CN118158118A

CN118158118A - Determination method, information transmission method, device and communication equipment of AI network model

Info

Publication number: CN118158118A
Application number: CN202211567286.2A
Authority: CN
Inventors: 孙布勒; 孙鹏
Original assignee: Vivo Mobile Communication Co Ltd
Current assignee: Vivo Mobile Communication Co Ltd
Priority date: 2022-12-07
Filing date: 2022-12-07
Publication date: 2024-06-07
Also published as: WO2024120409A1

Abstract

The application discloses a method for determining an AI network model, an information transmission method, an information transmission device and communication equipment, which belong to the technical field of communication, and the method for determining the AI network model in the embodiment of the application comprises the following steps: the terminal receives first information from network side equipment, wherein the first information comprises relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel; and the terminal determines a target AI network model according to the first information.

Description

Determination method, information transmission method, device and communication equipment of AI network model

Technical Field

The application belongs to the technical field of communication, and particularly relates to a determination method, an information transmission device and communication equipment of an AI network model.

Background

In the related art, artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) network models are deployed at transceivers in a communication network, respectively, to conduct model reasoning by means of AI network models.

However, when the AI network model is applied to a wireless communication system, various scenarios and various services are faced. Thus, since the AI network model needs to be trained separately for each service in each scenario, the training process of the AI network model is computationally intensive.

Disclosure of Invention

The embodiment of the application provides a determining method, an information transmission device and communication equipment for an AI network model, which can acquire a super network with high degree of freedom based on neural network architecture search (Neural Architecture Search, NAS), so that the AI network model applicable to various scenes and various services can be obtained through adjustment of the super network, and the calculated amount of training the AI network model can be reduced.

In a first aspect, there is provided an information transmission method, the method including:

The terminal receives first information from network side equipment, wherein the first information comprises relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel;

and the terminal determines a target AI network model according to the first information.

In a second aspect, there is provided an AI network model determination apparatus, including:

A first receiving module, configured to receive first information from a network side device, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, where the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel;

And the first determining module is used for determining a target AI network model according to the first information.

In a third aspect, there is provided an information transmission method, including:

The network side equipment sends first information to the terminal, wherein the first information comprises relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel.

In a fourth aspect, there is provided an information transmission apparatus comprising:

The terminal comprises a first sending module, a second sending module and a third sending module, wherein the first sending module is used for sending first information to the terminal, the first information comprises relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel.

In a fifth aspect, there is provided a communication device comprising a processor and a memory storing a program or instructions executable on the processor, which when executed by the processor, implement the steps of the method according to the first or third aspect.

In a sixth aspect, a terminal is provided, including a processor and a communication interface, where the communication interface is configured to receive first information from a network side device, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel; the processor is configured to determine a target AI network model based on the first information.

In a seventh aspect, a network side device is provided, including a processor and a communication interface, where the communication interface is configured to send first information to a terminal, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, and the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel.

In an eighth aspect, there is provided a communication system comprising: a terminal operable to perform the steps of the information transmission method as described in the first aspect, and a network side device operable to perform the steps of the information processing method as described in the third aspect.

In a ninth aspect, there is provided a readable storage medium having stored thereon a program or instructions which when executed by a processor, performs the steps of the method according to the first aspect or performs the steps of the method according to the third aspect.

In a tenth aspect, there is provided a chip comprising a processor and a communication interface, the communication interface and the processor being coupled, the processor being for running a program or instructions to implement the method according to the first aspect or to implement the method according to the third aspect.

In an eleventh aspect, there is provided a computer program/program product stored in a storage medium, the computer program/program product being executable by at least one processor to implement the steps of the information transmission method as described in the first aspect, or the computer program/program product being executable by at least one processor to implement the steps of the information processing method as described in the third aspect.

In the embodiment of the application, a terminal receives first information from network side equipment, wherein the first information comprises relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel; and the terminal determines a target AI network model according to the first information. The target AI network model acquired by the terminal is determined based on the super network model, and compared with the mode of training the target AI network model based on training samples in the related art, the calculation amount for obtaining the target AI network model can be reduced.

Drawings

Fig. 1 is a schematic diagram of a wireless communication system to which embodiments of the present application can be applied;

FIG. 2 is a graph of performance gains for predicted and non-predicted CSI based on a network model;

FIG. 3 is a flowchart of a method for determining an AI network model according to an embodiment of the application;

FIG. 4 is a schematic diagram of the structure of a super network model;

fig. 5 is a schematic diagram of the weighting coefficients of the super network model.

Fig. 6 is a flowchart of an information transmission method according to an embodiment of the present application;

Fig. 7 is a schematic structural diagram of a determining device for an AI network model according to an embodiment of the present application;

fig. 8 is a schematic structural diagram of an information transmission device according to an embodiment of the present application;

Fig. 9 is a schematic structural diagram of a communication device according to an embodiment of the present application;

fig. 10 is a schematic diagram of a hardware structure of a terminal according to an embodiment of the present application

Fig. 11 is a schematic structural diagram of a network side device according to an embodiment of the present application.

Detailed Description

The technical solutions of the embodiments of the present application will be clearly described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which are derived by a person skilled in the art based on the embodiments of the application, fall within the scope of protection of the application.

The terms first, second and the like in the description and in the claims, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the application are capable of operation in sequences other than those illustrated or otherwise described herein, and that the "first" and "second" distinguishing between objects generally are not limited in number to the extent that the first object may, for example, be one or more. Furthermore, in the description and claims, "and/or" means at least one of the connected objects, and the character "/" generally means a relationship in which the associated object is an "or" before and after.

It should be noted that the techniques described in the embodiments of the present application are not limited to long term evolution (Long Term Evolution, LTE)/LTE evolution (LTE-Advanced, LTE-a) systems, but may also be used in other wireless communication systems, such as code division multiple access (Code Division Multiple Access, CDMA), time division multiple access (Time Division Multiple Access, TDMA), frequency division multiple access (Frequency Division Multiple Access, FDMA), orthogonal frequency division multiple access (Orthogonal Frequency Division Multiple Access, OFDMA), single carrier frequency division multiple access (Single-carrier Frequency Division Multiple Access, SC-FDMA), and other systems. The terms "system" and "network" in embodiments of the application are often used interchangeably, and the techniques described may be used for both the above-mentioned systems and radio technologies, as well as other systems and radio technologies. The following description describes a New Radio (NR) system for exemplary purposes and NR terminology is used in much of the following description, but these techniques may also be applied to applications other than NR system applications, such as 6 th Generation (6G) communication systems.

