System for updating user model, device, and method

Abstract

A system for updating a user model, a device, and a method are disclosed. In one embodiment, an electronic system may be configured to: receive from a model management server model data representing an artificial intelligence model; send to the model management server model difference data representing model differences; and receive updated model data from the model management server, the updated model data representing an updated version of the artificial intelligence model, and the updated version being at least partially based on the model difference data.

Description

System, device and method for updating user profile

[0001] Priority Declaration

[0002] This application claims priority to the Chinese patent application filed on November 30, 2023, with application number 202311622741.9 and invention name “System, device and method for updating user model”, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present disclosure relates to the field of communications, and in particular, to a system, device, and method for updating a user model. Background Art

[0004] Artificial Intelligence (AI) technology is increasingly being used. Model providers can deploy trained AI models to users' devices or systems for their use. Summary of the Invention

[0005] The present disclosure provides systems, devices, and methods for updating user models.

[0006] One aspect of the present disclosure relates to an electronic system, comprising: at least one processing unit and at least one storage unit, the at least one storage unit comprising computer program code, wherein the at least one storage unit and the computer program code are configured to enable the electronic system to perform the following operations through the at least one processing unit: receive model data representing an artificial intelligence model from a model management server; send model difference data representing model differences to the model management server; receive updated model data from the model management server, the updated model data representing an updated version of the artificial intelligence model, and the updated version is at least partially based on the model difference data.

[0007] Another aspect of the present disclosure relates to a method, comprising: receiving model data representing an artificial intelligence model from a model management server; sending model difference data representing model differences to the model management server; receiving updated model data from the model management server, the updated model data representing an updated version of the artificial intelligence model, the updated version being at least partially based on the model difference data.

[0008] One aspect of the present disclosure relates to an electronic system, comprising: at least one processing unit and at least one storage unit, the at least one storage unit comprising computer program code, wherein the at least one storage unit and the computer program code are configured to cause the electronic system to perform the following operations through the at least one processing unit: send model data defining an artificial intelligence model to a user system; receive model difference data representing model differences from the user system; generate an updated version of the artificial intelligence model based at least in part on the model difference data; and send the updated model data to the user system, the updated model data representing the updated version of the artificial intelligence model.

[0009] Another aspect of the present disclosure relates to a method, comprising: sending model data defining an artificial intelligence model to a user system; receiving model difference data representing model differences from the user system; generating an updated version of the artificial intelligence model based at least in part on the model difference data; and sending the updated model data to the user system, the updated model data representing the updated version of the artificial intelligence model.

[0010] Another aspect of the present disclosure relates to a computer-readable storage medium storing one or more instructions, which, when executed by one or more processing circuits of an electronic device, causes the electronic device to perform any method as described in the present disclosure.

[0011] Another aspect of the present disclosure relates to a computer program product, comprising a computer program, which implements any method as described in the present disclosure when executed by a processor.

[0012] Another aspect of the present disclosure relates to an apparatus comprising means for performing any of the methods described in the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The above and other objects and advantages of the present disclosure will be further described below in conjunction with specific embodiments and with reference to the accompanying drawings. In the accompanying drawings, identical or corresponding technical features or components will be represented by identical or corresponding reference numerals.

[0014] FIG1 shows a schematic diagram of a system architecture according to an embodiment of the present disclosure.

[0015] FIG2 shows an exemplary block diagram of an electronic device according to some embodiments of the present disclosure.

[0016] FIG3 shows an exemplary block diagram of an electronic device according to some embodiments of the present disclosure.

[0017] FIG4 shows a flowchart of a method for updating a user model according to some embodiments of the present disclosure.

[0018] FIG5 shows a flowchart of a method for updating a user model according to some embodiments of the present disclosure.

[0019] FIG6 illustrates the application of techniques according to some embodiments of the present disclosure in a cellular network.

[0020] FIG7 shows a conventional beam selection process.

[0021] FIG8 illustrates an example of using a user model to predict an optimal beam pair according to some embodiments of the present disclosure.

[0022] FIG9 illustrates another example of using a user model to predict an optimal beam pair according to some embodiments of the present disclosure.

[0023] FIG10 illustrates the application of the technology according to some embodiments of the present disclosure in the Internet of Vehicles.

[0024] FIG11 is a block diagram illustrating a first example of an exemplary configuration of a gNB to which the techniques of this disclosure may be applied.

[0025] FIG12 is a block diagram illustrating a second example of an exemplary configuration of a gNB to which the techniques of this disclosure may be applied.

[0026] FIG. 13 is a block diagram illustrating an example of an exemplary configuration of a communication device to which the technology of the present disclosure may be applied.

[0027] FIG. 14 is a block diagram illustrating an example of an exemplary configuration of a car navigation device to which the technology of the present disclosure can be applied.

[0028] While the embodiments described in this disclosure may be susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and are herein described in detail. It should be understood that the drawings and detailed description thereof are not intended to limit the embodiments to the particular forms disclosed, but on the contrary, the intent is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the claims. DETAILED DESCRIPTION

[0029] Exemplary embodiments of the present disclosure will be described below with reference to the accompanying drawings. For the sake of clarity and conciseness, not all features of the embodiments are described in the specification. However, it should be understood that many implementation-specific settings must be made in the process of implementing the embodiments in order to achieve the developer's specific goals, such as meeting those restrictions related to equipment and services, and these restrictions may vary depending on the implementation. In addition, it should be understood that although the development work may be very complex and time-consuming, it is only a routine task for those skilled in the art who benefit from the contents of this disclosure.

[0030] It should also be noted here that in order to avoid obscuring the present disclosure due to unnecessary details, only the processing steps and / or equipment structures that are closely related to at least the scheme according to the present disclosure are shown in the accompanying drawings, while other details that are not very relevant to the present disclosure are omitted.

[0031] Over time, artificial intelligence (AI) has been applied across various industries. From innovative applications in healthcare to intelligent solutions in financial services, to efficient optimization in manufacturing and agriculture, the widespread use of AI has become indispensable. At this stage, businesses can not only provide products and services to customers but also integrate customized AI models into them to further enhance the functionality and performance of these products and services.

[0032] For example, in the communications sector, AI is transforming how communications networks are managed and optimized. By analyzing vast amounts of network data, AI can predict network congestion, optimize bandwidth allocation, and even proactively detect potential network failures. This helps provide more stable and efficient communications services, meeting modern society's ever-increasing demand for connectivity.

[0033] For example, in the field of autonomous driving, AI provides vehicles with intelligent decision-making and environmental perception capabilities. Through technologies such as deep learning and perception, vehicles can identify roads, vehicles, and pedestrians, enabling autonomous driving. This not only improves driving safety but also has the potential to enhance traffic flow and energy efficiency.

[0034] Communications and autonomous driving are just a few examples of areas where AI is being applied. As AI technology continues to advance, its applications will continue to expand, bringing more innovation and change to various industries. However, the resulting privacy, security, and ethical issues have also attracted widespread attention. The goal is to improve the performance of AI models while complying with privacy, security, and ethical requirements.

[0035] Typically, an AI model is generated and provided by a model provider. The model provider may be a product or service provider, or associated with a product or service provider. The model provider may train the AI ​​model based on a training dataset. The trained AI model may be deployed to a user's device or system for user use. The deployed AI model is referred to herein as a user model.

[0036] The training dataset used by the model provider to train the user model is usually limited. This training dataset may not be optimally matched to the actual environment of the user using the user model. Therefore, the user may not obtain the optimal performance of the user model in the actual environment. In some scenarios, it is desirable to change (e.g., adjust or update) the deployed user model based on the user's actual environment.

[0037] However, leaving changes to user models entirely up to the user is inappropriate. Direct user changes to a user model could degrade the model's performance. For example, if a user independently changes the user model used for autonomous driving, the changed user model could cause an accident. If a user independently changes the user model used for a communications network, the changed user model could degrade communication quality. The responsibility for such degradation would unreasonably fall on the user, while the model provider could escape any consequences of the degraded model.

[0038] On the other hand, it is not appropriate to directly provide user local data to the model provider in order to update the user model, because this may violate privacy or security requirements. Moreover, user local data may be large and redundant, which brings a significant data transmission burden.

[0039] Therefore, the present disclosure provides improved systems, devices, and methods for updating user models. One or more embodiments of the present disclosure enable targeted updates to user models using user-local data while maintaining user privacy and security. In addition, one or more embodiments of the present disclosure can also address one or more other problems.

[0040] 1. Exemplary System Architecture

[0041] FIG1 shows a schematic diagram of an architecture 100 according to an embodiment of the present disclosure. As shown in the figure, the architecture 100 may include a model management layer 110 and a model user layer 120 .

[0042] The model management layer 110 may be associated with the provider of the artificial intelligence model. In some embodiments, the model management layer 110 may be implemented as one or more model management servers. These model management servers may be physical servers or servers hosted in the cloud. The model management servers may be implemented, for example, using the electronic devices described in FIG. 2 .

[0043] According to some embodiments of the present disclosure, the model management layer 110 may be configured to maintain one or more artificial intelligence models. Maintenance may include generating, storing and / or updating one or more artificial intelligence models. In some examples, the model management layer 110 may initially train one or more artificial intelligence models based on the training data set collected by it. The trained artificial intelligence model may be provided by the model management layer 110 to the model user layer 120 for deployment in the model user layer 120. In addition, the model management layer 110 may also update the one or more artificial intelligence models based on feedback from the model user layer 120 (e.g., model differences 113). The model management layer 110 may provide the updated version of the artificial intelligence model to the model user layer 120.

[0044] In the embodiment shown in FIG1 , the model management layer 110 may be configured to maintain a large model 112 . The large model 112 may be an artificial intelligence model having a complexity exceeding a certain threshold. Such a large model 112 may not be suitable for running on a client device with limited processing power. Therefore, the model management layer 110 may also be configured to reduce the large model 112 to generate a small model 111 . The reduction may be performed by any suitable means, such as knowledge distillation or model pruning. The small model 111 may be provided to the model user layer 120 to be deployed and used as a user model 121 .

[0045] According to an embodiment of the present disclosure, the model management layer 110 may update the small model 111 based on feedback (e.g., model differences 113) obtained from the model user layer 120. In an embodiment where the model management layer 110 maintains the large model 112, the model management layer 110 may be configured to update the small model 111 based on an appropriate manner. In some examples, the model management layer 110 may first update the large model 112, and then generate an updated version of the small model 111 based on a reduction of the updated large model 112. In other examples, the model management layer 110 may not first update the large model 112, but may directly update the small model 111.

[0046] In an alternative embodiment (not shown), the model management layer 110 maintains one or more small models 111 rather than the large model 112. Respective small models 111 among these small models 111 can be provided to the model user layer 120 for deployment in corresponding application scenarios. In this case, the model management layer 110 can directly update the corresponding small model 111 based on the model differences 113 obtained in the corresponding application scenario.

