture provided in an embodiment of this application;

[0063] Figure 3B is a schematic diagram of another communication network architecture provided in an embodiment of this application;

[0064] Figure 4 is a flowchart illustrating a communication method provided in an embodiment of this application;

[0065] Figure 5A is a flowchart illustrating another communication method provided in an embodiment of this application;

[0066] Figure 5B is a flowchart illustrating another communication method provided in an embodiment of this application;

[0067] Figure 6 is a flowchart illustrating another communication method provided in an embodiment of this application;

[0068] Figure 7 is a flowchart illustrating another communication method provided in an embodiment of this application;

[0069] Figure 8 is a schematic diagram of a communication device provided in an embodiment of this application;

[0070] Figure 9 is a schematic diagram of another communication device provided in an embodiment of this application. Detailed Implementation

[0071] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings. This application will focus on various aspects, embodiments, or features of a system that may include multiple devices, components, modules, etc. It should be understood and appreciated that each system may include additional devices, components, modules, etc., and / or may not include all the devices, components, modules, etc. discussed in conjunction with the accompanying drawings. Furthermore, combinations of these solutions may also be used.

[0072] First, the relevant technical terms involved in the embodiments of this application will be explained. Unless otherwise specified, these explanations are for the purpose of making the embodiments of this application easier to understand, and should not be regarded as a strict limitation on the scope of protection claimed in this application.

[0073] I. Architecture of Communication Networks

[0074] Figure 1 is a schematic diagram of a communication network architecture provided in an embodiment of this application. This network architecture can be a 5th generation (5G) network architecture. The network architecture includes four components: terminal equipment, access network (AN), core network (CN), and data network (DN). The access network can be a radio access network (RAN). The terminal equipment, access network, and core network are the main components of the above network architecture. Logically, they can be divided into user plane and control plane. The control plane is responsible for the management of the mobile network, and the user plane is responsible for the transmission of service data.

[0075] (1) Terminal equipment

[0076] A terminal device is a device that provides voice and / or data connectivity to a user. Terminal devices may also be referred to as user equipment (UE), terminal, access terminal, terminal unit, terminal station, mobile station (MS), remote station, remote terminal, mobile terminal (MT), wireless communication equipment, terminal agent, or terminal equipment, etc.

[0077] For example, the terminal device can be a handheld device with wireless connectivity, or a vehicle with communication capabilities, such as in-vehicle equipment (e.g., in-vehicle communication device, in-vehicle communication chip). Examples of current terminal devices include: mobile phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistant (PDA) devices, handheld devices with wireless communication capabilities, computing devices or other processing devices connected to a wireless modem, tablet computers, computers with wireless transceiver capabilities, laptops, handheld computers, mobile internet devices (MIDs), wearable devices, virtual reality (VR) devices, augmented reality (AR) devices, wireless terminals in industrial control, wireless terminals in self-driving 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.

[0078] Terminal devices can be deployed on land, including indoors or outdoors, handheld, wearable, or vehicle-mounted; they can also be deployed on water (such as ships); and they can also be deployed in the air (such as airplanes, balloons, and satellites). This application does not limit the specific technologies, device forms, application scenarios, or names used in the terminal devices.

[0079] (2) Access Network

[0080] The access network is deployed close to the terminal equipment, providing network access functionality for authorized users in a specific area. It can determine different quality transmission tunnels to transmit user data based on user level, service requirements, and other factors. The access network manages and utilizes its own resources efficiently, providing access services to terminal equipment on demand, and is responsible for forwarding control signals and service data between the terminal equipment and the core network.

[0081] The access network can be the access network in the 3rd generation partnership project (3GPP), or it can be an open RAN (O-RAN or ORAN), a cloud radio access network (CRAN), or a communication network of two or more of the above.

[0082] Access network equipment is deployed in the access network to connect terminal devices to the wireless network. Access network equipment is typically connected to the core network via a wired link (e.g., fiber optic cable). Access network equipment can also be called access network devices or RAN equipment / nodes. Access network equipment can be a base station, an evolved NodeB (eNodeB), a transmission reception point (TRP), a next-generation NodeB (gNB) in 5G communication networks, or a base station in future communication networks.

[0083] Access network equipment can also be modules or units that perform some of the functions of a base station. For example, it can be a central unit (CU), a distributed unit (DU), or a radio unit (RU). The CU performs the functions of the base station's radio resource control (RRC) protocol and packet data convergence protocol (PDCP), and can also perform the functions of the service data adaptation protocol (SDAP). The CU can be further divided into a CU control plane (CP) (i.e., CU-CP) and a CU user plane (UP) (i.e., CU-UP). The DU performs the functions of the base station's radio link control (RLC) layer and medium access control (MAC) layer, and can also perform some or all of the physical layer functions. For specific descriptions of the above protocol layers, please refer to the relevant 3GPP technical specifications. The CU and DU can be set up separately or included in the same device, such as in the baseband unit (BBU). The RU can be included in radio frequency equipment or radio frequency units, such as in a remote radio unit (RRU), an active antenna unit (AAU), or a remote radio head (RRH). In different systems, CU, DU, or RU may have different names, but those skilled in the art will understand their meaning. For example, in an ORAN system, CU can also be called O-CU (open CU), DU can also be called O-DU (open DU), and RU can also be called O-RU (open RU). Any of the units among CU (or CU-CP, CU-UP), DU, and RU in this application can be implemented through software modules, hardware modules, or a combination of software and hardware modules. RA equipment can be a macro base station, a micro base station, an indoor station, a relay node, or a donor node, etc. The embodiments of this application do not limit the specific technology or equipment form used in the access network equipment.

[0084] (3) Core Network

[0085] The core network is responsible for maintaining the subscription data of the mobile network, managing the functions of the mobile network, and providing terminal devices with functions such as session management, mobility management, policy management, and security authentication.

[0086] The core network user plane includes user plane functions (UPF). The core network control plane includes, but is not limited to: access and mobility management functions (AMF), session management functions (SMF), authentication server functions (AUSF), network exposure functions (NEF), network function repository functions (NRF), policy control functions (PCF), unified data management (UDM), and application functions (AF).

[0087] UPF is primarily responsible for connecting to external networks and performing user data packet forwarding according to the routing rules of SMF, such as sending uplink data to the data network or other UPFs, and sending downlink data to other UPFs or access network devices.

[0088] AMF is primarily responsible for the access management and mobility management of terminal devices, such as maintaining the status of terminal devices, managing the reachability of terminal devices, forwarding mobility management non-access stratum (MM NAS) messages, and forwarding session management (SM) N2 messages.

[0089] SMF (Service Provider Function) is primarily responsible for session management in mobile networks, including establishing sessions for terminal devices, allocating and releasing resources for sessions, such as session quality of service (QoS), session paths, and forwarding rules. For example, it assigns Internet Protocol (IP) addresses to terminal devices and selects a UPF (User Provider Function) to provide packet forwarding functionality.

