Communication method and apparatus, related device, and storage medium

Abstract

The application discloses a communication method, device, network equipment, cloud platform and storage medium. The method comprises the following steps: a first application running on a network equipment receives a first model sent by a cloud platform; the first model is obtained by training sample data by the cloud platform; the network equipment is used for providing network access function at least; when the first application determines that data reasoning is needed, the first model is used for data reasoning, or a local preset reasoning strategy is used for data reasoning, and a reasoning result is obtained.

Description

Technical Field

[0001] This application relates to the field of wireless communication, and more particularly to a communication method, apparatus, related equipment, and storage medium. Background Technology

[0002] Operators can provide leased network services to industry users such as factories, industrial parks, and energy companies to meet their specific business needs, and deploy network equipment for these users to access these leased networks. On the one hand, the computing and storage resources of these access layer network devices are relatively idle, and a large amount of network resources are not being fully utilized (e.g., network resource utilization is below 50%). On the other hand, industry users also have a need to improve production efficiency. Therefore, how to improve the production efficiency of industry users through access layer network equipment has become an urgent problem to be solved. Summary of the Invention

[0003] To address the related technical problems, embodiments of this application provide a communication method, apparatus, related devices, and storage medium.

[0004] The technical solution of this application embodiment is implemented as follows:

[0005] This application provides a communication method applied to a network device, including:

[0006] A first application running on the network device receives a first model sent by a cloud platform; the first model is trained by the cloud platform using sample data; the network device is at least used to provide network access functionality.

[0007] When the first application determines that data reasoning is required, it performs data reasoning using the first model or by using a locally preset reasoning strategy to obtain the reasoning result.

[0008] In the above scheme, the first model sent by the cloud platform includes:

[0009] The first application sends a first request to the cloud platform through a first interface; the first request is used to request the cloud platform to distribute the first model;

[0010] The first application receives the first model sent by the cloud platform based on the first request through the first interface.

[0011] The above scheme, the method further includes:

[0012] The first application acquires first sample data and sends the first sample data to the cloud platform through a second interface, so that the cloud platform can at least use the first sample data to train the first model.

[0013] The above scheme, the method further includes:

[0014] The first application receives the updated first model sent by the cloud platform;

[0015] The first application uses the updated first model to perform data reasoning and obtain the reasoning result.

[0016] In the above scheme, receiving the updated first model sent by the cloud platform includes:

[0017] The first application receives first information sent by the cloud platform through a third interface; the first information indicates that the cloud platform has an updated first model.

[0018] The first application sends a second request to the cloud platform through the third interface; the second request is used to request the cloud platform to issue the updated first model;

[0019] The first application receives the updated first model sent by the cloud platform based on the second request through the third interface.

[0020] The above scheme, the method further includes:

[0021] The first application acquires the second sample data and sends the second sample data to the cloud platform through the second interface, so that the cloud platform can update the first model using at least the second sample data; the second sample data includes at least the reasoning result obtained by the first application using the first model for data reasoning.

[0022] The above scheme, the method further includes:

[0023] The first application obtains the first configuration information and uses the first configuration information to update the first model to obtain the updated first model;

[0024] The first application uses the updated first model to perform data reasoning and obtain the reasoning result.

[0025] The above scheme, the method further includes:

[0026] The first application sends second information to the cloud platform through the fourth interface; the second information at least represents the running status of the first application.

[0027] The first application receives management instruction information sent by the cloud platform based on the second information through the fourth interface;

[0028] The first application manages its own lifecycle according to the management instruction information.

[0029] The above scheme, the method further includes:

[0030] Obtain a third request; the third request is used to request the first service;

[0031] The third request is sent to the cloud platform via the fifth interface;

[0032] The system receives the second configuration information sent by the cloud platform based on the third request through the fifth interface, and uses the second configuration information to run the first application locally.

[0033] This application also provides a communication method applied to a cloud platform, including:

[0034] Train the first model using the sample data;

[0035] The first model is sent to a first application running on a network device; the network device is at least used to provide network access functionality; the first model is used for the first application to perform data inference.

[0036] In the above scheme, sending the first model to the first application running on the network device includes:

[0037] The system receives a first request sent by the first application through a first interface; the first request is used to request the cloud platform to issue the first model.

[0038] Based on the first request, the first model is sent to the first application through the first interface.

[0039] The above scheme, the method further includes:

[0040] Receive the first sample data sent by the first application through the second interface;

[0041] The first model is trained using at least the first sample data.

[0042] The above scheme, the method further includes:

[0043] The first model is updated using the new sample data to obtain the updated first model;

[0044] Send the updated first model to the first application.

[0045] In the above scheme, sending the updated first model to the first application includes:

[0046] The first information is sent to the first application via a third interface; the first information indicates that the cloud platform has an updated first model.

[0047] The third interface receives a second request sent by the first application; the second request is used to request the cloud platform to issue the updated first model.

[0048] Based on the second request, the updated first model is sent to the first application through the third interface.

[0049] The above scheme, the method further includes:

[0050] The second sample data sent by the first application is received through the second interface; the second sample data at least includes the reasoning result obtained by the first application using the first model for data reasoning.

[0051] The first model is updated using at least the second number of samples.

[0052] The above scheme, the method further includes:

[0053] The second information sent by the first application is received through the fourth interface; the second information at least represents the running status of the first application.

[0054] Based on the second information, management instruction information is determined and sent to the first application through the fourth interface; the management instruction information is used by the first application to manage its own lifecycle.

[0055] The above scheme, the method further includes:

[0056] The network device sends a third request via the fifth interface; the third request is used to request the first service.

[0057] Based on the third request, second configuration information is determined and sent to the network device through the fifth interface; the second configuration information is used for the network device to run the first application locally.

[0058] This application also provides a communication device, disposed on a network device, comprising: a first receiving unit and a first processing unit; the network device is at least used to provide network access functionality; wherein...

[0059] The first receiving unit is used by a first application running on the network device to receive a first model sent by the cloud platform; the first model is trained by the cloud platform using sample data.

[0060] The first processing unit is used to perform data reasoning using the first model or using a locally preset reasoning strategy when it is determined that data reasoning needs to be performed after being used by the first application, so as to obtain the reasoning result.

[0061] This application also provides a communication device, configured on a cloud platform, comprising:

[0062] The second processing unit is used to train the first model using sample data;

[0063] A first sending unit is configured to send the first model to a first application running on a network device; the network device is configured to provide network access functionality at least; the first model is configured to be used by the first application for data inference.

[0064] This application also provides a network device, including: a first communication interface and a first processor; wherein,

[0065] The first processor is configured to:

[0066] The first application is run to receive a first model sent by the cloud platform; the first model is trained by the cloud platform using sample data; the network device is at least used to provide network access functionality.

[0067] Run the first application, and when it is determined that data inference is required, perform data inference using the first model or by using a locally preset inference strategy to obtain the inference result.