Fig. 1 shows a block diagram of a wireless communication system to which an embodiment of the present application is applicable. The wireless communication system includes a terminal 11 and a network device 12. The terminal 11 may be a Mobile phone, a tablet Computer (Tablet Personal Computer), a Laptop (Laptop Computer) or a terminal-side device called a notebook, a Personal digital assistant (Personal DIGITAL ASSISTANT, PDA), a palm Computer, a netbook, an ultra-Mobile Personal Computer (ultra-Mobile Personal Computer, UMPC), a Mobile internet appliance (Mobile INTERNET DEVICE, MID), an augmented reality (augmented reality, AR)/Virtual Reality (VR) device, a robot, a wearable device (Wearable Device), a vehicle-mounted device (VUE), a pedestrian terminal (PUE), a smart home (home device with a wireless communication function, such as a refrigerator, a television, a washing machine, a furniture, etc.), a game machine, a Personal Computer (Personal Computer, a PC), a teller machine, or a self-service machine, etc., and the wearable device includes: intelligent wrist-watch, intelligent bracelet, intelligent earphone, intelligent glasses, intelligent ornament (intelligent bracelet, intelligent ring, intelligent necklace, intelligent anklet, intelligent foot chain etc.), intelligent wrist strap, intelligent clothing etc.. It should be noted that the specific type of the terminal 11 is not limited in the embodiment of the present application. The network-side device 12 may include an access network device or a core network device, where the access network device 12 may also be referred to as a radio access network device, a radio access network (Radio Access Network, RAN), a radio access network function, or a radio access network element. Access network device 12 may include a base station, which may be referred to as a node B, an evolved node B (eNB), an access Point, a base transceiver station (Base Transceiver Station, BTS), a radio base station, a radio transceiver, a Basic service set (Basic SERVICE SET, BSS), an Extended service set (Extended SERVICE SET, ESS), a home node B, a home evolved node B, a Transmission/Reception Point (TRP), or some other suitable terminology in the art, and the base station may be referred to as a node B, an evolved node B (eNB), an access Point, a base transceiver station (Base Transceiver Station, BTS), a radio base station, a radio transceiver, a Basic service set (Basic SERVICE SET, BSS), an Extended service set (Extended SERVICE SET, ESS), a home node B, a home evolved node B, a Transmission/Reception Point (TRP), or some other suitable terminology in the art, as long as the same technical effect is achieved, and it should be noted that the base station is not limited to a specific technical vocabulary, and that in the embodiments of the present application, only the base station in the NR system is described by way of example, and not limited to a specific type of base station.

Artificial intelligence is currently in wide-spread use in various fields. There are a number of implementations of AI models, such as neural networks, decision trees, support vector machines, bayesian classifiers, etc. The present application is described by way of example with respect to neural networks, but is not limited to a particular type of AI model.

The parameters of the neural network are optimized by an optimization algorithm. An optimization algorithm is a class of algorithms that can help us minimize or maximize an objective function (sometimes called a loss function). Whereas 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, the predicted output f (X) can be obtained from the input X, and the difference (f (X) -Y) between the predicted value and the true value, which is the loss function, can be calculated. Our aim is to find the appropriate weights and offsets to minimize the value of the above-mentioned loss function, the smaller the loss value, the closer our model is to reality.

The current common optimization algorithm is basically based on a Back Propagation (BP) algorithm. The basic idea of the BP algorithm is that the learning process consists of two processes, forward propagation of the signal and backward propagation of the error. In forward propagation, an input sample is transmitted from an input layer, is processed layer by each hidden layer, and is transmitted to an output layer. If the actual output of the output layer does not match the desired output, the back propagation phase of the error is shifted. The error back transmission is to make the output error pass through hidden layer to input layer in a certain form and to distribute the error to all units of each layer, so as to obtain the error signal of each layer unit, which is used as the basis for correcting the weight of each unit. The process of adjusting the weights of the layers of forward propagation and error back propagation of the signal is performed repeatedly. The constant weight adjustment process is the learning training process of the network. This process is continued until the error in the network output is reduced to an acceptable level or until a preset number of learnings is performed.

In general, the AI algorithm chosen and the model employed will also vary depending on the type of solution. According to the current published papers and published research results, the main method for improving the performance of the 5G network by means of AI is to enhance or replace the existing algorithm or processing module by using the algorithm and model based on the neural network. In certain scenarios, neural network-based algorithms and models may achieve better performance than deterministic-based algorithms. More common neural networks include deep neural networks, convolutional neural networks, recurrent neural networks, and the like. By means of the existing AI tool, the construction, training and verification work of the neural network can be realized.

The AI or machine learning (MACHINE LEARNING, ML) method can be used for replacing the modules in the existing system, so that the system performance can be effectively improved. For example: CSI prediction may be performed based on an AI network model, i.e., historical CSI is input to the AI model, which analyzes time-domain variation characteristics of the channel and outputs future CSI. As shown in fig. 2, when the AI network model is used to predict CSI at different future times, the performance gain (e.g., normalized mean square error (Normalized Mean Squared Error, NMSE)) obtained by the AI network model is greatly improved compared with the scheme of not predicting CSI, and the prediction accuracy that can be achieved is different at different future times.

When the AI network model is applied to a wireless communication system, a corresponding neural network needs to be run on a terminal. However, as the terminal moves, the model used by the terminal needs to be changed, evolved, updated, and the like due to a change in the wireless environment, a change in the execution service, and the like.

In the related art, when retraining or updating an AI network model, a large amount of training sample data with labels is required to train the latest AI network model, and the trained AI network model is synchronized to a terminal and a network side device, and a process of training the latest AI network model in the process needs to occupy a large amount of computing resources and has long time delay.

In the embodiment of the application, the network side device can issue the exceeding network model to the terminal, and when the AI network model of the terminal is not matched with the current communication environment or the execution service along with the movement of the terminal, the change of the wireless environment, the change of the execution service and the like, the terminal can update the terminal to obtain the final target AI network model on the basis of the super network model, or the network side device can determine the first AI network model needed by the terminal on the basis of the super network model according to the requirement of the terminal and issue the first AI network model to the terminal. In the process, an AI network model required by the terminal is generated in a new scene and a new service based on a super network model pre-trained by network side equipment. The embodiment of the application can reduce the calculation amount and time delay occupied by the process of training the AI network model required by the terminal.

