Model transmission method, terminal, base station and storage medium

Abstract

The embodiment of the invention discloses a model transmission method, a terminal, a base station and a storage medium. The model transmission method applied to the terminal comprises the following steps: receiving an NAS signaling sent by a core network; wherein the NAS signaling indicates whether the subsequently initiated transmission process is an artificial intelligence model transmission process or not; establishing a user plane connection used for bearing artificial intelligence model transmission between the terminal and the core network under the condition that the NAS signaling indicates that the artificial intelligence model transmission process is subsequently initiated; and receiving a first artificial intelligence model sent by the core network through the user plane connection.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of communication technology, and in particular to a model transmission method, a terminal, a base station and a storage medium. Background Art

[0002] At present, the mainstream use cases of physical layer artificial intelligence include channel state information compression feedback and beam management. The channel state information compression feedback use case involves a base station side model, which needs to be deployed on both the terminal and the base station side. The beam management use case involves a terminal side model, which needs to be deployed on the terminal. For artificial intelligence physical layer use cases, scenario generalization is an issue that must be faced. The use of pre-trained large models or directly trained large models with high generalization can be used to solve the scenario generalization problem.

[0003] Artificial intelligence models can be applied to a large number of base stations. In order to avoid a large amount of repeated training work when the model version is updated, it is possible to consider storing the model in the core network for unified training and updating. However, if the artificial intelligence model is stored in the core network, it is equivalent to making the core network undertake the task of model management. However, the actual deployment objects of the artificial intelligence model are base stations and terminals. There is a problem that the model management equipment and the model application equipment are inconsistent, making it difficult to implement model deployment. Summary of the invention

[0004] The embodiments of the present application provide a model transmission method, a terminal, a base station and a storage medium. The core network can trigger the terminal to establish a user plane connection for carrying the transmission of the artificial intelligence model through NAS signaling, so as to realize the model delivery process from the network side to the terminal side, and realize the model deployment when the model management device and the model application device are inconsistent.

[0005] The technical solution of the embodiment of the present application is implemented as follows:

[0006] The present application provides a model transmission method, which is applied to a terminal, and the method includes:

[0007] Receiving non-access layer NAS signaling sent by the core network; wherein the NAS signaling indicates whether the subsequent initiated process is an artificial intelligence model transmission process;

[0008] When the NAS signaling indicates that the subsequent initiation is an artificial intelligence model transmission process, establishing a user plane connection for carrying the artificial intelligence model transmission between the terminal and the core network;

[0009] The first artificial intelligence model sent by the core network is received through the user plane connection.

[0010] In the above method, it also includes:

[0011] The model version information of the currently stored second artificial intelligence model is provided to the core network, so that when the core network determines that the versions of the first artificial intelligence model and the second artificial intelligence model stored in the core network are different, the transmission of the artificial intelligence model to the terminal is triggered.

[0012] In the above method, after receiving the first artificial intelligence model sent by the core network, the method further includes:

[0013] When the model is received, a model download completion notification is sent to the base station.

[0014] In the above method, after receiving the first artificial intelligence model sent by the core network, the method further includes:

[0015] When the model is received, updated user context information is sent to the base station and the core network; wherein the user context information includes the model identifier of the first artificial intelligence model.

[0016] In the above method, the model identifier includes the model task identifier of the first artificial intelligence model, or the model task identifier and model version information of the first artificial intelligence model.

[0017] In the above method, after sending the updated user context information to the base station and the core network, the method further includes:

[0018] receiving a model processing signaling sent by the base station based on the model identifier; the model processing signaling is used to instruct the terminal whether to upload the first artificial intelligence model;

[0019] When the model processing signaling instructs the terminal to upload the first artificial intelligence model, synchronizing the first artificial intelligence model to the base station through a control plane connection established with the base station;

[0020] Receiving a model reception confirmation message sent by the base station;

[0021] Based on receiving a confirmation message from the model, the first artificial intelligence model is activated.

[0022] In the above method, after receiving the model processing signaling sent by the base station based on the model identifier, the method further includes:

[0023] When the model processing signaling instructs the terminal not to upload the first artificial intelligence model, the first artificial intelligence model is activated.

[0024] In the above method, after activating the first artificial intelligence model, the method further includes:

[0025] Sending a model activation confirmation message to the base station.

[0026] The present application provides a model transmission method, which is applied to a base station. The method includes:

[0027] Based on the model identifier of the first artificial intelligence model, determine whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model.

[0028] In the above method, the model identifier includes the model task identifier of the first artificial intelligence model, or the model task identifier and model version information of the first artificial intelligence model.