[0068] This application also provides a cloud platform, including: a second communication interface and a second processor; wherein,

[0069] The second processor is used for:

[0070] Train the first model using the sample data;

[0071] The first model is sent to a first application running on a network device; the network device is at least used to provide network access functionality; the first model is used for the first application to perform data inference.

[0072] This application also provides a network device, including: a first processor and a first memory for storing a computer program capable of running on the processor.

[0073] Wherein, when the first processor is used to run the computer program, it executes the steps of any of the methods described above on the network device side.

[0074] This application also provides a cloud platform, including: a second processor and a second memory for storing computer programs capable of running on the processor.

[0075] Wherein, when the second processor is running the computer program, it executes the steps of any of the methods described above on the cloud platform side.

[0076] This application embodiment also provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of any of the methods described above on the network device side, or implements the steps of any of the methods described above on the cloud platform side.

[0077] The communication method, apparatus, related devices, and storage medium provided in this application embodiment involve a cloud platform training a first model using sample data; sending the first model to a first application running on a network device; the network device at least providing network access functionality; the first application receiving the first model sent by the cloud platform; and when the first application determines that data inference is needed, performing data inference using the first model or using a locally preset inference strategy to obtain an inference result. In this application embodiment, the first application running on the network device (also known as a network transmission device, network access device, etc.) performs data inference using the first model or a locally preset inference strategy, thus integrating lightweight edge computing services on the network device. This fully utilizes idle network resources and improves network resource utilization. Furthermore, it can leverage the edge computing services provided by the network device to enhance the information processing capabilities of industry users, thereby improving their productivity. Simultaneously, the cloud platform trains the model and sends the trained model to the network device for data inference, thus enabling collaboration between the access layer network and cloud computing. Attached Figure Description

[0078] Figure 1 This is a flowchart illustrating a communication method according to an embodiment of this application;

[0079] Figure 2 This is a flowchart illustrating another communication method according to an embodiment of this application;

[0080] Figure 3 This is a schematic diagram of a business scenario in an application embodiment of this application;

[0081] Figure 4 This is a schematic diagram of the interface between the network device and the cloud platform in an application embodiment of this application;

[0082] Figure 5 This is a schematic diagram illustrating the interaction process between the network device and the cloud platform in an application embodiment of this application.

[0083] Figure 6 This is a schematic diagram of the structure of a communication device according to an embodiment of this application;

[0084] Figure 7 This is a schematic diagram of the structure of another communication device according to an embodiment of this application;

[0085] Figure 8This is a schematic diagram of the network device according to an embodiment of this application;

[0086] Figure 9 This is a schematic diagram of the cloud platform structure according to an embodiment of this application;

[0087] Figure 10 This is a schematic diagram of the communication system according to an embodiment of this application. Detailed Implementation

[0088] The present application will now be described in further detail with reference to the accompanying drawings and embodiments.

[0089] This application provides a communication method applied to network devices, such as... Figure 1 As shown, the method includes:

[0090] Step 101: The first application running on the network device receives the first model sent by the cloud platform;

[0091] Here, the first model is trained by the cloud platform using sample data; the network device is at least used to provide network access functionality.

[0092] Step 102: When the first application determines that data reasoning is required, it performs data reasoning using the first model or by using a locally preset reasoning strategy to obtain the reasoning result.

[0093] In practical applications, the network device is used to provide network access functionality. This can be understood as enabling users in industries such as factories, industrial parks, and energy companies to access leased networks provided by operators. The network device can also be called a network transmission device, network access device, access layer device, transmission access device, transmission device, transmission network access device, etc. This application embodiment does not limit the name of the network device, as long as it can provide network access functionality.

[0094] In practical applications, the cloud platform may also be called a public cloud platform, an edge computing management platform, etc. The name of the cloud platform is not limited in this application embodiment.

[0095] In practical applications, the data reasoning can be understood as the data processing and analysis by industry users in various business needs scenarios, such as identifying whether the products produced on the production line are qualified, analyzing whether the production environment is qualified, and analyzing the status of production machines.

[0096] In practical applications, when the first application determines that data inference is needed, it uses the first model or a locally preset inference strategy to perform data inference. Therefore, the network device can provide edge computing services based on data inference. It can be understood that the network device integrates an edge computing platform (also called an edge computing module) to support edge computing services, and the first application specifically runs on the edge computing platform.

[0097] Here, the edge computing platform is a lightweight edge computing platform. "Lightweight" means that the edge computing platform integrated on the network device has lower requirements for computing and storage resources than traditional edge computing platforms. Traditional edge computing platforms refer to edge computing platforms installed (i.e., deployed) on large servers or server clusters. Specifically, due to the significant differences in business needs among various industry users, the lack of information infrastructure, and limited campus space, they typically rely on manual experience to improve production line efficiency (i.e., increase production efficiency), or improve campus production efficiency by deploying traditional microservice-based edge computing platforms outside the campus, closest to the users. "Closest to the users" means that the edge computing network where the edge computing platform is located is the last network segment connecting the user. Compared to deploying traditional edge computing platforms outside the campus, the solution in this application, by integrating a lightweight edge computing platform on the network device, can fully utilize the idle resources on the network device and improve the production efficiency of industry users; at the same time, it eliminates the need to deploy additional equipment (such as traditional edge computing platforms) within the campus, ensuring the data security of industry users.

[0098] In step 101, in practical applications, the cloud platform can proactively send the first model to the first application after the first model has been trained; or, the first application can request the cloud platform to distribute the first model, and the cloud platform will distribute the first model after receiving the request from the first application.

[0099] Based on this, in one embodiment, the first model sent by the cloud platform may include:

[0100] The first application sends a first request to the cloud platform through a first interface; the first request is used to request the cloud platform to distribute the first model;

[0101] The first application receives the first model sent by the cloud platform based on the first request through the first interface.

[0102] In step 102, in practical applications, considering the different business needs of industry users, when the first application determines that data inference is required, it can first determine the data inference method, and then perform data inference using the determined data inference method. That is, it can perform data inference using the first model, or it can perform data inference using a locally preset inference strategy. Here, the first application can determine the data inference method based on the configuration file stored locally on the network device, or it can receive instruction information containing the data inference method sent by other electronic devices and determine the data inference method based on the instruction information. Industry users can determine the data inference method according to their specific needs, and this application embodiment does not limit this.

[0103] In practical applications, for business scenarios with high real-time requirements (such as requiring latency of less than or equal to 20 milliseconds), the first application can utilize a locally preset inference strategy for data inference. For example, in a business scenario involving controlling the grabbing of containers on a production line and determining whether the containers are qualified, the first application can use a first condition representing the container's qualification and / or a second condition representing the container's non-qualification to determine whether the container is qualified. That is, the locally preset inference strategy is: using the first condition and / or the second condition to determine whether the container is qualified.

[0104] In practical applications, for business scenarios where real-time requirements are low and industry users allow the uploading of park data (such as production line data, inference results of the first application, etc.) to the cloud platform, the first application can use the first model for data inference. Furthermore, before using the first model for data inference, the first application can collect park data as sample data for model training and send the collected park data to the cloud platform for the cloud platform to train the first model.