The method for determining an AI network model, the method for transmitting information, the device for determining an AI network model, the device for transmitting information, the communication device, and the like provided by the embodiments of the present application are described in detail below with reference to the accompanying drawings by some embodiments and application scenarios thereof.

Referring to fig. 3, in the method for determining an AI network model provided by the embodiment of the present application, an execution subject is a terminal, and as shown in fig. 3, the method for determining an AI network model performed by the terminal may include the following steps:

Step 301, a terminal receives first information from a network side device, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel.

And 302, the terminal determines a target AI network model according to the first information.

It is worth mentioning that the procedure of determining the AI network model based on the super network model may be a procedure of determining the AI network model with reference to a neural network architecture search (Neural Architecture Search, NAS) technique.

The NAS is described below:

The application of machine learning requires a great deal of manual intervention, which is manifested in: feature extraction, model selection, parameter adjustment, etc. Automated machine learning (AutoML) attempts to automatically learn these important steps related to features, models, optimization, and evaluation so that the machine learning model can be applied without human intervention. From a machine learning perspective AutoML can be seen as a very powerful system for learning and generalizing on given data and tasks. From an automation perspective AutoML can then be viewed as designing a series of advanced control systems to operate the machine learning model so that the model can automatically learn the appropriate parameters and configuration without human intervention.

NAS is a sub-area of AutoML, which shifts the line of sight from manual design of network architecture to network architecture design automation, freeing researchers from the complex network design effort and enabling classification accuracy to reach or even exceed that of manual design.

The NAS method most widely used at present is the differentiable method based on the super network model.

For example: the super network model as shown in fig. 4, the super network model includes a plurality of layers (e.g., layer1 to layer 20), and the total number of layers may be manually set. Each level of the super network model is composed of a plurality of neural network blocks (referred to as "network blocks" or blocks "in the following embodiments of the present application) connected in parallel, and the plurality of neural network blocks of different levels may be the same or different. Each neural network block is a pre-defined (typically validated, small neural network of excellent performance).

The above-mentioned parallel connection of network blocks within the same hierarchy can be understood as: the inputs of the network blocks within the same hierarchy are the same and the output of that hierarchy is the weighted sum of the outputs of all the network blocks within the present hierarchy.

For example: as shown in fig. 5, the output of each level is a weighted sum of the outputs of all the neural network blocks of the level, that is, the product of the output result of the nth neural network block of the i-th level and the weight coefficient of the nth neural network block of the i-th level, to obtain a first value, and then, the first values of the 9 neural network blocks in the i-th level are summed, so that the obtained result is the output of the i-th level. Where the weight coefficient α _n ⁽ⁱ⁾ is a variable to be learned, where α _n ⁽ⁱ⁾ represents the weight coefficient of the nth network block in the ith hierarchy. The training data is input into the super network model, and then the super network model is trained by means of supervised learning, semi-supervised learning and the like. There are two kinds of parameters that can be trained during training, namely the training parameters of each neural network block per se (which is a set of parameters, depending on the construction of the neural network block) and the weights of each neural network block per level (which is a real number). And after training the super network model, generating a final neural network according to the weight. For example, as shown in the following chart, a neural network block with the largest weight can be selected in each level, and the neural networks of all the selected levels are quickly connected in series to obtain the final neural network. And a single or a plurality of neural network blocks can be selected at each level according to the self demand and the distribution characteristic of the weight to form a final neural network.

It is worth mentioning that AI network modes required by the terminal device are continuously updated, for example: according to the method for determining the AI network model, which is provided by the embodiment of the application, the AI network model matched with the communication environment, the service and the like of the terminal can be obtained by training according to the actual needs of the terminal on the basis of the super network model, so that the NAS can be used for generating a neural network model in a new scene and a new service, and the NAS is used for selecting a final network model from the super network model.

In one embodiment, the target AI network model may include an AI network model updated as the terminal moves, the communication environment in which it is located changes, traffic changes, etc., i.e., the excess network model is the basis for the AI network model update, or the target AI network model may include an initial AI network model, i.e., the excess network model is the basis for determining the initial AI network model. In the embodiment of the present application, the target AI network model is mainly an updated AI network model, which is not specifically limited herein for illustration.

In one embodiment, the network side device may send relevant parameters of the super network model to the terminal, and the terminal may generate a required AI network model according to the super network model. For example: the terminal selects a neural network block with a maximum weight coefficient from each hierarchy of the super network model to form a target AI network model.

Wherein the super network model may represent a network model comprising I tiers, and each tier of the network model comprises at least one network block, wherein I is an integer greater than 1, and if a tier of the super network model comprises a plurality of network blocks, at least two network blocks of the tier are in parallel with each other.

In one embodiment, the network blocks in the super network model may be pre-trained or defined network blocks, such as a predefined set of network blocks including at least two network blocks, the network blocks in each hierarchy in the super network model may be selected from the set of network blocks, and different hierarchies in the super network model may select the same or different network blocks, such as: each level in the super network model includes all network blocks in the set of network blocks, and the weighting coefficients of the same network block at different levels may be different.

In one embodiment, the terminal may update the target AI network model based on the super network model according to the indication of the network side device.

In one embodiment, the network side device may determine a first AI network model based on a pre-trained super network model according to information such as an environment, a service, a requirement, and the like where the terminal is located, and send the first AI network model to the terminal, where the first AI network model may be an AI network model matched with the information such as the environment, the service, the requirement, and the like where the terminal is located. At this time, the terminal may determine the first AI network model as the target AI network model, or the terminal may perform fine tuning based on the first AI network model to obtain the target AI network model.

Optionally, the relevant parameters of the above-mentioned super network model may include at least one of the following:

the number of network blocks each hierarchy in the super network model has;

A first parameter for each network block in each hierarchy in the super network model;

the weight coefficient of each network block in each hierarchy in the super network model;

First indication information indicating a location or identity of a connection point between adjacent tiers in the super network model.

Optionally, the number of network blocks in each hierarchy in the super network model may be the same or different, for example: the ith hierarchy has Ni network blocks, where I may be an integer from 1 to I, I being the number of hierarchies that the super network model includes.

In option two, the first parameter may include parameters of the network block other than the weight coefficient, and the first parameter may be a parameter set, for example, if the network block includes a neural network model, the first parameter includes a weight (multiplicative coefficient), a bias (additive coefficient), an activation function, and the like of each neuron in the neural network model. Based on the first parameter, the corresponding network block may be restored.

