Machine learning model parameter transmission method and device

[0084] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0085] The introduction of machine learning in wireless mobile communication systems can solve complex problems or improve performance of mobile communication systems. However, due to the complex composition of the wireless mobile communication system and the numerous manufacturers, how to deploy/update the machine learning model trained on the network side to the terminal side has become a systematic problem, which requires a unified solution.

[0086] Therefore, the embodiments of the present application provide a method and apparatus for transferring machine learning model parameters, so as to realize transferring the machine learning model trained on the network side to the terminal side.

[0087] The method and the device are conceived based on the same application. Since the principles of the method and the device for solving the problem are similar, the implementation of the device and the method can be referred to each other, and repeated descriptions will not be repeated here.

[0088] The technical solutions provided in the embodiments of the present application may be applicable to various systems, especially a 5G system or a 6G system. For example, applicable systems may be global system of mobile communication (GSM) system, code division multiple access (CDMA) system, wideband code division multiple access (WCDMA) general packet radio service (general packet radioservice, GPRS) system, long term evolution (long term evolution, LTE) system, LTE frequency division duplex (frequency division duplex, FDD) system, LTE time division duplex (time division duplex, TDD), general mobile system ( Universal mobile telecommunication system, UMTS), worldwide interoperability for microwave access (WiMAX) system, 5G system and 5G NR system, etc. These various systems include terminal equipment and network equipment.

[0089] The terminal device involved in the embodiments of the present application may be a device that provides voice and/or data connectivity to a user, a handheld device with a wireless connection function, or other processing device connected to a wireless modem. In different systems, the names of the terminal equipment may be different. For example, in a 5G system, the terminal equipment may be called user equipment (user equipment, UE). The wireless end devices may communicate with one or more core networks via the RAN, and the wireless end devices may be mobile end devices such as mobile phones (or "cellular" phones) and computers with mobile end devices, for example, which may be portable , pocket, handheld, computer built-in or vehicle mounted mobile devices that exchange language and/or data with the radio access network. For example, personal communication service (PCS) phones, cordless phones, session initiated protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (personal digital assistants), PDA) and other devices. Wireless terminal equipment may also be referred to as a system, subscriber unit, subscriber station, mobile station, mobile station, remote station, access point, A remote terminal (remote terminal), an access terminal (access terminal), a user terminal (user terminal), a user agent (user agent), and a user device (user device) are not limited in the embodiments of the present application.

[0090] The network device involved in the embodiments of the present application may be a base station, and the base station may include multiple cells. Depending on the specific application, the base station may also be called an access point, or may refer to a device in the access network that communicates with wireless terminal devices through one or more sectors on the air interface, or other names. The network device may be used to convert received air frames to and from internet protocol (IP) packets, and act as a router between the wireless end device and the rest of the access network, which may include the Internet. Protocol (IP) communication network. The network devices may also coordinate attribute management for the air interface. For example, the network device involved in the embodiments of the present application may be a global system for mobile communications (GSM) or a network device (base transceiver station, BTS) in code division multiple access (code division multiple access, CDMA). , it can also be a network device (NodeB) in wide-band code division multiple access (WCDMA), or it can be an evolved network device in a long term evolution (long term evolution, LTE) system ( evolutional node B, eNB or e-NodeB), 5G base station in 5G network architecture (next generation system), home evolved node B (HeNB), relay node (relay node), home base station (femto ), a pico base station (pico), etc., which are not limited in the embodiments of the present application.

[0091] The various embodiments of the present application will be described in detail below with reference to the accompanying drawings. It should be noted that the display order of the embodiments of the present application only represents the sequence of the embodiments, and does not represent the advantages and disadvantages of the technical solutions provided by the embodiments.

[0092] The embodiment of the present application proposes a method for deploying/updating a machine learning model trained on the system side to the terminal side in a wireless mobile communication system by using a parameter transfer method. The main process of updating the machine learning model is as follows: figure 1 shown, including:

[0093] Step 1: The terminal reports to the base station the list of functional units that support the application of the machine learning model and the type of the machine learning model.

[0094] Table 1 below shows the list of functional units that support the application machine model on the terminal side and an example of the type of machine learning model:

[0095] Table 1:

[0096]

[0097] In particular, the terminal signal detection functional unit supports two types of machine learning models, DNN and RNN.