[0105] Based on this, in one embodiment, the method may further include:

[0106] The first application acquires first sample data and sends the first sample data to the cloud platform through a second interface, so that the cloud platform can at least use the first sample data to train the first model.

[0107] In practical applications, industry user parks can set up multiple network devices, and each network device can send the park data it collects to the cloud platform. In other words, the cloud platform can use the first sample data sent by at least one network device to train the first model.

[0108] In practical applications, after the cloud platform completes the training of the first model, it can send a notification message to the first application through the first interface to indicate that the training of the first model is complete. After receiving the notification message, the first application can send the first request to the cloud platform.

[0109] In practical applications, to conserve network resources, when the first application sends the first sample data to the cloud platform through the second interface, it can first send a first message to the cloud platform through the second interface. This first message notifies the cloud platform that the first application needs to send the first sample data. After receiving the first message, the cloud platform can send a second message to the first application through the second interface. This second message indicates that the cloud platform can receive the first sample data. After receiving the second message, the first application can begin sending the first sample data to the cloud platform through the second interface. After the first sample data is sent, the first application can send a third message to the cloud platform through a sixth interface. This third message notifies the cloud platform that the first sample data has been sent. After receiving the third message, the cloud platform can determine whether the first sample data has been received and then send a fourth message to the first application through the sixth interface. This fourth message indicates to the first application whether the cloud platform has received the first sample data.

[0110] In practical applications, for business scenarios with low real-time requirements and where industry users are not allowed to upload park data to the cloud platform, the first application can use the first model for data inference. In this case, the sample data required to train the first model can be collected by the cloud platform through public channels (such as the Internet), meaning the first model is a general-purpose model. To make the first model applicable to the business scenario corresponding to the first application, the first application can perform local parameter tuning on the first model before using it for data inference. That is, it can adjust the model parameters of the first model locally using locally preset configuration information or configuration information sent by other electronic devices to obtain an updated first model; then, it can use the updated first model for data inference.

[0111] Based on this, in one embodiment, the method may further include:

[0112] The first application obtains the first configuration information and uses the first configuration information to update the first model to obtain the updated first model;

[0113] Accordingly, the data reasoning using the first model may include:

[0114] The first application uses the updated first model to perform data reasoning and obtain the reasoning result.

[0115] In practical applications, in order to improve the performance of the first model, the first application can periodically optimize the local first model using the inference results, or the cloud platform can periodically update the first model and then distribute the updated first model to the first application.

[0116] Based on this, in one embodiment, the method may further include:

[0117] The first application receives the updated first model sent by the cloud platform;

[0118] Accordingly, the data reasoning using the first model may include:

[0119] The first application uses the updated first model to perform data reasoning and obtain the reasoning result.

[0120] In practical applications, after updating the first model, the cloud platform can proactively send the updated first model to the first application; or, the cloud platform can send an indication message to the first application indicating that the cloud platform has an updated first model. After receiving the indication message, the first application can request the cloud platform to send the updated first model. After receiving the request from the first application, the cloud platform will then send the updated first model.

[0121] Based on this, in one embodiment, receiving the updated first model sent by the cloud platform may include:

[0122] The first application receives first information sent by the cloud platform through a third interface; the first information indicates that the cloud platform has an updated first model.

[0123] The first application sends a second request to the cloud platform through the third interface; the second request is used to request the cloud platform to issue the updated first model;

[0124] The first application receives the updated first model sent by the cloud platform based on the second request through the third interface.

[0125] In practical applications, the cloud platform can update the first model using new sample data. Specifically, for business scenarios where industry users are not allowed to upload park data to the cloud platform, the new sample data can include public data collected by the cloud platform; for business scenarios where industry users are allowed to upload park data to the cloud platform, the new sample data can include park data collected by the first application.

[0126] Based on this, in one embodiment, the method may further include:

[0127] The first application acquires the second sample data and sends the second sample data to the cloud platform through the second interface, so that the cloud platform can update the first model using at least the second sample data; the second sample data includes at least the reasoning result obtained by the first application using the first model for data reasoning.

[0128] In practical applications, the cloud platform can update the first model using second sample data sent by at least one network device and public data it collects itself.

[0129] In practical applications, to conserve network resources, when the first application sends the second sample data to the cloud platform through the second interface, it can first send a fifth message to the cloud platform through the second interface. This fifth message notifies the cloud platform that the first application needs to send the second sample data. After receiving the fifth message, the cloud platform can send a sixth message to the first application through the second interface. This sixth message indicates that the cloud platform can receive the second sample data. After receiving the sixth message, the first application can begin sending the second sample data to the cloud platform through the second interface. After the second sample data is sent, the first application can send a seventh message to the cloud platform through the sixth interface. This seventh message notifies the cloud platform that the second sample data has been sent. After receiving the seventh message, the cloud platform can determine whether the second sample data has been received and then send an eighth message to the first application through the sixth interface. This eighth message indicates to the first application whether the cloud platform has received the second sample data.

[0130] In practical applications, the cloud platform can orchestrate the services provided by the network devices to industry users. For example, it can determine the service chain corresponding to the service and send the configuration information (such as the service chain) corresponding to the service to the network devices so that the network devices can run the corresponding application (such as the first application) and provide the corresponding service through the application.

[0131] Based on this, in one embodiment, the method may further include:

[0132] Obtain a third request; the third request is used to request the first service;

[0133] The third request is sent to the cloud platform via the fifth interface;

[0134] The system receives the second configuration information sent by the cloud platform based on the third request through the fifth interface, and uses the second configuration information to run the first application locally.

[0135] Here, it can be understood that the first application is used to provide the first service.

[0136] In practical applications, the third request may include the service requirements corresponding to the first service, so that the cloud platform can determine the second configuration information based on the service requirements corresponding to the first service.

[0137] In practical applications, the network device can download the first application from the cloud platform through the seventh interface.

[0138] In practical applications, the first application can periodically send its own operating status to the cloud platform, such as whether there is a fault; the cloud platform can instruct the first application to manage its own lifecycle based on the first application's operating status, such as updating or uninstalling.

[0139] Based on this, in one embodiment, the method may further include:

[0140] The first application sends second information to the cloud platform through the fourth interface; the second information at least represents the running status of the first application.

[0141] The first application receives management instruction information sent by the cloud platform based on the second information through the fourth interface;

[0142] The first application manages its own lifecycle according to the management instruction information.

[0143] Correspondingly, embodiments of this application also provide a communication method applied to a cloud platform, such as... Figure 2 As shown, the method includes:

[0144] Step 201: Train the first model using the sample data;

[0145] Step 202: Send the first model to the first application running on the network device;

[0146] Here, the network device is at least used to provide network access functionality; the first model is used for the first application to perform data inference.

[0147] In practical applications, the first model is used for data inference by the first application, which can be understood as the network device also having edge computing service capabilities.

[0148] In one embodiment, sending the first model to the first application running on the network device may include:

[0149] The system receives a first request sent by the first application through a first interface; the first request is used to request the cloud platform to issue the first model.

