Model parameter transmission methods, apparatus, terminal and network side device

Abstract

Disclosed in the present application are model parameter transmission methods, an apparatus, a terminal and a network side device, belonging to the technical field of wireless communications. A model parameter transmission method of the embodiments of the present application comprises: a first communication device determines model information of at least one first sub-model that requires an update of model parameters from amongst a plurality of sub-models of a target artificial intelligence (AI) model, the target AI model comprising the plurality of sub-models, and each sub-model comprising at least part of separate model information; and the first communication device transmits the model information of the at least one first sub-model to a second communication device, the model information comprising sub-model parameters of said first sub-model.

Description

Model parameter transmission method and apparatus, terminal, and network-side device

[0001] Cross-reference to Related Applications

[0002] This application claims priority to the Chinese Patent Application No. 202410984798.1, filed on July 22, 2024, and entitled "Model Parameter Transmission Method, Apparatus, Terminal, and Network-Side Device", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0003] The present application belongs to the technical field of wireless communication, and specifically relates to a model parameter transmission method, apparatus, terminal, and network-side device. BACKGROUND

[0004] Artificial intelligence (AI) is currently widely used in various fields. Integrating artificial intelligence into wireless communication networks significantly improves technical indicators such as throughput, latency, and user capacity, which is an important task for future wireless communication networks. An AI unit or AI model can be implemented through a neural network, and the parameters of the neural network can be optimized through a gradient optimization algorithm.

[0005] In related technologies, the transmission or update of an AI model is mainly based on the entire model. However, the overhead of transmitting or updating the entire model is large. SUMMARY

[0006] Embodiments of the present application provide a model parameter transmission method, apparatus, terminal, and network-side device, which can solve the problem of large overhead of model transmission or update.

[0007] In a first aspect, a model parameter transmission method is provided, including: a first communication device determining model information of at least one first sub-model in a plurality of sub-models of a target artificial intelligence (AI) model that needs to update model parameters, wherein the target AI model includes a plurality of sub-models, and each sub-model has at least partially separate model information; and the first communication device transmitting the model information of the at least one first sub-model to a second communication device, wherein the model information includes sub-model parameters of the first sub-model.

[0008] In a second aspect, a model parameter transmission method is provided, including: receiving, by a second communication device, model information of at least one first sub-model transmitted by a first communication device, wherein the at least one first sub-model is a sub-model of a plurality of sub-models of a target artificial intelligence (AI) model that needs to update a model parameter, the target AI model includes a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information includes a sub-model parameter of the first sub-model; and updating, by the second communication device, the model parameter of the at least one first sub-model based on the model information of the at least one first sub-model.

[0009] In a third aspect, a model parameter transmission apparatus is provided, including: a processing module configured to determine model information of at least one first sub-model of a plurality of sub-models of a target artificial intelligence (AI) model that needs to update a model parameter, wherein the target AI model includes a plurality of sub-models, each of the sub-models has at least partially separate model information; and a sending module configured to transmit the model information of the at least one first sub-model to a second communication device, wherein the model information includes a sub-model parameter of the first sub-model.

[0010] In a fourth aspect, a model parameter transmission apparatus is provided, including: a receiving module configured to receive model information of at least one first sub-model transmitted by a first communication device, wherein the at least one first sub-model is a sub-model of a plurality of sub-models of a target artificial intelligence (AI) model that needs to update a model parameter, the target AI model includes a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information includes a sub-model parameter of the first sub-model; and a processing module configured to update the model parameter of the at least one first sub-model based on the model information of the at least one first sub-model.

[0011] In a fifth aspect, a model parameter transmission apparatus is provided, which is configured to perform the steps of the method according to the first aspect or implement the steps of the method according to the second aspect.

[0012] In a sixth aspect, a communication device is provided, including a processor and a memory, the memory stores programs or instructions executable on the processor, and the programs or instructions are executed by the processor to implement the steps of the method according to the first aspect or implement the steps of the method according to the second aspect.

[0013] In a seventh aspect, a communication device is provided, including a processor and a communication interface, wherein the processor is configured to implement the steps of the method according to the first aspect or implement the steps of the method according to the second aspect.

[0014] In an eighth aspect, a readable storage medium is provided, and the readable storage medium stores a program or instructions, which, when executed by a processor, implement the steps of the method according to the first aspect or implement the steps of the method according to the second aspect.

[0015] In a ninth aspect, a wireless communication system is provided, and the wireless communication system includes a first communication device and a second communication device, the first communication device is configured to implement the steps of the method according to the first aspect, and the second communication device is configured to implement the steps of the method according to the second aspect.

[0016] In a tenth aspect, a chip is provided, and the chip includes a processor and a communication interface, the communication interface is coupled to the processor, and the processor is configured to run a program or instructions to implement the method according to the first aspect or implement the method according to the second aspect.

[0017] In an eleventh aspect, a computer program / program product is provided, and the computer program / program product is stored in a storage medium, and the computer program / program product is executed by at least one processor to implement the method according to the first aspect or implement the method according to the second aspect.

[0018] In the embodiments of the present application, in the case where it is necessary to update the model parameters of at least one of the plurality of sub-models of the target AI model, the model information of the at least one first sub-model is acquired, and the model information of the at least one first sub-model is transmitted to the second communication device, without the need to transmit all the model information of the target AI model, thereby reducing the transmission overhead and improving the transmission efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0019] FIG. 1 shows a block diagram of a wireless communication system to which the embodiments of the present application can be applied;

[0020] FIG. 2 shows a structural diagram of a neural network;

[0021] FIG. 3 shows a flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0022] FIG. 4 shows another flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0023] FIG. 5 shows another flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0024] FIG. 6 shows another flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0025] FIG. 7 shows another flow diagram of another model parameter transmission method according to an embodiment of the present application;

[0026] FIG. 8 shows a flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0027] FIG. 9 shows another flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0028] FIG. 10 shows still another flow diagram of a model parameter transmission method according to an embodiment of the present application;

[0029] FIG. 11 shows a structural diagram of a model parameter transmission apparatus according to an embodiment of the present application;

[0030] FIG. 12 shows another structural diagram of a model parameter transmission apparatus according to an embodiment of the present application;

[0031] FIG. 13 shows a structural diagram of a communication device according to an embodiment of the present application;

[0032] FIG. 14 shows a hardware structural diagram of a terminal according to an embodiment of the present application;

[0033] FIG. 15 shows a hardware structural diagram of a network-side device according to an embodiment of the present application;

[0034] FIG. 16 shows another hardware structural diagram of a network-side device according to an embodiment of the present application. DETAILED DESCRIPTION

[0035] The technical solutions in the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present application.

[0036] The terms "first", "second", and the like in the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than that illustrated or described herein, and the objects distinguished by "first", "second" are usually a category, not limited to the number of objects, for example, the first object can be one or more. In addition, "or" in the present application means at least one of the connected objects. For example, the protection scope of "A or B" at least covers three schemes, namely, scheme one: including A and not including B; scheme two: including B and not including A; scheme three: including A and B. In addition, the terms "A and / or B", "at least one of A and B", "at least one of A or B" also at least cover the above three schemes, respectively. The character " / " generally represents that the objects before and after are in an "or" relationship.

[0037] The term "indication" in this application can be a direct indication (or explicit indication) or an indirect indication (or implicit indication). The direct indication can be understood as that the sender explicitly informs the receiver of specific information, operations to be performed or requested results, etc. in the sent indication. The indirect indication can be understood as that the receiver determines the corresponding information according to the indication sent by the sender, or judges and determines the operations to be performed or the requested results according to the judgment result.

[0038] It is worth noting that the technology described in the embodiments of the present application is not limited to the Long Term Evolution (LTE) / LTE-Advanced (LTE-A) system, but can also be used in other wireless communication systems, such as Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), Orthogonal Frequency Division Multiple Access (OFDMA), Single-carrier Frequency-Division Multiple Access (SC-FDMA) or other systems. The terms "system" and "network" in the embodiments of the present application are often used interchangeably, and the described technology can be used in the above-mentioned systems and radio technologies, and also in other systems and radio technologies. The following description describes a New Radio (NR) system for example purposes, and NR terminology is used in most of the following description, but these technologies can also be applied to systems other than the NR system, such as a 6th Generation (6G) communication system. th

[0039] ​FIG. 1 shows a block diagram of a wireless communication system to which embodiments of the present application can be applied. The wireless communication system includes a terminal 11 and a network-side device 12. The terminal 11 can be a terminal-side device such as a mobile phone, a Tablet Personal Computer, a Laptop Computer, a notebook computer, a Personal Digital Assistant (PDA), a palmtop computer, a netbook, an Ultra-mobile Personal Computer (UMPC), a Mobile Internet Device (MID), an Augmented Reality (AR) device, a Virtual Reality (VR) device, a robot, a wearable device, a flight vehicle, a Vehicle User Equipment (VUE), a shipboard device, a Pedestrian User Equipment (PUE), a smart home (a home device with a wireless communication function such as a refrigerator, a television, a washing machine, or furniture), a game console, a Personal Computer (PC), a kiosk, or a self-service machine. The wearable device includes a smart watch, a smart bracelet, a smart earphone, smart glasses, smart jewelry (a smart bracelet, a smart necklace, a smart ring, a smart necklace, a smart anklet, a smart necklace, etc.), a smart wristband, smart clothes, etc. The vehicle-mounted device can also be referred to as a vehicle-mounted terminal, a vehicle-mounted controller, a vehicle-mounted module, a vehicle-mounted component, a vehicle-mounted chip, or a vehicle-mounted unit, etc. It should be noted that the specific type of the terminal 11 is not limited in the embodiments of the present application. The network-side device 12 can include an access network device or a core network device. The access network device can also be referred to as a Radio Access Network (RAN) device, a radio access network function, or a radio access network unit. The access network device can include a base station, a Wireless Local Area Network (WLAN) Access Point (AP), or a Wireless Fidelity (WiFi) node, etc.The base station can be referred to as a Node B (NB), an evolved Node B (eNB), a next generation Node B (gNB), a New Radio Node B (NR Node B), an access point, a relay station (RBS), a serving base station (SBS), a base transceiver station (BTS), a radio base station, a radio transceiver, a basic service set (BSS), an extended service set (ESS), a home Node B (HNB), a home evolved Node B, a transmit / receive point (TRP), or some other suitable terminology in the art, and is not limited to a particular technical terminology, provided that the same technical effect is achieved. It should be noted that in the embodiments of the present application, only the base station in the NR system is taken as an example for introduction, and the specific type of the base station is not limited.

[0040] The core network device can also be referred to as a core network node, a core network function, or a core network network element, etc., which includes but is not limited to at least one of the following: a mobility management entity (MME), an access and mobility management function (AMF), a session management function (SMF), a user plane function (UPF), a policy control function (PCF), a policy and charging rules function (PCRF), an edge application server discovery function (EASDF), a unified data management (UDM), a unified data repository (UDR), a home subscriber server (HSS), a centralized network configuration (CNC), a network repository function (NRF), a network exposure function (NEF), a local NEF (L-NEF), a binding support function (BSF), an application function (AF), a location management function (LMF), a gateway mobile location center (GMLC), a network data analytics function (NWDAF), etc. It should be noted that only the core network device in the NR system is taken as an example for introduction in the embodiments of the present application, and the specific type of the core network device is not limited. If the name of the core network device mentioned in the embodiments of the present application changes in the subsequent protocol version (for example, 6G), it is also within the protection scope of the present application.