[0047] The model user layer 120 may be associated with users of the artificial intelligence model. The model user layer 120 may be referred to as a user system. In some embodiments, the model user layer 120 may include one or more client devices. The client devices may include, for example, UEs, in-vehicle devices, media devices, etc. The client devices in the model user layer 120 may be implemented, for example, using the electronic devices described in FIG. 3 . In some embodiments, the model user layer 120 may also optionally include an intermediate device (e.g., a base station device) between the client device and the model management server.

[0048] The model user layer 120 can be configured to execute one or more trained user models 121 provided by the model management layer 110. The model user layer 120 can be configured to provide local data as input data to the user model 121. The local data may include actual data associated with the client device in the model user layer 120 (which is also referred to as user data herein). The local data may include, but is not limited to, environmental data, operational data, measurement data, multimedia data, etc. associated with the client device. The user model 121 is a trained artificial intelligence model that can generate one or more output results based on the input local data. These output results may include various types of outputs, such as classification of the input data or predictions generated based on the input data. The client device in the model user layer 120 can also use the output results of the user model 121 to perform one or more actions, such as controlling the operation of the client device.

[0049] According to some embodiments of the present disclosure, the model user layer 120 may also be configured to determine a model difference 122 associated with the user model 121. The model difference 122 may be used to characterize the difference between the actual performance of the deployed user model 121 and the expected performance of the user model 121 in the deployed environment. In some embodiments, the model user layer 120 may include two coaching models, such as coaching model A 123 and coaching model B 124. Each coaching model, coaching model A 123 and coaching model B 124, may be an artificial intelligence model different from the user model 121. Coaching model A 123 may be configured to simulate the actual performance of the deployed user model 121. Coaching model B 124 may be configured to simulate the expected performance of the user model 121. The expected performance may refer to the optimal performance of the user model 121 in the current environment. The difference between coaching model A 123 and coaching model B 124 may be used to characterize the model difference 122.

[0050] In some embodiments, the model user layer 120 may be configured to train one or more initial training models to obtain training model A 123 and training model B 124. For example, the model user layer 120 may be configured to perform model training using local data and actual output results generated by the user model 121 based on the local data, thereby obtaining training model A 123. The model user layer 120 may also be configured to perform model training using local data and expected output results of the user model 121 based on the local data, thereby obtaining training model B 124. The model user layer 120 may determine model differences 122 based on the obtained training models A 123 and B 124. Model differences 122 may represent the difference between the actual performance and the expected performance of the user model 121. Model differences 122 may be represented in any suitable manner. For example, model differences 122 may be represented as the difference between model parameters, performance indicators, or output results of the training models A 123 and B 124.

[0051] The model user layer 120 may also be configured to provide the determined model differences 122 to the model management layer 110 for updating the user model, as previously described. In some embodiments, the model management layer 110 may aggregate multiple model differences 122 from the model user layer 120 to update the user model.

[0052] It should be understood that the architecture 100 of FIG. 1 is merely exemplary. Various modifications may be made to the architecture 100 without departing from the scope of the present disclosure. For example, although the architecture 100 is shown as including only one small model 111 and a corresponding user model 121, in other embodiments, the architecture 100 may include any number of small models 111 and user models 121 for deployment to multiple client devices. In some embodiments, a corresponding pair of coaching models 123 and 124 may be trained for each client device in a plurality of client devices. In other embodiments, a corresponding pair of coaching models 123 and 124 may be trained for each subset of client devices in a plurality of client devices. In this case, the pair of coaching models 123 and 124 may be trained based on the actual output results and expected output results of the user models from all client devices in the subset. Other modifications are also possible without departing from the scope of the present disclosure.

[0053] 2. Exemplary Equipment

[0054] Figure 2 shows an exemplary block diagram of an electronic device 200 according to some embodiments of the present disclosure. The electronic device 200 can be implemented at the model management layer 110 of the architecture 100 as described in Figure 1. In some embodiments, the electronic device 200 can be implemented as a model management server. The electronic device 200 can be used to perform one or more operations related to the model management layer 110 or the model management server described herein. Specifically, the electronic device 200 can be implemented as the model management server itself, as a part of the model management server, or as a control device for controlling the model management server. For example, the electronic device 200 can be implemented as a chip for controlling the model management server. In some embodiments herein, the electronic device 200 is implemented as the model management server itself, which is merely for convenience of description and is not intended to constitute a limitation.

[0055] According to some embodiments of the present disclosure, the electronic device 200 may include a communication unit 210 , a storage unit 220 , and a processing circuit 230 .

[0056] The communication unit 210 of the electronic device 200 can be used to receive or send wired transmissions or radio transmissions. In some embodiments of the present disclosure, the communication unit 210 can perform functions such as up-conversion and digital-to-analog conversion on the transmitted signal and / or perform functions such as down-conversion and analog-to-digital conversion on the received signal. The communication unit 210 can be implemented using various technologies. For example, the communication unit 210 can be implemented as communication interface components such as an antenna device, a radio frequency circuit, and a portion of a baseband processing circuit. In Figure 2, the communication unit 210 is drawn with a dotted line because the communication unit 210 can alternatively be located within the processing circuit 230 or outside the electronic device 200.

[0057] The storage unit 220 of the electronic device 200 can store information generated by the processing circuit 230, information received from other devices through the communication unit 210 or information to be sent to other devices, computer programs, machine codes and data used for the operation of the electronic device 200, etc. According to some embodiments of the present disclosure, the storage unit 220 can store one or more artificial intelligence models, model differences, updates to artificial intelligence models, update logs or records and / or other associated data. The storage unit 220 can be a volatile memory and / or a non-volatile memory. For example, the storage unit 220 can include, but is not limited to, random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), read-only memory (ROM), and flash memory. The storage unit 220 is drawn with a dotted line because it can alternatively be located within the processing circuit 230 or outside the electronic device 200.

[0058] The processing circuit 230 of the electronic device 200 may be configured to perform one or more operations, thereby providing various functions of the electronic device 200. For example, the processing circuit 230 may be configured to perform the steps of the method described with respect to FIG4 . The processing circuit 230 may perform the corresponding operations by executing one or more executable instructions stored in the storage unit 220. The processing circuit 230 may be used to implement one or more steps performed by the model management server in various methods according to some embodiments of the present disclosure.

[0059] According to some embodiments of the present disclosure, the processing circuit 230 may be configured to execute the method for updating the user model described herein. For example, the processing circuit 230 may include a model management unit 231. Specifically, the model management unit 231 may be configured to send model data defining the artificial intelligence model to the user system. The user system is, for example, the model user layer 120 described in Figure 1. The model management unit 231 may also be configured to receive model difference data representing model differences from the user system. Then, the model management unit 231 may also be configured to generate an updated version of the artificial intelligence model based at least in part on the model difference data. The model management unit 231 may also be configured to send updated model data to the user system, and the updated model data represents the updated version of the artificial intelligence model.

[0060] It should be understood that the processing circuit 230 in Figure 2 is merely exemplary and may include one or more additional units and / or one or more additional steps for performing the techniques of the present disclosure.

[0061] Figure 3 shows an exemplary block diagram of an electronic device 300 according to some embodiments of the present disclosure. The electronic device 300 can be implemented at the model user layer 120 of the architecture 100 as described in Figure 1. In some embodiments, the electronic device 300 can be implemented as a client device. The electronic device 300 can be used to perform one or more operations related to the model user layer 120 or the client device described herein. Specifically, the electronic device 300 can be implemented as the client device itself, as a part of the client device, or as a control device for controlling the client device. For example, the electronic device 300 can be implemented as a chip for controlling the client device. In some embodiments herein, the electronic device 300 is implemented as the client device itself, which is merely for convenience of description and is not intended to be limiting.

[0062] According to some embodiments of the present disclosure, the electronic device 300 may include a communication unit 310 , a storage unit 320 , and a processing circuit 330 .

[0063] The communication unit 310 of the electronic device 300 can be used to receive or send wired transmissions or radio transmissions. In some embodiments of the present disclosure, the communication unit 310 can perform functions such as up-conversion and digital-to-analog conversion on the transmitted signal and / or perform functions such as down-conversion and analog-to-digital conversion on the received signal. The communication unit 310 can be implemented using various technologies. For example, the communication unit 310 can be implemented as communication interface components such as an antenna device, a radio frequency circuit, and a portion of the baseband processing circuit. In Figure 3, the communication unit 310 is drawn with a dotted line because the communication unit 310 can alternatively be located within the processing circuit 330 or outside the electronic device 300.

[0064] The storage unit 320 of the electronic device 300 can store information generated by the processing circuit 330, information received from other devices via the communication unit 310 or information to be transmitted to other devices, computer programs, machine code, and data used for the operation of the electronic device 300, and the like. According to some embodiments of the present disclosure, the storage unit 320 can store a user model as well as local data for the user model, actual output results, and expected output results. According to some embodiments of the present disclosure, the storage unit 320 can also store one or more coaching models and model differences. The storage unit 320 can be volatile memory and / or non-volatile memory. For example, the storage unit 320 can include, but is not limited to, random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), read-only memory (ROM), and flash memory. The storage unit 320 is depicted with dashed lines because it can alternatively be located within the processing circuit 330 or external to the electronic device 300.

[0065] The processing circuit 330 of the electronic device 300 can be configured to perform one or more operations, thereby providing various functions of the electronic device 300. For example, the processing circuit 330 can be configured to perform the steps of the method described with respect to FIG. 5 . The processing circuit 330 can perform the corresponding operations by executing one or more executable instructions stored in the storage unit 320. The processing circuit 330 can be used to implement one or more steps performed by the model user layer in various methods according to some embodiments of the present disclosure.

[0066] According to some embodiments of the present disclosure, the processing circuit 330 may be configured to perform the method for updating the user model described herein. For example, the processing circuit 330 may be configured to receive model data representing the artificial intelligence model from the model management server. The processing circuit 330 may also be configured to send model difference data representing model differences to the model management server. The processing circuit 330 may also be configured to receive updated model data from the model management server, the updated model data representing an updated version of the artificial intelligence model, the updated version being based at least in part on the model difference data.

[0067] According to some embodiments of the present disclosure, the processing circuit 330 may include a model execution unit 331 for executing or using the received artificial intelligence model. The model execution unit 331 may be configured to provide the user's local input data to the artificial intelligence model, thereby obtaining one or more output results of the artificial intelligence model. The obtained output result is the actual output result of the artificial intelligence model. The actual output result can be provided to the electronic device 300 for performing one or more actions. Optionally, as described below, the actual output result can also be used to train one or more tutoring models. The model execution unit 331 can also be configured to monitor the performance of the deployed artificial intelligence model. In response to the performance of the deployed artificial intelligence model being lower than a specific standard, the model execution unit 331 can send a message to trigger an update of the artificial intelligence model.