[0150] Based on the first request, the first model is sent to the first application through the first interface.

[0151] In practical applications, the first sample data may include park data of industry users collected by the network device, and may also include public data collected by the cloud platform.

[0152] Based on this, in one embodiment, the method may further include:

[0153] Receive the first sample data sent by the first application through the second interface;

[0154] Accordingly, training the first model using sample data may include:

[0155] The first model is trained using at least the first sample data.

[0156] In practical applications, the cloud platform can periodically update the first model to improve its performance.

[0157] Based on this, in one embodiment, the method may further include:

[0158] The first model is updated using the new sample data to obtain the updated first model;

[0159] Send the updated first model to the first application.

[0160] In one embodiment, sending the updated first model to the first application may include:

[0161] The first information is sent to the first application via a third interface; the first information indicates that the cloud platform has an updated first model.

[0162] The third interface receives a second request sent by the first application; the second request is used to request the cloud platform to issue the updated first model.

[0163] Based on the second request, the updated first model is sent to the first application through the third interface.

[0164] In practical applications, the new sample data may include park data of industry users collected by the network device, and may also include public data collected by the cloud platform.

[0165] Based on this, in one embodiment, the method may further include:

[0166] The second sample data sent by the first application is received through the second interface; the second sample data at least includes the reasoning result obtained by the first application using the first model for data reasoning.

[0167] Accordingly, updating the first model using new sample data may include:

[0168] The first model is updated using at least the second number of samples.

[0169] In one embodiment, the method may further include:

[0170] The second information sent by the first application is received through the fourth interface; the second information at least represents the running status of the first application.

[0171] Based on the second information, management instruction information is determined and sent to the first application through the fourth interface; the management instruction information is used by the first application to manage its own lifecycle.

[0172] In one embodiment, the method may further include:

[0173] The network device sends a third request via the fifth interface; the third request is used to request the first service.

[0174] Based on the third request, second configuration information is determined and sent to the network device through the fifth interface; the second configuration information is used for the network device to run the first application locally.

[0175] The communication method provided in this application embodiment involves a cloud platform training a first model using sample data; sending the first model to a first application running on a network device; the network device at least providing network access functionality; the first application receiving the first model sent by the cloud platform; and when the first application determines that data inference is needed, performing data inference using the first model or using a locally preset inference strategy to obtain an inference result. In this application embodiment, the first application running on the network device (also known as a network transmission device, network access device, etc.) performs data inference using the first model or a locally preset inference strategy, thus integrating lightweight edge computing services into the network device. This fully utilizes idle network resources and improves network resource utilization. Furthermore, it can leverage the edge computing services provided by the network device to enhance the information processing capabilities of industry users, thereby improving their productivity. Simultaneously, the cloud platform trains the model and sends the trained model to the network device for data inference, thus enabling collaboration between the access layer network and cloud computing.

[0176] The present application will be further described in detail below with reference to application examples.

[0177] In this application embodiment, a lightweight edge computing platform is integrated into the network devices at the access layer. This fully utilizes idle network resources within industry user parks and connects factories in multiple cities via the operator's transmission network to achieve large-scale network coverage. The lightweight edge computing platform uses a new interface to connect with the cloud computing platform (i.e., the cloud platform), enabling collaboration between cloud computing and network devices and significantly improving the information processing capabilities of park users. Specifically, model development and training are performed on the public cloud platform, while model inference is conducted on the network devices. The system supports incremental learning, iteration, publishing, and push functions for the model, thereby achieving a complete closed loop for optimal model performance. This allows the network devices to implement AI services such as industrial image recognition based on models trained on the cloud platform.

[0178] In practical applications, the network device can have Slicing Packet Network (SPN) capabilities and support data transmission with microsecond-level latency, thereby providing a good network foundation for data transmission between the network device and the cloud platform.

[0179] In this application example, the lightweight edge computing platform integrated into the network device supports the following three service scenarios:

[0180] 1) The network device collects sample data from the park and sends it to the cloud platform. The cloud platform trains the model and sends the trained model back to the network device for data inference.

[0181] 2) The network device collects public data to train the model, obtains a general model, and sends the general model to the network device; the network device can optimize the general model by local parameter tuning, that is, update the general model, and use the updated model for data inference;

[0182] 3) For services with high real-time requirements, network devices can use locally preset inference strategies to perform data inference.

[0183] For example, Figure 3 This refers to a business scenario involving cloud-network collaboration (i.e., collaboration between cloud platforms and network devices). For example... Figure 3 As shown, a lightweight edge computing platform is integrated into the network device. An application (APP) running on this platform provides image recognition services to the park, specifically determining whether product images captured by production line equipment are qualified (or good) or unqualified (or defective). The APP can collect and analyze data from the production line equipment (e.g., filtering data that does not meet preset conditions) and send the collected data to a public cloud AI platform (i.e., a cloud platform). The cloud platform can receive data collected from multiple network devices, using this data as sample data for centralized model training to obtain a trained model (e.g., the first model mentioned above). The APP running on the edge computing platform can retrieve the trained model from the cloud platform via an application programming interface (API) for data inference (i.e., image recognition) and push the recognition results to the park for filtering defective products. Furthermore, the APP can send recognition results and related inference data within a preset time range to the cloud platform. The cloud platform can analyze the entire data processing flow of the APP, determine if there are any malfunctions, and optimize the model using the APP's recognition results.

[0184] In this application example, such as Figure 4 As shown, the network devices and the cloud platform communicate through the following interfaces:

[0185] 1) Data collaboration interface, used to realize data transmission between network devices and cloud platform (such as sample data of the park collected by the network devices), may include StartTraffic interface (i.e. the second interface mentioned above), StopTraffic interface (i.e. the sixth interface mentioned above), etc.

[0186] 2) Intelligent collaboration interface, used to realize model transmission, may include DownloadModel interface for downloading models (i.e. the first interface mentioned above), UpdateModel interface for updating models (i.e. the third interface mentioned above), etc.; that is, when the APP on the edge computing platform needs to perform data inference, it downloads the corresponding model from the cloud platform through DownloadModel interface, or updates the corresponding model through UpdateModel interface.

[0187] 3) APP management collaboration interface, used to coordinate the lifecycle of the APP, which may include the APPManage interface (i.e. the fourth interface mentioned above);

[0188] 4) Business collaboration interface (i.e., the fifth interface mentioned above) is used to manage business instances, i.e., manage the APP.

[0189] In this application embodiment, based on Figure 4 The network device can specifically support the following services through the interface shown:

[0190] 1) The edge computing platform integrated by the network device provides modular, microservice-based APPs (such as the first application mentioned above). The cloud platform can provide on-demand business orchestration capabilities, that is, determine the configuration information (i.e., the second configuration information mentioned above) required to deploy the corresponding business instance (i.e. run the corresponding APP) according to the needs of industry users, and send the determined configuration information to the network device so that the network device can run the corresponding APP according to the configuration information.