[0090] AUSF is primarily responsible for performing security authentication of terminal devices.

[0091] The NEF is used to connect other internal functions of the core network with external devices (such as application servers) to provide network capability information to external devices or to provide information from external devices to core network functions.

[0092] NRF is primarily responsible for providing storage and selection functions for network function entity information for other functions.

[0093] PCF is primarily responsible for user policy management, including policy authorization, quality of service and generation of billing rules, and distributing the corresponding rules to UPF through SMF to complete the installation of the corresponding policies and rules.

[0094] UDM is primarily responsible for data management. For example, UDM can manage user subscription information, including acquiring subscription information and providing it to other functions (such as AMF); generating 3GPP authentication credentials for terminal devices; and registering and maintaining the functions currently serving the terminal devices.

[0095] The AF is primarily responsible for providing various application service data to the control plane functions of the operator's communication network, or obtaining network data and control information from the control plane functions of the communication network.

[0096] Although not shown, the core network may include other possible functions, such as network data analysis function (NWDAF), without any specific limitations.

[0097] (4) Data Network

[0098] A data network, also known as a packet data network (PDN), is a network located outside of the operator's network. An operator's network can connect to multiple data networks. Data networks can be private networks, such as local area networks (LANs), external networks not controlled by the operator, such as the Internet, or dedicated networks jointly deployed by operators; the specific type is not limited.

[0099] It is understood that the following description will use function names from 5G communication networks as examples, and the embodiments of this application do not limit the function names. Figure 1 illustrates a service-oriented architecture for the core network control plane. In this architecture, each control plane function is connected to a service bus, and the interaction between control plane functions is achieved through service calls. That is, a control plane function exposes its services to other control plane functions for them to call. In other possible implementations, the core network control plane can also adopt a point-to-point communication method. In point-to-point communication, the communication interfaces between control plane functions will have a specific set of messages. Of course, in future communication networks, the names of these interfaces may remain unchanged or may be replaced with other names; this application does not limit this. In future communication networks, the above-mentioned functions or devices can still use their names from the 5G communication network, or they may have other names.

[0100] The functions in the various possible network architectures described above 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). Optionally, these functions can be implemented by a single device, multiple devices working together, or different functional modules within a single device; this application embodiment does not specifically limit this. In actual deployment, these functions can be co-located. For example, AMF can be co-located with SMF; SMF can be co-located with UPF. When two functions are co-located, the interaction between these two functions provided in this application embodiment becomes the internal operation of the co-located function or can be omitted.

[0101] II. Data plane of future communication networks

[0102] Based on the 5G network architecture shown in Figure 1, the current 5G user plane is used to carry session data, which cannot meet the requirements of "in-the-path computing" and "arbitrary topology." Therefore, the 5G user plane cannot carry data for future communication networks. To systematically address the challenges of data services and solve the problem that the existing 5G user plane cannot carry data for future communication networks, an independent data plane has been introduced for future communication networks. Figure 2 is a schematic diagram of a data plane architecture for a future communication network provided in an embodiment of this application. As shown in Figure 2, the data plane architecture includes: a data orchestrator (DO) (and a data controller (DC)), a data agent (DA), and a data storage function (DSF). Optionally, it also includes a data communication proxy (DCP) and / or a data processing function (DPF).

[0103] (1) DO: Supports data service request translation, transforming data service requests into the construction of data bearers, and orchestrating programmable data pipelines to provide the data service. For example, the DO obtains global information about the DA logical network based on the data service capabilities reported by the DA and the logical connection status between DAs; then, the DO selects a suitable DA based on the received data service request, orchestrates the data pipeline, and calculates and constructs the data forwarding path to form the data bearer. The DO sends data forwarding information to the DA through the data forwarding control protocol (DFCP) and updates and deletes data forwarding information as needed.

[0104] DO can be a newly added function within the core network. For example, the newly added DO can communicate with other functions in the core network through a service interface, or it can communicate with other functions in the core network through point-to-point communication. Alternatively, DO can be built into existing functions (such as AMF or SMF), that is, DO can be an SMF or AMF with added data orchestration function.

[0105] In other examples, based on the real-time requirements and cross-domain nature of the data service tasks, the data orchestrator can be subdivided into DO and DC. DO is responsible for coarse-grained, non-real-time data orchestration, while DC is responsible for fine-grained, real-time data orchestration.

[0106] (2) DA: Performs data services such as data acquisition, data preprocessing, data processing, data storage, data analysis, and data sharing as assigned by the orchestration. For example, DA can implement a variety of data processing functions, which are reported to DO as capabilities of DA during DA registration, and capability updates can be reported in a timely manner.

[0107] DA can be built into existing devices (such as terminal devices, access network devices) or functions (such as AMF, SMF), or it can be deployed independently, without any specific limitations.

[0108] (3) DSF: A storage extension component that acts as DA when large-scale data storage or long-term storage is required. DSF can be built into existing devices or functions, or it can be deployed independently, without any specific limitation.

[0109] In addition to DSF, Figure 2 may also include other data storage functions, such as the analysis data repository function (ADRF) and / or the sensing data repository function (SDRF), without any specific limitations.

[0110] (4) DCP: Provides an efficient data transmission mechanism that supports multiple transmission protocols, such as Transmission Control Protocol (TCP), User Datagram Protocol (UDP), Quick UDP Internet Connection (QUIC), or other transmission protocols. Data producers can send data to the DCP, and data consumers can subscribe to and retrieve data from the DCP. The DCP can be built into existing devices or functions, or it can be deployed independently; there are no specific limitations.

[0111] (5) DPF: A special type of DA that performs data analysis and processing functions. DPF can be built into existing devices or functions, or it can be deployed independently, without any specific limitations.

[0112] III. Composite Data

[0113] Synthetic data is data generated using algorithms, rather than data directly obtained from actual observations, measurements, or records. It can simulate the characteristics and distribution of real-world data collected in real-world scenarios and can be used for various purposes, such as supplementing datasets, testing systems, or protecting privacy. For example, in the image domain, images can be synthesized using generative adversarial networks (GANs). A GAN consists of a generator and a discriminator. The generator attempts to produce realistic images, while the discriminator determines whether an image is real or generated. Through continuous training, the generator can produce synthetic images with similar features to real images, such as synthesized face images or landscape images.

[0114] There are several methods for generating synthetic data. For example, method 1 uses statistical models: this method generates data using statistical distributions. For instance, if the temperature data for a certain region is known to follow a normal distribution, synthetic temperature data conforming to that distribution can be generated by estimating the parameters of the normal distribution (mean and standard deviation). Method 2 uses machine learning models: in natural language processing, autoregressive language models (such as large language models (LLMs)) can also be used to synthesize text data. By learning from large amounts of text, these models can generate synthetic data based on given prompts.