[0068] According to some embodiments of the present disclosure, the processing circuit 330 may further include a difference determination unit 332 for determining model difference data. According to some embodiments of the present disclosure, the model difference data represents the difference between the actual performance of the deployed artificial intelligence model and the expected performance of the artificial intelligence model. In some embodiments, the difference may be represented by the difference between two tutoring models (e.g., a first tutoring model and a second tutoring model), where the first tutoring model simulates the actual performance of the artificial intelligence model and the second tutoring model simulates the expected performance of the artificial intelligence model. According to some embodiments of the present disclosure, the processing circuit 330 may further include a model training unit 333 for training the first tutoring model and the second tutoring model. The model training unit 333 may be configured to train the first tutoring model based at least in part on the input data and actual output results of the artificial intelligence model. The model training unit 333 may also be configured to train the second tutoring model based at least in part on the input data and expected output results of the artificial intelligence model. The difference determination unit 332 and the model training unit 333 are depicted with dashed lines because they may alternatively be located outside the processing circuit 330 or outside the electronic device 300.

[0069] It should be understood that the processing circuit 330 in Figure 3 is merely exemplary and may include one or more additional units and / or one or more additional steps for performing the techniques of the present disclosure.

[0070] 3. Exemplary Methods

[0071] FIG4 illustrates a flow chart of a method 400 for updating a user model according to some embodiments of the present disclosure. Method 400 may be performed at the model management layer. For example, method 400 may be performed by a model management server. The model management server may be implemented by the aforementioned electronic device 200. Accordingly, method 400 may be performed by the processing circuit 230 of the electronic device 200. Specifically, the processing circuit 230 of the electronic device 200 may execute method 400 by executing a computer program.

[0072] Method 400 may begin at step 410. In step 410, the model management server may be configured to send model data representing an artificial intelligence model to a user system. The artificial intelligence model sent to the user may be referred to as a user model.

[0073] According to embodiments of the present disclosure, the user model sent may be generated by a model management server. In some embodiments, the user model may be a small model generated by reducing a large model. A large model may be an artificial intelligence model with a complexity exceeding a certain threshold. The complexity of the model can be characterized, for example, by the complexity of the topology, the number of model parameters, or any other suitable metric, including but not limited to the number of layers or nodes in the model. For example, a large model of the model management server may have millions or more model parameters. Such a large model may not be suitable for running on a client device with limited processing power. The model management server may be configured to reduce the large model or a portion thereof through any suitable means, such as knowledge distillation or model pruning, to generate a small model. The small model may have a significantly lower complexity than the large model. For example, the small model may have a simpler model topology and / or fewer model parameters. The model management server may generate different small models for different application scenarios. Small models may be application-specific, for example, generated by reducing a large model specifically for a particular application scenario. Therefore, the model management server can generate multiple small models for different application scenarios based on the same large model. By way of example and not limitation, the large model may be a multimodal model capable of processing various media types (images, audio, text, video, etc.), while the multiple small models may include an image classification model, a face recognition model, a language translation model, a music recognition model, etc. Other types or sizes of large and small models are also possible.

[0074] It should be understood that the technology of the present disclosure can be applied to various types of artificial intelligence models, including models that have already been developed and models that may be developed in the future. Exemplary artificial intelligence models include, but are not limited to, models trained by convolutional neural networks (CNNs), deep neural networks (DNNs), recurrent neural networks (RNNs), long short-term memory (LSTMs), generative models, random forests, and their variants and enhancement algorithms.

[0075] According to an embodiment of the present disclosure, the model data defining the artificial intelligence model sent by the model management server may have various suitable forms. The model data enables the client device to deploy and use the artificial intelligence model based on the data. In some embodiments, the model data may include a set of model parameters of the artificial intelligence model. Each artificial intelligence model can be represented by a set of model parameters that characterize the model. For example, for a CNN model, the set of model parameters that characterizes the model may include, but is not limited to, one or more parameters that describe the number of layers of the neural network, the number and distribution of neurons in each layer, the connection relationship between neurons, the weight of each neuron connection, etc. For other types of models, other types of model parameter sets may be used for characterization. In some embodiments, the model data may include an identifier of the artificial intelligence model that uniquely identifies the model to be deployed. The user system that receives the model data can use the identifier to retrieve the corresponding model from the model repository. Other forms of model data are also possible.

[0076] In some embodiments, the model management server may also send an indication to the user system. This indication may inform the user system that the model sent in step 410 is an initial trained model. In addition, this indication may also inform the user system that the model management server can modify the model based on feedback from the user system.

[0077] Method 400 may then continue to step 420. In this step, the model management server may be configured to receive model difference data characterizing model differences from the user system.

[0078] According to some embodiments of the present disclosure, model difference data may represent the difference between the actual performance of an artificial intelligence model deployed at a user system and the expected performance of the artificial intelligence model. In some embodiments, the difference may be represented as the difference between a first tutoring model and a second tutoring model, where the first tutoring model simulates the actual performance of a user model deployed at the user system, and the second tutoring model simulates the expected performance of a user model deployed at the user system. Details regarding model difference data will be further described below.

[0079] Method 400 may then continue to step 430. In this step, the model management server may be configured to generate an updated version of the artificial intelligence model based at least in part on the model difference data.

[0080] According to some embodiments of the present disclosure, the model management server maintains a larger-scale model, while the user model is a reduced-scale model. In this case, the model management server can update the larger-scale model based at least in part on the model difference data, and generate a smaller-scale model based on the reduction of the updated larger-scale model as an updated version of the artificial intelligence model. Alternatively, the model management server can directly update the smaller-scale model based at least in part on the model difference data as an updated version of the artificial intelligence model. The update may include various adjustments to the artificial intelligence model, such as adjusting parameters (including topology and / or weights).

[0081] Method 400 may then proceed to step 440. In this step, the model management server may be configured to send updated model data representing an updated version of the artificial intelligence model to the user system. The updated version of the artificial intelligence model may be deployed to the client device to replace the previous version of the user model.

[0082] It should be understood that the above description is merely an exemplary embodiment of method 400. Method 400 may include one or more additional steps performed by the model management server. Details of various embodiments of method 400 and additional or alternative embodiments will be further described below.

[0083] Figure 5 shows a flow chart of a method 500 for updating a user model according to some embodiments of the present disclosure. Method 500 can be performed at the model user layer. For example, method 500 can be performed by a client device. The client device can be implemented by the aforementioned electronic device 300. Accordingly, method 500 can be performed by the processing circuit 330 of the electronic device 300. Specifically, the processing circuit 330 of the electronic device 300 can execute method 500 by executing a computer program.

[0084] Method 500 may begin at step 510. In step 510, a client device may be configured to receive model data representing an artificial intelligence model from a model management server.

[0085] As mentioned above, the model data representing the artificial intelligence model may include model parameters of the artificial intelligence model or an identifier of the artificial intelligence model. The client device can locally deploy the corresponding artificial intelligence model (i.e., user model) based on the model data. The short-circuit protection device can use local data as input data to execute the user model and obtain corresponding output results.

[0086] Method 500 may then continue to step 520. In this step, the client device may be configured to send model difference data characterizing the model differences to the model management server.

[0087] In the present disclosure, model difference data can be used to characterize the difference between the actual performance of the deployed user model and the expected performance of the user model in the deployed environment. The actual performance of the user model is associated with the actual output result provided by the user model based on local data. The expected performance of the user model is associated with the expected output result (e.g., correct output result) that the user model should provide based on the local data. As mentioned above, since the training data set used by the model provider (e.g., the model management layer) to train the user model is not necessarily the best match for the actual environment of the user using the user model, the user may not be able to obtain the optimal performance of the user model in the actual environment. In other words, since the performance of the user model is not perfect, the actual output result of the user model may be inconsistent with the expected output result of the user. This inconsistency is specific to the user's actual environment and reflects the defects of the deployed user model in the actual environment. The performance of the user model in the actual environment can be improved by discovering this inconsistency and providing it to the model management layer to update the user model.

[0088] According to an embodiment of the present disclosure, in order to characterize the aforementioned inconsistency from the perspective of the model, two coaching models can be introduced, namely, a first coaching model and a second coaching model. The first coaching model can be configured to simulate the actual performance of the user model. Given the same input data, the first coaching model is expected to produce output results that are the same as or similar to the actual output results of the user model. The second coaching model can be configured to simulate the expected performance of the user model. Given the same input data, the second coaching model is expected to produce output results that are the same as or similar to the expected output results of the user model. The model difference data provided in step 520 can be characterized as the model difference between the first coaching model and the second coaching model. Providing model difference data rather than user data to the model management layer can protect the privacy and security of the user because the model difference data does not contain data directly related to the user's actual environment.

[0089] According to some embodiments of the present disclosure, model difference data may be obtained based on one or more aspects of the first tutoring model and the second tutoring model. In some examples, the model difference data may represent the difference between one or more model parameters between the first tutoring model and the second tutoring model. In other examples, the model difference data may represent the difference between one or more performance indicators between the first tutoring model and the second tutoring model. In still other examples, the model difference data may represent the difference between one or more output results between the first tutoring model and the second tutoring model. In other examples, the model difference data may include a combination of the aforementioned various forms. Other forms of model difference data are also possible.

[0090] According to some embodiments of the present disclosure, various metrics can be used to characterize model differences. Exemplary metrics may include one or more of gradient, weight, divergence, distance, eigenvalue, etc. In some embodiments, the value of each metric characterizing the model difference may be the difference between the values ​​of two corresponding parameters of the first tutoring model and the second tutoring model. For example, in the case where the corresponding parameter is the gradient of the weight of the model, the value characterizing the model difference between the two models may include the difference in the gradient associated with the two models. In some embodiments, the value of each metric may be a distance metric or a similarity metric (such as Euclidean distance, KL divergence, normalized value, etc.) between the two parameters. In some embodiments, the features of the model difference can be used to represent the model difference to further reduce the amount of data. The features of the model difference can be represented by compression or enhancement of the model difference, including but not limited to distribution, eigenvalue, pruning of the model difference, model difference top-k weight, weight of the model difference above a certain threshold, certain layers of the model difference, and the like.

[0091] According to some embodiments of the present disclosure, the first tutoring model and the second tutoring model that meet the aforementioned requirements can be obtained in various ways. Preferably, the first tutoring model and the second tutoring model can be obtained through model training. For example, the first training data set can be constructed using the input data of the deployed user model and the actual output results of the user model. In the first training data set, the actual output results of the user model can be marked as facts. In addition, the second training data set can be constructed using the input data of the deployed user model and the expected output results of the user model. In the second training data set, the expected output results of the user model can be marked as facts. The first training data set can be used to perform model training to obtain the first tutoring model. The second training data set can be used to perform model training to obtain the second tutoring model.

[0092] For the purpose of explanation, as an example and not limitation, the input data can be a picture of a dog, and the actual output result of the user model classifies the picture as a cat, while the expected output result is a dog. Accordingly, in the first training data set, the picture is labeled as a cat (i.e., the actual output result of the user model) because it is expected that the output result of the first tutoring model for such a picture is the same as the actual output result of the user model (even if the result is incorrect). In the second training data set, the picture is labeled as a dog because it is expected that the output result of the second tutoring model for such a picture is the same as the expected output result. Using the constructed first training data set and second training data set, respectively, iteratively training the first tutoring model and the second tutoring model can make the two models simulate the actual performance and expected performance of the user model, respectively.