[0041] Optionally, the core network device can be implemented by one or more function modules in one device, or can be implemented by multiple devices together, and the embodiments of the present application do not make specific limitations. It can be understood that the above function modules can be network elements in a hardware device, software function modules running on a special hardware, or virtualized function modules instantiated on a platform (for example, a cloud platform).

[0042] Artificial intelligence (AI) is currently widely used in various fields. Integrating artificial intelligence into wireless communication networks significantly improves technical indicators such as throughput, latency, and user capacity, which is an important task for future wireless communication networks. AI modules have various implementation methods, such as neural networks, decision trees, support vector machines, and Bayesian classifiers. The present application takes neural networks as an example for illustration, but does not limit the specific type of AI module.

[0043] A schematic diagram of a neural network is shown in FIG. 2, where the neural network is composed of neurons, and a schematic diagram of a neuron is as follows. Where a1, a2, … a K are inputs, w is a weight (multiplicative coefficient), b is a bias (additive coefficient), and σ(.) is an activation function. Common activation functions include Sigmoid, tanh, ReLU (Rectified Linear Unit, linear rectification function, modified linear unit), etc.

[0044] The parameters of the neural network are optimized by a gradient optimization algorithm. Gradient optimization algorithms are a class of algorithms that minimize or maximize an objective function (sometimes also called a loss function), and the objective function is often a mathematical combination of model parameters and data. For example, given data X and its corresponding label Y, we construct a neural network model f(.). With the model, we can get the predicted output f(x) according to the input x, and we can calculate the difference between the predicted value and the true value (f(x)-Y), which is the loss function. Our goal is to find the appropriate W, b to minimize the value of the above loss function, and the smaller the loss value, the closer our model is to the true situation.

[0045] The common optimization algorithm is based on error back propagation (BP) algorithm. The basic idea of the BP algorithm is that the learning process consists of two processes of forward propagation of signals and backward propagation of errors. When the forward propagation is performed, the input sample is transmitted from the input layer to the output layer through the processing of each hidden layer. If the actual output of the output layer does not match the expected output, the backward propagation of errors is performed. The error back propagation is to transmit the output error to the input layer through the hidden layer in a certain form, and distribute the error to all units of each layer, so as to obtain the error signal of each unit, which is used as the basis for correcting the weight of each unit. The weight adjustment process of each layer is repeated. The process of continuously adjusting the weight is the learning and training process of the network. The process is performed until the error of the network output is reduced to an acceptable level, or the learning number reaches a preset value.

[0046] The common optimization algorithm includes gradient descent, stochastic gradient descent (SGD), mini-batch gradient descent, momentum method, Nesterov, adaptive gradient descent (Adagrad), Adadelta, root mean square prop (RMSprop), adaptive moment estimation (Adam), and the like.

[0047] When the error back propagation is performed, the error / loss obtained by the loss function is used to derive the current neuron, and the learning rate, the previous gradient / derivative / partial derivative, and the like are added to obtain the gradient, and the gradient is transmitted to the previous layer.

[0048] The model parameter transmission scheme provided by the embodiments of the present application will be described in detail in combination with the drawings and some embodiments and application scenarios.

[0049] FIG. 3 shows a flowchart of a model parameter transmission method in the embodiments of the present application. The method 300 can be performed by the first communication device. In other words, the method can be performed by software or hardware installed on the first communication device. As shown in FIG. 3, the method can include the following steps.

[0050] S310, the first communication device determines model information of at least one first sub-model in the plurality of sub-models of the target AI model that needs to update model parameters, wherein the target AI model includes a plurality of sub-models, and each of the sub-models has at least partially separate model information.

[0051] In the embodiments of the present application, the first communication device can be a network (Network) side device, for example, an access network device, or a core network device, or the first communication device can also be a terminal, and the specific embodiments of the present application are not limited.

[0052] In the embodiments of the present application, the first communication device, when determining the model information of the at least one first sub-model in the plurality of sub-models of the target AI model that needs to update model parameters, can also update the at least one first sub-model of the target AI model locally.

[0053] In the embodiments of the present application, the AI model (which can also be referred to as an AI unit, an AI structure, an AI function, an AI feature, a machine learning (ML) model, a machine learning unit, a neural network, a neural network function, a neural network function, etc.) can refer to a processing unit that can implement a specific algorithm, formula, processing flow, capability, etc. related to AI; or the AI unit can be a processing method, algorithm, function, module or unit for a specific data set; or the AI unit can be a processing method, algorithm, function, module or unit running on a graphics processing unit (GPU), neural network processor (NPU), tensor processing unit (TPU), application specific integrated circuit (ASIC) and other AI / machine learning (ML) related hardware, and the present application does not make specific limitations. Optionally, the specific data set includes the input or output of the AI unit.

[0054] Optionally, the identifier of the AI unit or AI model can be an AI model identifier, an AI structure identifier, an AI algorithm identifier, or an identifier of a specific data set associated with the AI unit / AI model, or an identifier of a specific scene, environment, area, cell, channel feature, device related to the AI / ML, or an identifier of a function, feature, capability or module related to the AI / ML, and the embodiments of the present application do not make specific limitations.

[0055] In the embodiments of the present application, in order to improve the efficiency of model transmission, the complete model is divided into multiple sub-models or sub-modules, each sub-model or sub-module is identified separately, and these sub-models can be transmitted or updated separately, thereby providing a specific scheme for updating part of the model parameters and simplifying the signaling process of model transmission.

[0056] In the embodiments of the present application, the target AI model can include N sub-models (or sub-modules), each sub-model has at least partially separate model information, that is, at least part of the model information of the multiple sub-models is not the same. For example, the N sub-models can have part of the model information determined based on each sub-model, and part of the model information shared by the N sub-models.

[0057] S312, the first communication device transmits the model information of the at least one first sub-model to the second communication device, wherein the model information includes the sub-model parameters of the first sub-model.

[0058] In the embodiments of the present application, the model information of the at least one first sub-model transmitted by the first communication device to the second communication device at least includes the sub-model parameters of each first sub-model. The sub-model parameters include but are not limited to neuron parameters, neuron information, and activation functions of the sub-model, etc. The sub-model parameters transmitted by the first communication device to the second communication device can include part of the sub-model parameters of the first sub-model, for example, part of the sub-model parameters that need to be updated, while the sub-model parameters that do not need to be updated can not be included in the model information transmitted by the first communication device, thereby saving transmission overhead and improving transmission efficiency. Alternatively, all sub-model parameters of the first sub-model can be included, for example, if all sub-model parameters of the first sub-model need to be updated, the first communication device can transmit all sub-model parameters of the first sub-model to the second communication device.

[0059] In some embodiments of the embodiments of the present application, the target AI model includes but is not limited to at least one of the following:

[0060] 1) the first AI model used by the first communication device or the second communication device;

[0061] 2) a reference AI model of the first AI model used by the first communication device or the second communication device; optionally, the protocol describes at least one of the model basic architecture, model structure, model parameter, training method, training data, or obtaining method of the reference AI model; optionally, the reference AI model can be used to obtain the first AI model, for example, the first AI model can be obtained by implementing the reference AI model, or obtained based on the optimization, training, compression, quantization, etc. operations of the reference AI model;

[0062] 3) a second AI model used by the first communication device, the second communication device or a test device in testing, wherein the test device is a device other than the first communication device and the second communication device;

[0063] 4) a reference model of the second AI model used by the first communication device, the second communication device or a test device in testing, wherein the test device is a device other than the first communication device and the second communication device;

[0064] 5) a third AI model used by the first communication device, the second communication device or a test device to match the second AI model used in testing; this optional implementation mainly targets a two-end communication use case of the first communication device and the second communication device, for example, channel state information (CSI) compression and recovery, encoding and decoding, etc. For example, the second AI model is used for CSI compression, and the third AI model is used for CSI recovery, or the second AI model is used for CSI recovery, and the third AI model is used for CSI compression; for example, the second AI model is used for encoding, and the third AI model is used for decoding, or the second AI model is used for decoding, and the third AI model is used for encoding.

[0065] 6) a reference model of the third AI model.

[0066] In the embodiments of the present application, the above AI model can be a plurality of sub-models, which are identified separately, so that these sub-models can be transmitted or updated separately, simplifying the signaling process of model transmission and the overhead of model information transmission.

[0067] In some embodiments of the embodiments of the present application, the model information can further include at least one of the following:

[0068] 1) model structure information of the first sub-model; for example, the number of layers of the first sub-model, the number of neurons contained in the first sub-model, the connection relationship between each or each layer of neurons, etc.

[0069] 2) identification information of the first sub-model; for example, the identification information of the first sub-model can include an ID of the first sub-model, or the identification information of the first sub-model can also include position information of the first sub-model in the target AI model.

[0070] 3) version information of the first sub-model.

[0071] In the embodiments of the present application, the model information (for example, sub-model parameters, sub-model structure, sub-model version information, etc.) of the first sub-model can also be used as the identification information of the first sub-model, or implicitly indicate the identification information of the first sub-model, or be used to distinguish different first sub-models.

[0072] Through the above embodiments, the first communication device can indicate at least one of the model structure information, the identification information, or the version information of the first sub-model that needs to be updated to the second communication device, so that the second communication device can determine the identification information of the first sub-model that needs to be updated.

[0073] In some embodiments, the version information of the first sub-model includes at least one of the following:

[0074] 1) The timestamp of the first sub-model, used to indicate at least one of the following: the update time of the first sub-model, the time of issuing the sub-model parameters of the first sub-model, the time of receiving the sub-model parameters of the first sub-model, the timestamp in the sub-model parameters of the first sub-model; optionally, the timestamp also considers time advance or time delay. The time advance is to advance a number of time units from the above time, and the time delay is to delay a number of time units from the above time. For example, the timestamp of the first sub-model is at least one of the following: the update time of the first sub-model is advanced by a number of time units, the time of issuing the sub-model parameters of the first sub-model is advanced by a number of time units, the time of receiving the sub-model parameters of the first sub-model is advanced by a number of time units, the timestamp in the sub-model parameters of the first sub-model is advanced by a number of time units, the update time of the first sub-model is delayed by a number of time units, the time of issuing the sub-model parameters of the first sub-model is delayed by a number of time units, the time of receiving the sub-model parameters of the first sub-model is delayed by a number of time units, the timestamp in the sub-model parameters of the first sub-model is delayed by a number of time units. Optionally, the time advance or the time delay can be indicated separately.

[0075] 2) The data set information associated with the target AI model; the version information of the first sub-model can include associated information of the target AI model or the identification (ID) of the associated information, for example, the version information of the first sub-model can include the data set information associated with the target AI model, wherein the data set information can be the identification information of the data set, such as the ID of the data set. Optionally, the data set information is information associated with data collection; for example, in the configuration of data collection (such as CSI reporting configuration, CSI resource configuration, reference signal resource set configuration, reference signal resource configuration, CSI-synchronization signal block resource set configuration), the data set information is configured.