Option three, the weight coefficient may represent a weight occupied by an output result of different network blocks in the same hierarchy, for example: the output result of a certain level is the sum of the products of the output result of each network block in the level and the corresponding weight coefficient.

In an alternative embodiment, in determining the AI network model based on the super network model, each hierarchical network block of the AI network model may be selected according to a weighting coefficient of the respective network block, for example: one or at least two of the highest weighting coefficients in each of the levels of the AI network model are selected and the final AI network model is constructed from the selected network blocks in all levels.

In option four, in one embodiment, a network block may include at least one layer of neural network, where the location or the identifier of the connection point between adjacent layers in the super network model is indicated by the first indication information, so that the adjacent layers in the super network model may be explicitly divided.

Optionally, the first indication information includes:

And a weighted sum of the output information of each network block and a corresponding weight coefficient in a first target hierarchy in the super network model, wherein a layer in the super network model comprises the first target hierarchy.

In this embodiment, the first coefficient is obtained by multiplying the output result of the network block by the weight coefficient of the network block, and then the result obtained by summing the first coefficients of all the network blocks in the same hierarchy is used as the position of the connection point between adjacent hierarchies in the super network model. In this way, when the terminal knows the relevant parameters of the super network model, the adjacent layers in the super network model can be divided according to the value indicated by the first indication information.

As an optional implementation manner, the terminal and the network side device agree on second information, where the second information includes at least one of the following:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

A set of network blocks comprising all levels of network blocks in the super network model;

An identification of each network block in the set of network blocks.

In an embodiment, the terminal and the network side device agree on the second information, which may be that the terminal and the network side device agree on the second information through at least one of a protocol convention, a network side device indication, and a terminal reporting.

In one embodiment, the identification of each network block in the set of network blocks may be a number of each network block in the set of network blocks, for example: the network block set includes 10 network blocks, and then the 10 network blocks may be numbered 1 to 10 in sequence.

Of course, in implementation, the identification of the network block may also be other types of identifications, such as: the type, parameters, number of layers of the neural network, etc. of the neural network model used by the network block can uniquely identify the identity of the network block in the network block set, and is not particularly limited herein.

In one embodiment, the network side device and the terminal can agree on the network block set, so that when the network side device issues the super network model, the identifier of the network block corresponding to each level in the super network model can be issued. For example: assuming that the super network model includes 3 tiers, the relevant parameters of the super network model may include that the network blocks of the first tier include { neural network block 3}, the network blocks of the second tier include { neural network block 1, neural network block 3, neural network block 8}, and the network blocks of the third tier include { neural network block 4}.

In one embodiment, by specifying the format of the input information of the super network model, the terminal can process the acquired data into information conforming to the format of the input information of the super network model according to the format of the input information of the super network model, so as to determine the target AI network model from the super network model by using the acquired data, and/or to input the acquired data into the AI network model determined based on the super network model.

In one embodiment, by specifying the format of the output information of the super network model, the terminal can be enabled to learn the meaning of the output information of the super network model or the AI network model determined based on the super network model, so as to facilitate the subsequent processing of the output information, for example: and determining a CSI prediction result based on the output information.

In one embodiment, the number of tiers included in the super network model may be determined according to at least one of terminal capabilities, terminal requirements, user settings, and the like, and is not specifically limited herein.

As an optional implementation manner, before the terminal receives the first information from the network side device, the method further includes:

The terminal sends first request information to the network side equipment, wherein the first request information is used for requesting the first information.

In this embodiment, when the terminal has a requirement of the target AI network model, for example, when the terminal has a new application scenario or service, the terminal may send first request information to the network side device to request the first information. In this way, the terminal can update or retrain the AI network model according to the first information requested to obtain the target AI network model matched with the new application scene or service.

As an alternative embodiment, before the terminal determines the target AI network model according to the first information, the method further includes:

The terminal receives second indication information from the network side equipment, wherein the second indication information is used for indicating the terminal to determine the target AI network model based on the super network model.

In this embodiment, the terminal performs an operation of determining the target AI network model based on the super network model under the direction of the network side device.

Of course, in another embodiment, the terminal may also actively request the network side device to allow the terminal to determine the operation of the target AI network model based on the super network model, which is not specifically limited herein.

As an optional implementation manner, the determining, by the terminal, the target AI network model according to the first information includes:

The terminal selects Ni network blocks from the ith hierarchy of the super network model, wherein I is an integer less than or equal to I, I is the number of the hierarchies included in the super network model, and Ni is a positive integer;

the terminal determines the target AI network model from network blocks selected from all tiers of the super network model.

In one embodiment, ni corresponding to each level of the super network model may be the same or different.

In one embodiment, ni for each level of the target AI network model is equal to 1. For example: one network block may be selected from each hierarchy of the super network model, i.e., only one network block per hierarchy of the target AI network model.

In one embodiment, the Ni network blocks include network blocks having a weight coefficient greater than a preset value among all network blocks in the i-th hierarchy. For example: one network block with the largest weight coefficient or at least two network blocks with the front weight coefficients may be selected from each hierarchy of the super network model.

In one embodiment, the terminal may select at least one network block from each level of the super network model and generate the target AI network model from the network blocks selected in all levels.

Optionally, the terminal may select at least one network block from each hierarchy of the super network model, and perform series connection on the selected network blocks according to the hierarchy, and then perform fine tuning according to sample data obtained by measurement of the terminal to obtain the target AI network model.

As an alternative embodiment, in case at least one hierarchy in the target AI network model comprises at least two network blocks, the method further comprises:

And the terminal sends third information to the network side equipment, wherein the third information comprises a scheme for determining the target AI network model based on the super network model.

In one embodiment, the determining the target AI network model based on the super network model may be selecting at least one network block from each level of the super network model, and connecting the network blocks selected from all levels of the super network model in series to obtain the target AI network model. At this time, determining the scheme of the target AI network model based on the super network model may be a scheme of how to select a network block from each hierarchy of the super network model, or include selecting which network block or blocks from each hierarchy of the super network model. For example: the number of network blocks selected from each level of the super network model and/or the selected network blocks are weighted and/or supported by terminal capabilities and/or service related etc. in the same level.

In one embodiment, the determination of the target AI network model based on the super network model may be described based on the identity of the network block. For example: the first-tier selected network block includes { neural network block 3}, the second-tier selected network block includes { neural network block 1, neural network block 3, neural network block 8}, the third-tier selected network block includes { neural network block 4}, and so on.

In one embodiment, the third information includes an identification of network blocks of each hierarchy in the target AI network model. Thus, through the third information, the network side device can clearly know the target AI network model actually used by the terminal.