[0115] Synthetic data has various applications. For example, it can be used for data augmentation. Specifically, synthetic datasets can serve as pre-training datasets, large-scale datasets used in machine learning, especially deep learning, to pre-train models. This pre-training process typically occurs before the model is fine-tuned for a specific task (such as text classification or image recognition). The goal is to allow the model to learn general feature representations, enabling faster convergence and improved performance on the specific task during subsequent fine-tuning stages. Alternatively, synthetic datasets can also serve as instruction datasets, containing a set of steps and parameters used in machine learning and deep learning to guide the fine-tuning of pre-trained models on specific tasks or datasets. Fine-tuning is a crucial process for adapting a general model to a specific application domain; through these instruction sets, the model can better adapt and optimize its performance on specific tasks.

[0116] IV. Data Quality Assessment

[0117] As shown in Table 1, the quality assessment of artificial intelligence (AI) data can evaluate its basic attributes, such as standardization, accuracy, completeness, consistency, timeliness, and effectiveness. Building upon this, and depending on the focus of attention regarding the AI ​​data, further evaluation can be conducted on its extended attributes, such as understandability, usability, relevance, scalability, and compliance.

[0118] For AI data geared towards business domains, in addition to the relatively general quality attributes mentioned above, quality attributes that are particularly important for business scenarios can also be constructed based on business characteristics to conduct business-related assessments of AI data quality.

[0119] Table 1

[0120] Currently, model training is a core element driving technological progress in the field of AI. For a long time, real-world data collected in real-world scenarios has been crucial for model training, enabling the model to learn patterns from the real world and thus achieve more accurate predictions and decisions.

[0121] Current data acquisition methods have shifted from data collection to a combination of data collection and data synthesis. This is because, on the one hand, obtaining real-world data is becoming increasingly difficult. High collection costs, time-consuming and laborious annotation processes, and reduced data availability due to strict privacy regulations all make it difficult to meet the ever-increasing performance demands of models by relying solely on real-world data. On the other hand, as model application scenarios continue to expand and diversify, real-world data often cannot fully cover certain specific, rare, but crucial scenarios. Synthesis, on the other hand, is less costly and can generate data for specific scenarios through algorithms, thereby enhancing the model's generalization ability and its ability to handle complex situations.

[0122] However, the quality of synthetic data varies greatly. If poor-quality synthetic data is used as training data for a model, it will affect the model's performance.

[0123] In view of this, embodiments of this application provide a communication method for effectively evaluating the quality of synthetic data in order to improve the performance of the trained model.

[0124] First, the network architecture to which the embodiments of this application are adapted will be introduced. The technical solutions in the embodiments of this application can be applied to various communication networks, such as wireless local area networks (WLAN), wireless fidelity (Wi-Fi) systems, 4th generation (4G) networks (such as long term evolution (LTE) networks), 5G networks (such as new radio (NR) systems), future communication networks, or other similar communication networks, without limitation.

[0125] Figure 3A is a schematic diagram of another communication network architecture provided in an embodiment of this application. As shown in Figure 3A, the network architecture may include a first communication device, a second communication device, and a third communication device. Optionally, the network architecture may also include a fourth communication device. Optionally, the network architecture may also include a fifth communication device.

[0126] The first communication device, the second communication device, the third communication device, the fourth communication device, and the fifth communication device can be communication equipment or components of communication equipment in a communication network. The specific communication equipment / functions included in the communication network can be found in the description in Figure 1 or Figure 2.

[0127] For example, the first communication device is a first communication equipment or a component of the first communication equipment, such as a chip or chip system disposed in the first communication equipment; the second communication device is a second communication equipment or a component of the second communication equipment, such as a chip or chip system disposed in the second communication equipment; the third communication device is a third communication equipment or a component of the third communication equipment, such as a chip or chip system disposed in the third communication equipment; the fourth communication device is a fourth communication equipment or a component of the fourth communication equipment, such as a chip or chip system disposed in the fourth communication equipment; and the fifth communication device is a fifth communication equipment or a component of the fourth communication equipment, such as a chip or chip system disposed in the fourth communication equipment.

[0128] For example, the first communication device is a communication device used to evaluate synthetic data (e.g., called the synthetic data evaluation function (SDEF)), the second communication device is a communication device used to generate synthetic data (e.g., called the synthetic data generation function (SDGF)), the third communication device is a DO / DC, the fourth communication device is an AF, and the fifth communication device can be a NEF or a gateway (GW).

[0129] The first communication device can be a newly added function within the core network. For example, the newly added first communication device can communicate with other functions in the core network through a service-oriented interface, or it can communicate with other functions in the core network through point-to-point communication. Alternatively, the first communication device can also be built into other functions in the core network (such as DSF), that is, the first communication device can be a DSF with added synthetic data evaluation function. This application does not limit this aspect.

[0130] The second communication device can be a newly added function within the core network. For example, the newly added second communication device can communicate with other functions in the core network through a service-oriented interface, or it can communicate with other functions in the core network through point-to-point communication. Alternatively, the second communication device can also be built into other functions in the core network (such as DSF), that is, the second communication device can be a DSF with added synthetic data generation function. This application does not limit this aspect.

[0131] The first communication device and the second communication device can be the same communication device or different communication devices, and the embodiments of this application do not limit this.

[0132] For example, as shown in Figure 3B(1), SDEF and SDGF are built into DSF, that is, SDEF and SDGF are DSF with added synthetic data generation and synthetic data evaluation functions. In other words, the first communication device and the second communication device are the same communication device.

[0133] For example, as shown in Figure 3B(2), SDEF is a newly added function in the core network for evaluating synthetic data, and SDGF is built into DSF, that is, SDGF is DSF with added synthetic data generation function. In other words, the first communication device and the second communication device are different communication devices.

[0134] For example, as shown in Figure 3B(3), SDEF is built into DSF, that is, SDEF is DSF with added synthetic data evaluation function, and SDGF is a new function in the core network for generating synthetic data. In other words, the first communication device and the second communication device are different communication devices.

[0135] It is understood that the network architecture shown in Figures 3A and 3B may also include other possible devices / functions, and no specific limitations are imposed. The network architecture shown in Figures 3A and 3B is only one possible example. The network architecture and business scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of communication network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0136] It is understood that in the embodiments of this application, "when," "if," and "if" all refer to the device taking corresponding actions under certain objective circumstances, and are not time-limited, nor do they require the device to perform a judgment action, nor do they imply any other limitations. Unless otherwise specified, "if" and "if" can be substituted, and "when" and "in the case of" can be substituted. "When" and "if" / "if" can be substituted.

[0137] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0138] In this document, "used for indication" can include both direct and indirect indication. For example, when descriptive information I is used to indicate information J, it can mean that information I directly indicates information J or indirectly indicates information J, but it does not necessarily mean that information I carries information J.