[0029] In the above method, after determining whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model based on the model identifier of the first artificial intelligence model, the method further includes:

[0030] According to the determination result, a model processing signaling is sent to the terminal; the model processing signaling is used to instruct the terminal whether to upload the first artificial intelligence model.

[0031] In the above method, after sending the model processing signaling to the terminal, the method further includes:

[0032] In a case where the model processing signaling instructs the terminal to upload the first artificial intelligence model, the first artificial intelligence model synchronized with the terminal is received through a control plane connection established with the terminal.

[0033] In the above method, after receiving the first artificial intelligence model synchronized by the terminal, the method further includes:

[0034] When the model is received, a model reception confirmation message is sent to the terminal, so that the terminal activates the first artificial intelligence model in the terminal based on the model reception confirmation message.

[0035] An embodiment of the present application provides a terminal, including: a first processor, a first memory, and a first communication bus;

[0036] The first communication bus is used to implement a communication connection between the first processor and the first memory;

[0037] The first processor is used to execute one or more computer programs stored in the first memory to implement the above-mentioned model transmission method applied to the terminal.

[0038] An embodiment of the present application provides a base station, including: a second processor, a second memory, and a second communication bus;

[0039] The second communication bus is used to realize the communication connection between the second processor and the second memory;

[0040] The second processor is used to execute one or more computer programs stored in the second memory to implement the above-mentioned model transmission method applied to the base station.

[0041] An embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned model transmission method is implemented.

[0042] The embodiment of the present application provides a model transmission method, terminal, base station and storage medium. The model transmission method applied to the terminal includes: receiving NAS signaling sent by the core network; wherein the NAS signaling indicates whether the subsequent process initiated is an artificial intelligence model transmission process; when the NAS signaling indicates that the subsequent process initiated is an artificial intelligence model transmission process, establishing a user plane connection for carrying the transmission of the artificial intelligence model between the terminal and the core network; receiving the first artificial intelligence model sent by the core network through the user plane connection. The technical solution provided by the embodiment of the present application is that the core network can trigger the terminal to establish a user plane connection for carrying the transmission of the artificial intelligence model through NAS signaling, so as to realize the model distribution process from the network side to the terminal side, and realize the model deployment when the model management device and the model application device are inconsistent. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of a model transmission method provided in an embodiment of the present application Figure 1 ;

[0044] Figure 2 A schematic diagram of a model transmission method provided in an embodiment of the present application Figure 2 ;

[0045] Figure 3 An exemplary schematic diagram of interaction between a base station and a terminal provided in an embodiment of the present application;

[0046] Figure 4 An exemplary interactive diagram of model transmission provided in the embodiment of the present application Figure 1 ;

[0047] Figure 5 An exemplary interactive diagram of model transmission provided in the embodiment of the present application Figure 2 ;

[0048] Figure 6 A schematic diagram of a terminal structure provided in an embodiment of the present application Figure 1 ;

[0049] Figure 7 A schematic diagram of a terminal structure provided in an embodiment of the present application Figure 2 ;

[0050] Figure 8 A schematic diagram of a base station structure provided in an embodiment of the present application Figure 1 ;

[0051] Fig. 9 A schematic diagram of a base station structure provided in an embodiment of the present application Figure 2 . DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0053] The following will specifically describe the technical solution of the present application and how the technical solution of the present application solves the above technical problems through embodiments and in conjunction with the accompanying drawings. The following embodiments may be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0054] In addition, the technical solutions described in the embodiments of the present application can be combined arbitrarily without conflict.

[0055] An embodiment of the present application provides a model transmission method, which is applied to a terminal. Figure 1 A schematic diagram of a model transmission method provided in an embodiment of the present application Figure 1 .like Figure 1 As shown, in an embodiment of the present application, the model transmission method applied to the terminal mainly includes the following steps:

[0056] S101. Receive NAS signaling sent by the core network; wherein the NAS signaling indicates whether the subsequent process initiated is an artificial intelligence model transmission process.

[0057] In an embodiment of the present application, the terminal can receive NAS signaling sent by the core network; wherein the NAS signaling indicates whether the subsequent initiated process is an artificial intelligence model transmission process.

[0058] It can be understood that in the embodiments of the present application, various artificial intelligence models can be stored in the core network, the core network can initiate an artificial intelligence model transmission process to the terminal, and the terminal can receive the NAS signaling sent by the core network, thereby knowing whether the subsequent process initiated by the core network is an artificial model transmission process according to the NAS signaling, and clarifying the subsequent task process to trigger the corresponding operation. Among them, the core network triggers the artificial intelligence model transmission process to the terminal, which is determined based on the artificial intelligence capability of the terminal.