[0191] 2) Industry users can log in to the cloud platform and deploy their self-developed APP to the edge computing platform of the network device through the APPManage interface of the cloud platform management interface; or, industry users can pass API interface parameters to the edge computing platform. The API interface parameters are used to realize cloud inference. In other words, when local inference of the network device is not required, the network device can realize cloud inference by calling the API interface of the cloud platform.

[0192] 3) The APP can call the data collaboration interface (i.e., StartTraffic, StopTraffic, etc.) to drive the control module of the network device to transmit data to the cloud platform, so that the cloud platform can carry out centralized model training of artificial intelligence (AI) models (such as the first model mentioned above); at the same time, the APP can call the DownloadModel interface to drive the control module to download the model trained by the cloud platform to the lightweight edge computing platform of the network device.

[0193] 4) The APP can collect production line data, perform preliminary processing (such as filtering) on ​​the data according to local preset strategies and feature libraries, and use models trained on the cloud platform for data inference.

[0194] 5) The APP can periodically send sample data to the cloud platform. The cloud platform continuously receives sample data, iteratively trains a more optimized model, and sends the optimized model (i.e. the updated model) to the APP through the UpdateModel interface so that the APP can use the updated model for data inference.

[0195] 6) For businesses with high real-time requirements, the APP can perform data inference directly using local preset inference strategies without needing models trained on the cloud platform.

[0196] In this application example, such as Figure 5 As shown, the interaction process between network devices and the cloud platform may include the following steps:

[0197] Step 501: The network device integrates a lightweight edge computing module; the edge computing module is used to provide edge computing capabilities;

[0198] Step 502: Industry users log in to the cloud platform and deploy the developed APP to the edge computing module;

[0199] Step 503: The APP transmits sample data to the cloud platform by calling the data collaboration interface of the edge computing platform (i.e., StartTraffic, StopTraffic, etc.) for the cloud platform to train the model; and downloads the trained model from the cloud platform by calling the DownloadModel interface.

[0200] Step 504: The APP drives the edge computing platform to collect production line data and uses the model trained on the cloud platform to perform data inference.

[0201] Step 505: The cloud platform continuously receives sample data (e.g., network devices periodically send sample data to the cloud platform), uses the received sample data to iteratively train a more optimized model, and distributes it to the network devices so that the APP can perform inference based on the updated model.

[0202] The solution provided in this application embodiment has the following advantages:

[0203] 1) It solves the problem of extending public cloud capabilities to the park. Network devices in the park can act as proxies for the public cloud, interact with the cloud platform through data collaboration interfaces, and extend public cloud capabilities to the park by relying on the low-latency SPN network of the network devices themselves.

[0204] 2) It solves the problem of preventing user data from leaving the factory (i.e., whether users allow users in the park to upload to the cloud platform). Network devices can download the model trained by the cloud platform through the intelligent collaboration interface. This model can be a general model or a model trained using user data in the park. Network devices can use this model locally to perform business-related data inference in real time.

[0205] 3) It solves the problem of large-scale networking of edge computing nodes, that is, it realizes networking between edge computing platforms through networking between network devices; the edge computing nodes after networking can leverage N squared advantages, such as applying the fault characteristics of plant A to plant B to achieve rapid improvement in business capabilities;

[0206] 4) It solves the problem of single traditional network services. Network devices provide artificial intelligence services such as fault prediction and industrial recognition through APP management and collaboration interfaces and service collaboration interfaces, which is beneficial to the fierce network service market competition.

[0207] To implement the network device-side method of this application embodiment, this application embodiment also provides a communication device, disposed on the network device, such as... Figure 6 As shown, the device includes: a first receiving unit 601 and a first processing unit 602; the network device is at least used to provide network access functionality; wherein,

[0208] The first receiving unit 601 is used to receive a first model sent by the cloud platform after being used by a first application running on the network device; the first model is obtained by the cloud platform using sample data.

[0209] The first processing unit 602 is used to perform data reasoning using the first model or using a locally preset reasoning strategy when it is determined that data reasoning needs to be performed after being used by the first application, so as to obtain the reasoning result.

[0210] In one embodiment, the device further includes a second sending unit, which, after being used by the first application, sends a first request to the cloud platform through a first interface; the first request is used to request the cloud platform to issue the first model;

[0211] Accordingly, the first receiving unit 601 is also used to receive the first model sent by the cloud platform based on the first request through the first interface after being used by the first application.

[0212] In one embodiment, the device further includes an acquisition unit for acquiring first sample data after being used by the first application;

[0213] Accordingly, the second sending unit is further configured to, after being used by the first application, send the first sample data to the cloud platform through the second interface, so that the cloud platform can at least use the first sample data to train the first model.

[0214] In one embodiment, the first receiving unit 601 is further configured to receive the updated first model sent by the cloud platform after being used by the first application;

[0215] Accordingly, the first processing unit 602 is also used to perform data reasoning using the updated first model after being used by the first application, and to obtain reasoning results.

[0216] In one embodiment, the first receiving unit 601 is further configured to receive first information sent by the cloud platform through a third interface after being used by the first application; the first information indicates that the cloud platform has an updated first model.

[0217] The second sending unit is further configured to, after being used by the first application, send a second request to the cloud platform through the third interface; the second request is used to request the cloud platform to issue the updated first model;

[0218] Accordingly, the first receiving unit 601 is also used to receive the updated first model sent by the cloud platform based on the second request through the third interface after being used by the first application.

[0219] In one embodiment, the acquisition unit is further configured to acquire second sample data after being used by the first application;

[0220] Accordingly, the second sending unit is further configured to send the second sample data to the cloud platform through the second interface after being used by the first application, so that the cloud platform can update the first model using at least the second sample data; the second sample data includes at least the reasoning result obtained by the first application using the first model for data reasoning.

[0221] In one embodiment, the acquisition unit is further configured to acquire first configuration information after being used by the first application;

[0222] Accordingly, the first processing unit 602 is further configured to, after being used by the first application, update the first model using the first configuration information to obtain an updated first model; and perform data reasoning using the updated first model to obtain a reasoning result.

[0223] In one embodiment, the second sending unit is further configured to send second information to the cloud platform via a fourth interface after being utilized by the first application; the second information at least characterizes the running status of the first application.

[0224] The first receiving unit 601 is also used to receive management instruction information sent by the cloud platform based on the second information through the fourth interface after being used by the first application;

[0225] The first processing unit 602 is also used to manage its own lifecycle according to the management instruction information after being used by the first application.

[0226] In one embodiment, the acquiring unit is further configured to acquire a third request; the third request is used to request the first service;

[0227] The second sending unit is also configured to send the third request to the cloud platform via the fifth interface;

[0228] The first receiving unit 601 is further configured to receive second configuration information sent by the cloud platform based on the third request through the fifth interface;

[0229] The first processing unit 602 is also configured to run the first application locally using the second configuration information.

[0230] In practical applications, the first receiving unit 601 and the second sending unit can be implemented by the communication interface in the communication device; the first processing unit 602 can be implemented by the processor in the communication device; and the acquisition unit can be implemented by the processor in the communication device in combination with the communication interface.