[0093] According to an embodiment of the present disclosure, the data used to construct the first training dataset or the second training dataset may come from one or more clients that have deployed a user model. For example, in some embodiments, the input data and / or output results (e.g., the actual output results and expected output results of the user model) may come from a single client that has deployed a user model (e.g., a single UE or an in-vehicle device). In other embodiments, the input data and / or output results from multiple clients that have deployed the user model may be aggregated to construct a first training dataset for training the first tutoring model and a second training dataset for training the second tutoring model. For example, the multiple clients may include all clients that have deployed the user model. Alternatively, the multiple clients may include a specific subset of all clients that have deployed the user model, such as some clients that share the same real-world environment. In some examples, the specific subset may be identified based on the location of each client (e.g., cellular cell, driving area, etc.). In some embodiments, the output results (e.g., the actual output results and expected output results of the user model) used to construct the training dataset may come from a specific client, while the input data used to construct the training dataset may come from one or more other clients of the subset to which the specific client belongs.

[0094] According to some embodiments of the present disclosure, the first and second tutoring models can be based on the same initial tutoring model. In other words, the two tutoring models are obtained by training the same initial tutoring model using different first and second training data sets. In some embodiments, the model management layer can provide the initial tutoring model to the model user layer. As previously described, providing the initial tutoring model can include providing model parameters or an identifier of the model. The initial tutoring model can be of the same type as the user model.

[0095] In some embodiments, the training of the first tutoring model and the second tutoring model is initiated in response to the actual performance of the user model not meeting specific criteria. For example, the client device may monitor the performance of the deployed user model. When the performance of the deployed user model does not meet specific criteria, the client device may send a message to the model management layer to, for example, request an update of the deployed user model. In response to the message, the model management layer may trigger the training of the tutoring model. For example, the model management layer may provide the initial tutoring model to the model user layer. In addition, the model management layer may send a training data configuration to the model user layer, which may specify the composition and format of the first training data set and the second training data set. Based on the training data configuration, the model user layer may collect and save the actual output results, expected output results, and input data of the user model accordingly for constructing the first training data set and the second training data set.

[0096] According to some embodiments of the present disclosure, the first tutoring model and the second tutoring model may have a complexity different from that of the deployed user model. In some examples, the complexity of the first tutoring model and the second tutoring model may be lower than that of the user model, which can save the overhead of model training. In some examples, the complexity of the first tutoring model and the second tutoring model may be the same as that of the user model. For example, the initial tutoring model may be the same as the initial model on which the user model is based. In some examples, the first tutoring model and the second tutoring model may have a higher complexity than that of the deployed user model. For example, in the case of using a CNN model, the initial tutoring model used to train the first tutoring model and the second tutoring model may have more layers of neural networks and / or the number of neurons per layer than the deployed user model. The increased complexity allows the model differences to be characterized with higher resolution.

[0097] Method 500 may then continue to step 530 . In this step, the client device may be configured to receive updated model data from the model management server. The updated model data represents an updated version of the user model that is based at least in part on the model difference data from step 520 .

[0098] According to some embodiments of the present disclosure, the user model deployed at the client is a smaller-scale model generated by reducing a larger-scale model. In this case, the model management server can update the larger-scale model based at least in part on the model difference data, and generate the smaller-scale model based on the reduction of the updated larger-scale model as an updated version of the user model. Alternatively, the model management server can directly update the smaller-scale model based at least in part on the model difference data as an updated version of the user model.

[0099] According to some embodiments of the present disclosure, a client device may be configured to deploy an updated version of the user model using updated model data. The client device may enable the updated version. If the performance of the updated version of the user model still does not meet certain criteria, the client device may send a message to the model management layer to request further updates to the user model. Further updates to the user model may include repeating some or all of the aforementioned steps, which are not described in detail here. The model management server may be configured to log each update to the model.

[0100] It should be understood that the above description is merely an exemplary embodiment of method 500. Method 500 may include one or more additional steps performed by the model user layer. Details of various embodiments of method 500 and additional or alternative embodiments will be further described below.

[0101] The functions of the elements disclosed herein can be implemented using circuits or processing circuits that include general-purpose processors, special-purpose processors, integrated circuits, ASICs ("application-specific integrated circuits"), conventional circuits, and / or combinations thereof that are configured or programmed to perform the disclosed functions. Processors are considered processing circuits or circuits because they include transistors and other circuits therein. In the present disclosure, a circuit, unit, or device is hardware that performs or is programmed to perform the functions described. The hardware can be any hardware disclosed herein or otherwise known that is programmed or configured to perform the functions described. When the hardware is a processor that can be considered a type of circuit, the circuit, device, or unit is a combination of hardware and software, with the software being used to configure the hardware and / or processor.

[0102] 4. Example Scenarios

[0103] The following describes the application of the technology of the present disclosure in multiple exemplary scenarios in conjunction with Figures 6 to 10. It should be understood that these exemplary scenarios are illustrative and not intended to be limiting.

[0104] FIG6 illustrates an application of techniques according to some embodiments of the present disclosure in a cellular network 600. The illustrated application scenario is associated with beam steering and is illustrative only and not intended to be limiting.

[0105] The cellular network 600 may include a core network 610, a base station 620 (e.g., a gNB), and a UE 630. The model management layer 110 may be deployed at the core network 610. For example, the core network 610 may include a 5G core network (5GC). The model management layer 110 may be deployed in the cloud at the 5GC. In some embodiments, the model management layer 110 may be deployed at a network data analysis function (NWDAF).

[0106] The base station 620 may receive the user model 121 from the model management layer 110. The initial user model 121 may be trained by the model management layer 110 using training data. The base station 620 may also forward the user model 121 to the UE 630 for deployment at the UE 630. For example, the base station 620 may receive model data representing the user model 121 from the model management layer 110 and forward the model data to the UE 630.

[0107] The UE 630 may deploy the user model 121 based on the received model data. The UE 630 may use data specific to the UE 630 as input to run the deployed user model 121. The data specific to the UE 630 may be referred to as local data of the UE 630, which includes but is not limited to environmental data, operational data, measurement data, multimedia data, etc. Different UEs may have different local data.

[0108] The actual output results of the user model 121 deployed at the UE 630 can be used for one or more operations of the UE 630. In the example shown in FIG6 , the one or more operations include beam steering. In a cellular network, the base station 620 may have multiple candidate beams, and the UE 630 may also have multiple candidate beams. These beams form multiple candidate beam pairs. Beam steering is used to use the best beam pair among the multiple candidate beam pairs for communication between the base station 620 and the UE 630.

[0109] Traditionally, the best beam pair is selected through a beam selection process. FIG7 illustrates a conventional beam selection process. UE 630 may include a conventional beam selection module 632 for performing the conventional beam selection process. In this example, the base station has five candidate beams, each pointing in a different direction. In addition, the UE has four candidate beams, each pointing in a different direction. Therefore, there are 20 candidate beam pairs between the base station and the UE. The number of candidate beams is exemplary and not limiting. During the beam selection process, the base station and the UE need to perform measurements for each of these candidate beam pairs. For example, the reference signal received power (RSRP) or any other suitable parameter at the UE may be measured. Specifically, for each of the four candidate beams of the UE, the base station sequentially transmits five candidate beams. The conventional beam selection module 632 may perform 20 measurements to obtain a measurement result (e.g., RSRP) associated with each candidate beam pair. The beam with the best measurement result (e.g., the highest RSRP) may be considered the best beam pair. The conventional beam selection module 632 may provide the best beam pair to the beam control module 631. The beam control module 631 may control the UE 630 to communicate with the base station 620 using the best beam pair.

[0110] An artificial intelligence model (e.g., user model 121) can be used to improve the selection process of the best beam pair. Figure 8 shows an example 800 of using the user model 121 to predict the best beam pair according to some embodiments of the present disclosure. In this example, the user model 121 is trained to predict the best beam pair based on a small number of beam measurements. In this example, the number of beam measurements can be reduced. For example, the base station may not send the candidate beams shown as blank, but only transmit candidate beams in a few predetermined specific directions. The results of the small number of beam measurements performed can be provided to the user model 121 as input data. The user model 121 can infer the best beam pair based on the input data. This can reduce the overhead for beam measurements.

[0111] FIG9 shows another example 900 of using the user model 121 to predict the best beam pair according to some embodiments of the present disclosure. In this example, the user model 121 is trained to predict the best beam pair in the future based on a previous beam selection sequence. The model is, for example, an LSTM model. The previous beam selection sequence may include the results of beam measurements at one or more previous time points (e.g., T, T+1, etc.) and the determined best beam pair. In this example, the previous beam selection sequence may be provided to the user model 121 as input data. The user model 121 may predict the best beam pair at the current time or at a future time based on the input data.

[0112] In the examples of Figures 8-9, the prediction results (i.e., predicted beam pairs) of the user model 121 can be provided to the beam control module 631 of the UE 630. The beam control module 631 can control the UE 630 to use the predicted beam pair to communicate with the base station 620. It should be understood that the examples of Figures 8-9 are merely illustrative and not intended to be limiting. In other examples, the user model 121 can predict the optimal beam pair based on other types of input data.

[0113] Although using the user model 121 to predict the best beam pair may have some advantages over conventional methods, the predicted best beam pair is affected by the actual environment in which the UE is located (e.g., a specific cell environment). The user model 121 trained and provided by the model management layer may not completely match the actual environment. Therefore, the technology disclosed in this article can be used to update the user model. For example, when it is detected that the performance of the user model 121 is lower than a specific standard, the update of the user model can be triggered. As an example, when the RSRP of the best beam pair predicted by the user model 121 is lower than a first predetermined threshold, or when the communication quality using the predicted best beam pair is lower than a second predetermined threshold, the UE 630 (or the base station 620) can send a message or request to the model management layer 110 to trigger an update. Other specific standards are also possible.

[0114] In response to triggering an update of the user model, the model management layer 110 may provide the model user layer with initial coaching models for training a first coaching model (e.g., coaching model A) and a second coaching model (e.g., coaching model B). The coaching models may be trained at the UE 630 or the base station 620. In the example of FIG6 , coaching model A 123 and coaching model B are trained at the base station 620 because the base station 620 has more processing power than the UE 630. In this case, the UE 630 may construct the input data and actual results of the user model 121 as a first training dataset for training the coaching model A. As described above, the input data may be the results of beam measurements (e.g., from beam measurements performed by the regular beam selection module 632), a previous beam selection sequence (e.g., from stored data), or other local data (e.g., measurement data associated with the channel environment) that is input to the user model 121 to predict the optimal beam pair. The actual results may be the predicted beam pair generated by the user model 121 based on the corresponding input data. Additionally, the UE 630 may also construct the input data and the expected result of the user model 121 into a second training data set for training the coaching model B. For example, the expected result may be the best beam pair with the best beam measurement result. The best beam measurement result may be, for example, the highest RSRP determined by measuring all candidate beam pairs. The best beam pair with the best beam measurement result may be provided by the regular beam selection module 632.