[0076] 3) The data feature information associated with the target AI model; for example, the data feature ID associated with the target AI model.

[0077] 4) Base station hardware information associated with the target AI model; for example, base station hardware ID associated with the target AI model.

[0078] 5) Base station configuration information associated with the target AI model; for example, base station configuration ID associated with the target AI model.

[0079] 6) Physical cell information associated with the target AI model; for example, physical cell ID associated with the target AI model.

[0080] 7) Serving cell information associated with the target AI model; for example, serving cell ID associated with the target AI model.

[0081] 8) Area information associated with the target AI model; for example, area ID associated with the target AI model.

[0082] 9) Cell group information associated with the target AI model; for example, cell group ID associated with the target AI model.

[0083] 10) Cell list information associated with the target AI model; for example, cell list ID associated with the target AI model.

[0084] 11) Function associated with the target AI model; for example, function ID associated with the target AI model.

[0085] 12) Characteristics associated with the target AI model; for example, characteristic ID associated with the target AI model.

[0086] Through the version information of the first sub-model described above, the second communication model can know the version corresponding to the first sub-model to be updated, and then update the corresponding version.

[0087] In some embodiments, after S312, the method can further include: the first communication device receiving third feedback information sent by the second communication device, wherein the third feedback information includes at least one of:

[0088] 1) Identification information of the sub-model to be updated by the second communication device; through the identification information, the first communication device can know the sub-model to be updated by the second communication device, so as to keep consistent understanding with the second communication device.

[0089] 2) Identification information of the sub-model completed by the second communication device; through the identification information, the first communication device can know the sub-model completed by the second communication device, so as to keep consistent understanding with the second communication device.

[0090] 3) the identification information of the sub-module that the second communication device has not completed updating; through the identification information, the first communication device can know the sub-module that the second communication device has not completed updating, so as to keep consistent understanding with the second communication device.

[0091] 4) the identification information of the sub-module that the second communication device has failed to update; through the identification information, the first communication device can know the sub-module that the second communication device has failed to update, so as to keep consistent understanding with the second communication device.

[0092] 5) signaling or information of the completion of the sub-model updating; through the signaling or information, the first communication device can know the completion of the updating of the second communication device, so as to keep consistent understanding with the second communication device.

[0093] 6) signaling or information of the failure of the sub-model updating; through the signaling or information, the first communication device can know the failure of the updating of the second communication device, so as to keep consistent understanding with the second communication device.

[0094] In some embodiments, the method can further include: updating, based on target information, the version information of the at least one first sub-model during the updating of the at least one first sub-model based on the sub-model parameters of the at least one first sub-model or in the case of completion of the updating. In these embodiments, the version information of the at least one first sub-model can be updated based on the satisfied condition during the updating of the at least one first sub-model or in the case of completion of the updating, so as to avoid the situation of error updating caused by the out-of-time updating of the version information of the at least one first sub-model.

[0095] Optionally, the above target information can include at least one of the following:

[0096] 1) the first communication device transmits the model information of the at least one first sub-model; for example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of dataset information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of data feature information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of base station hardware information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of base station configuration information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of physical cell information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of serving cell information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of area information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of cell group information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of cell list information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of a function associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of a characteristic associated with the target AI model.

[0097] 2) the first communication device transmits the model information of the at least one first sub-model; for example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of dataset information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of data feature information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of base station hardware information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of base station configuration information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of physical cell information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of serving cell information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of area information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of cell group information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of cell list information associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of a function associated with the target AI model. For another example, the first communication device transmits the model information of the at least one first sub-model at a time of transmission of a characteristic associated with the target AI model.

[0098] 3) information related to the at least one first sub-model update; for example, the time when the at least one first sub-model is updated, or the data set information associated with the target AI model when the at least one first sub-model is updated, etc.

[0099] 4) information related to the completion of the at least one first sub-model update; for example, the time when the at least one first sub-model update is completed, or at least one of the data set information, data feature information, base station hardware information, base station configuration information, physical cell information, serving cell information, area information, cell group information, cell list information, functions and characteristics associated with the target AI model after the at least one first sub-model update is completed.

[0100] 5) information related to the feedback of the second communication device indicating the success of the update; for example, the time when the second communication device feeds back the success of the at least one sub-model update, or the time when the second communication device sends the third feedback information indicating the completion of the at least one first sub-model update, or the time when the first communication device successfully receives the third feedback information indicating the completion of the at least one first sub-model update.

[0101] Through the above embodiments, the version information of the at least one first sub-model can be updated based on the above target information, so that the version information of the at least one first sub-model is more accurate, and the management of the version information of the at least one first sub-model is avoided.

[0102] In some embodiments, the method can further include the following steps:

[0103] Step 1, the first communication device updates the model identifier of the target AI model, wherein the updated model identifier of the target AI model is different from the model identifier of the target AI model before the update;

[0104] Step 2, the first communication device indicates the updated model identifier of the target AI model to the second communication device.

[0105] In the above embodiments, the first communication device can indicate the new model identifier or the global model identifier to the second communication device when / at before / at after the at least one first sub-model is updated. Optionally, the new model identifier or the global model identifier is transmitted to the second communication device together with the model parameters of the sub-model update, so that the model identifiers of the target AI models of the first communication device and the second communication device are consistent.

[0106] In some embodiments, the second communication device can update the model identifier of the target AI model after receiving the model information of the at least one first sub-model, before / after updating the at least one first sub-model, and then send the updated model identifier of the target AI model to the first communication device. Therefore, in these embodiments, the method can further include: the first communication device receiving fourth indication information sent by the second communication device, wherein the fourth indication information is used to indicate the updated model identifier of the target AI model, and the updated model identifier of the target AI model is different from the model identifier of the target AI model before the update. Optionally, the updated model identifier of the target AI model can be transmitted to the first communication device together with the third feedback information. Thus, the model identifiers of the target AI models of the first communication device and the second communication device are consistent.

[0107] Through the technical solutions provided by the embodiments of the present application, in the case where it is necessary to update the model parameters of at least one sub-model in the plurality of sub-models of the target AI model, the model information of the at least one first sub-model is obtained, and the model information of the at least one first sub-model is transmitted to the second communication device without the need to transmit all the model information of the target AI model, thereby reducing the transmission overhead and improving the transmission efficiency.

[0108] FIG. 4 shows another flowchart of a model parameter transmission method provided by an embodiment of the present application, which can be executed by a first communication device. In other words, the method can be executed by software or hardware installed on the first communication device. As shown in FIG. 4, the method can include the following steps.

[0109] S410, the first communication device sends first indication information to the second communication device.

[0110] The first indication information is used to indicate one of the following: model information of a second sub-model in the target AI model that needs to be updated (for example, identification information of the second sub-model), and model information of a third sub-model in the target AI model that does not need to be updated (for example, identification information of the third sub-model).

[0111] Through the first indication information, the first communication device can indicate the second sub-model that needs to be updated or the third sub-model that does not need to be updated to the second communication device. Based on the indication of the first communication device, the second communication device can determine the second sub-model in the target AI model that needs to be updated.

[0112] Optionally, the at least one first sub-model can be a subset of the second sub-model determined based on the first indication information.

[0113] In some embodiments, after S410, the method can further include: receiving, by the first communication device, first feedback information sent by the second communication device, wherein the first feedback information includes at least one of the following: identification information of the updated first sub-model supported by the second communication device; and confirmation information indicating that the update of the second sub-model indicated by the first indication information is supported. Through the first feedback information, the first communication device can learn that the second communication device supports the updated first sub-model, or learn that the second communication device supports the update of the second sub-model, so that the first communication device can determine the at least one first sub-model that needs to be updated.

[0114] Optionally, the second communication device supporting the updated first sub-model indicated by the first feedback information can be a first sub-model selected by the second communication device from the second sub-model indicated by the first indication information.

[0115] S412, determining, by the first communication device, model information of at least one first sub-model that needs to update model parameters in the plurality of sub-models of the target AI model.

[0116] In the case where the first communication device receives the first feedback information, the at least one first sub-model can be the first sub-model indicated by the first feedback information that the second communication device supports to update. Or, in the case where the first feedback information indicates that the update of the second sub-model indicated by the first indication information is supported, the at least one first sub-model can be the second sub-model indicated by the first indication information.

[0117] Optionally, the model information is the same as the model information in method 300, and the same optional embodiments as in method 300 can be adopted. For details, refer to the description in method 300.

[0118] S414, transmitting, by the first communication device, the model information of the at least one first sub-model to the second communication device, wherein the model information includes sub-model parameters of the first sub-model.

[0119] This step is the same as S312 described above. For details, refer to the related description in method 300 described above.

[0120] In some embodiments of the embodiments of the present application, after S414, the first communication device can perform the steps performed after S312 in some embodiments of method 300. For details, refer to the related description in method 300, which will not be repeated here.

[0121] By means of the technical solutions provided in the embodiments of the present application, the first communication device can indicate the model information of the second sub-models that need to be updated, and the second communication device can feed back the first sub-models that support updating or whether the second sub-models support updating based on the indication of the first communication device, so that the model information of the at least one first sub-model sent by the first communication device to the second communication device is the model information of the sub-models that the second communication device supports updating, and the effectiveness of model information transmission is improved.

[0122] FIG. 5 shows another flowchart of a model parameter transmission method provided by an embodiment of the present application. The method 500 can be performed by a first communication device. In other words, the method can be performed by software or hardware installed on the first communication device. As shown in FIG. 5, the method can include the following steps.

[0123] S510, the first communication device receives second indication information sent by the second communication device, wherein the second indication information includes model information (for example, identification information or version information) of at least one fourth sub-model of the target AI model.

[0124] In this embodiment, the second communication device can send the second indication information to the first communication device in the case where at least one fourth sub-model needs to be updated, so that the first communication device can learn the identification information or version information of the fourth sub-model that needs to be updated by the second communication device.

[0125] S512, the first communication device sends first indication information to the second communication device, wherein the first indication information is used to indicate one of the following: identification information of a second sub-model that needs to be updated in the target AI model, identification information of a third sub-model that does not need to be updated in the target AI model.

[0126] Optionally, the second sub-model indicated in the first indication information is a subset of the at least one fourth sub-model. That is, the first communication device determines at least one second sub-model that needs to be updated in the at least one fourth sub-model based on the model information of the at least one fourth sub-model.

[0127] Optionally, S512 can be the same as S410 described above, and specific details can be referred to the description in method 300.

[0128] Optionally, after S512, the first communication device receives first feedback information sent by the second communication device, wherein the first feedback information includes at least one of the following: identification information of a first sub-model that the second communication device supports updating; confirmation information used to indicate that the second sub-model that needs to be updated indicated by the first indication information is supported to be updated. Specific details can be referred to the related description in method 400 described above.

[0129] S514, the first communication device determines model information of at least one first sub-model in the plurality of sub-models of the target AI model that needs to update model parameters.