In this embodiment, the terminal sends the third information to the network side device, so that the network side device can obtain, according to the third information, a scheme of determining the target AI network model by the terminal based on the super network model, and further obtain the target AI network model of the terminal.

It should be noted that, the terminal may send the third information to the network side device without model hiding, or the terminal defaults to select a network block with the largest weight coefficient from each level of the super network model.

In the case that the terminal needs to perform model hiding, the terminal may select at least two network blocks from at least one hierarchy of the super network model, and does not send third information to the network side device. At this time, the target AI network model finally determined by the terminal is unknown to the network side device, so that model hiding is realized.

For example: the UE side generates a neural network to be finally used based on the super network model. The general principle is to select a neural network block with the largest weight at each level, and connect the neural networks of all the levels in series quickly to obtain the final neural network. However, model hiding cannot be achieved in this way, i.e. the model used by the UE is completely exposed to the network side. In this regard, the UE may select a single or multiple neural network blocks at each level to form a final neural network according to the distribution characteristics of the weights. For example, the first level selects one neural network block with the largest weight, the second level selects three neural network blocks with the first three weights to be connected in parallel, and the third level selects one neural network block with the second largest weight until all the levels of neural network blocks are selected, so that if the UE does not report its own model generation scheme (i.e., the scheme of generating the finally used model from the super network), the network side cannot exactly master the AI network model used by the UE side.

As an alternative embodiment, the method further comprises:

and the terminal sends fourth information to the network side equipment, wherein the fourth information indicates network blocks supported by the terminal and/or network blocks not supported by the terminal, and the network blocks in the super network model are the network blocks supported by the terminal.

In one embodiment, the terminal may determine which network blocks it is capable of supporting and/or determine which network blocks it is not capable of supporting according to its hardware and/or software configuration.

In this embodiment, the terminal sends the fourth information to the network side device, so that the network side device can determine the super network model by selecting only the network blocks supported by the terminal in the process of determining the super network model, and thus, the adaptation of the super network model and the target AI network model determined based on the super network model and the terminal capability can be improved.

As an alternative embodiment, the method further comprises:

And the terminal sends fifth information to the network side equipment, wherein the fifth information indicates the model size and/or complexity of a target AI network model required by the terminal, and the first AI network model is an AI network model meeting the requirement of the terminal.

In one embodiment, the model size may be described in terms of bits.

Alternatively, the fifth information may indicate the maximum number of bits of the target AI network model required by the terminal.

In one embodiment, the complexity may be described in terms of floating point numbers, such as bits per second.

Alternatively, the fifth information may indicate the maximum floating point number of the target AI network model required by the terminal.

In this embodiment, the terminal sends the fifth information to the network side device, which may be that the network side device determines a first AI network model that meets the requirements of the terminal based on the super network model according to the fifth information, and sends the first AI network model to the terminal, and the terminal may use the first AI network model as the target AI network model, or fine tune the first AI network model to obtain the target AI network model. Thus, the target AI network model acquired by the terminal can be light.

It should be noted that, in the embodiment of the present application, the information interaction between the terminal and the network side device may be air interface information interaction.

Optionally, in the case that the information sending end is a terminal and the information receiving end is a network side device, the information (such as the first request information, the third information, the fourth information, and the fifth information) in the interaction process may be carried in at least one of the following signaling or information:

Layer (layer) 1 signaling of a physical uplink control channel (Physical Uplink Control Channel, PUCCH);

MSG 1 of Physical Random access channel (Physical Random ACCESS CHANNEL, PRACH);

MSG 3 of PRACH;

MSG A of PRACH;

information carried by a Physical Uplink shared channel (Physical Uplink SHARED CHANNEL, PUSCH).

Optionally, in the case that the information sending end is a network side device and the information receiving end is a terminal, the information (such as the first information and the second indication information) in the interaction process may be carried in at least one of the following signaling or information:

A medium access control element (Medium Access Control Control Element, MAC CE);

A radio resource control (Radio Resource Control, RRC) message;

Non-Access Stratum (NAS) messages;

Managing the orchestration message;

User plane data;

Downlink control information (Downlink Control Information, DCI) information;

System information blocks (System Information Block, SIBs);

Layer 1 signaling of the physical downlink control channel (Physical Downlink Control Channel, PDCCH);

information of a physical downlink shared channel (Physical Downlink SHARED CHANNEL, PDSCH);

MSG 2 of PRACH;

MSG 4 of PRACH;

MSG B of PRACH.

Referring to fig. 6, the execution body of the information processing method provided in the embodiment of the present application may be a network side device, and as shown in fig. 6, the information transmission method may include the following steps:

step 601, a network side device sends first information to a terminal, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel.

The first information, the super network model, the relevant parameters of the super network model, the first AI network model, the hierarchy, and the network block in the embodiment of the present application are the same as the first information, the super network model, the relevant parameters of the super network model, the first AI network model, the hierarchy, and the network block in the embodiment of the method shown in fig. 3, and are not specifically limited herein.

The information transmission method in this embodiment corresponds to the method embodiment shown in fig. 3, and is different in that the information transmission method in this embodiment is a network side device, and the execution subject of the method embodiment shown in fig. 3 is a terminal, and the explanation of the information transmission method in this embodiment may refer to the related explanation in the method embodiment shown in fig. 3, which is not described herein, and the terminal and the network side device are mutually configured, for example: the network side equipment determines a super network model and/or a first AI network model and sends first information to the terminal so as to jointly realize that the terminal obtains the target AI network model determined based on the super network model, and further reduce the calculated amount of the terminal for determining the target AI network model.

As an alternative embodiment, the relevant parameters of the super network model include at least one of the following:

the number of network blocks each hierarchy in the super network model has;

first indication information indicating a location or identity of a connection point between adjacent layers in the super network model.

As an optional embodiment, the first indication information includes:

and a weighted sum of the output information of each network block and a corresponding weight coefficient within a first target hierarchy in the super network model, wherein the hierarchy in the super network model comprises the first target hierarchy.

As an optional implementation manner, the network side device and the terminal agree on second information, where the second information includes at least one of the following:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

An identification of each network block in the set of network blocks.

As an optional implementation manner, before the network side device sends the first information to the terminal, the method further includes:

The network side equipment receives first request information from the terminal, wherein the first request information is used for requesting the first information.

As an alternative embodiment, the method further comprises:

The network side equipment sends second indication information to the terminal, wherein the second indication information is used for indicating the terminal to determine a target AI network model based on the super network model.