[0139] Let information J, indicated by information I, be called the information to be indicated. In practice, there are many ways to indicate the information to be indicated, such as, but not limited to, directly indicating the information to be indicated, such as the information itself or its index. It can also be indirectly indicated by indicating other information, where there is a relationship between the other information and the information to be indicated. It can also indicate only a part of the information to be indicated, while the other parts are known or pre-agreed upon. For example, the indication of specific information can be achieved by using a pre-agreed (e.g., protocol-defined) order of various pieces of information, thereby reducing indication overhead to some extent. Simultaneously, common parts of various pieces of information can be identified and indicated uniformly to reduce the indication overhead caused by individually indicating the same information.

[0140] Furthermore, the specific instruction method can also be any existing instruction method, such as, but not limited to, the above-mentioned instruction methods and their various combinations. As described above, for example, when multiple pieces of information of the same type need to be indicated, the instruction methods for different pieces of information may differ. In specific implementation, the required instruction method can be selected according to specific needs. This application embodiment does not limit the selected instruction method. Therefore, the instruction methods involved in this application embodiment should be understood to cover various methods that enable the party to be instructed to obtain the information to be indicated.

[0141] In the embodiments of this application, "send" and "receive" indicate the direction of signal transmission. For example, "send information to XX" can be understood as the destination of the information being XX, which may include direct transmission via the air interface or indirect transmission via the air interface by other units or modules. "Receive information from YY" can be understood as the source of the information being YY, which may include direct reception from YY via the air interface or indirect reception from YY via the air interface by 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.

[0142] Information may undergo necessary processing, such as encoding and modulation, between the source and destination ends, but the destination end can understand the valid information from the source end. Similar statements in the embodiments of this application can be understood in a similar way, and will not be repeated here.

[0143] In this application embodiment, the number of nouns, unless otherwise specified, refers to "singular nouns or plural nouns," that is, "one or more." "At least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, or B exists alone, where A and B can be singular or plural. The character " / " can indicate that the related objects before and after are in an "or" relationship. For example, A / B means: A or B. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c means: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0144] In this application, the ordinal numbers such as "first" and "second" are used to distinguish multiple objects, and are not used to limit the size, content, order, timing, priority, or importance of the multiple objects. For a technical feature, the technical features within that technical feature are distinguished by "A," "B," "C," and "D," and there is no sequential or hierarchical order among the technical features described by "A," "B," "C," and "D." For example, scenario A and scenario B are only used to distinguish different content, and do not limit the sequential or hierarchical order, priority, or importance between scenario A and scenario B.

[0145] Based on the network architecture shown in Figures 3A and 3B, the communication method provided by the embodiments of this application is described below with reference to specific embodiments. The communication method provided by the embodiments of this application involves the interaction between multiple communication devices, such as a first communication device, a second communication device, a third communication device, a fourth communication device, and a fifth communication device.

[0146] The first communication device is SDEF or a component of SDEF, which is used to evaluate the synthesized data. The following description uses "the first communication device is SDEF" as an example. It is understood that the name of SDEF is not limited in the embodiments of this application.

[0147] The second communication device is an SDGF or a component of an SDGF, which is used to generate composite data. The following description uses "the second communication device is an SDGF" as an example. It is understood that the name SDEF is not limited in the embodiments of this application.

[0148] The third communication device is a DO / DC or a component of a DO / DC, and will be described below as "the third communication device is a DO / DC".

[0149] The fourth communication device is an AF or a component of the AF, and will be described below as "the fourth communication device is an AF".

[0150] The fifth communication device is a component of NEF / GW or NEF / GW, and will be described below as "the fifth communication device is NEF / GW".

[0151] Figure 4 is a schematic flowchart of a communication method provided in an embodiment of this application. As shown in Figure 4, the communication method includes the following steps.

[0152] S401 and SDEF acquire the first data, which is data synthesized based on the simulation scenario.

[0153] In the embodiments of this application, SDEF and SDGF can be the same communication device or different communication devices. This application does not limit this.

[0154] For example, take SDEF and SDGF as the same communication device. As shown in Figure 3B (1) above, SDEF and SDGF are built into DSF, that is, SDEF and SDGF are DSF with added synthetic data generation and synthetic data evaluation functions.

[0155] For example, consider SDEF and SDGF as different communication devices. As shown in Figure 3B(2) above, SDEF is a new function added to the core network, while SDGF is built into DSF, that is, SDGF is DSF with added synthetic data generation function. Or, as shown in Figure 3B(3) above, SDEF is built into DSF, that is, SDGF is DSF with added synthetic data evaluation function, while SDGF is a new function added to the core network.

[0156] In practice, depending on whether SDEF and SDGF are the same communication device, there are several scenarios in which SDEF obtains the first data, which will be introduced below.

[0157] In scenario A, SDEF and SDGF are the same communication device. As shown in Figure 5A, S401 may include the following steps.

[0158] Step A0, SDEF synthesizes the first data.

[0159] It is understood that SDEF can synthesize the first data based on the simulation scenario using a model. The model can be a statistical model, a machine learning model (such as a large language model (LLM)), or other possible models; this application does not limit the specific models used in this embodiment.

[0160] In one possible implementation, as shown in FIG5A, the present application may also perform the following steps before performing step A0.

[0161] Step A1: AF sends a synthetic data generation request 1.

[0162] Accordingly, the DO / DC receives Synthetic Data Generation Request 1. Synthetic Data Generation Request 1 is used to request the synthesis of first data.

[0163] It is understood that the synthetic data generation request 1 may include one or more of the following: the type of the first data, such as the first data being terminal location data; the quality requirements of the first data, such as the difference between the first data and real data collected in a real scene (or actual collected data), and / or the sample data used to synthesize the first data (or seed data) needing to be less than 20%, and the real scene being related to the simulation scene; or, the size of the first data. This application embodiment does not limit this.

[0164] It is understood that the DO / DC can determine the synthesis method of the first data based on the type and / or quality requirements of the first data included in the synthetic data generation request 1, such as synthesizing the first data according to the simulation scenario using LLM. The DO / DC can also determine the synthesis time of the first data based on the size of the first data included in the synthetic data generation request 1. This application embodiment does not limit this aspect.

[0165] Step A2: DO / DC sends a synthetic data generation request 2.

[0166] Accordingly, SDEF receives Synthetic Data Generation Request 2. Synthetic Data Generation Request 2 is used to request the synthesis of the first data.

[0167] It is understood that the synthetic data generation request 2 may include one or more of the following: the type of the first data; the quality requirements of the first data; the size of the first data; the synthesis method of the first data; or, the synthesis time of the first data. This application embodiment does not limit this.

[0168] It is understood that SDEF can determine the synthesis method of the first data based on the type and / or quality requirements of the first data included in the Synthetic Data Generation Request 2. SDEF can also determine the synthesis time of the first data based on the size of the first data included in the Synthetic Data Generation Request 2. This application embodiment does not limit this aspect.

[0169] In other words, the method and / or timing of the synthesis of the first data can be determined by the DO / DC instructing the SDEF, or it can be determined by the SDEF itself. This application does not limit this aspect.