[0059] In an embodiment of the present application, the terminal may also perform the following steps: provide the core network with model version information of the currently stored second artificial intelligence model, so that when the core network determines that the versions of the first artificial intelligence model and the second artificial intelligence model stored in the core network are different, it triggers the transmission of the artificial intelligence model to the terminal.

[0060] It should be noted that in the embodiment of the present application, the first artificial intelligence model is an artificial intelligence model stored in the core network, and the second artificial intelligence model is an artificial intelligence model currently stored in the terminal. The specific first artificial intelligence model and the second artificial intelligence model are not limited in the embodiment of the present application.

[0061] It should be noted that in the embodiments of the present application, the terminal can register artificial intelligence capabilities with the core network. Specifically, the terminal can register artificial intelligence capabilities at the authentication management function (AMF) in the core network, and the core network triggers the artificial intelligence model transmission process through AMF. Among them, the terminal can supplement the registration of artificial intelligence capabilities when registering with AMF, mainly to synchronize the model version under each model task identity (IdentityDocument, ID). In this way, when AMF detects that the model version is inconsistent, it will trigger the artificial intelligence model transmission process. Based on this, the terminal will receive NAS signaling sent from the core network indicating that the subsequent initiation is the artificial intelligence model transmission process. Of course, when AMF detects that the model version is consistent, the artificial intelligence model transmission process will not be triggered. The NAS signaling received by the terminal may be related NAS signaling of other tasks. The NAS signaling indicates that the subsequent initiation is not an artificial intelligence model transmission process. Specifically, it can indicate the process of other tasks, which is not limited by the embodiments of the present application.

[0062] S102. When the NAS signaling indicates that the subsequent process to be initiated is an artificial intelligence model transmission process, a user plane connection is established for carrying the transmission of the artificial intelligence model between the terminal and the core network.

[0063] In an embodiment of the present application, after the terminal receives the NAS signaling sent by the core network, when the NAS signaling indicates that the subsequent process to be initiated is an artificial intelligence model transmission process, a user plane connection for carrying the transmission of the artificial intelligence model between the terminal and the core network is established.

[0064] It can be understood that in an embodiment of the present application, when the NAS signaling indicates that the subsequent process to be initiated is an artificial intelligence model transmission process, in order to receive the artificial intelligence model from the core network, the terminal needs to establish a user plane connection for carrying the transmission of the artificial intelligence model between the terminal and the core network.

[0065] It should be noted that in an embodiment of the present application, the core network can trigger the transmission of the artificial intelligence model to the terminal through AMF. AMF transmits the address of the model to be downloaded obtained from the application function (AF) to the terminal along with the NAS signaling. This NAS signaling not only indicates whether the subsequent initiation is an artificial intelligence model transmission process as in the above step S101, but also clarifies the subsequent establishment of user plane connections and other services for artificial intelligence model transmission, so as to activate the base station and the terminal to complete subsequent operations such as model ID judgment, model transmission and model activation. The terminal establishes a user plane connection according to the address of the model to be downloaded, and supports the core network to transmit the artificial intelligence model to the terminal through the user plane connection.

[0066] It should be noted that, in the embodiment of the present application, the user plane connection established by the terminal may specifically be a protocol data unit (PDU) session.

[0067] S103. Receive the first artificial intelligence model sent by the core network through the user plane connection.

[0068] In an embodiment of the present application, after establishing a user plane connection, the terminal can receive the first artificial intelligence model sent by the core network through the user plane connection.

[0069] In an embodiment of the present application, after the terminal receives the first artificial intelligence model sent by the core network, it can also perform the following steps: when the model is received, send a model download completion notification to the base station.

[0070] It can be understood that in an embodiment of the present application, after the terminal detects that the model has been received, it will send a model download completion notification to the base station. This is because, in actual applications, both the terminal side model and the base station side model require the cooperation of the base station, so it is necessary to inform the base station whether the model has been transmitted.

[0071] In an embodiment of the present application, after the terminal receives the first artificial intelligence model sent by the core network, it can also perform the following steps: when the model is received, send updated user context information to the base station and the core network; wherein the user context information includes the model identifier of the first artificial intelligence model.

[0072] It should be noted that, in an embodiment of the present application, the model identifier includes the model task identifier of the first artificial intelligence model, or the model task identifier and model version information of the first artificial intelligence model.

[0073] It can be understood that in an embodiment of the present application, the terminal also needs to update the model version information registered by the terminal in the AMF of the core network to ensure the accuracy of subsequent model transmission decisions.