[0098] Step 2: When the base station decides to deploy/update the terminal-side machine learning model, the base station searches the machine learning model library in the machine learning model library according to the learned functional unit list of the terminal machine learning model and the type of the machine learning model and determines the to-be-deployed/updated according to the deployment status. The updated target functional unit and its corresponding machine learning model parameters.

[0099] The following table 2 gives an example of the machine learning model library, in which the 2nd and 3rd columns are used for search, that is, according to the information of the 2nd and 3rd columns in table 1 row by row, the 2nd and 3rd columns of table 2 are searched for the corresponding Consistent information, and determine the corresponding number. Specifically, the numbers found in this embodiment are #81, #83, #84, #88, #90, #92, and #95. The fourth column is the deployment status. If its status is "Deployed", it means that its corresponding machine learning model parameters have been deployed on the terminal side. If its status is "Not deployed", it means that its corresponding machine learning model has been deployed. The parameters have not been deployed on the terminal side. Specifically, in this embodiment, the functional units whose deployment status is displayed as "undeployed" are further screened and determined as target functional units, that is, functional units corresponding to numbers #81, #84, #88, and #90 respectively. "Terminal channel estimation", "Terminal signal detection", "Terminal channel decoding", "Terminal CQI measurement". The fifth column is the specific machine learning model parameters in the machine learning model library.

[0100] Table 2

[0101]

[0102]

[0103] Step 3: The base station sends the determined parameters of the machine learning model (including the target functional unit, the type of the machine learning model, and the detailed parameters of the machine learning model) to the terminal.

[0104] Step 4: After receiving the machine learning model parameters, the terminal transmits the machine learning model parameters to the corresponding target functional unit.

[0105] Step 5: The terminal side supports the target functional unit of the application machine learning model, and configures and generates a corresponding machine learning model according to the received machine learning model parameters.

[0106] Take the following machine learning model parameters as an example (assuming it has 3 inputs and 6 outputs):

[0107] ● Target functional unit identification: terminal channel decoding;

[0108] ◆Machine learning model type: DNN;

[0109] ●The detailed parameters of the machine model are as follows:

[0110] ■Number of hidden layers: 2

[0111] ◆Parameters of hidden layer #1: the number of neurons is 5;

[0112] ● Parameters of neuron #1-1:

[0113] ■Weight: 0.1, 0.3, 0.5;

[0114] ■ Deviation: 0.1;

[0115] ■Activation function: Sigmoid;

[0116] ● Parameters of neurons #1-2:

[0117] ■Weight: 0.1, 0.1, 0.2;

[0118] Offset: 0;

[0119] ■Activation function: Sigmoid;

[0120] ● Parameters of neurons #1-3:

[0121] ■Weight: 0.2, 0.3, 0.7;

[0122] ■ Deviation: 0.2;

[0123] ■Activation function: Sigmoid;

[0124] ● Parameters of neurons #1-4:

[0125] ■Weight: 0.1, 0, 0;

[0126] Offset: 0;

[0127] ■Activation function: Sigmoid;

[0128] ● Parameters of neurons #1-5:

[0129] ■Weight: 0.2, 0.3, 0.3;

[0130] ■Deviation: 0.8;

[0131] ■Activation function: ReLU;

[0132] ◆Parameters of hidden layer #2: the number of neurons is 5;

[0133] ● Parameters of neuron #2-1:

[0134] ■Weight: 0.1, 0.4, 0.5, 0.7, 0;

[0135] Offset: 0;

[0136] ■Activation function: ReLU;

[0137] ● Parameters of neuron #2-2:

[0138] ■Weight: 0.1, 0.1, 0.1, 0.7, 0.1;

[0139] ■Deviation: 0.5;

[0140] ■Activation function: Sigmoid;

[0141] ● Parameters of neurons #2-3:

[0142] ■Weight: 0.2, 0.4, 0.6, 0.3, 0.7;

[0143] ■Deviation: 0.24;

[0144] ■Activation function: ReLU;

[0145] ● Parameters of neurons #2-4:

[0146] ■Weight: 0, 0.8, 0.1, 0, 0.1;

[0147] ■Deviation: 0.7;

[0148] ■Activation function: Tanh;

[0149] ● Parameters of neurons #2-5:

[0150] ■Weight: 0.2, 0.1, 0.8, 0.3, 0.3;

[0151] ■ Deviation: 0.18;

[0152] ■Activation function: Sigmoid;

[0153] ◆The parameters of the output layer:

[0154] ● Parameters of neuron #O-1:

[0155] ■Weight: 0.1, 0.7, 0, 0.4, 0.5;

[0156] ■ Deviation: 0.6;

[0157] ■Activation function: ReLU;

[0158] ● Parameters of neuron #O-2:

[0159] ■Weight: 0.1, 0.2, 0.1, 0.5, 0.1;

[0160] ■ Deviation: 0.2;

[0161] ■Activation function: Sigmoid;

[0162] Parameters of neuron #O-3:

[0163] ■Weight: 0.2, 0.3, 0.6, 0.13, 0.7;

[0164] ■ Deviation: 0.4;

[0165] ■Activation function: ReLU;

[0166] Parameters of neuron #O-4:

[0167] ■Weight: 0.1, 0.8, 0.2, 0, 0.1;

[0168] ■Deviation: 0.17;

[0169] ■Activation function: Sigmoid;

[0170] Parameters for neuron #O-5:

[0171] ■Weight: 0.2, 0.3, 0.18, 0.3, 0.3;

[0172] Offset: 0;

[0173] ■Activation function: ReLU;

[0174] ● Parameters of neuron #O-6:

[0175] ■Weight: 0.2, 0.2, 0.8, 0.2, 0.3;

[0176] ■ Deviation: 0.2;

[0177] ■Activation function: Sigmoid.

[0178] After receiving the machine learning model parameters, the terminal can determine the target functional unit to which the machine learning model is specifically applied according to the information provided by the target function identifier: terminal channel decoding; Machine learning model type: DNN; according to the description of the detailed parameters of the machine learning model and the following formula, the specific model can be determined, that is, all parameters in the formula are known.

[0179] f(x [0] )=f [L] (f [L-1] (…f [1] (x [0] )))

[0180] Among them, the superscript indicates its corresponding layer number, x [0] is the input of the target functional unit, f(x [0] ) is the output of the target functional unit. The corresponding output f of each layer [n] (x [n-1] ) is a vector consisting of the outputs of all neurons in the layer, where the output of each neuron is where w iis the weight of the neuron in the previous layer connected to it, b is a bias, and f(x) is a nonlinear function, also known as the activation function. Common activation functions include: Sigmoid function, hyperbolic tangent (Tanh: hyperbolic tangent) function, linear rectification (ReLU: Rectified Linear Unit) function, etc.:

[0181] Sigmoid:

[0182] ReLU: f ReLU (z)=max{0,z}

[0183]

[0184] Wherein, in step 1 and step 3, when the base station and the terminal exchange information, high-layer signaling, such as RRC signaling, MAC signaling, etc., may be used.

[0185] For a specific functional unit, in order to apply the above method provided by the embodiment of the present application, its function needs to be identified.

[0186] by figure 2 The function division located in the receiver part is given as an example, in which, in addition to using text identification, such as "channel estimation", "signal detection", "channel decoding", the identification can also be numerically numbered, such as #2, #3, #4.

[0187] In the case that the functional processing unit in the wireless mobile communication system is clearly identified, the machine learning model parameters to be transmitted and delivered include one or a combination of the following parameters:

[0188] target functional unit identification;

[0189] Machine learning model types, such as DNN, CNN, RNN, etc.;

[0190] The detailed parameters of the machine learning model are different for specific machine learning model types.

[0191] Taking the function division and identification of the above receiver part as an example, it is assumed that the machine learning model adopts DNN, and the parameters passed by it include at least one or a combination of the following parameters:

[0192] Target functional unit identification: #3;

[0193] Machine learning model type: DNN;

[0194] Machine learning model detailed parameters:

[0195] number of hidden layers;

[0196] The parameters of each hidden layer: the number of neurons, and the weight, bias, and activation function of each neuron.

[0197] Output layer parameters: the weight, bias, activation function of each neuron.

[0198] After receiving the parameters of the machine learning model, the functional unit on the terminal side can configure and generate the corresponding machine learning model. Taking the block diagram of the aforementioned receiver as an example, assuming that each module is constructed by the DNN model, the schematic diagram of the neural network of all modules connected together is as follows image 3 (the number of hidden layers and neurons is only for illustration). In fact, the wire frame represents the overall model of the neural network designed for the receiver. The realization circle represents the neurons participating in the calculation in the current model, and the dotted circle represents the neurons that do not participate in the calculation in the current model. It can be seen that which are controlled by configuring the activation function Neurons participate in the calculation, that is, the specific implementation of the machine learning model can be realized by passing parameters. The operations represented by each neuron and its connections are only addition and a limited number of types of nonlinear functions, and the operations of multiple neurons in the same layer can be performed in parallel, so that the types of processing operations are simplified and the computational efficiency is improved. promote.