[0231] To implement the cloud platform-side method of this application embodiment, this application embodiment also provides a communication device, which is set on the cloud platform, such as... Figure 7 As shown, the device includes:

[0232] The second processing unit 701 is used to train the first model using sample data;

[0233] The first sending unit 702 is used to send the first model to a first application running on a network device; the network device is at least used to provide network access functionality; the first model is used for the first application to perform data inference.

[0234] In one embodiment, the device further includes a second receiving unit, configured to receive a first request sent by the first application through a first interface; the first request is used to request the cloud platform to issue the first model.

[0235] Accordingly, the first sending unit 701 is specifically used to send the first model to the first application through the first interface based on the first request.

[0236] In one embodiment, the second receiving unit is further configured to receive first sample data sent by the first application through a second interface;

[0237] Accordingly, the second processing unit 701 is also configured to train the first model using at least the first sample data.

[0238] In one embodiment, the second processing unit 701 is further configured to update the first model using new sample data to obtain an updated first model;

[0239] The first sending unit 701 is also used to send the updated first model to the first application.

[0240] In one embodiment, the first sending unit 701 is further configured to send first information to the first application via a third interface; the first information indicates that the cloud platform has an updated first model;

[0241] The second receiving unit is further configured to receive a second request sent by the first application through the third interface; the second request is used to request the cloud platform to issue the updated first model;

[0242] The first sending unit 701 is specifically used to send the updated first model to the first application through the third interface based on the second request.

[0243] In one embodiment, the second receiving unit is further configured to receive second sample data sent by the first application through a second interface; the second sample data at least includes the reasoning result obtained by the first application using the first model for data reasoning;

[0244] Accordingly, the second processing unit 701 is also configured to update the first model using at least the second number of samples.

[0245] In one embodiment, the second receiving unit is further configured to receive second information sent by the first application through a fourth interface; the second information at least characterizes the running state of the first application;

[0246] The second processing unit 701 is further configured to determine management instruction information based on the second information; the management instruction information is used for the first application to manage its own lifecycle;

[0247] The first sending unit 702 is also used to send the management instruction information to the first application through the fourth interface.

[0248] In one embodiment, the second receiving unit is further configured to receive a third request sent by the network device through a fifth interface; the third request is used to request the first service.

[0249] The second processing unit 701 is further configured to determine second configuration information based on the third request; the second configuration information is used to enable the network device to run the first application locally;

[0250] The first sending unit 702 is also configured to send the second configuration information to the network device through the fifth interface.

[0251] In practical applications, the second processing unit 701 can be implemented by a processor in the communication device; the first sending unit 702 and the second receiving unit can be implemented by a communication interface in the communication device.

[0252] It should be noted that the communication device provided in the above embodiments is only illustrated by the division of the above program modules when transmitting and processing information. In actual applications, the above processing can be assigned to different program modules as needed, that is, the internal structure of the device can be divided into different program modules to complete all or part of the processing described above. In addition, the communication device and communication method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiments, which will not be repeated here.

[0253] Based on the hardware implementation of the above program modules, and in order to implement the method on the network device side of the embodiments of this application, the embodiments of this application also provide a network device, such as... Figure 8 As shown, the network device 800 includes:

[0254] The first communication interface 801 is capable of exchanging information with the cloud platform;

[0255] The first processor 802 is connected to the first communication interface 801 to enable information interaction with the cloud platform. When running a computer program, it executes the methods provided by one or more technical solutions on the network device side. The computer program is stored in the first memory 803.

[0256] Specifically, the first processor 802 is used for:

[0257] The first application (i.e., a computer program stored on the first memory 803) is run to receive a first model sent by the cloud platform; the first model is trained by the cloud platform using sample data; the network device 800 is at least used to provide network access functionality.

[0258] Run the first application, and when it is determined that data inference is required, perform data inference using the first model or by using a locally preset inference strategy to obtain the inference result.

[0259] In one embodiment, the first processor 802 is further configured to:

[0260] Run the first application to send a first request to the cloud platform through a first interface; the first request is used to request the cloud platform to distribute the first model;

[0261] Run the first application to receive the first model sent by the cloud platform based on the first request through the first interface.

[0262] In one embodiment, the first processor 802 is further configured to run the first application to obtain first sample data and send the first sample data to the cloud platform via a second interface, so that the cloud platform can at least use the first sample data to train the first model.

[0263] In one embodiment, the first processor 802 is further configured to:

[0264] Run the first application to receive the updated first model sent by the cloud platform;

[0265] Run the first application to perform data reasoning using the updated first model and obtain the reasoning result.

[0266] In one embodiment, the first processor 802 is further configured to:

[0267] The first application is run to receive first information sent by the cloud platform through a third interface; the first information indicates that the cloud platform has an updated first model.

[0268] The first application is run to send a second request to the cloud platform through the third interface; the second request is used to request the cloud platform to issue the updated first model.

[0269] Run the first application to receive the updated first model sent by the cloud platform based on the second request through the third interface.

[0270] In one embodiment, the first processor 802 is further configured to run the first application to obtain second sample data and send the second sample data to the cloud platform through a second interface, so that the cloud platform can update the first model using at least the second sample data; the second sample data includes at least the reasoning result obtained by the first application using the first model for data reasoning.

[0271] In one embodiment, the first processor 802 is further configured to:

[0272] Run the first application to obtain the first configuration information, and use the first configuration information to update the first model to obtain the updated first model;

[0273] Run the first application to perform data reasoning using the updated first model and obtain the reasoning result.

[0274] In one embodiment, the first processor 802 is further configured to:

[0275] The first application is run to send second information to the cloud platform via a fourth interface; the second information at least represents the running status of the first application.

[0276] Run the first application to receive management instruction information sent by the cloud platform based on the second information through the fourth interface;

[0277] Run the first application to manage its own lifecycle according to the management instructions.

[0278] In one embodiment, the first processor 802 is further configured to:

[0279] Obtain a third request; the third request is used to request the first service;

[0280] The third request is sent to the cloud platform via the fifth interface;

[0281] The system receives the second configuration information sent by the cloud platform based on the third request through the fifth interface, and uses the second configuration information to run the first application locally.

[0282] It should be noted that the specific processing procedure of the first processor 802 can be understood by referring to the above method.

[0283] Of course, in practical applications, the various components in network device 800 are coupled together through bus system 804. It can be understood that bus system 804 is used to implement communication between these components. In addition to a data bus, bus system 804 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 8 The general labeled all buses as Bus System 804.

[0284] The first memory 803 in this embodiment is used to store various types of data to support the operation of the network device 800. Examples of such data include any computer program used to operate on the network device 800.

[0285] The methods disclosed in the above embodiments of this application can be applied to the first processor 802, or implemented by the first processor 802. The first processor 802 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware or by instructions in the form of software in the first processor 802. The first processor 802 may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The first processor 802 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the methods disclosed in the embodiments of this application can be directly reflected as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in a storage medium, which is located in the first memory 803. The first processor 802 reads the information in the first memory 803 and completes the steps of the aforementioned method in combination with its hardware.