[0130] In a case where the first communication device receives the first feedback information, the at least one first sub-model can indicate, for the first feedback information, that the second communication device supports updating of the first sub-model. Alternatively, in a case where the first feedback information indicates support for updating of the second sub-model that needs to be updated as indicated by the first indication information, the at least one first sub-model can be the second sub-model indicated by the first indication information.

[0131] Optionally, the model information is the same as the model information in the method 300, and the same optional implementation as in the method 300 can be adopted. For details, refer to the description in the method 300.

[0132] S516, the first communication device transmits, to the second communication device, model information of the at least one first sub-model, where the model information includes sub-model parameters of the first sub-model.

[0133] This step is the same as S312 described above. For details, refer to the related description in the method 300 described above.

[0134] In some embodiments of the embodiments of the present application, after S516, the first communication device can perform the steps performed after S312 in some embodiments of the method 300. For details, refer to the related description in the method 300, which will not be repeated here.

[0135] Through the technical solutions provided in the embodiments of the present application, the second communication device can report model information of at least one fourth sub-model, and the first communication device can indicate model information of a second sub-model that needs to be updated based on the model information of the at least one fourth sub-model. The second communication device can feed back support for updating of the first sub-model or whether to support updating of the second sub-model based on the indication of the first communication device, so that the model information of the at least one first sub-model transmitted by the first communication device to the second communication device based on the report of the second communication device is model information of a sub-model that the second communication device supports updating, thereby improving the effectiveness of model information transmission.

[0136] FIG. 6 shows another flowchart of a model parameter transmission method according to an embodiment of the present application. The method 600 can be performed by a first communication device. In other words, the method can be performed by software or hardware installed on the first communication device. As shown in FIG. 6, the method can include the following steps.

[0137] S610, the first communication device sends model information (e.g., identification information or version information) of at least one fifth sub-model of the target AI model to the second communication device.

[0138] For example, the first communication device can send the model information of the at least one fifth sub-model after updating the at least one fifth sub-model of the target AI model, so that the second communication device can determine whether the local at least one fifth sub-model needs to be updated based on the received model information of the at least one fifth sub-model.

[0139] In some embodiments, the first communication device broadcasts the model information of the at least one fifth sub-model of the target AI model, so that the communication devices within the communication range of the first communication device can all know the model information of the at least one fifth sub-model.

[0140] In other embodiments, the first communication device sends the identification information or version information of the at least one fifth sub-model of the target AI model through a dedicated channel or dedicated signaling with the second communication device. In some scenarios, the target AI model can not be used by all communication devices within the communication range of the first communication device, but only by some specific communication devices. In order to avoid interference to other communication devices, the first communication device can send the identification information or version information of the at least one fifth sub-model of the target AI model through a dedicated channel or dedicated signaling with the second communication device.

[0141] In some embodiments, the at least one fifth sub-model is a sub-model that needs to be updated.

[0142] In other embodiments, the at least one fifth sub-model is a sub-model that does not need to be updated.

[0143] S612, the first communication device receives second feedback information sent by the second communication device, wherein the second feedback information includes model information (e.g., identification information or version information) of a first sub-model that needs to be updated.

[0144] Optionally, the at least one first sub-model is a sub-model in the at least one fifth sub-model. After receiving the model information of the at least one fifth sub-model, the second communication device can determine at least one first sub-model that needs to be updated from the at least one fifth sub-model based on the model information of the at least one fifth sub-model, and then send second feedback information to the first communication device by including the model information of the first sub-model that needs to be updated in the second feedback information.

[0145] S614, the first communication device determines model information of at least one first sub-model in the plurality of sub-models of the target AI model that needs to update model parameters.

[0146] This step is the same as S310, and specific reference can be made to the related description in method 300, which will not be repeated here.

[0147] S616, the first communication device transmits the model information of the at least one first sub-model to the second communication device, wherein the model information includes the sub-model parameters of the first sub-model.

[0148] This step is the same as S312, and specific reference can be made to the related description in method 300, which will not be repeated here.

[0149] In some embodiments of the embodiments of the present application, after S616, the first communication device can perform the steps performed after S312 in some embodiments of method 300, and specific reference can be made to the related description in method 300, which will not be repeated here.

[0150] Through the technical solutions provided by the embodiments of the present application, the first communication device can transmit the model information of the sub-models of the target AI model, the second communication device feeds back the model information of the sub-models that need to be updated based on the received model information of the sub-models, and the first communication device can send the model parameters of the at least one first sub-model to the second communication device based on the feedback of the second communication device, so that the model information of the at least one first sub-model sent by the first communication device is the model information of the sub-models that need to be updated by the second communication device, and the effectiveness of the model information transmission is improved.

[0151] Based on the same technical concept, the embodiments of the present application also provide another model parameter transmission method, which is executed by the second communication device.

[0152] It should be noted that the following embodiments only describe the operation of the second communication device, and other details can be referred to the above related description of methods 300 to 600.

[0153] FIG. 7 shows a flowchart of another model parameter transmission method provided by the embodiments of the present application, which can be executed by the second communication device. In other words, the method can be executed by software or hardware installed on the second communication device. As shown in FIG. 7, the method mainly includes the following steps.

[0154] S710, the second communication device receives model information of at least one first sub-model transmitted by the first communication device, wherein the at least one first sub-model is a sub-model of a plurality of sub-models of a target AI model that needs to update model parameters, the target AI model includes a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information includes sub-model parameters of the first sub-model;

[0155] S712, the second communication device updates the model parameters of the at least one first sub-model based on the model information of the at least one first sub-model.

[0156] Through the technical solutions provided by the embodiments of the present application, the second communication device can receive the model information of at least one first sub-model transmitted by the first communication device, and update the model parameters of the at least one first sub-model based on the received model information, thereby realizing the transmission of part of the model parameters of the target AI model, saving the transmission cost of the model parameters, and providing the transmission efficiency of the model parameters.

[0157] In an optional implementation, the model information further includes at least one of the following: model structure information of the first sub-model, identification information of the first sub-model, and version information of the first sub-model.

[0158] Optionally, the identification information of the first sub-model includes at least one of the following:

[0159] an identification ID of the first sub-model;

[0160] position information of the first sub-model in the target AI model.

[0161] In an optional implementation, the version information of the first sub-model includes at least one of the following:

[0162] 1) a timestamp of the first sub-model, used to indicate at least one of the following: an update time of the first sub-model, a time of issuing the sub-model parameters of the first sub-model, a time of receiving the sub-model parameters of the first sub-model, and a timestamp in the sub-model parameters of the first sub-model;

[0163] 2) dataset information associated with the target AI model;

[0164] 3) data feature information associated with the target AI model;

[0165] 4) base station hardware information associated with the target AI model;

[0166] 5) base station configuration information associated with the target AI model;

[0167] 6) Physical cell information associated with the target AI model;

[0168] 7) Serving cell information associated with the target AI model;

[0169] 8) Area information associated with the target AI model;

[0170] 9) Cell group information associated with the target AI model;

[0171] 10) Cell list information associated with the target AI model;

[0172] 11) Function associated with the target AI model;

[0173] 12) Characteristics associated with the target AI model.

[0174] Optionally, the target AI model comprises at least one of the following:

[0175] a first AI model used by the first communication device or the second communication device;

[0176] a reference AI model of the first AI model used by the first communication device or the second communication device;

[0177] a second AI model used by the first communication device, the second communication device, or a test device in testing, wherein the test device is a device other than the first communication device and the second communication device;

[0178] a reference model of the second AI model used by the first communication device, the second communication device, or a test device in testing, wherein the test device is a device other than the first communication device and the second communication device;

[0179] a third AI model used by the first communication device, the second communication device, or a test device to match the second AI model used in testing;

[0180] a reference model of the third AI model.

[0181] In some embodiments, before the second communication device receives the model information of at least one first sub-model transmitted by the first communication device, the method can further comprise: the second communication device receiving first indication information sent by the first communication device, wherein the first indication information is used to indicate one of the following: identification information of a second sub-model that needs to be updated in the target AI model, identification information of a third sub-model that does not need to be updated in the target AI model.

[0182] In some embodiments, after the second communication device receives the first indication information sent by the first communication device, the method can further include: the second communication device sending first feedback information to the first communication device, wherein the first feedback information includes at least one of the following: identification information of the second communication device supporting an updated first sub-model; and confirmation information indicating support for updating the second sub-model indicated by the first indication information.

[0183] In some embodiments, before the second communication device receives the first indication information sent by the first communication device, the method can further include: the second communication device sending second indication information to the first communication device, wherein the second indication information includes identification information or version information of at least one fourth sub-model of the target AI model.

[0184] Optionally, the second sub-model indicated in the first indication information is a subset of the at least one fourth sub-model.

[0185] In some embodiments, before the second communication device receives the model information of the at least one first sub-model transmitted by the first communication device, the method can further include the following steps:

[0186] Step 1: The second communication device receives identification information or version information of at least one fifth sub-model of the target AI model sent by the first communication device.

[0187] Step 2: The second communication device sends second feedback information to the first communication device, wherein the second feedback information includes identification information or version information of the first sub-model that needs to be updated.

[0188] In some embodiments, the second communication device receives identification information or version information of at least one fifth sub-model of the target AI model sent by the first communication device, including one of the following:

[0189] The second communication device receives identification information or version information of at least one fifth sub-model of the target AI model broadcasted by the first communication device;

[0190] The second communication device receives identification information or version information of at least one fifth sub-model of the target AI model sent by the first communication device through a dedicated channel or dedicated signaling with the second communication device.

[0191] In some embodiments, after the second communication device receives the model information of the at least one first sub-model transmitted by the first communication device, the method can further include: the second communication device sending third feedback information to the first communication device, wherein the third feedback information includes at least one of the following:

[0192] 1) identification information of a sub-model to be updated by the second communication device;

[0193] 2) identification information of a completed updated sub-model by the second communication device;

[0194] 3) identification information of an uncompleted updated sub-model by the second communication device;

[0195] 4) identification information of a failed updated sub-model by the second communication device;

[0196] 5) signaling or information of completion of sub-model updating;

[0197] 6) signaling or information of failure of sub-model updating.

[0198] In an optional implementation, the method can further include: during updating of the at least one first sub-model based on the sub-model parameters of the at least one first sub-model or in the case of completion of updating, updating version information of the at least one first sub-model based on target information.

[0199] Optionally, the target information includes at least one of the following:

[0200] 1) related information when the first communication device transmits the model information of the at least one first sub-model;

[0201] 2) related information included in or associated with the model information of the at least one first sub-model transmitted by the first communication device;

[0202] 3) related information of updating of the at least one first sub-model;

[0203] 4) related information of completion of updating of the at least one first sub-model;

[0204] 5) related information of feedback of success of updating by the second communication device.

[0205] In some embodiments, the method can further include one of the following:

[0206] 1) the second communication device receives third indication information sent by the first communication device, and updates the model identification of the target AI model based on the third indication information, wherein the third indication information indicates the model identification of the updated target AI model;

[0207] 2) The second communication device updates the model identifier of the target AI model and sends fourth indication information to the first communication device, wherein the fourth indication information is used to indicate the model identifier of the updated target AI model, and the model identifier of the updated target AI model is different from the model identifier of the target AI model before the update.