[0115] In some embodiments, the UE 630 may receive a training data configuration from the base station 620, which may specify how the UE constructs the first training data set and the second training data set. Specifically, the training data configuration may specify the input data, actual output results, and / or expected output results of the user model for transmission from the UE 630 to the base station 620. The training data configuration may be carried in one or more signalings from the base station to the UE, such as RRC, MAC CE, or DCI. For example, in a conventional beam selection process, the UE 630 may only feed back a limited number of beam measurement results (e.g., the four beam pairs with the highest RSRP) to the base station 620. In order to use the technology of the present disclosure, the base station 620 may instruct the UE 630 to feed back more beam measurement results (e.g., all measurement results used as input data for the user model 121). This may be indicated by a training data configuration sent to the UE 630.

[0116] The base station 620 can train the coaching model A and coaching model B based on the received first training data set and the received second training data set, respectively. If there are multiple UEs 630 deployed with the same user model 121 in the cell served by the base station 620, the base station 620 can optionally aggregate the input data and output results of the user models from the multiple UEs 630 as training data to train the coaching model. Various appropriate techniques can be used to train the corresponding coaching model, including training techniques that have already been developed and training techniques that may be developed in the future. Training can be an iterative process that ends when a termination condition (e.g., convergence) is reached.

[0117] Because coaching model A is trained based on the input data and actual output results of user model 121, coaching model A can simulate the actual performance of user model 121. Coaching model A may have the same or similar model parameters or topology as user model 121. Because coaching model B is trained based on the input data and expected output results of user model 121, coaching model B can simulate the expected performance of user model 121. Coaching model B may have the same or similar model parameters or topology as the expected version of user model 121 in the current environment. In actual applications, due to security and confidentiality requirements, user model 121 may be opaque to UE 630, so UE 630 cannot access or understand the internal structure of user model 121. However, by training coaching model A and coaching model B, UE 630 can simulate the actual and expected versions of user model 121 without modifying user model 121 itself. Differences between the actual and expected versions of simulated user model 121 can, to a certain extent, reflect the deficiencies of user model 121 in the current environment. This defect is specific to the environment and can therefore be used to update the user model so that the updated user model better matches the current environment.

[0118] The model difference 122 can be determined based on the trained coaching model A and coaching model B. In some embodiments, the model difference 122 can be represented as the difference between one or more model parameters of the coaching model A and the coaching model B. In some embodiments, the model difference 122 can be represented as the difference between one or more output results of the coaching model A and the coaching model B. For example, when the same input data is provided, the difference between the first best beam pair predicted by the coaching model A and the second best beam pair predicted by the coaching model B can constitute part of the model difference 122. In some embodiments, the model difference 122 can be represented as the difference between one or more performance indicators of the coaching model A and the coaching model B. For example, the difference between the RSRP of the first best beam pair and the second best beam pair can constitute part of the model difference 122. The model difference data may include a combination of the various forms described above. Other forms of model difference data are also possible.

[0119] The base station 620 may send the determined model differences 122 to the model management layer 110 for updating the user model 121. As previously described, the update of the user model 121 by the model management layer 110 may be direct or indirect (e.g., updating a larger model first). When the model management layer 110 is connected to multiple base stations 620, the model management layer 110 may perform an update based on multiple model differences 122 from the multiple base stations 620. For example, these model differences 122 may be combined. In some examples, the model management layer 110 may perform the update in a federated learning manner. The obtained updated version of the user model 121 may be deployed to the UE 620 via the base station 620 to replace the previously deployed user model. The updated version of the user model 121 may be sent at an appropriate time. For example, the updated version of the user model 121 may be sent at a time when the UE 630 is not currently using the user model 121 to complete the replacement of the user model 121.

[0120] It should be understood that the embodiments shown in Figures 6-9 are merely illustrative and not intended to be limiting. In other embodiments, the AI ​​model may be used for aspects other than beam steering. For example, in some embodiments, the AI ​​model may be used to select a target cell. In some embodiments, the AI ​​model may be used to schedule transmission resources. In some embodiments, the AI ​​model may be used to predict network congestion. In some embodiments, the AI ​​model may be used for power adjustment. The techniques disclosed herein are not limited by the type or functionality of the AI ​​model used. Each type of AI model may have corresponding input data and corresponding actual output results. Furthermore, the expected output result of the AI ​​model may be obtained through various suitable means. For example, as described above for beam steering, a UE or base station may have or execute a conventional method, the result of which may be considered the expected output result of the AI ​​model. In some examples, the expected output result of the AI ​​model may be provided by a person or other device that is aware of the expected output result. For example, in an example where the AI ​​model is used to predict network speed, the actual network speed may be measured and provided as the expected output result of the AI ​​model.

[0121] Furthermore, it should be understood that the distribution of the various components or elements shown in FIG6 is merely illustrative and not intended to be limiting. In other embodiments, the distribution of components or elements associated with the techniques of this disclosure may be different. In some examples, one or both of the two coaching models may alternatively be located at the UE. In other examples, one or both of the two coaching models may alternatively be located at other devices (e.g., other network functions) in the cellular network 600 other than the base station and the UE.

[0122] It should be understood that the cellular network 600 shown in Figures 6-9 may include cellular communication technologies that are currently developed, under development, or to be developed, such as 2G, 3G, 4G, 5G, 6G, etc. Similar technologies can also be implemented in other types of wireless networks, such as Wi-Fi networks.

[0123] FIG10 illustrates the application of the technology according to some embodiments of the present disclosure in a connected vehicle 1000. The application scenario shown is associated with vehicle control and is merely illustrative and not intended to be limiting.

[0124] The connected vehicle network 1000 may include a cloud 1010 , an access point (AP) 1020 , and an in-vehicle device 1030 . The model management layer 110 may be deployed at the cloud 1010 .

[0125] AP 1020 may be a Wi-Fi access point, a cellular base station, or any other suitable network access point. AP 1020 may receive user model 121 from model management layer 110. AP 1020 may also forward user model 121 to in-vehicle device 1030 for deployment therein. For example, AP 1020 may receive model data representing user model 121 from model management layer 110 and forward the model data to in-vehicle device 1030.

[0126] The in-vehicle device 1030 may deploy the user model 121 based on the received model data. The in-vehicle device 1030 may use data specific to the in-vehicle device 1030 as input to run the deployed user model 121. The data specific to the in-vehicle device 1030 may be referred to as local data of the in-vehicle device 1030, which includes but is not limited to environmental data, operational data, measurement data, multimedia data, etc. Different in-vehicle devices 1030 may have different local data.

[0127] The actual output results of the user model 121 deployed at the in-vehicle device 1030 can be used for one or more operations of the in-vehicle device 1030. In the example shown in FIG10 , the one or more operations include vehicle control. Vehicle control may include, but is not limited to, various vehicle actions such as starting, accelerating, decelerating, braking, lane changing, and steering. As an example and not a limitation, the user model 121 can be configured to control lane changing based on measurement results from sensors installed on the vehicle. In other embodiments, the actual output results of the user model 121 can be used in other aspects of the in-vehicle device 1030 without limitation.

[0128] The actual output results of the user model 121 can be provided to the vehicle control module 1031. The vehicle control module 1031 can control the vehicle based on the results. At the same time, the vehicle control module 1031 can also receive control commands from other sources, such as driving decisions from the driver of the vehicle or other control devices on the vehicle. If the actual output results of the user model 121 do not meet specific standards (for example, the degree or frequency of manual intervention by the driver exceeds a threshold), the performance of the user model 121 may not reach the expected level. Accordingly, the in-vehicle device 1030 can initiate an update to the user model 121.

[0129] In the example shown in FIG10 , coaching model A 123 and coaching model B 124 can be trained on an in-vehicle device 1030. The in-vehicle device 1030 can be configured to construct different training datasets to train coaching model A 123 and coaching model B 124, respectively. For example, the in-vehicle device 1030 can use input data from the user model 121 deployed on the in-vehicle device 1030 and the corresponding actual output results to train coaching model A 123. In the example of vehicle control, the input data can be sensory data associated with the vehicle, such as captured environmental images, current location, current speed, etc. The actual output results can include autonomous driving decisions given by the user model 121, such as lane change instructions, which can include, for example, lane change direction, lane change speed, turn signal activation, etc. In some embodiments, the in-vehicle device 1030 can use human driving decisions as the desired output results to train coaching model B 124. The human decision module 1032 of the in-vehicle device 1030 can collect human driving decisions. In the example of vehicle control, the human driving decisions can be decisions actually made by the vehicle driver. For example, the manual driving decision may include whether the driver allowed the vehicle to operate based on the autonomous driving decision given by the user model 121, or whether the driver took a driving action that differed from the autonomous driving decision. The in-vehicle device 1030 may also be configured to determine a model difference 122 based on the trained coaching model A 123 and coaching model B 124, and send the determined model difference 122 to the model management layer 110 via the AP 1020 for updating the user model 121. The updated version of the user model 121 may also be forwarded to the in-vehicle device 1030 via the AP 1020 for deployment.

[0130] It should be understood that the embodiment shown in Figure 10 is merely illustrative and not intended to be limiting. In other embodiments, the artificial intelligence model may be used in other aspects of vehicle control. For example, in some embodiments, the artificial intelligence model may be used for air conditioning control. In some embodiments, the artificial intelligence model may be used for gasoline or battery control. Other functions of the artificial intelligence model are also possible. In addition, the distribution of the various components or elements shown in Figure 10 is merely illustrative and not intended to be limiting. In other embodiments, the distribution of components or elements associated with the technology of the present disclosure may be different.

[0131] In another example, the technology according to some embodiments of the present disclosure can be used in a media device. For example, an artificial intelligence model can be deployed in a media device and configured to classify media data. For example, the media device can be a camera with an artificial intelligence model that can classify captured images. Accordingly, the input data of the artificial intelligence model can include media data, and the actual output result of the artificial intelligence model includes a predicted classification generated based on the media data. The expected data result of the artificial intelligence model includes the true classification of the media data, which can be obtained in various ways. For example, the true classification can be obtained through user feedback.

[0132] It should be understood that the exemplary scenarios described above are illustrative and not intended to be limiting. The technology disclosed herein can also be used in other suitable scenarios.

[0133] 5. Application product examples

[0134] The technology disclosed herein can be applied to various products.