[0074] In an embodiment of the present application, after the terminal sends updated user context information to the base station and the core network, the following steps may also be performed: receiving model processing signaling sent by the base station based on the model identifier; the model processing signaling is used to indicate whether the terminal uploads the first artificial intelligence model; when the model processing signaling instructs the terminal to upload the first artificial intelligence model, the first artificial intelligence model is synchronized to the base station through the control plane connection established with the base station; receiving a model reception confirmation message sent by the base station; and activating the first artificial intelligence model based on the model reception confirmation message.

[0075] It can be understood that in the embodiment of the present application, the terminal sends updated user uplink information to the base station, and the base station can determine whether the first artificial intelligence model is a terminal side model or a base station side model based on the model identifier of the first artificial intelligence model in the user context information, and determine whether the first artificial intelligence model needs to be obtained, thereby sending a model processing signaling to the terminal based on the determination result, and instructing the terminal whether to upload the first artificial intelligence model through the model processing signaling. When the terminal receives the model processing signaling, it can perform corresponding operations according to the model processing signaling.

[0076] It should be noted that in the embodiment of the present application, a control plane connection is established between the terminal and the base station, and the terminal can synchronize the first artificial intelligence model to the base station through the control plane connection. After receiving the model, the base station can send a model reception confirmation message to the terminal, and the terminal reactivates the first artificial intelligence model based on the model reception confirmation message, and then the first artificial intelligence model can provide services for the communication system normally.

[0077] In an embodiment of the present application, after the terminal receives the model processing signaling sent by the base station based on the model identifier, the following steps may also be performed: when the model processing signaling indicates that the terminal does not upload the first artificial intelligence model, activate the first artificial intelligence model.

[0078] It can be understood that in an embodiment of the present application, the first artificial intelligence model may be a terminal side model. In this regard, the model processing signaling received by the terminal instructs the terminal not to upload the first artificial intelligence model. The terminal can directly activate the first artificial intelligence model, thereby providing services for the communication system normally.

[0079] In an embodiment of the present application, after the terminal activates the first artificial intelligence model, the following steps may also be performed: sending a model activation confirmation message to the base station so that the base station knows that the first artificial intelligence model has entered the use stage.

[0080] The embodiment of the present application provides a model transmission method. For the scenario where the model management device (core network) of the artificial intelligence model is inconsistent with the model application device (terminal), the core network can indicate through NAS signaling whether the subsequent initiation is an artificial intelligence model transmission process, so that when the process of initiating the artificial intelligence model transmission is initiated, the terminal can promptly establish a user plane connection that carries the artificial intelligence model transmission, thereby receiving the first artificial intelligence model sent by the core network through the user plane connection, realizing the model distribution process from the network side to the terminal side, and realizing the model deployment when the model management device and the model application device are inconsistent. In addition, for the artificial intelligence model as a base station side model, it can be synchronized to the base station by the terminal through the control plane connection between the terminal and the base station for base station deployment, solving the problem that the user plane transmission has no anchor point at the base station and cannot be directly transmitted.

[0081] An embodiment of the present application provides a model transmission method, which is applied to a base station. Figure 2 A schematic diagram of a model transmission method provided in an embodiment of the present application Figure 2 .like Figure 2 As shown, in an embodiment of the present application, the model transmission method applied to the base station mainly includes the following steps:

[0082] S201. Based on a model identifier of the first artificial intelligence model, determine whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model.

[0083] In an embodiment of the present application, the base station may determine whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model based on the model identifier of the first artificial intelligence model.

[0084] It should be noted that in an embodiment of the present application, as described in the model transmission method on the terminal side, when the terminal receives the first artificial intelligence model sent by the core network and the model is received, the terminal sends updated user context information to the base station, which includes the model identifier of the first artificial intelligence model. Based on this, the base station can determine whether to obtain the first artificial intelligence model based on the model identifier of the first artificial intelligence model.

[0085] It should be noted that, in an embodiment of the present application, the model identifier includes the model task identifier of the first artificial intelligence model, or the model task identifier and model version information of the first artificial intelligence model.

[0086] It can be understood that in the embodiments of the present application, the model identifier of the first artificial intelligence model includes at least the model task identifier of the first artificial intelligence model, and of course, may also include the model version information of the first artificial intelligence model. The base station can determine whether the first artificial intelligence model is a terminal side model or a base station side model based on this information. If it is a base station side model, it can be determined that the first artificial intelligence model needs to be obtained from the terminal.

[0087] In an embodiment of the present application, after the base station determines whether to obtain the first artificial intelligence model from the terminal storing the first artificial intelligence model based on the model identifier of the first artificial intelligence model, it can also perform the following steps: sending a model processing signaling to the terminal based on the determination result; the model processing signaling is used to indicate whether the terminal uploads the first artificial intelligence model.