As an alternative embodiment, the method further comprises:

The network side equipment receives third information from the terminal, wherein the third information comprises a scheme for determining a target AI network model based on the super network model, the target AI network model is an AI network model used by the terminal, and at least one layer in the target AI network model comprises at least two network blocks.

As an alternative embodiment, the third information includes an identification of network blocks of each hierarchy in the target AI network model.

As an alternative embodiment, the method further comprises:

The network side equipment receives fourth information from the terminal, wherein the fourth information indicates network blocks supported by the terminal and/or network blocks not supported by the terminal;

And the network side equipment determines the super network model according to the fourth information and the network block set, wherein the network blocks in the super network model are network blocks supported by the terminal in the network block set.

As an alternative embodiment, the method further comprises:

The network side equipment receives fifth information from the terminal, wherein the fifth information indicates the model size and/or complexity of a target AI network model required by the terminal;

and the network side equipment determines the first AI network model according to the fifth information and the super network model, wherein the first AI network model is an AI network model meeting the requirements of the terminal.

The information transmission method provided by the embodiment of the application corresponds to the determination method of the AI network model in the implementation as shown in fig. 3, and can jointly realize that the terminal obtains the target AI network model determined based on the super network model, thereby reducing the calculation amount of determining the target AI network model by the terminal.

According to the method for determining the AI network model provided by the embodiment of the application, the execution subject can be a device for determining the AI network model. In the embodiment of the present application, a method for determining an AI network model by using a determining device for determining an AI network model is taken as an example, and the determining device for an AI network model provided in the embodiment of the present application is described.

Referring to fig. 7, the apparatus for determining an AI network model provided by the embodiment of the application may be an apparatus in a terminal, and as shown in fig. 7, the apparatus 700 for determining an AI network model may include the following modules:

A first receiving module 701, configured to receive first information from a network side device, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, where the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel;

A first determining module 702 is configured to determine a target AI network model according to the first information.

Optionally, the relevant parameters of the super network model include at least one of the following:

the number of network blocks each hierarchy in the super network model has;

Optionally, the first indication information includes:

Optionally, the terminal and the network side device agree on second information, where the second information includes at least one of:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

An identification of each network block in the set of network blocks.

Optionally, the determining apparatus 700 of the AI network model further includes:

and the second sending module is used for sending first request information to the network side equipment, wherein the first request information is used for requesting the first information.

and the second receiving module is used for receiving second indication information from the network side equipment, wherein the second indication information is used for indicating the terminal to determine the target AI network model based on the super network model.

Optionally, the first determining module 702 is specifically configured to:

Selecting Ni network blocks from the ith hierarchy of the super network model, wherein I is an integer less than or equal to I, I is the number of the hierarchies included in the super network model, and Ni is a positive integer;

The target AI network model is determined from network blocks selected from all tiers of the super network model.

Optionally, ni corresponding to each level of the target AI network model is equal to 1.

Optionally, the Ni network blocks include network blocks with weight coefficients greater than a preset value in all network blocks in the ith hierarchy.

Optionally, in the case that at least one hierarchy in the target AI network model includes at least two network blocks, the AI network model determination apparatus 700 further includes:

And the third sending module is used for sending third information to the network side equipment, wherein the third information comprises a scheme for determining the target AI network model based on the super network model.

Optionally, the third information includes an identification of network blocks of each hierarchy in the target AI network model.

And the fourth sending module is used for sending fourth information to the network side equipment, wherein the fourth information indicates the network blocks supported by the terminal and/or the network blocks not supported by the terminal, and the network blocks in the super network model are the network blocks supported by the terminal.

And a fifth sending module, configured to send fifth information to the network side device, where the fifth information indicates a model size and/or complexity of a target AI network model required by the terminal, and the first AI network model is an AI network model that meets a requirement of the terminal.

The determination device of the AI network model in the embodiment of the application may be an electronic device, for example, an electronic device with an operating system, or may be a component in the electronic device, for example, an integrated circuit or a chip. The electronic device may be a terminal, or may be other devices than a terminal. By way of example, the terminals may include, but are not limited to, the types of terminals 11 listed above, other devices may be servers, network attached storage (Network Attached Storage, NAS), etc., and embodiments of the present application are not limited in detail.

The determination device 700 of the AI network model provided in the embodiment of the present application can implement each process implemented by the terminal in the method embodiment shown in fig. 3, and can obtain the same beneficial effects, so that repetition is avoided, and detailed description is omitted here.

According to the information transmission method provided by the embodiment of the application, the execution main body can be an information transmission device. In the embodiment of the present application, an information transmission device is described by taking an information transmission method performed by an information transmission device as an example.

Referring to fig. 8, an information transmission apparatus provided in an embodiment of the present application may be an apparatus in a network side device, and as shown in fig. 8, the information transmission apparatus 800 may include the following modules:

A first sending module 801, configured to send first information to a terminal, where the first information includes relevant parameters of a super network model or information of a first AI network model generated based on the super network model, where the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel.

the number of network blocks each hierarchy in the super network model has;

Optionally, the first indication information includes:

Optionally, the network side device and the terminal agree on second information, where the second information includes at least one of the following:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

An identification of each network block in the set of network blocks.

Optionally, the information transmission apparatus 800 further includes:

and the third receiving module is used for receiving first request information from the terminal, wherein the first request information is used for requesting the first information.

Optionally, the information transmission apparatus 800 further includes:

and a sixth sending module, configured to send second instruction information to the terminal, where the second instruction information is used to instruct the terminal to determine a target AI network model based on the super network model.

Optionally, the information transmission apparatus 800 further includes:

And a fourth receiving module, configured to receive third information from the terminal, where the third information includes a scheme for determining a target AI network model based on the super network model, the target AI network model being an AI network model used by the terminal, and at least one hierarchy in the target AI network model includes at least two network blocks.

Optionally, the information transmission apparatus 800 further includes:

A fifth receiving module, configured to receive fourth information from the terminal, where the fourth information indicates a network block supported by the terminal and/or a network block not supported by the terminal;

And the second determining module is used for determining the super network model according to the fourth information and the network block set, wherein the network blocks in the super network model are network blocks supported by the terminal in the network block set.

Optionally, the information transmission apparatus 800 further includes:

a sixth receiving module, configured to receive fifth information from the terminal, where the fifth information indicates a model size and/or complexity of a target AI network model required by the terminal;

and the third determining module is used for determining the first AI network model according to the fifth information and the super network model, wherein the first AI network model is an AI network model meeting the requirements of the terminal.