[0135] For example, the control device / base station mentioned in this disclosure can be implemented as any type of base station, such as an eNB, such as a macro eNB and a small eNB. A small eNB can be an eNB that covers a cell smaller than a macro cell, such as a pico eNB, a micro eNB, and a home (femto) eNB. For another example, it can be implemented as a gNB, such as a macro gNB and a small gNB. A small gNB can be a gNB that covers a cell smaller than a macro cell, such as a pico gNB, a micro gNB, and a home (femto) gNB. Alternatively, the base station can be implemented as any other type of base station, such as a NodeB and a base transceiver station (BTS). A base station can include: a main body configured to control wireless communications (also called a base station device); and one or more remote radio heads (RRHs) located at a location different from the main body. In addition, the various types of terminals described below can all operate as a base station by temporarily or semi-permanently performing base station functions. For example, the terminal devices mentioned in the present disclosure may be implemented as mobile terminals (such as smart phones, tablet personal computers (PCs), notebook PCs, portable game terminals, portable / dongle-type mobile routers, and digital camera devices) or vehicle-mounted terminals (such as car navigation devices) in some embodiments. The terminal device may also be implemented as a terminal that performs machine-to-machine (M2M) communication (also known as a machine-type communication (MTC) terminal). In addition, the terminal device may be a wireless communication module (such as an integrated circuit module comprising a single chip) installed on each of the above-mentioned terminals.

[0136] Application examples according to the present disclosure will be described below with reference to the accompanying drawings.

[0137] [Example about base stations]

[0138] It should be understood that the term "base station" in the present disclosure has the full breadth of its usual meaning and includes at least a wireless communication station used as part of a wireless communication system or radio system to facilitate communication. Examples of base stations may include, but are not limited to, the following: a base station may be one or both of a base transceiver station (BTS) and a base station controller (BSC) in a GSM system, one or both of a radio network controller (RNC) and a Node B in a WCDMA system, an eNB in ​​an LTE and LTE-Advanced system, or a corresponding network node in a future communication system (such as a gNB, eLTE eNB, etc. that may appear in a 5G communication system). Some of the functions in the base station of the present disclosure may also be implemented as an entity that has a control function for communication in D2D, M2M, and V2V communication scenarios, or as an entity that plays a spectrum coordination role in a cognitive radio communication scenario.

[0139] First example

[0140] FIG11 is a block diagram illustrating a first exemplary configuration of a gNB to which the techniques of this disclosure may be applied. The gNB 2100 includes multiple antennas 2110 and a base station device 2120. The base station device 2120 and each antenna 2110 may be connected to each other via an RF cable. In one implementation, the gNB 2100 (or base station device 2120) herein may correspond to the control-side electronic device described above.

[0141] Each antenna 2110 includes a single or multiple antenna elements (such as multiple antenna elements included in a multiple-input multiple-output (MIMO) antenna) and is used for base station device 2120 to transmit and receive wireless signals. As shown in Figure 11, gNB 2100 may include multiple antennas 2110. For example, multiple antennas 2110 may be compatible with multiple frequency bands used by gNB 2100.

[0142] The base station device 2120 includes a controller 2121 , a memory 2122 , a network interface 2123 , and a wireless communication interface 2125 .

[0143] The controller 2121 may be, for example, a CPU or a DSP, and operates various functions of the higher layers of the base station device 2120. For example, the controller 2121 determines the location information of a target terminal device in at least one terminal device based on the positioning information of at least one terminal device on the terminal side in the wireless communication system acquired by the wireless communication interface 2125 and the specific location configuration information of at least one terminal device. The controller 2121 may have a logical function of performing the following controls: the control may be, for example, radio resource control, radio bearer control, mobility management, access control, and scheduling. The control may be performed in conjunction with a nearby gNB or core network node. The memory 2122 includes RAM and ROM, and stores programs executed by the controller 2121 and various types of control data (such as a terminal list, transmission power data, and scheduling data).

[0144] The network interface 2123 is a communication interface for connecting the base station device 2120 to the core network 2124. The controller 2121 can communicate with the core network node or another gNB via the network interface 2123. In this case, the gNB 2100 and the core network node or other gNB can be connected to each other via a logical interface (such as an S1 interface and an X2 interface). The network interface 2123 can also be a wired communication interface or a wireless communication interface for wireless backhaul. If the network interface 2123 is a wireless communication interface, the network interface 2123 can use a higher frequency band for wireless communication than the frequency band used by the wireless communication interface 2125.

[0145] The wireless communication interface 2125 supports any cellular communication scheme, such as Long Term Evolution (LTE) and LTE-Advanced, and provides wireless connectivity to terminals located in the gNB 2100 cell via the antenna 2110. The wireless communication interface 2125 may typically include, for example, a baseband (BB) processor 2126 and RF circuitry 2127. The BB processor 2126 can perform various signal processing functions, such as encoding / decoding, modulation / demodulation, and multiplexing / demultiplexing, and performs various signal processing for layers such as Layer 1 (L1), Medium Access Control (MAC), Radio Link Control (RLC), and Packet Data Convergence Protocol (PDCP). In place of the controller 2121, the BB processor 2126 may perform some or all of the aforementioned logical functions. The BB processor 2126 may be a memory storing communication control programs, or a module including a processor configured to execute programs and associated circuitry. Program updates can modify the functionality of the BB processor 2126. This module may be a card or blade inserted into a slot in the base station device 2120. Alternatively, it may be a chip mounted on the card or blade. Meanwhile, the RF circuit 2127 may include, for example, a mixer, a filter, and an amplifier, and transmits and receives wireless signals via the antenna 2110. Although FIG11 shows an example in which one RF circuit 2127 is connected to one antenna 2110, the present disclosure is not limited to this illustration, and one RF circuit 2127 may be connected to multiple antennas 2110 at the same time.

[0146] As shown in Figure 11 , the wireless communication interface 2125 may include multiple BB processors 2126. For example, multiple BB processors 2126 may be compatible with multiple frequency bands used by the gNB 2100. As shown in Figure 11 , the wireless communication interface 2125 may include multiple RF circuits 2127. For example, multiple RF circuits 2127 may be compatible with multiple antenna elements. While Figure 11 illustrates an example in which the wireless communication interface 2125 includes multiple BB processors 2126 and multiple RF circuits 2127, the wireless communication interface 2125 may also include a single BB processor 2126 or a single RF circuit 2127.

[0147] Second example

[0148] FIG12 is a block diagram illustrating a second exemplary configuration of a gNB to which the techniques of this disclosure can be applied. gNB 2200 includes multiple antennas 2210, RRHs 2220, and base station equipment 2230. RRHs 2220 and each antenna 2210 can be connected to each other via an RF cable. Base station equipment 2230 and RRHs 2220 can be connected to each other via a high-speed line such as an optical fiber cable. In one implementation, gNB 2200 (or base station equipment 2230) herein may correspond to the control-side electronic device described above.

[0149] Each antenna 2210 includes a single or multiple antenna elements (such as multiple antenna elements included in a MIMO antenna) and is used for RRH 2220 to transmit and receive wireless signals. As shown in Figure 12, gNB 2200 may include multiple antennas 2210. For example, multiple antennas 2210 may be compatible with multiple frequency bands used by gNB 2200.

[0150] Base station device 2230 includes a controller 2231, a memory 2232, a network interface 2233, a wireless communication interface 2234, and a connection interface 2236. Controller 2231, memory 2232, and network interface 2233 are the same as controller 2121, memory 2122, and network interface 2123 described with reference to FIG.

[0151] The wireless communication interface 2234 supports any cellular communication scheme (such as LTE and LTE-Advanced) and provides wireless communication to terminals located in the sector corresponding to the RRH 2220 via the RRH 2220 and the antenna 2210. The wireless communication interface 2234 may generally include, for example, a BB processor 2235. The BB processor 2235 is identical to the BB processor 2126 described with reference to FIG. 11 , except that the BB processor 2235 is connected to the RF circuit 2222 of the RRH 2220 via the connection interface 2236. As shown in FIG. 12 , the wireless communication interface 2234 may include multiple BB processors 2235. For example, the multiple BB processors 2235 may be compatible with multiple frequency bands used by the gNB 2200. Although FIG. 12 illustrates an example in which the wireless communication interface 2234 includes multiple BB processors 2235, the wireless communication interface 2234 may also include a single BB processor 2235.

[0152] The connection interface 2236 is an interface for connecting the base station device 2230 (wireless communication interface 2234) to the RRH 2220. The connection interface 2236 may also be a communication module for connecting the base station device 2230 (wireless communication interface 2234) to the RRH 2220 for communication in the high-speed line.

[0153] The RRH 2220 includes a connection interface 2223 and a wireless communication interface 2221 .

[0154] The connection interface 2223 is an interface for connecting the RRH 2220 (wireless communication interface 2221) to the base station device 2230. The connection interface 2223 may also be a communication module for communication in the above-mentioned high-speed line.

[0155] The wireless communication interface 2221 transmits and receives wireless signals via the antenna 2210. The wireless communication interface 2221 may generally include, for example, an RF circuit 2222. The RF circuit 2222 may include, for example, a mixer, a filter, and an amplifier, and transmits and receives wireless signals via the antenna 2210. Although FIG12 illustrates an example in which one RF circuit 2222 is connected to one antenna 2210, the present disclosure is not limited to this illustration, and one RF circuit 2222 may be connected to multiple antennas 2210 simultaneously.

[0156] As shown in FIG12 , the wireless communication interface 2221 may include multiple RF circuits 2222. For example, the multiple RF circuits 2222 may support multiple antenna elements. Although FIG12 shows an example in which the wireless communication interface 2221 includes multiple RF circuits 2222, the wireless communication interface 2221 may also include a single RF circuit 2222.

[0157] [Example of User Equipment / Terminal Equipment]

[0158] First example

[0159] 13 is a block diagram illustrating an example of an exemplary configuration of a communication device 2300 (e.g., a smart phone, a contact, etc.) to which the technology of the present disclosure may be applied. The communication device 2300 includes a processor 2301, a memory 2302, a storage device 2303, an external connection interface 2304, a camera 2306, a sensor 2307, a microphone 2308, an input device 2309, a display device 2310, a speaker 2311, a wireless communication interface 2312, one or more antenna switches 2315, one or more antennas 2316, a bus 2317, a battery 2318, and an auxiliary controller 2319. In one implementation, the communication device 2300 (or processor 2301) herein may correspond to the aforementioned transmitting device or terminal-side electronic device.

[0160] The processor 2301 may be, for example, a CPU or a system on a chip (SoC), and controls the functions of the application layer and other layers of the communication device 2300. The memory 2302 includes RAM and ROM, and stores data and programs executed by the processor 2301. The storage device 2303 may include storage media such as semiconductor memories and hard disks. The external connection interface 2304 is an interface for connecting an external device (such as a memory card and a universal serial bus (USB) device) to the communication device 2300.