[0088] In an embodiment of the present application, after the base station sends the model processing signaling to the terminal, the following steps may also be performed: when the model processing signaling instructs the terminal to upload the first artificial intelligence model, the first artificial intelligence model synchronized with the terminal is received through a control plane connection established with the terminal.

[0089] In an embodiment of the present application, after the base station receives the first artificial intelligence model synchronized by the terminal, the following steps may also be performed: when the model is received, a model reception confirmation message is sent to the terminal, so that the terminal activates the first artificial intelligence model in the terminal based on the model reception confirmation message.

[0090] It can be understood that in an embodiment of the present application, if the first artificial intelligence model is a base station side model, the base station needs to instruct the terminal to activate the model after receiving the model. If the first artificial intelligence model is a terminal side model, the model processing signaling instructs the terminal not to upload the first artificial intelligence model. In this case, the terminal can directly activate the model.

[0091] It can be understood that in the embodiment of the present application, the artificial intelligence model in the current physical layer use case is mainly deployed on a single side of the terminal or between the terminal and the base station. After the first artificial intelligence model is transmitted from the core network to the terminal through the user plane, the base station side model also needs to be transmitted to the base station. Whether the first artificial intelligence model received by the terminal is a base station side model needs to be determined by the base station according to the model identifier. The base station can obtain the model identifier through the synchronization process of the user context information. When the model identifier is identified as a base station side model, since the user plane transmission has no anchor point in the base station, the terminal will use the control plane connection transmission when uploading the model to the base station. Considering that the amount of model data transmitted is significantly higher than the control signaling transmitted by the current control plane, the segmentation mechanism in the existing control plane transmission is reused. When all segmented transmissions are completed, the base station activates the first artificial intelligence model on the base station side and sends a model reception confirmation message to the terminal. After receiving the confirmation message, the terminal activates the first artificial intelligence model on the terminal side and sends a model activation confirmation message to the base station. After that, the first artificial intelligence model will be able to provide services for the communication system normally; when the model identifier is identified as a terminal side model, the model processing signaling sent by the base station to the terminal is used to instruct the terminal not to upload the first artificial intelligence model. The terminal can directly activate the model and send a model activation confirmation message to the base station.

[0092] In an embodiment of the present application, a multi-user scenario is considered. For example, if each terminal uploads the model to the base station after receiving the base station side model, it may cause data transmission redundancy. Because for terminals connected to the same base station, the artificial intelligence model used for the same artificial intelligence service use case is likely to be the same. Therefore, as long as one terminal uploads the latest version of the base station side model to the base station, there is no need to upload it again. Therefore, when determining whether to obtain the first artificial intelligence model in step S201, the base station can further determine whether the first artificial intelligence model has been stored when determining whether the first artificial intelligence model is a terminal side model or a base station side model. If it has been stored, it is determined that the first artificial intelligence model will no longer be obtained.

[0093] It should be noted that in an embodiment of the present application, the base station can determine the model processing signaling to be fed back to the terminal based on the previously synchronized user context information with the model identifier. If it is not an updated version, the model processing signaling sent by the base station to the terminal instructs the terminal not to upload the first artificial intelligence model. After receiving the notification, the terminal activates the model on this side and sends a model activation confirmation to the base station to start the artificial intelligence service (at this time, the model on the base station side is already in an activated state and does not need to be activated again); if it is an updated version, the first artificial intelligence model is synchronized to the base station through the control plane, and then the base station activates the model and sends a model reception confirmation notification to the terminal. After receiving the model reception confirmation notification, the terminal activates the first artificial intelligence model on the terminal side and sends a model activation confirmation message to the base station, thereby starting the artificial intelligence service based on the first artificial intelligence model.

[0094] It should be noted that, in the embodiments of the present application, the model version information in the model identifier is not limited to the pre-set model version number, but may also be related to the weight parameters of the model itself. Considering that the models sent from the core network to the base station are generally updated models, it is only necessary to distinguish whether the model weights are different. It can be considered to use a hash algorithm to calculate the hash value at the terminal, and use the hash value as the model version information. When the hash value is inconsistent, the base station believes that the terminal has received an updated version of the model, thereby triggering the transmission of the model from the terminal to the base station.