The information transmission device 800 provided in the embodiment of the present application can implement each process implemented by the network side device in the method embodiment shown in fig. 6, and can obtain the same beneficial effects, so that repetition is avoided, and no further description is given here.

Optionally, as shown in fig. 9, the embodiment of the present application further provides a communication device 900, including a processor 901 and a memory 902, where the memory 902 stores a program or instructions that can be executed on the processor 901, for example, when the communication device 900 is a terminal, the program or instructions implement, when executed by the processor 901, the steps of the method embodiment shown in fig. 3, and achieve the same technical effects. When the communication device 900 is a network side device, the program or the instruction, when executed by the processor 901, implements the steps of the method embodiment shown in fig. 6, and the same technical effects can be achieved, so that repetition is avoided, and no further description is given here.

The embodiment of the application also provides a terminal, which comprises a processor and a communication interface, wherein the communication interface is used for receiving first information from network side equipment, the first information comprises relevant parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel; the processor is configured to determine a target AI network model based on the first information.

The terminal embodiment can implement each process executed by the AI network model determining device 700 shown in fig. 7, and can achieve the same technical effects, which are not described herein. Specifically, fig. 10 is a schematic diagram of a hardware structure of a terminal for implementing an embodiment of the present application.

The terminal 1000 includes, but is not limited to: at least some of the components of the radio frequency unit 1001, the network module 1002, the audio output unit 1003, the input unit 1004, the sensor 1005, the display unit 1006, the user input unit 1007, the interface unit 1008, the memory 1009, and the processor 1010, etc.

Those skilled in the art will appreciate that terminal 1000 can also include a power source (e.g., a battery) for powering the various components, which can be logically connected to processor 1010 by a power management system so as to perform functions such as managing charge, discharge, and power consumption by the power management system. The terminal structure shown in fig. 10 does not constitute a limitation of the terminal, and the terminal may include more or less components than shown, or may combine some components, or may be arranged in different components, which will not be described in detail herein.

It should be appreciated that in embodiments of the present application, the input unit 1004 may include a graphics processing unit (Graphics Processing Unit, GPU) 10041 and a microphone 10042, where the graphics processor 10041 processes image data of still pictures or video obtained by an image capturing device (e.g., a camera) in a video capturing mode or an image capturing mode. The display unit 1006 may include a display panel 10061, and the display panel 10061 may be configured in the form of a liquid crystal display, an organic light emitting diode, or the like. The user input unit 1007 includes at least one of a touch panel 10071 and other input devices 10072. The touch panel 10071 is also referred to as a touch screen. The touch panel 10071 can include two portions, a touch detection device and a touch controller. Other input devices 10072 may include, but are not limited to, a physical keyboard, function keys (e.g., volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, and so forth, which are not described in detail herein.

In the embodiment of the present application, after receiving downlink data from the network side device, the radio frequency unit 1001 may transmit the downlink data to the processor 1010 for processing; in addition, the radio frequency unit 1001 may send uplink data to the network side device. In general, the radio frequency unit 1001 includes, but is not limited to, an antenna, an amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, and the like.

The memory 1009 may be used to store software programs or instructions and various data. The memory 1009 may mainly include a first memory area storing programs or instructions and a second memory area storing data, wherein the first memory area may store an operating system, application programs or instructions (such as a sound playing function, an image playing function, etc.) required for at least one function, and the like. Further, the memory 1009 may include volatile memory or nonvolatile memory, or the memory 1009 may include both volatile and nonvolatile memory. The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable EPROM (EEPROM), or a flash Memory. The volatile memory may be random access memory (Random Access Memory, RAM), static random access memory (STATIC RAM, SRAM), dynamic random access memory (DYNAMIC RAM, DRAM), synchronous Dynamic Random Access Memory (SDRAM), double data rate Synchronous dynamic random access memory (Double DATA RATE SDRAM, DDRSDRAM), enhanced Synchronous dynamic random access memory (ENHANCED SDRAM, ESDRAM), synchronous link dynamic random access memory (SYNCH LINK DRAM, SLDRAM), and Direct random access memory (DRRAM). Memory 1009 in embodiments of the application includes, but is not limited to, these and any other suitable types of memory.

The processor 1010 may include one or more processing units; optionally, the processor 1010 integrates an application processor that primarily processes operations involving an operating system, user interface, application programs, and the like, and a modem processor that primarily processes wireless communication signals, such as a baseband processor. It will be appreciated that the modem processor described above may not be integrated into the processor 1010.

The radio frequency unit 1001 is configured to receive first information from a network side device, where the first information includes related parameters of a super network model or information of a first AI network model generated based on the super network model, and the super network model includes at least one hierarchy, each hierarchy has at least one network block, and when there are multiple network blocks in the same hierarchy, at least two network blocks of the multiple network blocks are connected in parallel;

A processor 1010 for determining a target AI network model from the first information.

the number of network blocks each hierarchy in the super network model has;

Optionally, the first indication information includes:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

An identification of each network block in the set of network blocks.

Optionally, before performing the receiving the first information from the network side device, the radio frequency unit 1001 is further configured to send first request information to the network side device, where the first request information is used to request the first information.

Optionally, before the processor 1010 executes the determining the target AI network model according to the first information:

The radio frequency unit 1001 is further configured to receive second indication information from the network side device, where the second indication information is used to instruct the terminal to determine the target AI network model based on the super network model.

Optionally, the determining, by the processor 1010, the target AI network model according to the first information includes:

Optionally, in the case that at least one hierarchy in the target AI network model includes at least two network blocks:

The radio frequency unit 1001 is further configured to send third information to the network side device, where the third information includes a scheme for determining the target AI network model based on the super network model.

Optionally, the radio frequency unit 1001 is further configured to send fourth information to the network side device, where the fourth information indicates a network block supported by the terminal and/or a network block not supported by the terminal, and the network block in the super network model is the network block supported by the terminal.

Optionally, the radio frequency unit 1001 is further configured to send fifth information to the network side device, where the fifth information indicates a model size and/or complexity of a target AI network model required by the terminal, and the first AI network model is an AI network model that meets a requirement of the terminal.

The terminal 1000 provided in the embodiment of the present application can implement each process executed by the AI network model determining device 700 shown in fig. 7, and can obtain the same beneficial effects, and for avoiding repetition, the description is omitted herein.