[0170] In scenario B, SDEF and SDGF are different communication devices, as shown in Figure 5B. S401 may include the following steps.

[0171] Step B0: SDGF synthesizes the first data and sends the first data.

[0172] Accordingly, SDEF receives the first data.

[0173] It is understood that SDGF can synthesize first data based on the simulation scenario through a model. The model can be a statistical model, a machine learning model (such as LLM), or other possible models; this application does not limit this type of model.

[0174] In one possible implementation, as shown in FIG5B, the present application may also perform the following steps before performing step B0.

[0175] Step B1: AF sends a synthetic data generation request 1.

[0176] Accordingly, the DO / DC receives Synthetic Data Generation Request 1. Synthetic Data Generation Request 1 is used to request the synthesis of first data.

[0177] It is understood that the specific description of step B1 can be referred to step A1 above, and will not be repeated here.

[0178] Step B2: DO / DC sends a synthetic data generation request 2.

[0179] Accordingly, SDGF receives Synthetic Data Generation Request 2. Synthetic Data Generation Request 2 is used to request the synthesis of the first data.

[0180] It is understood that the specific description of step B2 can be referred to step A2 above, and will not be repeated here.

[0181] S402, SDEF evaluates the first data according to the evaluation rules of the first data. The evaluation rules of the first data include the evaluation parameters of the first data and the conditions that the evaluation parameters of the first data need to meet. The evaluation rules of the first data are related to the performance requirements of the first model. The first data is used to train the first model.

[0182] In this embodiment, the evaluation rules for the first data may include evaluation parameters for the first data and the conditions that the evaluation parameters for the first data must satisfy. These will be described below.

[0183] 1) The evaluation parameters of the first data may include a first parameter and / or a second parameter.

[0184] The first parameter can be used to indicate the degree of difference between the first data and the second and / or third data. The second data is real data collected in a real scene (or actual collected data), and the real scene is related to the simulation scene. The third data is sample data (or seed data) used to synthesize the first data. The first parameter can also be called a realism parameter, and the name of the first parameter is not limited in this embodiment.

[0185] The second parameter can be used to indicate the distribution characteristics of the first data. The first parameter can also be called a diversity parameter, and the name of the second parameter is not limited in the embodiments of this application.

[0186] As can be understood, as shown in Table 2, the evaluation parameters for primary data can include basic parameters such as normative parameters, accuracy parameters, completeness parameters, consistency parameters, timeliness parameters, effectiveness parameters, authenticity parameters, and diverse new parameters. Furthermore, the evaluation parameters for primary data can also include extended parameters such as understandability parameters, usability parameters, relevance parameters, scalability parameters, and compliance parameters.

[0187] It is understandable that the quality assessment of primary data can be based on evaluating its fundamental parameters, such as standardization, accuracy, completeness, consistency, timeliness, validity, authenticity, and diversity. Building upon this, and depending on the focus of attention for the primary data, further evaluation can be conducted on its extended parameters, such as understandability, usability, relevance, scalability, and compliance. This application does not limit this approach. Furthermore, for primary data oriented towards a business domain, in addition to the aforementioned relatively general evaluation parameters, it is also possible to construct evaluation parameters that are particularly relevant to the business scenario, thereby conducting a business-related assessment of the primary data's quality.

[0188] Table 2

[0189] 2) The conditions that the evaluation parameters of the first data must meet.

[0190] For example, when the evaluation parameters of the first data include the first parameter, the condition that the first parameter needs to meet is that the difference between the first data and the second and / or third data is less than 20%.

[0191] For example, when the evaluation parameters of the first data include the second parameter, the condition that the second parameter needs to satisfy is that the distribution characteristics of the first data conform to distribution characteristic 1.

[0192] In one possible implementation, the evaluation rule for the first data may further include one or more of the following: a first method, wherein the first method is a method for obtaining evaluation parameters of the first data. For example, the validity parameter of the first data can be obtained by methods such as linear discriminant analysis and principal component analysis; the authenticity parameter of the first data (i.e., the first parameter) can be obtained by methods such as distance metric and binary classifier; the diversity parameter of the first data (the second parameter) can be obtained by clustering algorithm; the type of the first data; or, the size of the first data. These will be described below.

[0193] It is understood that the first method may be determined based on the type of the first data and / or the conditions that the evaluation parameters of the first data need to meet.

[0194] For example, when the first data is of type 1, the validity parameters can be obtained through linear discriminant analysis; when the first data is of type 2, the validity parameters can be obtained through principal component analysis. As another example, when the validity parameters of the first data require an accuracy of 95% or higher, they can be obtained through linear discriminant analysis; when the validity parameters require an accuracy of 80%-90%, they can be obtained through principal component analysis.

[0195] It is understood that the relationship between the first method and the type of the first data and / or the evaluation parameters of the first data may be pre-configured, or may be defined by a standard, or may be negotiated by SDEF, SDGF or DO / DC. This application embodiment does not limit this.

[0196] It is understood that the performance requirements of the first model refer to the performance requirements of the first model, such as its accuracy, generalization ability, runtime, and resource consumption. This application does not limit these aspects.

[0197] Since the first data is used to train the first model, the quality requirements of the first data are related to the performance requirements of the first model. Furthermore, the evaluation rules for the first data used to assess its quality are also related to the performance requirements of the first model. For example, when the accuracy of the first model needs to be greater than 70%, the difference between the first data and the real data and / or sample data needs to be less than 20%, and the distribution characteristics of the first data need to conform to distribution characteristic 1, etc.

[0198] If the quality of the first data meets the performance requirements of the first model, then the first data can be used to train the first model; or, if the quality of the first data does not meet the performance requirements of the first model, the first data can be updated until the quality of the updated first data meets the performance requirements, and then the updated first data can be used to train the first model. This reduces the probability of using synthetic data whose quality does not meet the model's performance requirements as training data, which is beneficial for improving the model's performance.

[0199] In practice, depending on whether SDEF and SDGF are the same communication device, there are multiple scenarios for the evaluation rules of SDEF to obtain the first data, which will be introduced below.

[0200] In scenario A, SDEF and SDGF are the same communication device, as shown in Figure 5A. Before executing S402, this application may also perform the following steps.

[0201] Step A3: DO / DC obtains the evaluation rules for the first data and sends the evaluation rules for the first data.

[0202] Accordingly, SDEF receives the evaluation rules for the first data.

[0203] It is understood that the DO / DC can determine the evaluation parameters of the first data and the conditions that the evaluation parameters of the first data must meet based on the quality requirements of the first data included in the synthetic data generation request 1. Optionally, the DO / DC can determine the evaluation parameters of the first data and the conditions that the evaluation parameters of the first data must meet based on the type of the first data included in the synthetic data generation request 1 and the quality requirements of the first data.

[0204] It is understood that step A3 can be executed before step A2, after step A2, or simultaneously with step A2. This application embodiment does not limit this. Figure 5A shows an example where step A3 is executed after step A2.