[0095] Figure 3 The following is a schematic diagram of an exemplary interaction between a base station and a terminal provided in an embodiment of the present application. Figure 3 As shown, it mainly includes the following steps:

[0096] S301. The base station identifies a model identifier of a first artificial intelligence model and determines whether to acquire the first artificial intelligence model;

[0097] S302, the base station sends a model processing instruction to the terminal;

[0098] S303. When the model processing instruction instructs the terminal to upload the first artificial intelligence model, the terminal synchronizes the first artificial intelligence model to the base station through the control plane connection;

[0099] S304, the base station receives and activates the first artificial intelligence model;

[0100] S305, the base station sends a model reception confirmation message to the terminal;

[0101] S306. The terminal activates the first artificial intelligence model based on the model reception confirmation message;

[0102] S307: The terminal sends a model activation confirmation message to the base station.

[0103] It can be understood that in the embodiments of the present application, the characteristics of control plane and user plane transmission are combined, and a model transmission method is provided on the basis of avoiding major modifications to the protocol stack architecture. Control plane transmission is used for the parts that are difficult to reuse, and a transmission mechanism that can minimize the control plane transmission frequency is designed in combination with the characteristics of artificial intelligence model transmission on the base station side. The artificial intelligence model transmission from the core network to the base station and the terminal is completed by utilizing the existing protocol stack architecture while minimizing transmission redundancy.

[0104] The following combination Figure 4 and Figure 5 The deployment of the terminal side model and the base station side model are respectively illustrated as examples, wherein the core network includes AMF, AF and user plane function (UPF).

[0105] Figure 4 An exemplary interactive diagram of model transmission provided in the embodiment of the present application Figure 1 .like Figure 4 As shown, in the embodiment of the present application, the following steps are mainly included:

[0106] S401. The terminal registers the artificial intelligence capability in the AMF and stores the current artificial intelligence model version of the terminal in the user context of the AMF;

[0107] S402, AMF finds that the terminal's artificial intelligence model version is updated, triggering the model download process;

[0108] S403, AMF initiates model addressing to AF;

[0109] S404, AF feeds back the model address to AMF and confirms it;

[0110] S405. AMF notifies the base station and the terminal to start model download through downlink NAS signaling, and carries the AF address of the storage model;

[0111] S406, the terminal initiates a model download PDU session;

[0112] S407, a user plane connection is established between the terminal, the base station, the UPF and the AF (a new 5G quality of service (QoS) identifier 5QI needs to be defined for the model transmission task);

[0113] S408. Transmitting the artificial intelligence model from the AF to the terminal through the established user plane connection;

[0114] S409: The terminal detects the model end mark and determines that the model download is complete;

[0115] S410, the terminal sends a model download completion notification (including model version information) to the base station;

[0116] S411. The terminal and AMF update user context information (including model version information) through NAS signaling;

[0117] S412: The base station identifies the model identifier and determines that the artificial intelligence model is a terminal side model, or determines that the artificial intelligence model is a base station side model, and the base station has stored the artificial intelligence model;

[0118] S413, the base station sends a message to the terminal to start activating the artificial intelligence model;

[0119] S414: The terminal activates the artificial intelligence model and sends a model activation confirmation message to the base station.

[0120] Figure 5An exemplary interactive diagram of model transmission provided in the embodiment of the present application Figure 2 .like Figure 5 As shown, in the embodiment of the present application, the steps mainly include the following steps:

[0121] S501. The terminal registers the artificial intelligence capability in the AMF and stores the current artificial intelligence model version of the terminal in the user context of the AMF;

[0122] S502, AMF finds that the artificial intelligence model version of the terminal is updated, triggering the model download process;

[0123] S503, AMF initiates model addressing to AF;

[0124] S504, AF feeds back the model address to AMF and confirms it;

[0125] S505. AMF notifies the base station and the terminal to start model download through downlink NAS signaling, and carries the AF address of the storage model;

[0126] S506, the terminal initiates a model download PDU session;

[0127] S507: Establish a user plane connection between the terminal, the base station, the UPF and the AF;

[0128] S508. Transmitting the artificial intelligence model from the AF to the terminal through the established user plane connection;

[0129] S509: The terminal detects the model end mark and determines that the model download is complete;

[0130] S510, the terminal sends a model download completion notification (including model version information) to the base station;

[0131] S511. The terminal and AMF update user context information (including model version information) through NAS signaling;

[0132] S512: The base station identifies the model identifier and determines that the artificial intelligence model is a base station side model, and the base station has not stored the artificial intelligence model;

[0133] S513. The base station sends a message to the terminal to inform that the transmission of the artificial intelligence model to the base station needs to be started;

[0134] S514. The terminal transmits the artificial intelligence model to the base station through the control plane connection;

[0135] S515: The base station receives the artificial intelligence model and activates it, and sends a model reception confirmation message to the terminal so that the terminal activates the artificial intelligence model;

[0136] S516: The terminal sends a model activation confirmation message to the base station.