[0286] In an exemplary embodiment, the network device 800 may be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to perform the aforementioned method.

[0287] Based on the hardware implementation of the above program modules, and in order to implement the cloud platform method of this application embodiment, this application embodiment also provides a cloud platform, such as... Figure 9 As shown, the cloud platform 900 includes:

[0288] The second communication interface 901 is capable of exchanging information with network devices;

[0289] The second processor 902 is connected to the second communication interface 901 to enable information interaction with network devices and to execute the methods provided by one or more technical solutions on the cloud platform side when running computer programs. The computer program is stored in the second memory 903.

[0290] Specifically, the second processor 902 is used for:

[0291] Train the first model using the sample data;

[0292] The first model is sent to a first application running on a network device; the network device is at least used to provide network access functionality; the first model is used for the first application to perform data inference.

[0293] In one embodiment, the second processor 902 is further configured to:

[0294] The first request sent by the first application is received through the first interface; the first request is used to request the cloud platform 900 to issue the first model.

[0295] Based on the first request, the first model is sent to the first application through the first interface.

[0296] In one embodiment, the second processor 902 is further configured to:

[0297] Receive the first sample data sent by the first application through the second interface;

[0298] The first model is trained using at least the first sample data.

[0299] In one embodiment, the second processor 902 is further configured to:

[0300] The first model is updated using the new sample data to obtain the updated first model;

[0301] Send the updated first model to the first application.

[0302] In one embodiment, the second processor 902 is further configured to:

[0303] The first information is sent to the first application through the third interface; the first information indicates that the cloud platform 900 has an updated first model.

[0304] The third interface receives a second request sent by the first application; the second request is used to request the cloud platform 900 to issue the updated first model.

[0305] Based on the second request, the updated first model is sent to the first application through the third interface.

[0306] In one embodiment, the second processor 902 is further configured to:

[0307] The second sample data sent by the first application is received through the second interface; the second sample data at least includes the reasoning result obtained by the first application using the first model for data reasoning.

[0308] The first model is updated using at least the second number of samples.

[0309] In one embodiment, the second processor 902 is further configured to:

[0310] The second information sent by the first application is received through the fourth interface; the second information at least represents the running status of the first application.

[0311] Based on the second information, management instruction information is determined and sent to the first application through the fourth interface; the management instruction information is used by the first application to manage its own lifecycle.

[0312] In one embodiment, the second processor 902 is further configured to:

[0313] The network device sends a third request via the fifth interface; the third request is used to request the first service.

[0314] Based on the third request, second configuration information is determined and sent to the network device through the fifth interface; the second configuration information is used for the network device to run the first application locally.

[0315] It should be noted that the specific processing procedure of the second processor 902 can be understood by referring to the above method.

[0316] Of course, in practical applications, the various components in the cloud platform 900 are coupled together through the bus system 904. It can be understood that the bus system 904 is used to implement communication between these components. In addition to the data bus, the bus system 904 also includes a power bus, a control bus, and a status signal bus. However, for clarity, in... Figure 9 The general designated all buses as Bus System 904.

[0317] The second memory 903 in this embodiment is used to store various types of data to support the operation of the cloud platform 900. Examples of such data include any computer program used to operate on the cloud platform 900.

[0318] The methods disclosed in the embodiments of this application can be applied to, or implemented by, the second processor 902. The second processor 902 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuitry of the hardware or by instructions in the form of software within the second processor 902. The second processor 902 may be a general-purpose processor, a DSP, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The second processor 902 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the methods disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software modules may be located in a storage medium, specifically a second memory 903. The second processor 902 reads information from the second memory 903 and, in conjunction with its hardware, completes the steps of the aforementioned method.

[0319] In an exemplary embodiment, the cloud platform 900 may be implemented by one or more ASICs, DSPs, PLDs, CPLDs, FPGAs, general-purpose processors, controllers, MCUs, microprocessors, or other electronic components to perform the aforementioned method.

[0320] It is understood that the memories (first memory 803, second memory 903) in the embodiments of this application can be volatile memory or non-volatile memory, or both. 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), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM); magnetic surface memory can be disk storage or magnetic tape storage. 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), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memories described in the embodiments of this application are intended to include, but are not limited to, these and any other suitable types of memories.

[0321] To implement the method provided in the embodiments of this application, the embodiments of this application also provide a communication system, such as... Figure 10 As shown, the system includes: network device 1001 and cloud platform 1002.

[0322] It should be noted that the specific processing procedures of the network device 1001 and the cloud platform 1002 have been described in detail above and will not be repeated here.

[0323] In an exemplary embodiment, this application also provides a storage medium, namely a computer storage medium, specifically a computer-readable storage medium, such as a first memory 803 storing a computer program, which can be executed by a first processor 802 of a network device 800 to complete the steps described in the aforementioned network device-side method. Another example is a second memory 903 storing a computer program, which can be executed by a second processor 902 of a cloud platform 900 to complete the steps described in the aforementioned cloud platform-side method. The computer-readable storage medium can be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

[0324] It should be noted that terms such as "first" and "second" are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0325] Furthermore, the technical solutions described in the embodiments of this application can be combined arbitrarily without conflict.

[0326] The above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application.

Claims

1. A communication method, characterized in that, Applied to network devices, including: A first application running on the network device receives a first model sent by a cloud platform; the first model is trained by the cloud platform using sample data; the network device is at least used to provide network access functionality, and the network device is able to utilize the network resources of the network device to perform data inference through the first application to provide edge computing services; When the first application determines that data inference is required, it performs data inference using the first model or using a locally preset inference strategy to obtain the inference result; wherein, The network device communicates with the cloud platform through the following interface: A data collaboration interface is used to enable data transmission between the network device and the cloud platform. An intelligent collaboration interface is used to transmit the model; An application management collaboration interface is used to coordinate the lifecycle of the first application. A business collaboration interface is used to manage the first application.

2. The method according to claim 1, characterized in that, The first model received from the cloud platform includes: The first application sends a first request to the cloud platform through a first interface; the first request is used to request the cloud platform to issue the first model, and the intelligent collaboration interface includes the first interface; The first application receives the first model sent by the cloud platform based on the first request through the first interface.

3. The method according to claim 1, characterized in that, The method further includes: The first application acquires first sample data and sends the first sample data to the cloud platform through a second interface, so that the cloud platform can at least use the first sample data to train the first model. The data collaboration interface includes the second interface.

4. The method according to claim 1, characterized in that, The method further includes: The first application receives the updated first model sent by the cloud platform; The first application uses the updated first model to perform data reasoning and obtain the reasoning result.

5. The method according to claim 4, characterized in that, The process of receiving the updated first model sent by the cloud platform includes: The first application receives first information sent by the cloud platform through a third interface; the first information indicates that the cloud platform has an updated first model, and the intelligent collaboration interface includes the third interface. The first application sends a second request to the cloud platform through the third interface; the second request is used to request the cloud platform to issue the updated first model; The first application receives the updated first model sent by the cloud platform based on the second request through the third interface.