The embodiment of the application also provides network side equipment, which comprises a processor and a communication interface, wherein the communication interface is used for sending first information to a terminal, the first information comprises related parameters of a super network model or information of a first AI network model generated based on the super network model, the super network model comprises at least one hierarchy, each hierarchy is provided with at least one network block, and when a plurality of network blocks exist in the same hierarchy, at least two network blocks of the plurality of network blocks are connected in parallel.

The network side device embodiment can implement each process executed by the information transmission apparatus 800 shown in fig. 8, and achieve the same technical effects, which are not described herein. Specifically, the embodiment of the application also provides network side equipment. As shown in fig. 11, the network side device 1100 includes: an antenna 1101, a radio frequency device 1102, a baseband device 1103, a processor 1104 and a memory 1105. The antenna 1101 is connected to a radio frequency device 1102. In the uplink direction, the radio frequency device 1102 receives information via the antenna 1101, and transmits the received information to the baseband device 1103 for processing. In the downlink direction, the baseband device 1103 processes information to be transmitted, and transmits the processed information to the radio frequency device 1102, and the radio frequency device 1102 processes the received information and transmits the processed information through the antenna 1101.

The method performed by the network-side device in the above embodiment may be implemented in the baseband apparatus 1103, where the baseband apparatus 1103 includes a baseband processor.

The baseband apparatus 1103 may, for example, include at least one baseband board, where a plurality of chips are disposed, as shown in fig. 11, where one chip, for example, a baseband processor, is connected to the memory 1105 through a bus interface, so as to call a program in the memory 1105 to perform the network device operation shown in the above method embodiment.

The network-side device may also include a network interface 1106, such as a common public radio interface (Common Public Radio Interface, CPRI).

Specifically, the network side device 1100 of the embodiment of the present invention further includes: instructions or programs stored in the memory 1105 and executable on the processor 1104, the processor 1104 invokes the instructions or programs in the memory 1105 to perform the method performed by the modules shown in fig. 8 and achieve the same technical effects, so repetition is avoided and will not be described here.

The embodiment of the present application further provides a readable storage medium, where a program or an instruction is stored, where the program or the instruction realizes each process of the method embodiment shown in fig. 3 or fig. 6 when being executed by a processor, and the process can achieve the same technical effect, so that repetition is avoided and no further description is given here.

Wherein the processor is a processor in the terminal described in the above embodiment. The readable storage medium includes computer readable storage medium such as computer readable memory ROM, random access memory RAM, magnetic or optical disk, etc.

The embodiment of the application further provides a chip, 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, so as to implement each process of the method embodiment shown in fig. 3 or fig. 6, and achieve the same technical effect, so that repetition is avoided, and no further description is provided herein.

It should be understood that the chips referred to in the embodiments of the present application may also be referred to as system-on-chip chips, or the like.

Embodiments of the present application further provide a computer program/program product stored in a storage medium, where the computer program/program product is executed by at least one processor to implement the respective processes of the method embodiments shown in fig. 3 or fig. 6, and achieve the same technical effects, and are not repeated herein.

The embodiment of the application also provides a communication system, which comprises: a terminal operable to perform the steps of the method for determining an AI network model shown in fig. 3, and a network side device operable to perform the steps of the method for transmitting information shown in fig. 6.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element. Furthermore, it should be noted that the scope of the methods and apparatus in the embodiments of the present application is not limited to performing the functions in the order shown or discussed, but may also include performing the functions in a substantially simultaneous manner or in an opposite order depending on the functions involved, e.g., the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a computer software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present application.

The embodiments of the present application have been described above with reference to the accompanying drawings, but the present application is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and many forms may be made by those having ordinary skill in the art without departing from the spirit of the present application and the scope of the claims, which are to be protected by the present application.

Claims

1. A method for determining an AI network model, comprising:

2. The method of claim 1, wherein the relevant parameters of the super network model include at least one of:

the number of network blocks each hierarchy in the super network model has;

3. The method of claim 2, wherein the first indication information comprises:

4. The method of claim 1, wherein the terminal agrees with the network-side device on second information, wherein the second information comprises at least one of:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

An identification of each network block in the set of network blocks.

5. The method according to any one of claims 1 to 4, characterized in that before the terminal receives the first information from the network side device, the method further comprises:

6. The method according to any one of claims 1 to 4, characterized in that before the terminal determines a target AI network model from the first information, the method further comprises:

7. The method according to any one of claims 1 to 4, wherein the terminal determining a target AI network model from the first information comprises:

8. The method of claim 7, wherein Ni for each level of the target AI network model is equal to 1.

9. The method of claim 7, wherein the Ni network blocks include network blocks having a weight coefficient greater than a preset value among all network blocks within the i-th hierarchy.

10. The method of claim 7, wherein, in the case where at least one level in the target AI network model includes at least two network blocks, the method further comprises:

11. The method of claim 10, wherein the third information comprises an identification of network blocks of each tier in the target AI network model.

12. The method according to any one of claims 1 to 4, further comprising:

13. The method according to any one of claims 1 to 4, further comprising:

14. An information transmission method, comprising:

15. The method of claim 14, wherein the parameters associated with the super network model include at least one of:

the number of network blocks each hierarchy in the super network model has;

16. The method of claim 15, wherein the first indication information comprises:

17. The method of claim 14, wherein the network-side device agrees with the terminal on second information, wherein the second information comprises at least one of:

the number of tiers comprised by the super network model;

the format of the input information of the super network model;

The format of the output information of the super network model;

An identification of each network block in the set of network blocks.

18. The method according to any one of claims 14 to 17, wherein before the network side device sends the first information to the terminal, the method further comprises:

19. The method according to any one of claims 14 to 17, further comprising:

20. The method according to any one of claims 14 to 17, further comprising:

21. The method of claim 20, wherein the third information comprises an identification of network blocks of each tier in the target AI network model.

22. The method according to any one of claims 14 to 17, further comprising:

23. The method according to any one of claims 14 to 17, further comprising:

24. A determination apparatus of an AI network model, applied to a terminal, the apparatus comprising:

25. An information transmission apparatus, characterized by being applied to a network-side device, comprising:

26. A communication device comprising a processor and a memory storing a program or instructions executable on the processor, which when executed by the processor, implement the steps of the AI network model determination method of any of claims 1-13, or the steps of the information transfer method of any of claims 14-23.

27. A readable storage medium, wherein a program or instructions is stored on the readable storage medium, which when executed by a processor, implements the steps of the AI network model determination method of any of claims 1 to 13, or implements the steps of the information transmission method of any of claims 14 to 23.