[0205] It is understood that the DO / DC can request the SDEF to evaluate the first data through the evaluation rules of the first data, which can be understood as an indirect request. Alternatively, the DO / DC can also request the SDEF to evaluate the first data through a synthetic data evaluation request, which can be understood as a direct request. This application does not limit this approach.

[0206] It is understood that the synthetic data evaluation request can also be called the first request, and the name of the synthetic data evaluation request is not limited in the embodiments of this application.

[0207] When the DO / DC requests the SDEF to evaluate the first data through the first request, the evaluation rules of the first data may be encapsulated or carried in the first request, or they may not be encapsulated or carried in the first request. This application embodiment does not limit this.

[0208] In scenario B, SDEF and SDGF are different communication devices, as shown in Figure 5B. Before executing S402, this application may also perform the following steps.

[0209] Step B3: DO / DC obtains the evaluation rules for the first data and sends the evaluation rules for the first data.

[0210] Accordingly, SDEF receives the evaluation rules for the first data.

[0211] It is understood that step B3 can be executed before step B2, after step B2, or simultaneously with step B2. This application embodiment does not limit this. Figure 5B takes the example of step B3 being executed after step B2.

[0212] It is understood that the specific description of step B3 can be referred to step A3 above, and will not be repeated here.

[0213] In one possible implementation, if the evaluation result of the first data is used to indicate that the evaluation parameters of the first data do not meet the required conditions, as shown in FIG5A, after executing S402, this application may also perform the following steps.

[0214] Step A4: SDEF updates the first data based on the evaluation results of the first data to obtain the fourth data, wherein the evaluation parameters of the fourth data meet the required conditions.

[0215] Step A5: SDEF sends the fourth data.

[0216] Correspondingly, AF receives the fourth data.

[0217] Alternatively, as shown in Figure 5B, after executing S402, this application may also perform the following steps.

[0218] Step B4: SDEF sends the evaluation results of the first data.

[0219] Accordingly, SDGF receives the evaluation results of the first data.

[0220] Step B5: SDGF updates the first data based on the evaluation results of the first data to obtain the fourth data, wherein the evaluation parameters of the fourth data meet the required conditions.

[0221] Step B6: SDGF sends the fourth data.

[0222] Correspondingly, AF receives the fourth data.

[0223] It is understood that all or some of the evaluation parameters of the first data may not meet the required conditions. When some of the evaluation parameters of the first data do not meet the required conditions, the evaluation result of the first data may only indicate that the relevant evaluation parameters do not meet the required conditions, or the evaluation result of the first data may also indicate that the relevant evaluation parameters do not meet the required conditions, while the other evaluation parameters besides the relevant evaluation parameters meet the required conditions. This application does not limit this aspect.

[0224] For example, the evaluation parameters of the first data include evaluation parameter 1, evaluation parameter 2, and evaluation parameter 3. If evaluation parameter 1 and evaluation parameter 2 do not meet the required conditions, and evaluation parameter 3 meets the required conditions, then the evaluation result of the first data may only indicate that evaluation parameter 1 and evaluation parameter 2 do not meet the required conditions; or, the evaluation result of the first data may indicate that evaluation parameter 1 and evaluation parameter 2 do not meet the required conditions, and evaluation parameter 3 meets the required conditions.

[0225] Figure 6 is a schematic flowchart of a communication method provided in an embodiment of this application. As shown in Figure 6, the communication method includes the following steps.

[0226] S601, The synthetic data evaluation platform acquires the first data, wherein the first data is data synthesized based on the simulation scenario.

[0227] In this embodiment, the synthetic data evaluation platform can be a third-party platform outside the core network used to evaluate synthetic data. The name of the synthetic data evaluation platform is not limited in this embodiment. The synthetic data evaluation platform can provide a synthetic data evaluation application programming interface (API) to functions within the core network used to generate synthetic data (e.g., SDGF).

[0228] In the specific implementation process, as shown in Figure 7, S601 may include the following steps. That is, the synthetic data evaluation platform can obtain the first data through the following steps.

[0229] Step C0: SDGF synthesizes the first data and sends the first data.

[0230] Correspondingly, the synthetic data evaluation platform receives the first data.

[0231] It is understood that SDGF can synthesize the first data through a model. The model can be a statistical model, a machine learning model (such as LLM), or other possible models, which are not limited in this application.

[0232] It is understandable that SDGF can encrypt the first data before sending it.

[0233] It is understandable that SDGF can send the first data to the synthetic data evaluation platform through NEF / GW, and correspondingly, the synthetic data evaluation platform can receive the first data from SDGF through NEF / GW.

[0234] In one possible implementation, as shown in FIG7, the present application may also perform the following steps before performing step C0.

[0235] Step C1: AF sends a synthetic data generation request 1.

[0236] Accordingly, the DO / DC receives Synthetic Data Generation Request 1. Synthetic Data Generation Request 1 is used to request the synthesis of first data.

[0237] It is understood that the specific description of step C1 can be referred to step A1 above, and will not be repeated here.

[0238] Step C2: DO / DC sends a synthetic data generation request 2.

[0239] Accordingly, SDGF receives Synthetic Data Generation Request 2. Synthetic Data Generation Request 2 is used to request the synthesis of first data. Synthetic Data Generation Request 2 includes a second method, which is a method for calling a first API, and the first API is used to evaluate the first data.

[0240] It is understood that the synthetic data generation request 2 can be referred to as the second request, and the name of the synthetic data generation request 2 is not limited in this embodiment. The first API can be referred to as the synthetic data evaluation API, and the name of the first API is not limited in this embodiment.

[0241] It is understood that the specific description of step C2 can be referred to step A2 above, and will not be repeated here.

[0242] S602. The synthetic data evaluation platform evaluates the first data according to the evaluation rules of the first data. The evaluation rules of the first data include the evaluation parameters of the first data and the conditions that the evaluation parameters of the first data need to meet. The evaluation rules of the first data are related to the performance requirements of the first model. The first data is used to train the first model.

[0243] In the embodiments of this application, the specific description of the evaluation rules for the first data can be referred to in S402 above, and will not be repeated here.

[0244] In the specific implementation process, as shown in Figure 7, before executing S402, this application may also perform the following steps, that is, the synthetic data evaluation platform can obtain the evaluation rules of the first data through the following steps.

[0245] Step C3: DO / DC obtains the evaluation rules for the first data and sends the evaluation rules for the first data.

[0246] Accordingly, the synthetic data evaluation platform receives the evaluation rules for the first data.

[0247] It is understandable that the DO / DC can send the evaluation rules of the first data to the synthetic data evaluation platform through the NEF / GW, and correspondingly, the synthetic data evaluation platform can receive the evaluation rules of the first data from the DO / DC through the NEF / GW.

[0248] It is understood that step C3 can be executed before step C2, after step C2, or simultaneously with step C2. This application embodiment does not limit this. Figure 7 shows an example where step C3 is executed after step C2.