6. The method according to claim 4, characterized in that, The method further includes: The first application acquires the second sample data and sends the second sample data to the cloud platform through the second interface, so that the cloud platform can update the first model using at least the second sample data; the second sample data includes at least the reasoning result obtained by the first application using the first model for data reasoning, and the data collaboration interface includes the second interface.

7. The method according to claim 1, characterized in that, The method further includes: The first application obtains the first configuration information and uses the first configuration information to update the first model to obtain the updated first model; The first application uses the updated first model to perform data reasoning and obtain the reasoning result.

8. The method according to any one of claims 1 to 6, characterized in that, The method further includes: The first application sends second information to the cloud platform through the fourth interface; the second information at least represents the running status of the first application, and the application management collaboration interface includes the fourth interface; The first application receives management instruction information sent by the cloud platform based on the second information through the fourth interface; The first application manages its own lifecycle according to the management instruction information.

9. The method according to any one of claims 1 to 6, characterized in that, The method further includes: Obtain a third request; the third request is used to request the first service; The third request is sent to the cloud platform through the fifth interface, and the business collaboration interface includes the fifth interface; The system receives the second configuration information sent by the cloud platform based on the third request through the fifth interface, and uses the second configuration information to run the first application locally.

10. A communication method, characterized in that, Applied to cloud platforms, including: Train the first model using the sample data; The first model is sent to a first application running on a network device; the network device is at least configured to provide network access functionality, and the network device is capable of utilizing its network resources to perform data inference through the first application to provide edge computing services; the first model is used for the first application to perform data inference; wherein... The network device communicates with the cloud platform through the following interface: A data collaboration interface is used to enable data transmission between the network device and the cloud platform. An intelligent collaboration interface is used to transmit the model; An application management collaboration interface is used to coordinate the lifecycle of the first application. A business collaboration interface is used to manage the first application.

11. The method according to claim 10, characterized in that, Sending the first model to the first application running on the network device includes: The first request sent by the first application is received through the first interface; the first request is used to request the cloud platform to issue the first model, and the intelligent collaboration interface includes the first interface. Based on the first request, the first model is sent to the first application through the first interface.

12. The method according to claim 10, characterized in that, The method further includes: The data collaboration interface includes the second interface, which receives the first sample data sent by the first application through the second interface. The first model is trained using at least the first sample data.

13. The method according to claim 10, characterized in that, The method further includes: The first model is updated using the new sample data to obtain the updated first model; Send the updated first model to the first application.

14. The method according to claim 13, characterized in that, Sending the updated first model to the first application includes: The first information is sent to the first application through the third interface; the first information indicates that the cloud platform has an updated first model, and the intelligent collaboration interface includes the third interface. The third interface receives a second request sent by the first application; the second request is used to request the cloud platform to issue the updated first model. Based on the second request, the updated first model is sent to the first application through the third interface.

15. The method according to claim 13, characterized in that, The method further includes: The second sample data sent by the first application is received through the second interface; the second sample data at least includes the reasoning result obtained by the first application using the first model for data reasoning, and the data collaboration interface includes the second interface. The first model is updated using at least the second number of samples.

16. The method according to any one of claims 10 to 15, characterized in that, The method further includes: The application management collaboration interface receives second information sent by the first application through a fourth interface; the second information at least characterizes the running status of the first application, and the application management collaboration interface includes the fourth interface. Based on the second information, management instruction information is determined and sent to the first application through the fourth interface; the management instruction information is used by the first application to manage its own lifecycle.

17. The method according to any one of claims 10 to 15, characterized in that, The method further includes: The network device sends a third request through the fifth interface; the third request is used to request the first service, and the service coordination interface includes the fifth interface. Based on the third request, second configuration information is determined and sent to the network device through the fifth interface; the second configuration information is used for the network device to run the first application locally.

18. A communication device, characterized in that, The device is configured on a network device and includes: a first receiving unit and a first processing unit; the network device is at least used to provide network access functionality, and the network device is capable of utilizing its network resources to perform data inference through a first application to provide edge computing services; wherein... The first receiving unit is used by a first application running on the network device to receive a first model sent by the cloud platform; the first model is trained by the cloud platform using sample data. The first processing unit, after being used by the first application, determines that data reasoning is required, and performs data reasoning using the first model or using a locally preset reasoning strategy to obtain a reasoning result; wherein, The network device communicates with the cloud platform through the following interface: A data collaboration interface is used to enable data transmission between the network device and the cloud platform. An intelligent collaboration interface is used to transmit the model; An application management collaboration interface is used to coordinate the lifecycle of the first application. A business collaboration interface is used to manage the first application.

19. A communication device, characterized in that, Configured on a cloud platform, including: The second processing unit is used to train the first model using sample data; A first sending unit is configured to send the first model to a first application running on a network device; the network device is at least configured to provide network access functionality, and the network device is capable of utilizing its network resources to perform data inference through the first application to provide edge computing services; the first model is used for data inference by the first application; wherein... The network device communicates with the cloud platform through the following interface: A data collaboration interface is used to enable data transmission between the network device and the cloud platform. An intelligent collaboration interface is used to transmit the model; An application management collaboration interface is used to coordinate the lifecycle of the first application. A business collaboration interface is used to manage the first application.

20. A network device, characterized in that, include: A first communication interface and a first processor; wherein... The first processor is configured to: The first application is run to receive a first model sent by the cloud platform; the first model is trained by the cloud platform using sample data; the network device is at least used to provide network access functions, and the network device can utilize the network resources of the network device to perform data inference through the first application to provide edge computing services. The first application is run, and when data inference is required, data inference is performed using the first model or a locally preset inference strategy to obtain the inference result; wherein, The network device communicates with the cloud platform through the following interface: A data collaboration interface is used to enable data transmission between the network device and the cloud platform. An intelligent collaboration interface is used to transmit the model; An application management collaboration interface is used to coordinate the lifecycle of the first application. A business collaboration interface is used to manage the first application.

21. A cloud platform, characterized in that, include: The second communication interface and the second processor; wherein... The second processor is used for: Train the first model using the sample data; The first model is sent to a first application running on a network device; the network device is at least configured to provide network access functionality, and the network device is capable of utilizing its network resources to perform data inference through the first application to provide edge computing services; the first model is used for the first application to perform data inference; wherein... The network device communicates with the cloud platform through the following interface: A data collaboration interface is used to enable data transmission between the network device and the cloud platform. An intelligent collaboration interface is used to transmit the model; An application management collaboration interface is used to coordinate the lifecycle of the first application. A business collaboration interface is used to manage the first application.

22. A network device, characterized in that, include: A first processor and a first memory for storing computer programs capable of running on the processor. Wherein, when the first processor is used to run the computer program, it performs the steps of the method according to any one of claims 1 to 9.

23. A cloud platform, characterized in that, include: A second processor and a second memory for storing computer programs that can run on the processor. Wherein, when the second processor is used to run the computer program, it performs the steps of the method according to any one of claims 10 to 17.

24. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 9, or the steps of the method according to any one of claims 10 to 17.

Images