[0208] Through the above implementation, the model identifier of the target AI model can be updated in time, and confusion of the model identifier of the target AI model is avoided.

[0209] Hereinafter, the technical solutions provided by the embodiments of the present application are described by taking the first communication device as a network side device (NW, for example, an access network device) and the second communication device as a terminal as an example. Of course, it is not limited thereto, and in actual application, the first communication device can also be a terminal or a core network device, or the second communication device can be a network side device (for example, an access network device) or a core network device.

[0210] Embodiment one

[0211] In this embodiment, the identifier information of the sub-models that need to be updated is indicated by the NW, as shown in FIG. 8, the model parameter transmission method in this embodiment includes the following steps:

[0212] S801, the NW indicates the identifier information of the sub-models that need to be updated, or the NW indicates the identifier information of the sub-models that do not need to be updated.

[0213] S802, the UE feeds back first feedback information.

[0214] Optionally, the UE feeding back the first feedback information can include at least one of the following:

[0215] The UE feeds back the identifier information of the sub-models that support updating or the version information of these sub-models, wherein the UE feeding back the sub-models that support updating can be in the set of the updating sub-models indicated by the NW, or outside the set of the sub-models that do not need to be updated indicated by the NW.

[0216] The UE feeds back confirmation information, which is used to indicate whether the UE supports all the set of the updating sub-models indicated by the NW.

[0217] Wherein, S802 is an optional step.

[0218] S803, the NW transmits the model parameters of the sub-models that need to be updated.

[0219] S804, the UE sends second feedback information.

[0220] Optionally, the second feedback information can be at least one of the following:

[0221] Type 1: The UE feeds back identification information of sub-models to be updated or version information of the sub-models.

[0222] Type 2: The UE feeds back identification information of sub-models completed in updating or version information of the sub-models.

[0223] Type 3: The UE feeds back identification information of sub-models not completed in updating or version information of the sub-models. The sub-models correspond to sub-modules not needing updating, and the UE feeds back identification information of the sub-models or version information of the sub-models after the UE completes updating the sub-models.

[0224] Type 4: The UE feeds back identification information of sub-models failed in updating or version information of the sub-models. The sub-models correspond to sub-modules needing updating but failed in updating, and the UE feeds back identification information of the sub-models or version information of the sub-models after the UE can complete updating the sub-models.

[0225] Type 5: The UE feeds back signaling or information of completion or failure of updating the sub-models. The UE feeds back signaling or information of completion or failure of updating the sub-models after the UE can complete updating the sub-models.

[0226] Type 6: The UE does not feed back.

[0227] Optionally, the S804 can be an optional step.

[0228] Embodiment Two

[0229] In this embodiment, the NW issues version information of sub-models. As shown in FIG. 9, in this embodiment, the model parameter transmission method mainly includes the following steps:

[0230] S901: The NW issues version information of sub-models or identification information of the sub-models.

[0231] Optionally, the NW can broadcast version information of the sub-models or separately send the version information of the sub-models.

[0232] Optionally, the NW issues the sub-models with version information, and by default, the sub-models are sub-models needing updating.

[0233] Optionally, the NW does not issue the sub-models without version information, and by default, the sub-models are sub-models not needing updating.

[0234] S902: The UE feeds back identification information of sub-models needing updating or version information of the sub-models.

[0235] Optionally, the sub-models fed back by the UE are in a set of sub-models with version information issued by the NW.

[0236] S903: The NW transmits model parameters of the sub-models needing updating.

[0237] Optionally, the NW transmits all the sub-models whose version information is issued, or the NW itself selects the sub-models to be issued, when the UE does not feedback or before the UE feedback.

[0238] S904, the UE feeds back the second feedback information.

[0239] Optionally, the second feedback information can be at least one of the following:

[0240] Type 1: the UE feeds back the identification information of the sub-models to be updated or the version information of the sub-models.

[0241] Type 2: the UE feeds back the identification information of the sub-models completed updating or the version information of the sub-models.

[0242] Type 3: the UE feeds back the identification information of the sub-models not completed updating or the version information of the sub-models. The sub-models correspond to the sub-models not needing updating, and the UE feeds back the identification information of the sub-models not completed updating or the version information of the sub-models after the UE completes updating the sub-models.

[0243] Type 4: the UE feeds back the identification information of the sub-models failed updating or the version information of the sub-models. The sub-models correspond to the sub-models needing updating but failed updating, and the UE feeds back the identification information of the sub-models or the version information after the UE can complete updating the sub-models.

[0244] Type 5: the UE feeds back the signaling or information of the sub-models completed or failed updating. The UE feeds back the signaling or information of the sub-models completed or failed updating after the UE can complete updating the sub-models.

[0245] Type 6: the UE does not feedback.

[0246] Optionally, the above S904 can be an optional step.

[0247] Embodiment Three

[0248] In this embodiment, the UE reports the version information of the sub-models, as shown in FIG. 10, and the model parameter transmission method in this embodiment mainly includes the following steps:

[0249] S1001, the UE reports the version information of the sub-models or the identification information of the sub-models.

[0250] S1002, the NW indicates the identification information of the sub-models needing updating or the version information of the sub-models.

[0251] Optionally, the sub-models indicated by the NW are in the set of the sub-models whose version information is reported by the UE.

[0252] S1003, the UE feeds back first feedback information.

[0253] The UE feeding back the first feedback information can include one of the following:

[0254] The UE feeds back identification information of the updated sub-models or version information of the sub-models, wherein the sub-models can be in the set of updated sub-models indicated by the NW or outside the set of sub-models not requiring update indicated by the NW.

[0255] The UE feeds back confirmation information indicating whether the UE supports the set of updated sub-models indicated by the NW.

[0256] The step is optional.

[0257] S1004, the NW transmits model parameters of the sub-models requiring update.

[0258] S1005, the UE feeds back second feedback information.

[0259] The second feedback information can include at least one of the following:

[0260] Optionally, the second feedback information can be at least one of the following:

[0261] Type 1: the UE feeds back identification information of the sub-models to be updated or version information of the sub-models.

[0262] Type 2: the UE feeds back identification information of the sub-models completed update or version information of the sub-models.

[0263] Type 3: the UE feeds back identification information of the sub-models not completed update or version information of the sub-models. The sub-models correspond to the sub-models not requiring update, and the UE feeds back the identification information of the sub-models not completed update or version information of the sub-models after the UE completes the update of the sub-models.

[0264] Type 4: the UE feeds back identification information of the sub-models failed in update or version information of the sub-models. The sub-models correspond to the sub-models requiring update but failed in update, and the UE feeds back the identification information of the sub-models or version information after the UE can complete the update of the sub-models.

[0265] Type 5: the UE feeds back signaling or information of the completion or failure of the update of the sub-models. The UE feeds back the signaling or information of the completion or failure of the update after the UE can complete the update of the sub-models.

[0266] Type 6: the UE does not feed back the AI model

[0267] The above S1005 can be an optional step.

[0268] The technical solution provided by the embodiments of the present application can update part of the parameters of the model, reduce the air interface overhead of model transmission, and simplify the signaling process of model transmission.

[0269] The model parameter transmission method provided by the embodiments of the present application is executed by the model parameter transmission device. The model parameter transmission device provided by the embodiments of the present application is described by taking the model parameter transmission device executing the model parameter transmission method as an example.

[0270] The model parameter transmission device provided by the embodiments of the present application can be a communication device or a component in the communication device, for example, a chip. The communication device can be a terminal, a network side device, a server, or the like. For example, the terminal can include, but is not limited to, the types of the terminal 11 listed above, the network side device can include, but is not limited to, the types of the network side device 12 listed above, and the embodiments of the present application are not limited in this regard.

[0271] The model parameter transmission device includes a receiving module, a sending module, and a processing module. The receiving module, the sending module, and the processing module can be implemented by software or hardware. When implemented by hardware, the processing module can be implemented by a processor. For example, the processor can include a general processor, a special-purpose processor, or the like, such as a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), an artificial intelligent (AI) processor, a graphics processing unit (GPU), an application specific integrated circuit (ASIC), a network processor (NP), a field programmable gate array (FPGA), or other programmable logic devices, gate circuits, transistors, discrete hardware components, or the like. The receiving module and the sending module can be implemented by a communication interface, which can include one or more of a transceiver, a pin, a circuit, a bus, a radio frequency unit, or the like.

[0272] Specifically, referring to FIG. 11, when the model parameter transmission apparatus is a first communication device or a component in the first communication device, the model parameter transmission apparatus 1100 includes a processing module 1101 configured to determine model information of at least one first sub-model that needs to update model parameters in a plurality of sub-models of a target AI model, wherein the target AI model includes a plurality of sub-models, and each of the sub-models has at least partially separate model information; and a sending module 1102 configured to transmit the model information of the at least one first sub-model to a second communication device, wherein the model information includes sub-model parameters of the first sub-model.

[0273] In an optional implementation, the model information further includes at least one of the following: model structure information of the first sub-model, identification information of the first sub-model, and version information of the first sub-model.

[0274] In an optional implementation, the identification information of the first sub-model includes at least one of the following:

[0275] an identification ID of the first sub-model;

[0276] position information of the first sub-model in the target AI model.

[0277] In an optional implementation, the version information of the first sub-model includes at least one of the following:

[0278] a timestamp of the first sub-model, used to indicate at least one of the following: an update time of the first sub-model, a time of issuing the sub-model parameters of the first sub-model, a time of receiving the sub-model parameters of the first sub-model, and a timestamp in the sub-model parameters of the first sub-model;

[0279] dataset information associated with the target AI model;

[0280] data feature information associated with the target AI model;

[0281] base station hardware information associated with the target AI model;

[0282] base station configuration information associated with the target AI model;

[0283] physical cell information associated with the target AI model;

[0284] serving cell information associated with the target AI model;

[0285] area information associated with the target AI model;

[0286] cell group information associated with the target AI model;

[0287] The cell list information associated with the target AI model;

[0288] The function associated with the target AI model;

[0289] The characteristics associated with the target AI model.

[0290] In an optional implementation, the target AI model includes at least one of the following:

[0291] The first AI model used by the first communication device or the second communication device;

[0292] The reference AI model of the first AI model used by the first communication device or the second communication device;

[0293] The second AI model used by the first communication device, the second communication device, or a test device in a test, wherein the test device is a device other than the first communication device and the second communication device;

[0294] The reference model of the second AI model used by the first communication device, the second communication device, or a test device in a test, wherein the test device is a device other than the first communication device and the second communication device;

[0295] The third AI model used by the first communication device, the second communication device, or a test device to match the second AI model used in the test;

[0296] The reference model of the third AI model.

[0297] In an optional implementation, the sending module 1102 is further configured to send first indication information to the second communication device, wherein the first indication information is used to indicate one of the following: identification information of a second sub-model that needs to be updated in the target AI model, and identification information of a third sub-model that does not need to be updated in the target AI model.