[0161] The camera 2306 includes an image sensor (such as a charge coupled device (CCD) and a complementary metal oxide semiconductor (CMOS)) and generates a captured image. The sensor 2307 may include a group of sensors such as a measurement sensor, a gyroscope sensor, a geomagnetic sensor, and an acceleration sensor. The microphone 2308 converts the sound input to the communication device 2300 into an audio signal. The input device 2309 includes, for example, a touch sensor, a keypad, a keyboard, a button, or a switch configured to detect a touch on the screen of the display device 2310, and receives an operation or information input from the user. The display device 2310 includes a screen (such as a liquid crystal display (LCD) and an organic light emitting diode (OLED) display) and displays the output image of the communication device 2300. The speaker 2311 converts the audio signal output from the communication device 2300 into sound.

[0162] The wireless communication interface 2312 supports any cellular communication scheme (such as LTE and LTE-Advanced) and performs wireless communication. The wireless communication interface 2312 may generally include, for example, a BB processor 2313 and an RF circuit 2314. The BB processor 2313 may perform, for example, encoding / decoding, modulation / demodulation, and multiplexing / demultiplexing, and perform various types of signal processing for wireless communication. Meanwhile, the RF circuit 2314 may include, for example, a mixer, a filter, and an amplifier, and transmit and receive wireless signals via an antenna 2316. The wireless communication interface 2312 may be a chip module on which the BB processor 2313 and the RF circuit 2314 are integrated. As shown in FIG13 , the wireless communication interface 2312 may include multiple BB processors 2313 and multiple RF circuits 2314. Although FIG13 shows an example in which the wireless communication interface 2312 includes multiple BB processors 2313 and multiple RF circuits 2314, the wireless communication interface 2312 may also include a single BB processor 2313 or a single RF circuit 2314.

[0163] In addition, in addition to the cellular communication scheme, the wireless communication interface 2312 can support other types of wireless communication schemes, such as a short-range wireless communication scheme, a near-field communication scheme, and a wireless local area network (LAN) scheme. In this case, the wireless communication interface 2312 can include a BB processor 2313 and an RF circuit 2314 for each wireless communication scheme.

[0164] Each of the antenna switches 2315 switches the connection destination of the antenna 2316 between a plurality of circuits (eg, circuits for different wireless communication schemes) included in the wireless communication interface 2312 .

[0165] Each of the antennas 2316 includes a single or multiple antenna elements (such as multiple antenna elements included in a MIMO antenna) and is used for transmitting and receiving wireless signals via the wireless communication interface 2312. As shown in FIG13 , the communication device 2300 may include multiple antennas 2316. Although FIG13 shows an example in which the communication device 2300 includes multiple antennas 2316, the communication device 2300 may also include a single antenna 2316.

[0166] In addition, the communication device 2300 may include an antenna 2316 for each wireless communication scheme. In this case, the antenna switch 2315 may be omitted from the configuration of the communication device 2300.

[0167] The bus 2317 connects the processor 2301, the memory 2302, the storage device 2303, the external connection interface 2304, the camera 2306, the sensor 2307, the microphone 2308, the input device 2309, the display device 2310, the speaker 2311, the wireless communication interface 2312, and the auxiliary controller 2319. The battery 2318 supplies power to the various blocks of the communication device 2300 shown in FIG13 via a feeder line, which is partially shown as a dotted line in the figure. The auxiliary controller 2319 operates the minimum necessary functions of the communication device 2300, for example, in sleep mode.

[0168] Second example

[0169] Figure 14 is a block diagram showing an example of an exemplary configuration of a car navigation device 2400 to which the technology of the present disclosure can be applied. The car navigation device 2400 includes a processor 2401, a memory 2402, a global positioning system (GPS) module 2404, a sensor 2405, a data interface 2406, a content player 2407, a storage medium interface 2408, an input device 2409, a display device 2510, a speaker 2411, a wireless communication interface 2413, one or more antenna switches 2416, one or more antennas 2417, and a battery 2418. In one implementation, the car navigation device 2400 (or processor 2401) herein may correspond to a transmitting device or a terminal-side electronic device.

[0170] The processor 2401 may be, for example, a CPU or an SoC, and controls a navigation function and other functions of the car navigation apparatus 2400. The memory 2402 includes a RAM and a ROM, and stores data and programs executed by the processor 2401.

[0171] The GPS module 2404 uses GPS signals received from GPS satellites to measure the position (such as latitude, longitude, and altitude) of the car navigation device 2400. The sensor 2405 may include a group of sensors such as a gyroscope sensor, a geomagnetic sensor, and an air pressure sensor. The data interface 2406 is connected to, for example, the vehicle network 2421 via a terminal not shown, and acquires data generated by the vehicle (such as vehicle speed data).

[0172] The content player 2407 reproduces content stored in a storage medium (such as a CD or DVD) inserted into the storage medium interface 2408. The input device 2409 includes, for example, a touch sensor, button, or switch configured to detect a touch on the screen of the display device 2510, and receives an operation or information input from the user. The display device 2510 includes a screen such as an LCD or OLED display and displays an image of a navigation function or reproduced content. The speaker 2411 outputs the sound of the navigation function or the reproduced content.

[0173] The wireless communication interface 2413 supports any cellular communication scheme (such as LTE and LTE-Advanced) and performs wireless communication. The wireless communication interface 2413 may generally include, for example, a BB processor 2414 and an RF circuit 2415. The BB processor 2414 may perform, for example, encoding / decoding, modulation / demodulation, and multiplexing / demultiplexing, and perform various types of signal processing for wireless communication. Meanwhile, the RF circuit 2415 may include, for example, a mixer, a filter, and an amplifier, and transmit and receive wireless signals via an antenna 2417. The wireless communication interface 2413 may also be a chip module on which the BB processor 2414 and the RF circuit 2415 are integrated. As shown in Figure 14, the wireless communication interface 2413 may include multiple BB processors 2414 and multiple RF circuits 2415. Although Figure 14 shows an example in which the wireless communication interface 2413 includes multiple BB processors 2414 and multiple RF circuits 2415, the wireless communication interface 2413 may also include a single BB processor 2414 or a single RF circuit 2415.

[0174] In addition, in addition to the cellular communication scheme, the wireless communication interface 2413 can support other types of wireless communication schemes, such as short-range wireless communication schemes, near field communication schemes, and wireless LAN schemes. In this case, for each wireless communication scheme, the wireless communication interface 2413 can include a BB processor 2414 and an RF circuit 2415.

[0175] Each of the antenna switches 2416 switches the connection destination of the antenna 2417 between a plurality of circuits included in the wireless communication interface 2413 , such as circuits for different wireless communication schemes.

[0176] Each of the antennas 2417 includes a single or multiple antenna elements (such as multiple antenna elements included in a MIMO antenna) and is used for transmitting and receiving wireless signals with the wireless communication interface 2413. As shown in Figure 14, the car navigation device 2400 may include multiple antennas 2417. Although Figure 14 shows an example in which the car navigation device 2400 includes multiple antennas 2417, the car navigation device 2400 may also include a single antenna 2417.

[0177] In addition, the car navigation device 2400 may include an antenna 2417 for each wireless communication scheme. In this case, the antenna switch 2416 may be omitted from the configuration of the car navigation device 2400.

[0178] The battery 2418 supplies power to the respective blocks of the car navigation device 2400 shown in Fig. 14 via a feeder line, which is partially shown as a dotted line in the figure. The battery 2418 accumulates the power supplied from the vehicle.

[0179] The technology of the present disclosure can also be implemented as an in-vehicle system (or vehicle) 2420 including a car navigation device 2400, an in-vehicle network 2421, and one or more blocks of a vehicle module 2422. The vehicle module 2422 generates vehicle data (such as vehicle speed, engine speed, and fault information) and outputs the generated data to the in-vehicle network 2421.

[0180] The exemplary embodiments of the present disclosure are described above with reference to the accompanying drawings, but the present disclosure is certainly not limited to the above examples. Those skilled in the art may obtain various changes and modifications within the scope of the appended claims, and it should be understood that these changes and modifications will naturally fall within the technical scope of the present disclosure.

[0181] It should be understood that the machine-executable instructions in the machine-readable storage medium or program product according to the embodiments of the present disclosure can be configured to perform operations corresponding to the above-mentioned device and method embodiments. When referring to the above-mentioned device and method embodiments, the embodiments of the machine-readable storage medium or program product are clear to those skilled in the art and are therefore not described again. Machine-readable storage media and program products for carrying or including the above-mentioned machine-executable instructions also fall within the scope of the present disclosure. Such storage media may include, but are not limited to, floppy disks, optical disks, magneto-optical disks, memory cards, memory sticks, and the like.

[0182] In addition, it should be understood that the above series of processes and devices can also be implemented by software and / or firmware. In the case of implementation by software and / or firmware, the storage medium of the relevant device stores the corresponding program constituting the corresponding software, and when the program is executed, various functions can be performed.

[0183] For example, a plurality of functions included in one unit in the above embodiments may be implemented by separate devices. Alternatively, a plurality of functions implemented by a plurality of units in the above embodiments may be implemented by separate devices, respectively. In addition, one of the above functions may be implemented by a plurality of units. Needless to say, such a configuration is included in the technical scope of the present disclosure.

[0184] In this specification, the steps described in the flowchart include not only processing that is performed in order in time series, but also processing that is performed in parallel or individually rather than necessarily in time series. In addition, even in the steps processed in time series, it goes without saying that the order can be appropriately changed.

[0185] 6. Exemplary embodiments of the present disclosure

[0186] 1. An electronic system comprising: at least one processing unit; and at least one storage unit, the at least one storage unit comprising computer program code, wherein the at least one storage unit and the computer program code are configured to cause the electronic system to perform the following operations through the at least one processing unit: receive model data representing an artificial intelligence model from a model management server; send model difference data representing model differences to the model management server; and receive updated model data from the model management server, the updated model data representing an updated version of the artificial intelligence model, the updated version being at least partially based on the model difference data.

[0187] 2. The electronic system of embodiment 1, wherein the model difference data represents a difference between actual performance of the artificial intelligence model and expected performance of the artificial intelligence model.

[0188] 3. An electronic system as described in Example 2, wherein the model difference data represents a model difference between a first tutoring model and a second tutoring model, wherein the first tutoring model simulates the actual performance of the artificial intelligence model, and wherein the second tutoring model simulates the expected performance of the artificial intelligence model.

[0189] 4. The electronic system of embodiment 3, wherein the model difference data represents at least one of the following differences between the first tutoring model and the second tutoring model: a difference between one or more model parameters; a difference between one or more performance indicators; or a difference between one or more output results.

[0190] 5. An electronic system as described in Example 3, wherein the at least one storage unit and the computer program code are further configured to enable the electronic system to perform the following operations through the at least one processing unit: train the first tutoring model based at least in part on the input data and actual output results of the artificial intelligence model; and train the second tutoring model based at least in part on the input data and expected output results of the artificial intelligence model.

[0191] 6. The electronic system of embodiment 5, wherein the first tutoring model and the second tutoring model are based on the same initial tutoring model.

[0192] 7. The electronic system of embodiment 5, wherein the first tutoring model and the second tutoring model have a higher complexity than the artificial intelligence model.