[0137] An embodiment of the present application provides a terminal. Figure 6 A schematic diagram of a terminal structure provided in an embodiment of the present application Figure 1 .like Figure 6 As shown, the terminal includes:

[0138] The first communication module 601 is used to receive a non-access layer NAS signaling sent by a core network; wherein the NAS signaling indicates whether the subsequent initiated process is an artificial intelligence model transmission process;

[0139] A connection establishment module 602 is configured to establish a user plane connection for carrying the transmission of the artificial intelligence model between the terminal and the core network when the NAS signaling indicates that the subsequent initiation is an artificial intelligence model transmission process;

[0140] The first communication module 601 is also used to receive the first artificial intelligence model sent by the core network through the user plane connection.

[0141] In one embodiment of the present application, the first communication module 601 is also used to provide the core network with model version information of the currently stored second artificial intelligence model, so that when the core network determines that the versions of the first artificial intelligence model and the second artificial intelligence model stored in the core network are different, it triggers the transmission of the artificial intelligence model to the terminal.

[0142] In an embodiment of the present application, the first communication module 601 is further configured to send a model download completion notification to the base station when the model is received.

[0143] In one embodiment of the present application, the first communication module 601 is also used to send updated user context information to the base station and the core network when the model is received; wherein the user context information includes the model identifier of the first artificial intelligence model.

[0144] In one embodiment of the present application, the model identifier includes a model task identifier of the first artificial intelligence model, or a model task identifier and model version information of the first artificial intelligence model.

[0145] In an embodiment of the present application, the terminal further includes a model activation module 603 (not shown in the figure);

[0146] The first communication module 601 is further configured to receive a model processing signaling sent by the base station based on the model identifier; the model processing signaling is used to indicate whether the terminal uploads the first artificial intelligence model; when the model processing signaling indicates that the terminal uploads the first artificial intelligence model, synchronize the first artificial intelligence model to the base station through a control plane connection established with the base station; and receive a model reception confirmation message sent by the base station;

[0147] The model activation module 603 is used to activate the first artificial intelligence model based on receiving a confirmation message from the model.

[0148] In one embodiment of the present application, the model activation module 603 is further used to activate the first artificial intelligence model when the model processing signaling instructs the terminal not to upload the first artificial intelligence model.

[0149] In an embodiment of the present application, the first communication module 601 is further configured to send a model activation confirmation message to the base station.

[0150] Figure 7 A schematic diagram of a terminal structure provided in an embodiment of the present application Figure 2 .like Figure 7 As shown, the terminal includes: a first processor 701, a first memory 702 and a first communication bus 703;

[0151] The first communication bus 703 is used to implement a communication connection between the first processor 701 and the first memory 702;

[0152] The first processor 701 is used to execute one or more computer programs stored in the first memory 702 to implement the above-mentioned model transmission method applied to the terminal.

[0153] An embodiment of the present application provides a base station. Figure 8 A schematic diagram of a base station structure provided in an embodiment of the present application Figure 1 .like Figure 8 As shown, the base station includes:

[0154] The determination module 801 is used to determine whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model based on the model identifier of the first artificial intelligence model.

[0155] In one embodiment of the present application, the model identifier includes a model task identifier of the first artificial intelligence model, or a model task identifier and model version information of the first artificial intelligence model.

[0156] In an embodiment of the present application, the base station further includes a second communication module 802 (not shown in the figure);

[0157] The determination module 801 is also used to send a model processing signaling to the terminal through the second communication module 802 according to the determination result; the model processing signaling is used to indicate whether the terminal uploads the first artificial intelligence model.

[0158] In one embodiment of the present application, the second communication module 802 is further used to receive the first artificial intelligence model synchronized by the terminal through a control plane connection established with the terminal when the model processing signaling instructs the terminal to upload the first artificial intelligence model.

[0159] In one embodiment of the present application, the second communication module 802 is also used to send a model reception confirmation message to the terminal when the model reception is completed, so that the terminal activates the first artificial intelligence model in the terminal based on the model reception confirmation message.

[0160] Fig. 9 A schematic diagram of a base station structure provided in an embodiment of the present application Figure 2 .like Fig. 9 As shown, the base station includes: a second processor 901, a second memory 902 and a second communication bus 903;

[0161] The second communication bus 903 is used to realize the communication connection between the second processor and the second memory;

[0162] The second processor is used to execute one or more computer programs stored in the second memory to implement the model transmission method applied to the base station.