[0298] In an optional implementation, as shown in FIG. 11, the apparatus can further include a receiving module 1103 configured to receive first feedback information sent by the second communication device, wherein the first feedback information includes at least one of the following: identification information of an updated first sub-model supported by the second communication device; and confirmation information used to indicate support for updating of the second sub-model that needs to be updated as indicated by the first indication information.

[0299] In an optional implementation, as shown in FIG. 11, the apparatus can further include a receiving module 1103 configured to receive second indication information sent by the second communication device, wherein the second indication information includes identification information or version information of at least one fourth sub-model of the target AI model.

[0300] In an optional implementation, the second sub-model indicated in the first indication information is a subset of the at least one fourth sub-model.

[0301] In an optional implementation, the sending module 1102 is further configured to send, to the second communication device, identification information or version information of at least one fifth sub-model of the target AI model; and the apparatus further includes a receiving module 1103 configured to receive second feedback information sent by the second communication device, wherein the second feedback information includes identification information or version information of a first sub-model that needs to be updated.

[0302] In an optional implementation, the sending module 1102 sends, to the second communication device, identification information or version information of at least one fifth sub-model of the target AI model, including one of the following:

[0303] broadcasting the identification information or version information of the at least one fifth sub-model of the target AI model;

[0304] sending the identification information or version information of the at least one fifth sub-model of the target AI model through a dedicated channel or dedicated signaling with the second communication device.

[0305] In an optional implementation, the at least one fifth sub-model is a sub-model that needs to be updated.

[0306] In an optional implementation, the at least one fifth sub-model is a sub-model that does not need to be updated.

[0307] In an optional implementation, the at least one first sub-model is a sub-model in the at least one fifth sub-model.

[0308] In an optional implementation, the apparatus further includes a receiving module 1103 configured to receive third feedback information sent by the second communication device, wherein the third feedback information includes at least one of the following:

[0309] identification information of a sub-model to be updated by the second communication device;

[0310] identification information of a sub-model updated by the second communication device;

[0311] identification information of a sub-model not updated by the second communication device.

[0312] the second communication device updates the identification information of the failed sub-module;

[0313] signaling or information indicating that the sub-model updating is completed;

[0314] signaling or information indicating that the sub-model updating fails.

[0315] In an optional implementation, the processing module 1101 is further configured to update version information of the at least one first sub-model based on target information during the process of updating the at least one first sub-model based on the sub-model parameters of the at least one first sub-model or in the case where the updating is completed.

[0316] In an optional implementation, the target information includes at least one of the following:

[0317] related information when the model information of the at least one first sub-model is transmitted;

[0318] related information included in or associated with the transmitted model information of the at least one first sub-model;

[0319] related information of the updating of the at least one first sub-model;

[0320] related information of the completion of the updating of the at least one first sub-model;

[0321] related information of the feedback of the second communication device indicating that the updating succeeds.

[0322] In an optional implementation, the processing module 1101 is further configured to update the model identification of the target AI model, where the model identification of the updated target AI model is different from the model identification of the target AI model before the updating; and the sending module 1102 is further configured to indicate the model identification of the updated target AI model to the second communication device.

[0323] In an optional implementation, the apparatus further includes a receiving module 1103 configured to receive fourth indication information sent by the second communication device, where the fourth indication information is used to indicate the model identification of the updated target AI model, and the model identification of the updated target AI model is different from the model identification of the target AI model before the updating.

[0324] Referring to FIG. 12, when the model parameter transmission apparatus is a second communication device or a component in the second communication device, the model parameter transmission apparatus 1200 includes a receiving module 1201 configured to receive model information of at least one first sub-model transmitted by a first communication device, where the at least one first sub-model is a sub-model in a plurality of sub-models of a target AI model that needs to update a model parameter, the target AI model includes a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information includes a sub-model parameter of the first sub-model; and a processing module 1202 configured to update a model parameter of the at least one first sub-model based on the model information of the at least one first sub-model.

[0325] In an optional implementation, the model information further includes at least one of the following: model structure information of the first sub-model, identification information of the first sub-model, and version information of the first sub-model.

[0326] In an optional implementation, the identification information of the first sub-model includes at least one of the following:

[0327] an identification ID of the first sub-model;

[0328] position information of the first sub-model in the target AI model.

[0329] In an optional implementation, the version information of the first sub-model includes at least one of the following:

[0330] a timestamp of the first sub-model, used to indicate at least one of the following: an update time of the first sub-model, a time of issuing the sub-model parameter of the first sub-model, a time of receiving the sub-model parameter of the first sub-model, and a timestamp in the sub-model parameter of the first sub-model;

[0331] dataset information associated with the target AI model;

[0332] data feature information associated with the target AI model;

[0333] base station hardware information associated with the target AI model;

[0334] base station configuration information associated with the target AI model;

[0335] physical cell information associated with the target AI model;

[0336] serving cell information associated with the target AI model;

[0337] area information associated with the target AI model;

[0338] cell group information associated with the target AI model;

[0339] a cell list information associated with the target AI model;

[0340] a function associated with the target AI model;

[0341] a characteristic associated with the target AI model.

[0342] In an optional implementation, the target AI model includes at least one of the following:

[0343] a first AI model used by the first communication device or the second communication device;

[0344] a reference AI model of the first AI model used by the first communication device or the second communication device;

[0345] a second AI model used by the first communication device, the second communication device, or a test device in a test, wherein the test device is a device other than the first communication device and the second communication device;

[0346] a reference model of the second AI model used by the first communication device, the second communication device, or a test device in a test, wherein the test device is a device other than the first communication device and the second communication device;

[0347] a third AI model used by the first communication device, the second communication device, or a test device to match the second AI model used in the test;

[0348] a reference model of the third AI model.

[0349] In an optional implementation, the receiving module 1201 is further configured to receive first indication information sent by the first communication device, wherein the first indication information is used to indicate one of the following: identification information of a second sub-model in the target AI model that needs to be updated, and identification information of a third sub-model in the target AI model that does not need to be updated.

[0350] In an optional implementation, as shown in FIG. 12, the apparatus further includes a sending module 1203 configured to send first feedback information to the first communication device, wherein the first feedback information includes at least one of the following: identification information of an updated first sub-model supported by the second communication device; and confirmation information used to indicate support for updating of the second sub-model that needs to be updated as indicated by the first indication information.

[0351] In an optional implementation, the apparatus further includes a sending module 1203 configured to send second indication information to the first communication device, where the second indication information includes identification information or version information of at least one fourth sub-model of the target AI model.

[0352] In an optional implementation, the second sub-model indicated in the first indication information is a subset of the at least one fourth sub-model.

[0353] In an optional implementation, the receiving module 1201 is further configured to receive identification information or version information of at least one fifth sub-model of the target AI model sent by the first communication device; and the apparatus further includes a sending module 1303 configured to send second feedback information to the first communication device, where the second feedback information includes identification information or version information of a first sub-model that needs to be updated.

[0354] In an optional implementation, the receiving of the identification information or version information of the at least one fifth sub-model of the target AI model sent by the first communication device includes one of the following:

[0355] Receiving the identification information or version information of the at least one fifth sub-model of the target AI model broadcasted and sent by the first communication device;

[0356] Receiving the identification information or version information of the at least one fifth sub-model of the target AI model sent by the first communication device through a dedicated channel or dedicated signaling with the second communication device.

[0357] In an optional implementation, the apparatus further includes a sending module 1203 configured to send third feedback information to the first communication device, where the third feedback information includes at least one of the following:

[0358] Identification information of a sub-model to be updated by the second communication device;

[0359] Identification information of a sub-model updated by the second communication device;

[0360] Identification information of a sub-model not updated by the second communication device;

[0361] Identification information of a sub-model updated unsuccessfully by the second communication device;

[0362] Signaling or information of a sub-model update completion;

[0363] Signaling or information of a sub-model update failure.

[0364] In an optional implementation, the processing module 1202 is further configured to update version information of the at least one first sub-model based on the target information, in a process of updating the at least one first sub-model based on the sub-model parameters of the at least one first sub-model or in a case where the updating is completed.

[0365] In an optional implementation, the target information includes at least one of the following:

[0366] related information when the first communication device transmits the model information of the at least one first sub-model;

[0367] related information included in or associated with the model information of the at least one first sub-model transmitted by the first communication device;

[0368] related information of the updating of the at least one first sub-model;

[0369] related information of the completion of the updating of the at least one first sub-model;

[0370] related information of the feedback of the success of the updating by the second communication device.

[0371] In an optional implementation, the receiving module 1201 is further configured to receive third indication information sent by the first communication device, and the processing module 1202 is further configured to update the model identifier of the target AI model based on the third indication information, where the third indication information indicates the model identifier of the updated target AI model; or, the processing module 1202 is further configured to update the model identifier of the target AI model, and the apparatus further includes a sending module 1203 configured to send fourth indication information to the second communication device, where the fourth indication information is used to indicate the model identifier of the updated target AI model, and the model identifier of the updated target AI model is different from the model identifier of the target AI model before the updating.

[0372] The model parameter transmission apparatus provided by the embodiments of the present application can implement each process implemented by the method embodiments of FIGS. 3 to 10 and achieve the same technical effects. To avoid repetition, details are not described herein.

[0373] As shown in FIG. 13, the embodiments of the present application further provide a communication device 1300, comprising a processor 1301 and a memory 1302, wherein the memory 1302 stores programs or instructions executable on the processor 1301. For example, when the communication device 1300 is a first communication device, the programs or instructions are executed by the processor 1301 to implement each step of the model parameter transmission method 300 to 600 embodiments and achieve the same technical effects. When the communication device 1300 is a second communication device, the programs or instructions are executed by the processor 1301 to implement each step of the model parameter transmission method 700 embodiment and achieve the same technical effects. To avoid repetition, details are not described herein.

[0374] The embodiments of the present application further provide a terminal, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is configured to run programs or instructions to implement the steps in the method embodiments shown in FIG. 3 to FIG. 7. The terminal embodiments correspond to the terminal-side method embodiments described above, and each implementation process and implementation manner of the method embodiments can be applied to the terminal embodiments and achieve the same technical effects. The terminal can be the model parameter transmission apparatus shown in FIG. 11 or 12. Specifically, FIG. 14 is a schematic diagram of a hardware structure of a terminal implementing the embodiments of the present application.

[0375] The terminal 1400 includes, but is not limited to, at least part of the components such as a radio frequency unit 1401, a network module 1402, an audio output unit 1403, an input unit 1404, a sensor 1405, a display unit 1406, a user input unit 1407, an interface unit 1408, a memory 1409, and a processor 1410.

[0376] Those skilled in the art can understand that the terminal 1400 can further include a power supply (such as a battery) for supplying power to each component, and the power supply can be logically connected to the processor 1410 through a power management system, so as to realize functions such as power management, discharge management, and power consumption management through the power management system. The terminal structure shown in FIG. 14 does not constitute a limitation on the terminal, and the terminal can include more or fewer components than those shown, or combine certain components, or different component arrangements, which are not described herein.