[0193] 8. The electronic system of embodiment 5, wherein the training of the first tutoring model and the second tutoring model is initiated in response to the actual performance of the artificial intelligence model not meeting a specific standard.

[0194] 9. The electronic system as shown in Example 1, wherein the model data representing the artificial intelligence model includes at least one of the following: a model parameter of the artificial intelligence model; or an identifier of the artificial intelligence model.

[0195] 10. The electronic system as shown in Example 1, wherein the artificial intelligence model is a smaller-scale model generated by reducing a larger-scale model.

[0196] 11. An electronic system as shown in Example 10, wherein the updated version is generated based on one of the following operations: directly updating the artificial intelligence model based at least in part on the model difference data; or updating the larger-scale model based at least in part on the model difference data, and generating the updated version of the artificial intelligence model based on a reduction of the updated larger-scale model.

[0197] 12. An electronic system as described in Example 5, wherein the artificial intelligence model is deployed at a user equipment UE in a cellular network.

[0198] 13. The electronic system of embodiment 12, wherein the first coaching model and the second coaching model are trained at a network device of the cellular network.

[0199] 14. An electronic system as described in Example 13, wherein the UE receives a training data configuration from the cellular network, and the training data configuration specifies the input data, the actual output result and / or the expected output result of the artificial intelligence model for transmission from the UE to the network device.

[0200] 15. An electronic system as described in Example 13, wherein the artificial intelligence model is configured to predict the optimal beam pair between the UE and the base station of the cellular network, and wherein: the actual output result includes the predicted beam pair output by the artificial intelligence model; and the expected output result includes the optimal beam pair with the optimal beam measurement result.

[0201] 16. An electronic system as in Example 5, wherein the electronic system is associated with a vehicle and the artificial intelligence model is configured to generate driving decisions for the vehicle, and wherein: the input data includes at least sensory data associated with the vehicle; the actual output result includes the autonomous driving decision generated by the artificial intelligence model based on the environmental sensory data; and the expected output result includes the manual driving decision made by the vehicle driver.

[0202] 17. An electronic system as in Example 5, wherein the electronic system is associated with a media device, and the artificial intelligence model is configured to generate a classification of media data of the media device, and wherein: the input data includes the media data; the actual output result includes the predicted classification generated by the artificial intelligence model based on the media data; and the expected output result includes the true classification of the media data.

[0203] 18. An electronic system comprising: at least one processing unit; and at least one storage unit, the at least one storage unit comprising computer program code, wherein the at least one storage unit and the computer program code are configured to cause the electronic system to perform the following operations through the at least one processing unit: send model data defining an artificial intelligence model to a user system; receive model difference data representing model differences from the user system; generate an updated version of the artificial intelligence model based at least in part on the model difference data; and send updated model data to the user system, the updated model data representing the updated version of the artificial intelligence model.

[0204] 19. The electronic system of embodiment 18, wherein the model difference data represents a difference between actual performance of the artificial intelligence model and expected performance of the artificial intelligence model.

[0205] 20. An electronic system as described in Example 19, wherein the model difference data represents a model difference between a first tutoring model and a second tutoring model, wherein the first tutoring model simulates the actual performance of the artificial intelligence model deployed on the user system, and wherein the second tutoring model simulates the expected performance of the artificial intelligence model deployed on the user system.

[0206] 21. The electronic system of embodiment 20, wherein the model difference data represents at least one of the following differences between the first coaching model and the second coaching model: a difference between one or more model parameters; a difference between one or more performance indicators; or a difference between one or more output results.

[0207] 22. An electronic system as described in Example 20, wherein generating the updated version of the artificial intelligence model includes one of the following operations: directly updating the artificial intelligence model based at least in part on the model difference data; or updating a larger-scale model based at least in part on the model difference data, and generating a smaller-scale model based on a reduction of the updated larger-scale model as the updated version of the artificial intelligence model.

[0208] 23. A method comprising: receiving model data representing an artificial intelligence model from a model management server; sending model difference data representing model differences to the model management server; and receiving updated model data from the model management server, wherein the updated model data represents an updated version of the artificial intelligence model, and the updated version is based at least in part on the model difference data.

[0209] 24. A method comprising: sending model data defining an artificial intelligence model to a user system; receiving model difference data representing model differences from the user system; generating an updated version of the artificial intelligence model based at least in part on the model difference data; and sending updated model data to the user system, wherein the updated model data represents the updated version of the artificial intelligence model.

[0210] 25. A computer-readable storage medium storing one or more instructions, wherein when the one or more instructions are executed by one or more processing circuits of an electronic device, the electronic device performs the method of any one of embodiments 23-24.

[0211] 26. A computer program product comprising a computer program, which, when executed by a processor, performs the method of any one of embodiments 23-24.

[0212] 27. An apparatus comprising means for performing the method of any one of embodiments 23-24.

Claims

1. An electronic system comprising: at least one processing unit; and at least one storage unit, the at least one storage unit comprising computer program code, wherein the at least one storage unit and the computer program code are configured to cause the electronic system to perform the following operations through the at least one processing unit: receiving model data representing an artificial intelligence model from a model management server; Sending model difference data representing model differences to a model management server; and Updated model data is received from the model management server, the updated model data representing an updated version of the artificial intelligence model, the updated version being based at least in part on the model difference data.

2. The electronic system as claimed in claim 1, wherein: The model difference data represents the difference between the actual performance of the artificial intelligence model and the expected performance of the artificial intelligence model.

3. The electronic system as claimed in claim 2, wherein: The difference is described by a difference between a first tutoring model and a second tutoring model, wherein the first tutoring model simulates an actual performance of the artificial intelligence model, and wherein the second tutoring model simulates an expected performance of the artificial intelligence model.

4. The electronic system as claimed in claim 3, wherein: The model difference data represents at least one of the following differences between the first tutoring model and the second tutoring model: differences between one or more model parameters; Differences between one or more performance indicators; or The difference between one or more output results.

5. The electronic system as claimed in claim 3, wherein: The at least one storage unit and the computer program code are further configured to cause the electronic system to perform the following operations through the at least one processing unit: training the first tutoring model based at least in part on the input data and actual output results of the artificial intelligence model; and The second tutoring model is trained based at least in part on the input data and desired output results of the artificial intelligence model.

6. The electronic system as claimed in claim 5, wherein: The first tutoring model and the second tutoring model are based on the same initial tutoring model.

7. The electronic system as claimed in claim 5, wherein: The first tutoring model and the second tutoring model have higher complexity than the artificial intelligence model.

8. The electronic system as claimed in claim 5, wherein: The training of the first tutoring model and the second tutoring model is initiated in response to the actual performance of the artificial intelligence model not meeting a specific standard.

9. The electronic system as claimed in claim 1, wherein: The model data characterizing the artificial intelligence model includes at least one of the following: model parameters of the artificial intelligence model; or An identifier for the artificial intelligence model.

10. The electronic system as claimed in claim 1, wherein: The artificial intelligence model is a smaller-scale model generated based on the reduction of a larger-scale model.

11. The electronic system as claimed in claim 10, wherein: The updated version is generated based on one of the following operations: directly updating the artificial intelligence model based at least in part on the model difference data; or The larger scale model is updated based at least in part on the model difference data, and the updated version of the artificial intelligence model is generated based on a reduction of the updated larger scale model.

12. The electronic system as claimed in claim 5, wherein: The artificial intelligence model is deployed at a user equipment UE in a cellular network.

13. The electronic system of claim 12, wherein: The first coaching model and the second coaching model are trained at a network device of the cellular network.

14. The electronic system of claim 13, wherein: The UE receives a training data configuration from the cellular network, the training data configuration specifying the input data, the actual output result and / or the expected output result of the artificial intelligence model for transmission from the UE to the network device.

15. The electronic system of claim 13, wherein: The artificial intelligence model is configured to predict an optimal beam pair between the UE and a base station of the cellular network, and wherein: The actual output result includes the predicted beam pair output by the artificial intelligence model; The desired output results include the best beam pair with the best beam measurement results.

16. The electronic system of claim 5, wherein: The electronic system is associated with a vehicle, and the artificial intelligence model is configured to generate driving decisions for the vehicle, and wherein: The input data includes at least sensory data associated with the vehicle; The actual output result includes an autonomous driving decision generated by the artificial intelligence model based on the environmental sensing data; The desired output results include manual driving decisions made by the vehicle driver.

17. The electronic system of claim 5, wherein: The electronic system is associated with a media device, and the artificial intelligence model is configured to generate a classification of media data of the media device, and wherein: The input data includes the media data; The actual output result includes a predicted classification generated by the artificial intelligence model based on the media data; The expected output result includes the true classification of the media data.

18. An electronic system comprising: at least one processing unit; and at least one storage unit, the at least one storage unit comprising computer program code, wherein the at least one storage unit and the computer program code are configured to cause the electronic system to perform the following operations through the at least one processing unit: Sending model data defining the artificial intelligence model to the user system; receiving model difference data representing a model difference from a user system; generating an updated version of the artificial intelligence model based at least in part on the model difference data; and Updated model data is sent to the user system, the updated model data representing the updated version of the artificial intelligence model.

19. The electronic system of claim 18, wherein: The model difference data represents the difference between the actual performance of the artificial intelligence model and the expected performance of the artificial intelligence model.

20. The electronic system of claim 19, wherein: The model difference data represents a model difference between a first tutoring model and a second tutoring model, wherein the first tutoring model simulates an actual performance of the artificial intelligence model deployed on the user system, and wherein the second tutoring model simulates an expected performance of the artificial intelligence model deployed on the user system.

21. The electronic system of claim 20, wherein: The model difference data represents at least one of the following differences between the first tutoring model and the second tutoring model: differences between one or more model parameters; Differences between one or more performance indicators; or The difference between one or more output results.

22. The electronic system of claim 20, wherein: Generating an updated version of the artificial intelligence model includes one of the following operations: directly updating the artificial intelligence model based at least in part on the model difference data; or The larger-scale model is updated based at least in part on the model difference data, and a smaller-scale model is generated based on a reduction of the updated larger-scale model as the updated version of the artificial intelligence model.

23. A method comprising: receiving model data representing an artificial intelligence model from a model management server; Sending model difference data representing model differences to a model management server; and Updated model data is received from the model management server, the updated model data representing an updated version of the artificial intelligence model, the updated version being based at least in part on the model difference data.

24. A method comprising: Sending model data defining the artificial intelligence model to the user system; receiving model difference data representing a model difference from a user system; generating an updated version of the artificial intelligence model based at least in part on the model difference data; and Updated model data is sent to the user system, the updated model data representing the updated version of the artificial intelligence model.

25. A computer-readable storage medium storing one or more instructions, which, when executed by one or more processing circuits of an electronic device, causes the electronic device to perform the method of any one of claims 23-24.

26. A computer program product comprising a computer program which, when executed by a processor, performs the method according to any one of claims 23 to 24.

27. An apparatus comprising means for performing the method of any one of claims 23-24.

Images