[0249] It is understood that the specific description of step C3 can be referred to step A3 above, and will not be repeated here.

[0250] In one possible implementation, if the evaluation result of the first data is used to indicate that the evaluation parameters of the first data do not meet the required conditions, as shown in FIG7, after executing S602, this application may also perform the following steps.

[0251] Step C4: The synthetic data evaluation platform sends the evaluation results of the first data.

[0252] Accordingly, SDGF receives the evaluation results of the first data.

[0253] It is understood that the specific description of step C4 can be referred to step B4 above, and will not be repeated here.

[0254] Step C5: SDGF updates the first data based on the evaluation results of the first data to obtain the fourth data, wherein the evaluation parameters of the fourth data meet the required conditions.

[0255] Step C6: SDGF sends the fourth data.

[0256] Correspondingly, AF receives the fourth data.

[0257] It is understood that the above embodiments of this application can be implemented individually or in combination with each other, and the embodiments of this application are not limited.

[0258] The methods provided by the embodiments of this application have been described above with reference to the accompanying drawings. The apparatus provided by the embodiments of this application will be described below with reference to the accompanying drawings.

[0259] Based on the same technical concept, embodiments of this application provide a communication device, which includes a module / unit / means for executing the method performed by the device in the above-described method embodiments. This module / unit / means can be implemented in software, or in hardware, or implemented by hardware executing corresponding software.

[0260] For example, see Figure 8, which is a schematic diagram of a communication device 800, which includes a transceiver module 801 and a processing module 802.

[0261] When the device 800 is a first communication device (e.g., SDEF), the functions of each module of the device 800 are as follows:

[0262] The processing module 802 is used to acquire first data, which is data synthesized based on a simulation scenario; to evaluate the first data according to the evaluation rules of the first data, the evaluation rules including the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet, the evaluation rules being related to the performance requirements of the first model, and the first data being used to train the first model.

[0263] In one possible implementation, the evaluation rule further includes one or more of the following: a first method, wherein the first method is a method for obtaining the evaluation parameters; the type of the first data; or, the size of the first data.

[0264] In one possible implementation, the processing module 802 is configured to obtain the evaluation parameters according to a first method, wherein the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters are required to satisfy.

[0265] In one possible implementation, the evaluation parameters include one or more of the following: a first parameter, which indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, the real scene being related to the simulation scene, and the third data is sample data used to synthesize the first data; or, a second parameter, which indicates the distribution characteristics of the first data.

[0266] In one possible implementation, the transceiver module 801 is configured to receive a first request, the first request being used to request evaluation of the first data.

[0267] In one possible implementation, the transceiver module 801 is used to receive the evaluation rules.

[0268] In one possible implementation, the processing module 802 is used to synthesize the first data.

[0269] In one possible implementation, the evaluation result of the first data is used to indicate that the evaluation parameter does not meet the required conditions. The processing module 802 is used to update the first data according to the evaluation result to obtain fourth data, wherein the evaluation parameter of the fourth data meets the required conditions.

[0270] In one possible implementation, the transceiver module 801 is used to receive the first data from a second communication device, the second communication device being used to synthesize the first data.

[0271] In one possible implementation, the evaluation result of the first data is used to indicate that the evaluation parameter does not meet the required conditions. After receiving the first data from the second communication device, the transceiver module 801 is used to send the evaluation result to the second communication device, and the evaluation result is used by the second communication device to update the first data.

[0272] Alternatively, when the device 800 is a second communication device (e.g., SDGF), the functions of each module of the device 800 are as follows:

[0273] The transceiver module 801 is used to receive the evaluation result of the first data. The evaluation result of the first data is obtained by evaluating the first data according to the evaluation rules of the first data. The evaluation rules include the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet. The evaluation rules are related to the performance requirements of the first model. The first data is used to train the first model. The first data is data synthesized according to the simulation scenario.

[0274] The processing module 802 is configured to update the first data according to the evaluation result when the evaluation result indicates that the evaluation parameter does not meet the required conditions, thereby obtaining fourth data, wherein the evaluation parameter of the fourth data meets the required conditions.

[0275] In one possible implementation, the evaluation rule is further configured to include one or more of the following: a first method, wherein the first method is a method for obtaining the evaluation parameters; the type of the first data; or, the size of the first data.

[0276] In one possible implementation, the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to satisfy.

[0277] In one possible implementation, the evaluation parameters include one or more of the following: a first parameter, which indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, the real scene being related to the simulation scene, and the third data is sample data used to synthesize the first data; or, a second parameter, which indicates the distribution characteristics of the first data.

[0278] In one possible implementation, a processing module 802 is used to synthesize the first data, and a transceiver module 801 is used to send the first data.

[0279] In one possible implementation, the transceiver module 801 is configured to receive a second request, the second request being a request to synthesize the first data, the second request including a second method, the second method being a method for calling a first API, the first API being used to evaluate the first data.

[0280] In one possible implementation, before sending the first data, the processing module 802 encrypts the first data.

[0281] In one possible implementation, the transceiver module 801 is used to send the evaluation rules.

[0282] Alternatively, when the device 800 is a third communication device (e.g., DO / DC), the functions of each module of the device 800 are as follows:

[0283] The transceiver module 801 is used to send a first request, which is used to request the evaluation of first data, which is data synthesized based on the simulation scenario.

[0284] In one possible implementation, the transceiver module 801 is used to send the evaluation rules of the first data. The evaluation rules include the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet. The evaluation rules are related to the performance requirements of the first model, and the first data is used to train the first model.

[0285] In one possible implementation, the evaluation rule further includes one or more of the following: a first method, wherein the first method is a method for obtaining the evaluation parameters; the type of the first data; or, the size of the first data.

[0286] In one possible implementation, the processing module 802 is configured to obtain the evaluation parameters according to a first method, wherein the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters are required to satisfy.

[0287] In one possible implementation, the evaluation parameters include one or more of the following: a first parameter, which indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, the real scene being related to the simulation scene, and the third data is sample data used to synthesize the first data; or, a second parameter, which indicates the distribution characteristics of the first data.

[0288] In one possible implementation, the transceiver module 801 is configured to send a second request, the second request being used to request the synthesis of the first data, the second request including a second method, the second method being a method for calling a first API, the first API being used to evaluate the first data.

[0289] In practical implementation, the above-mentioned device 800 can have various product forms. Several possible product forms are introduced below.

[0290] Referring to Figure 9, which is a schematic diagram of another communication device, the communication device 900 includes a processor 901 and an interface circuit 902. The interface circuit 902 is used to receive signals from other communication devices outside the communication device and transmit them to the processor 901, or to send signals from the processor 901 to other communication devices outside the communication device. The processor 901 is used to implement the methods performed by the first communication device, the second communication device, and the third communication device in the above method embodiments through logic circuits or execution instructions.