[0377] It should be understood that in the embodiments of the present application, the input unit 1404 can include a graphics processor 14041 and a microphone 14042, and the graphics processor 14041 processes image data of a still picture or a video obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The display unit 1406 can include a display panel 14061, which can be configured in the form of a liquid crystal display, an organic light-emitting diode, etc. The user input unit 1407 includes at least one of a touch panel 14071 and other input devices 14072. The touch panel 14071 is also called a touch screen. The touch panel 14071 can include two parts of a touch detection device and a touch controller. The other input devices 14072 can include, but are not limited to, a physical keyboard, function keys (such as volume control keys, on-off keys, etc.), a trackball, a mouse, a joystick, and the like, which will not be described here.

[0378] In the embodiments of the present application, after the radio frequency unit 1401 receives the downlink data from the network side device, it can be transmitted to the processor 1410 for processing. In addition, the radio frequency unit 1401 can send uplink data to the network side device. Generally, the radio frequency unit 1401 includes, but is not limited to, an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc.

[0379] The memory 1409 can be used to store software programs or instructions and various data. The memory 1409 can mainly include a first storage area storing programs or instructions and a second storage area storing data, wherein the first storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), and the like. In addition, the memory 1409 can include a volatile memory or a non-volatile memory. The non-volatile memory can be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically EPROM (EEPROM), or a flash memory. The volatile memory can be a Random Access Memory (RAM), a Static RAM (SRAM), a Dynamic RAM (DRAM), a Synchronous DRAM (SDRAM), a Double Data Rate SDRAM (DDR SDRAM), an Enhanced SDRAM (ESDRAM), a Synch link DRAM (SLDRAM), and a Direct Rambus RAM (DRRAM). The memory 1409 in the embodiments of the present application includes but is not limited to these and any other suitable types of memory.

[0380] The processor 1410 can include one or more processing units; optionally, the processor 1410 integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to an operating system, a user interface, and an application program, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It can be understood that the above-mentioned modem processor can also not be integrated into the processor 1410.

[0381] The processor 1410 is configured to determine model information of at least one first sub-model in a plurality of sub-models of a target AI model which needs to update model parameters, wherein the target AI model includes a plurality of sub-models, and each sub-model has at least partially separate model information.

[0382] The radio frequency unit 1401 is configured to transmit model information of the at least one first sub-model to the second communication device, where the model information comprises sub-model parameters of the first sub-model. Alternatively,

[0383] The radio frequency unit 1401 is configured to receive model information of at least one first sub-model transmitted by the first communication device, where the at least one first sub-model is a sub-model in a plurality of sub-models of a target AI model that needs to update model parameters, the target AI model comprises a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information comprises sub-model parameters of the first sub-model.

[0384] The processor 1410 is configured to update model parameters of the at least one first sub-model based on the model information of the at least one first sub-model.

[0385] It can be understood that the implementation process of each implementation manner mentioned in the embodiment can refer to the related description of the method embodiments 300 to 700 and achieve the same or corresponding technical effects. To avoid repetition, details are not described here.

[0386] The embodiment of the application further provides a network side device, which comprises a processor and a communication interface, the communication interface is coupled with the processor, and the processor is configured to run programs or instructions to implement the steps of the method embodiments shown in FIGS. 3 to 7. The network side device embodiment corresponds to the network side device method embodiment described above. Each implementation process and implementation manner of the method embodiments described above can be applied to the network side device embodiment and can achieve the same technical effects.

[0387] Specifically, the embodiment of the application further provides a network side device, which can be the model parameter transmission apparatus shown in FIGS. 11 or 12. As shown in FIG. 15, the network side device 1500 comprises an antenna 151, a radio frequency device 152, a baseband device 153, a processor 154 and a memory 155. The antenna 151 is connected with the radio frequency device 152. In the uplink direction, the radio frequency device 152 receives information through the antenna 151 and sends the received information to the baseband device 153 for processing. In the downlink direction, the baseband device 153 processes the information to be sent and sends it to the radio frequency device 152. The radio frequency device 152 processes the received information and sends it out through the antenna 151.

[0388] The method performed by the network side device in the above embodiment can be implemented in the baseband device 153, which comprises a baseband processor.

[0389] The baseband device 153 can include at least one baseband board on which a plurality of chips are disposed, as shown in FIG. 15, one of the chips being, for example, a baseband processor connected with the memory 155 through a bus interface to invoke a program in the memory 155 to perform the network device operation shown in the above method embodiments.

[0390] The network side device can further include a network interface 156, for example, a common public radio interface (CPRI).

[0391] Specifically, the network side device 1500 of the embodiments of the present application further includes instructions or programs stored in the memory 155 and executable on the processor 154, the processor 154 invoking the instructions or programs in the memory 155 to perform the method executed by the modules shown in FIG. 11 or 12 and achieve the same technical effects, and thus the details are not repeated here.

[0392] Specifically, the embodiments of the present application further provide a network side device. As shown in FIG. 16, the network side device 1600 includes a processor 1601, a network interface 1602 and a memory 1603. The network side device can be the model parameter transmission device shown in FIG. 11 or 12. The network interface 1602 is, for example, a common public radio interface (CPRI).

[0393] Specifically, the network side device 1600 of the embodiments of the present application further includes instructions or programs stored in the memory 1603 and executable on the processor 1601, the processor 1601 invoking the instructions or programs in the memory 1603 to perform the method executed by the modules shown in FIG. 11 or 12 and achieve the same technical effects, and thus the details are not repeated here.

[0394] The embodiments of the present application further provide a readable storage medium having a program or instructions stored thereon, the program or instructions being executed by a processor to implement each process of the above model parameter transmission method embodiments and achieve the same technical effects, and thus the details are not repeated here.

[0395] The processor is the processor in the terminal in the above embodiments. The readable storage medium includes a computer readable storage medium, such as a computer readable only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc. In some examples, the readable storage medium can be a non-transitory readable storage medium.

[0396] The chip provided by the embodiment of the present application also can be called a system chip, a chip system, a system on chip, or the like.

[0397] It should be understood that the chip mentioned in the embodiment of the present application can also be called a system chip, a chip system, a system on chip, or the like.

[0398] The embodiment of the present application further provides a computer program / program product stored in a storage medium, which is executed by at least one processor to implement the processes of the above-mentioned model parameter transmission method embodiments and achieve the same technical effects. To avoid repetition, details are not described herein.

[0399] The embodiment of the present application further provides a model parameter transmission system, which comprises a first communication device and a second communication device. The first communication device can be used to execute the steps of the model parameter transmission method 300-600 described above, and the second communication device can be used to execute the steps of the model parameter transmission method 700 described above.

[0400] It should be noted that, in this document, the terms "comprising", "including", or any other variant thereof are intended to cover non-exclusive inclusions, so that processes, methods, articles, or devices that include a series of elements not only include those elements, but also include other elements not explicitly listed, or include elements inherent to such processes, methods, articles, or devices. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article, or device that includes the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to the order of performing the functions shown or discussed, and can also include performing the functions in a substantially simultaneous manner or in reverse order, for example, the described method can be performed in an order different from that described, and various steps can also be added, omitted, or combined. In addition, the features described with reference to certain examples can be combined in other examples.

[0401] From the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment method can be realized by means of computer software product and general hardware platform, of course, it can also be realized by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, magnetic disk, optical disk, etc.), which includes a plurality of instructions for making the terminal or network side device execute the method described in each embodiment of the present application.

[0402] The embodiments of the present application are described above with reference to the accompanying drawings, but the present application is not limited to the above-described specific embodiments, and the above-described specific embodiments are merely illustrative, but not restrictive, and a person of ordinary skill in the art can make many forms of embodiments under the inspiration of the present application without departing from the purpose of the present application and the scope protected by the claims, and these embodiments all belong to the protection of the present application.

Claims

1. A model parameter transmission method, comprising: determining, by a first communication device, model information of at least one first sub-model in a plurality of sub-models of a target artificial intelligence (AI) model that needs to update model parameters, wherein the target AI model comprises a plurality of sub-models, each of which has at least partially separate model information; transmitting, by the first communication device, the model information of the at least one first sub-model to a second communication device, wherein the model information comprises sub-model parameters of the first sub-model.

2. The method of claim 1, wherein, The model information further comprises at least one of the following: model structure information of the first sub-model, identification information of the first sub-model, and version information of the first sub-model.

3. The method of claim 2, wherein, The identification information of the first sub-model comprises at least one of the following: an identification (ID) of the first sub-model; location information of the first sub-model in the target AI model.

4. The method of claim 2, wherein, The version information of the first sub-model comprises at least one of the following: a timestamp of the first sub-model, used to indicate at least one of the following: an update time of the first sub-model, a time of issuing the sub-model parameters of the first sub-model, a time of receiving the sub-model parameters of the first sub-model, and a timestamp in the sub-model parameters of the first sub-model; dataset information associated with the target AI model; data feature information associated with the target AI model; base station hardware information associated with the target AI model; base station configuration information associated with the target AI model; physical cell information associated with the target AI model; serving cell information associated with the target AI model; area information associated with the target AI model; cell group information associated with the target AI model; cell list information associated with the target AI model; a function associated with the target AI model; a characteristic associated with the target AI model.

5. The method according to any one of claims 1 to 4, wherein, The target AI model comprises at least one of the following: a first AI model used by the first communication device or the second communication device; a reference AI model of the first AI model used by the first communication device or the second communication device; a second AI model used in testing by the first communication device, the second communication device, or a test device, wherein the test device is a device other than the first communication device and the second communication device; a reference model of the second AI model used in testing by the first communication device, the second communication device, or the test device, wherein the test device is a device other than the first communication device and the second communication device; a third AI model used by the first communication device, the second communication device, or the test device to match the second AI model used in testing; a reference model of the third AI model.

6. The method according to any one of claims 1 to 5, wherein, Before the first communication device determines the model information of at least one first sub-model in a plurality of sub-models of a target artificial intelligence (AI) model that needs to update model parameters, the method further comprises: The first communication device sends first indication information to the second communication device, wherein the first indication information is used to indicate one of the following: identification information of a second sub-model in the target AI model that needs to be updated, and identification information of a third sub-model in the target AI model that does not need to be updated.

7. The method of claim 6, wherein, After the first communication device sends the first indication information to the second communication device, the method further includes: The first communication device receives first feedback information sent by the second communication device, wherein the first feedback information includes at least one of the following: identification information of a first sub-model supported by the second communication device for updating; and confirmation information used to indicate support for updating of the second sub-model that needs to be updated as indicated by the first indication information.

8. The method of claim 6 or 7, wherein, Before the first communication device sends the first indication information to the second communication device, the method further includes: The first communication device receives second indication information sent by the second communication device, wherein the second indication information includes identification information or version information of at least one fourth sub-model of the target AI model.

9. The method of claim 8, wherein, The second sub-model indicated in the first indication information is a subset of the at least one fourth sub-model.

10. The method according to any one of claims 1 to 5, wherein, Before the first communication device determines model information of at least one first sub-model that needs to update model parameters in a plurality of sub-models of a target artificial intelligence (AI) model, the method further includes: The first communication device sends identification information or version information of at least one fifth sub-model of the target AI model to the second communication device; The first communication device receives second feedback information sent by the second communication device, wherein the second feedback information includes identification information or version information of the first sub-model that needs to be updated.