[0163] The embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and the computer program implements the above-mentioned model transmission method when executed by a processor. The computer-readable storage medium may be a volatile memory (volatile memory), such as a random access memory (Random-Access Memory, RAM); or a non-volatile memory (non-volatile memory), such as a read-only memory (Read-Only Memory, ROM), a flash memory (flash memory), a hard disk (Hard Disk Drive, HDD) or a solid-state drive (Solid-State Drive, SSD); or may be a respective device including one or any combination of the above-mentioned memories, such as a mobile phone, a computer, a tablet device, a personal digital assistant, etc.

[0164] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of hardware embodiments, software embodiments, or embodiments in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.

[0165] The present application is described with reference to implementation flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0166] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which is implemented in the implementation flow diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0167] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing the steps in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0168] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed in the utility application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A model transmission method, characterized in that: Applied to a terminal, the method comprises: Receiving non-access layer NAS signaling sent by the core network; wherein the NAS signaling indicates whether the subsequent initiated process is an artificial intelligence model transmission process; When the NAS signaling indicates that the subsequent initiation is an artificial intelligence model transmission process, establishing a user plane connection for carrying the artificial intelligence model transmission between the terminal and the core network; The first artificial intelligence model sent by the core network is received through the user plane connection.

2. The method according to claim 1, characterized in that The method further comprises: The model version information of the currently stored second artificial intelligence model is provided to the core network, so that when the core network determines that the versions of the first artificial intelligence model and the second artificial intelligence model stored in the core network are different, the transmission of the artificial intelligence model to the terminal is triggered.

3. The method according to claim 1, characterized in that After receiving the first artificial intelligence model sent by the core network, the method further includes: When the model is received, a model download completion notification is sent to the base station.

4. The method according to claim 1, characterized in that After receiving the first artificial intelligence model sent by the core network, the method further includes: When the model is received, updated user context information is sent to the base station and the core network; wherein the user context information includes the model identifier of the first artificial intelligence model.

5. The method according to claim 4, characterized in that The model identifier includes a model task identifier of the first artificial intelligence model, or a model task identifier and model version information of the first artificial intelligence model.

6. The method according to claim 4, characterized in that After sending the updated user context information to the base station and the core network, the method further includes: receiving a model processing signaling sent by the base station based on the model identifier; the model processing signaling is used to instruct the terminal whether to upload the first artificial intelligence model; When the model processing signaling instructs the terminal to upload the first artificial intelligence model, synchronizing the first artificial intelligence model to the base station through a control plane connection established with the base station; Receiving a model reception confirmation message sent by the base station; Based on receiving a confirmation message from the model, the first artificial intelligence model is activated.

7. The method according to claim 6, characterized in that After receiving the model processing signaling sent by the base station based on the model identifier, the method further includes: When the model processing signaling instructs the terminal not to upload the first artificial intelligence model, the first artificial intelligence model is activated.

8. The method according to claim 6 or 7, characterized in that: After activating the first artificial intelligence model, the method further includes: Sending a model activation confirmation message to the base station.

9. A model transmission method, characterized in that: Applied to a base station, the method comprises: Based on the model identifier of the first artificial intelligence model, determine whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model.

10. The method according to claim 9, characterized in that The model identifier includes a model task identifier of the first artificial intelligence model, or a model task identifier and model version information of the first artificial intelligence model.

11. The method according to claim 9 or 10, characterized in that: After determining whether to obtain the first artificial intelligence model from a terminal storing the first artificial intelligence model based on the model identifier of the first artificial intelligence model, the method further includes: According to the determination result, a model processing signaling is sent to the terminal; the model processing signaling is used to instruct the terminal whether to upload the first artificial intelligence model.

12. The method according to claim 11, characterized in that After sending the model processing signaling to the terminal, the method further includes: In a case where the model processing signaling instructs the terminal to upload the first artificial intelligence model, the first artificial intelligence model synchronized with the terminal is received through a control plane connection established with the terminal.

13. The method according to claim 12, characterized in that After receiving the first artificial intelligence model synchronized by the terminal, the method further includes: When the model is received, a model reception confirmation message is sent to the terminal, so that the terminal activates the first artificial intelligence model in the terminal based on the model reception confirmation message.

14. A terminal, characterized in that: include: a first processor, a first memory, and a first communication bus; The first communication bus is used to implement a communication connection between the first processor and the first memory; The first processor is used to execute one or more computer programs stored in the first memory to implement the model transmission method according to any one of claims 1 to 8.

15. A base station, characterized in that: include: a second processor, a second memory, and a second communication bus; The second communication bus is used to realize the communication connection between the second processor and the second memory; The second processor is used to execute one or more computer programs stored in the second memory to implement the model transmission method described in any one of claims 9 to 13.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the model transmission method according to any one of claims 1 to 13 is implemented.

Images