[0291] The processor 901 and the interface circuit 902 are coupled to each other. It is understood that the interface circuit 902 can be a transceiver or an input / output interface. Optionally, the communication device 900 may also include a memory 903 for storing instructions executed by the processor 901, or storing input data required by the processor 901 to execute instructions, or storing data generated after the processor 901 executes instructions.

[0292] It should be understood that the processor mentioned in the embodiments of this application can be implemented in hardware or software. When implemented in hardware, the processor can be a logic circuit, integrated circuit, etc. When implemented in software, the processor can be a general-purpose processor, implemented by reading software code stored in memory.

[0293] For example, 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), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0294] It should be understood that the memory mentioned in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0295] It should be noted that when the processor is a general-purpose processor, DSP, ASIC, FPGA, or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, the memory (storage module) can be integrated into the processor.

[0296] It should be noted that the memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0297] Based on the same technical concept, embodiments of this application also provide a computer-readable storage medium storing a computer program or instructions, which, when executed by a processor, causes the methods performed by the first communication device, the second communication device, and the third communication device in the above method embodiments to be implemented.

[0298] Based on the same technical concept, this application also provides a computer program product, which includes a computer program or instructions that, when executed by a processor, cause the methods performed by the first communication device, the second communication device, and the third communication device in the above method embodiments to be implemented.

[0299] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0300] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0301] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0302] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

Claims

1. A communication method, characterized in that, Applied to a first communication device, the method includes: Acquire first data, which is data synthesized based on the simulation scenario; The first data is evaluated according to the evaluation rules of the first data. The evaluation rules include the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet. The evaluation rules are related to the performance requirements of the first model. The first data is used to train the first model.

2. The method according to claim 1, characterized in that, The evaluation rules also include one or more of the following: A first method, wherein the first method is a method for obtaining the evaluation parameters; The type of the first data; or, The size of the first data.

3. The method according to claim 1 or 2, characterized in that, The method further includes: The evaluation parameters are obtained according to a first method, wherein the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to satisfy.

4. The method according to any one of claims 1-3, characterized in that, The evaluation parameters include one or more of the following: The first parameter indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, which is related to the simulation scene, and the third data is sample data used to synthesize the first data; or, The second parameter is used to indicate the distribution characteristics of the first data.

5. The method according to any one of claims 1-4, characterized in that, The method further includes: Receive a first request, which is used to request the evaluation of the first data.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Receive the evaluation rules.

7. The method according to any one of claims 1-6, characterized in that, The acquisition of the first data includes: Synthesize the first data.

8. The method according to claim 7, characterized in that, The evaluation result of the first data is used to indicate that the evaluation parameter does not meet the required condition, and the method further includes: The first data is updated based on the evaluation results to obtain the fourth data, wherein the evaluation parameters of the fourth data meet the required conditions.

9. The method according to any one of claims 1-6, characterized in that, The acquisition of the first data includes: The first data is received from a second communication device, which is used to synthesize the first data.

10. The method according to claim 9, characterized in that, The evaluation result of the first data is used to indicate that the evaluation parameter does not meet the required condition. After receiving the first data from the second communication device, the method further includes: The evaluation result is sent to the second communication device, and the evaluation result is used by the second communication device to update the first data.

11. A communication method, characterized in that, Applied to a second communication device, the method includes: The evaluation result of the first data is received. The evaluation result of the first data is obtained by evaluating the first data according to the evaluation rules of the first data. The evaluation rules include the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet. The evaluation rules are related to the performance requirements of the first model. The first data is used to train the first model. The first data is data synthesized based on the simulation scenario. If the evaluation result indicates that the evaluation parameter does not meet the required conditions, the first data is updated according to the evaluation result to obtain fourth data, wherein the evaluation parameter of the fourth data meets the required conditions.

12. The method according to claim 11, characterized in that, The evaluation rules also include one or more of the following: A first method, wherein the first method is a method for obtaining the evaluation parameters; The type of the first data; or, The size of the first data.

13. The method according to claim 12, characterized in that, The first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to meet.

14. The method according to any one of claims 11-13, characterized in that, The evaluation parameters include one or more of the following: The first parameter indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, which is related to the simulation scene, and the third data is sample data used to synthesize the first data; or, The second parameter is used to indicate the distribution characteristics of the first data.

15. The method according to any one of claims 11-14, characterized in that, The method further includes: The first data is synthesized and then sent.

16. The method according to claim 15, characterized in that, The method further includes: A second request is received, the second request being for requesting the synthesis of the first data, the second request including a second method, the second method being for calling a first application programming interface (API), the first API being for evaluating the first data.

17. The method according to claim 16, characterized in that, Before sending the first data, the method further includes: The first data is encrypted.

18. The method according to claim 16 or 17, characterized in that, The method further includes: Send the evaluation rules.

19. A communication method, characterized in that, Applied to a third communication device, the method includes: Send a first request, the first request being used to request the evaluation of first data, the first data being data synthesized based on the simulation scenario.

20. The method according to claim 19, characterized in that, The method further includes: The evaluation rules for the first data are sent. The evaluation rules include the evaluation parameters of the first data and the conditions that the evaluation parameters need to meet. The evaluation rules are related to the performance requirements of the first model. The first data is used to train the first model.

21. The method according to claim 20, characterized in that, The evaluation rules also include one or more of the following: A first method, wherein the first method is a method for obtaining the evaluation parameters; The type of the first data; or, The size of the first data.

22. The method according to claim 20 or 21, characterized in that, The method further includes: The evaluation parameters are obtained according to a first method, wherein the first method is determined based on the type of the first data and / or the conditions that the evaluation parameters need to satisfy.

23. The method according to any one of claims 20-22, characterized in that, The evaluation parameters include one or more of the following: The first parameter indicates the degree of difference between the first data and the second and / or third data, wherein the second data is real data collected in a real scene, which is related to the simulation scene, and the third data is sample data used to synthesize the first data; or, The second parameter is used to indicate the distribution characteristics of the first data.

24. The method according to any one of claims 19-23, characterized in that, The method further includes: A second request is sent, which requests the synthesis of the first data. The second request includes a second method, which is a method for calling a first API, which is used to evaluate the first data.

25. A communication device, characterized in that, The communication device includes a module for performing the method as described in any one of claims 1-10, or a module for performing the method as described in any one of claims 11-18, or a module for performing the method as described in any one of claims 19-24.

26. A communication device, characterized in that, The communication device includes a processor configured to perform the method as described in any one of claims 1-10, or the method as described in any one of claims 11-18, or the method as described in any one of claims 19-24.

27. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program that, when run on a computer, causes the method as described in any one of claims 1-10 to be performed, or causes the method as described in any one of claims 11-18 to be performed, or causes the method as described in any one of claims 19-24 to be performed.

28. A computer program product, characterized in that, The computer program product includes a computer program that, when run on a computer, causes the method as described in any one of claims 1-10 to be performed, or causes the method as described in any one of claims 11-18 to be performed, or causes the method as described in any one of claims 19-24 to be performed.

Images