11. The method of claim 10, wherein, The first communication device sends identification information or version information of at least one fifth sub-model of the target AI model to the second communication device, including one of the following: The first communication device broadcasts the identification information or version information of the at least one fifth sub-model of the target AI model; The first communication device sends the identification information or version information of the at least one fifth sub-model of the target AI model through a dedicated channel or dedicated signaling with the second communication device.

12. The method of claim 10, wherein, The at least one fifth sub-model is a sub-model that needs to be updated.

13. The method of claim 10, wherein, The at least one fifth sub-model is a sub-model that does not need to be updated.

14. The method of claim 12, wherein, The at least one first sub-model is a sub-model in the at least one fifth sub-model.

15. The method according to any one of claims 1 to 14, wherein, After the first communication device transmits model information of the at least one first sub-model to the second communication device, the method further includes: The first communication device receives third feedback information sent by the second communication device, wherein the third feedback information includes at least one of the following: Identification information of a sub-model to be updated by the second communication device; Identification information of a completed updated sub-model of the second communication device; Identification information of a sub-model that the second communication device has not completed updating; Identification information of a failed updated sub-model of the second communication device; Signaling or information of a completed sub-model update; Signaling or information of a failed sub-model update.

16. The method of any one of claims 1 to 15, wherein, The method further comprises: updating, based on the target information, version information of the at least one first sub-model during updating of the at least one first sub-model based on the sub-model parameter of the at least one first sub-model or in the case of completion of the updating.

17. The method of claim 16, wherein, The target information comprises at least one of: related information when the first communication device transmits the model information of the at least one first sub-model; related information included in or associated with the model information of the at least one first sub-model transmitted by the first communication device; related information of updating of the at least one first sub-model; related information of completion of updating of the at least one first sub-model; related information of feedback of the second communication device on success of the updating.

18. The method of any one of claims 1 to 15, wherein, The method further comprises: The first communication device updates the model identifier of the target AI model, wherein the model identifier of the updated target AI model is different from the model identifier of the target AI model before the updating. The first communication device indicates the model identifier of the updated target AI model to the second communication device.

19. The method according to any one of claims 1 to 15, wherein, The method further comprises: The first communication device receives fourth indication information sent by the second communication device, wherein the fourth indication information is used to indicate the model identifier of the updated target AI model, and the model identifier of the updated target AI model is different from the model identifier of the target AI model before the updating.

20. A model parameter transmission method, comprising: The second communication device receives model information of at least one first sub-model transmitted by the first communication device, wherein the at least one first sub-model is a sub-model in a plurality of sub-models of a target AI model that needs to update a model parameter, the target AI model comprises a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information comprises a sub-model parameter of the first sub-model. The second communication device updates the model parameter of the at least one first sub-model based on the model information of the at least one first sub-model.

21. The method of claim 20, wherein, The model information further comprises at least one of: model structure information of the first sub-model, identification information of the first sub-model, and version information of the first sub-model.

22. The method of claim 21, wherein, The identification information of the first sub-model comprises at least one of: an identification ID of the first sub-model; position information of the first sub-model in the target AI model.

23. The method of claim 21, wherein, The version information of the first sub-model comprises at least one of: a timestamp of the first sub-model, used to indicate at least one of: an updating time of the first sub-model, a time of issuing the sub-model parameter of the first sub-model, a time of receiving the sub-model parameter of the first sub-model, and a timestamp in the sub-model parameter of the first sub-model; dataset information associated with the target AI model; data feature information associated with the target AI model; base station hardware information associated with the target AI model; base station configuration information associated with the target AI model; physical cell information associated with the target AI model; service cell information associated with the target AI model; area information associated with the target AI model; Cell group information associated with the target AI model; Cell list information associated with the target AI model; Function associated with the target AI model; Characteristics associated with the target AI model.

24. The method of any one of claims 20 to 23, wherein, The target AI model comprises at least one of: A first AI model used by the first communication device or the second communication device; A reference AI model of the first AI model used by the first communication device or the second communication device; A second AI model used by the first communication device, the second communication device, or a test device in testing, wherein the test device is a device other than the first communication device and the second communication device; A reference model of the second AI model used by the first communication device, the second communication device, or a test device in testing, wherein the test device is a device other than the first communication device and the second communication device; A third AI model used by the first communication device, the second communication device, or a test device to match the second AI model used in testing; A reference model of the third AI model.

25. The method of any one of claims 20 to 24, wherein, Before the second communication device receives the model information of at least one first sub-model transmitted by the first communication device, the method further comprises: The second communication device receives first indication information sent by the first communication device, wherein the first indication information is used to indicate one of the following: identification information of a second sub-model that needs to be updated in the target AI model, identification information of a third sub-model that does not need to be updated in the target AI model.

26. The method of claim 25, wherein, After the second communication device receives the first indication information sent by the first communication device, the method further comprises: The second communication device sends first feedback information to the first communication device, wherein the first feedback information comprises at least one of the following: identification information of an updated first sub-model supported by the second communication device; confirmation information used to indicate support for updating of the second sub-model that needs to be updated indicated by the first indication information.

27. The method of claim 25 or 26, wherein, Before the second communication device receives the first indication information sent by the first communication device, the method further comprises: The second communication device sends second indication information to the first communication device, wherein the second indication information comprises identification information or version information of at least one fourth sub-model of the target AI model.

28. The method of claim 27, wherein, The second sub-model indicated in the first indication information is a subset of the at least one fourth sub-model.

29. The method of any one of claims 20 to 24, wherein, Before the second communication device receives the model information of at least one first sub-model transmitted by the first communication device, the method further comprises: The second communication device receives identification information or version information of at least one fifth sub-model of the target AI model sent by the first communication device; The second communication device sends second feedback information to the first communication device, wherein the second feedback information comprises identification information or version information of the first sub-model that needs to be updated.

30. The method of claim 29, wherein, The second communication device receives the identification information or version information of at least one fifth sub-model of the target AI model sent by the first communication device, comprising one of: The second communication device receives the identification information or version information of the at least one fifth sub-model of the target AI model broadcasted by the first communication device; The second communication device receives the identification information or version information of the at least one fifth sub-model of the target AI model sent by the first communication device through a dedicated channel or dedicated signaling with the second communication device.

31. The method of any one of claims 20 to 30, wherein, After the second communication device receives the model information of the at least one first sub-model transmitted by the first communication device, the method further comprises: The second communication device sends third feedback information to the first communication device, wherein the third feedback information comprises at least one of the following: Identification information of the sub-model to be updated by the second communication device; Identification information of the updated sub-model completed by the second communication device; Identification information of the sub-model not completed for updating by the second communication device; Identification information of the sub-model failed to update by the second communication device; Signaling or information of the sub-model update completion; Signaling or information of the sub-model update failure.

32. The method of any one of claims 20 to 30, wherein, The method further comprises: During the process of updating the at least one first sub-model based on the sub-model parameters of the at least one first sub-model or in the case of update completion, the version information of the at least one first sub-model is updated based on target information.

33. The method of claim 32, wherein, The target information comprises at least one of the following: Related information when the first communication device transmits the model information of the at least one first sub-model; Related information included in or associated with the model information of the at least one first sub-model transmitted by the first communication device; Related information of the update of the at least one first sub-model; Related information of the update completion of the at least one first sub-model; Related information of the update success feedback by the second communication device.

34. The method of any one of claims 20 to 30, wherein, The method further comprises one of the following: The second communication device receives third indication information sent by the first communication device, and updates the model identification of the target AI model based on the third indication information, wherein the third indication information indicates the model identification of the updated target AI model; The second communication device updates the model identification of the target AI model and sends fourth indication information to the first communication device, wherein the fourth indication information is used to indicate the model identification of the updated target AI model, and the model identification of the updated target AI model is different from the model identification of the target AI model before updating.

35. A model parameter transmission apparatus, comprising: A processing module for determining model information of at least one first sub-model in a plurality of sub-models of a target AI model that needs to update model parameters, wherein the target AI model comprises a plurality of sub-models, and each sub-model has at least partially separate model information; A sending module for transmitting the model information of the at least one first sub-model to a second communication device, wherein the model information comprises sub-model parameters of the first sub-model.

36. The apparatus of claim 35, wherein, The sending module is further configured to send first indication information to the second communication device, where the first indication information is used to indicate one of the following: identification information of a second sub-model in the target AI model that needs to be updated, and identification information of a third sub-model in the target AI model that does not need to be updated.

37. The apparatus of claim 36, wherein, Further comprising: The receiving module is configured to receive first feedback information sent by the second communication device, where the first feedback information includes at least one of the following: identification information of a first sub-model that the second communication device supports to update, and confirmation information used to indicate that the second communication device supports to update the second sub-model that needs to be updated and is indicated by the first indication information.

38. The apparatus of claim 36 or 37, wherein, Further comprising: The receiving module is configured to receive second indication information sent by the second communication device, where the second indication information includes identification information or version information of at least one fourth sub-model of the target AI model.

39. The apparatus of claim 35, wherein, The sending module is further configured to send, to the second communication device, identification information or version information of at least one fifth sub-model of the target AI model; The apparatus further includes a receiving module configured to receive second feedback information sent by the second communication device, where the second feedback information includes identification information or version information of the first sub-model that needs to be updated.

40. An apparatus for model parameter transmission, comprising: A receiving module configured to receive model information of at least one first sub-model transmitted by a first communication device, where the at least one first sub-model is a sub-model of a target AI model that needs to update a model parameter, the target AI model includes a plurality of sub-models, each of the sub-models has at least partially separate model information, and the model information includes a sub-model parameter of the first sub-model; A processing module configured to update a model parameter of the at least one first sub-model based on the model information of the at least one first sub-model.

41. The apparatus of claim 40, wherein, The receiving module is further configured to receive first indication information sent by the first communication device, where the first indication information is used to indicate one of the following: identification information of a second sub-model in the target AI model that needs to be updated, and identification information of a third sub-model in the target AI model that does not need to be updated.

42. The device of claim 41, wherein, Further comprising: The sending module is further configured to send first feedback information to the first communication device, where the first feedback information includes at least one of the following: identification information of a first sub-model that the first communication device supports to update, and confirmation information used to indicate that the first communication device supports to update the second sub-model that needs to be updated and is indicated by the first indication information.

43. The apparatus of claim 41 or 42, wherein, Further comprising: The sending module is further configured to send second indication information to the first communication device, where the second indication information includes identification information or version information of at least one fourth sub-model of the target AI model.

44. The apparatus of claim 40, wherein, The receiving module is further configured to receive, from the first communication device, identification information or version information of at least one fifth sub-model of the target AI model; and The apparatus further includes a sending module configured to send second feedback information to the first communication device, wherein the second feedback information includes identification information or version information of the first sub-model that needs to be updated. 45.A communication device comprising a processor and a memory, the memory storing programs or instructions executable on the processor, the programs or instructions, when executed by the processor, implement the steps of the model parameter transmission method according to any one of claims 1 to 34. 46.A readable storage medium, the readable storage medium storing programs or instructions, the programs or instructions, when executed by a processor, implement the steps of the model parameter transmission method according to any one of claims 1 to 34.

Images