[0199] Correspondingly, an embodiment of the present application provides a system for transferring machine learning model parameters in a mobile communication system, which at least includes a base station side and a terminal side, wherein the base station side includes a control signaling transceiver unit and a machine learning model library; The terminal side includes a control signaling transceiver unit and several functional units that can apply machine learning models.

[0200] In addition to using the signaling of the Uu interface between the terminal and the base station to transfer machine learning model parameters, when there are multiple units on the system side, such as the central unit (Centralized Unit, CU) and the distributed unit (Distributed Unit, DU) , and the signaling between the CU and the DU can also be used to transfer machine learning model parameters.

[0201] In summary, see Figure 4 , on the first device side, a method for transferring machine learning model parameters in a mobile communication system provided by an embodiment of the present application includes:

[0202] S101. Report to the second device a list of functional units of the first device that supports the application of the machine learning model and the type of the machine learning model;

[0203] The second device may be, for example, a base station on the network side, and the first device may be, for example, a terminal; the second device may be, for example, a CU, or the like, and the first device, for example, may also be DU et al.

[0204] S102. Receive machine learning model parameters sent by the second device;

[0205] S103. Send the machine learning model parameters to the target functional unit in the first device.

[0206] Through this method, report to the second device a list of functional units supported by the first device to apply the machine learning model and the type of the machine learning model; receive the machine learning model parameters sent by the second device; and send the machine learning model parameters to the second device. The target functional unit in the first device is described, so that when the machine learning derivation model is inside the wireless mobile communication system, the deployment/update of the machine learning model is realized.

[0207] Optionally, the machine learning model parameter is a list of functional units and machine learning model types that the second device supports to apply the machine learning model according to the first device, in the first device determined according to the deployment state in the machine learning model library. Machine learning model parameters corresponding to the target functional unit of the machine learning model to be deployed.

[0208] Optionally, the method further includes:

[0209] The target functional unit configures and generates a machine learning model according to the machine learning model parameters.

[0210] Optionally, the machine learning model parameters include one or a combination of the following parameters:

[0211] target functional unit identification;

[0212] machine learning model type;

[0213] Machine learning model configuration parameters.

[0214] Optionally, the functional unit list, machine learning model type, and/or machine learning model parameters are carried by control signaling in the mobile communication system.

[0215] Optionally, the first device is a terminal device in a mobile communication system, and the second device is a base station device in a mobile communication system; or the first device is a base station distributed unit in a mobile communication system, so The second device is a centralized unit of a base station in a mobile communication system.

[0216] On the second device side, see Figure 5 , a method for transferring machine learning model parameters in a mobile communication system provided by an embodiment of the present application includes:

[0217] S201. Receive a list of functional units that support the application of the machine learning model and the type of the machine learning model reported by the first device;

[0218] S202: Determine the target functional unit of the machine learning model to be deployed in the first device and the target functional unit of the machine learning model to be deployed in the first device according to the list of functional units supporting the applied machine learning model and the deployment status of the machine learning model type in the machine learning model library reported by the first device The corresponding machine learning model parameters;

[0219] S203. Send the machine learning model parameters to the first device.

[0220] Optionally, the machine learning model parameters include one or a combination of the following parameters:

[0221] target functional unit identification;

[0222] machine learning model type;

[0223] Machine learning model configuration parameters.

[0224] Optionally, the functional unit list, machine learning model type, and/or machine learning model parameters are carried by control signaling in the mobile communication system.

[0225] Optionally, the first device is a terminal device in a mobile communication system, and the second device is a base station device in a mobile communication system; or the first device is a base station distributed unit in a mobile communication system, so The second device is a centralized unit of a base station in a mobile communication system.

[0226] see Image 6 , on the first device side, an apparatus for transferring machine learning model parameters in a mobile communication system provided by an embodiment of the present application includes:

[0227] a memory 620 for storing program instructions;

[0228] The processor 600 is configured to call the program instructions stored in the memory, and execute according to the obtained program:

[0229] reporting to the second device a list of functional units that the first device supports applying the machine learning model and the type of the machine learning model;

[0230] receiving machine learning model parameters sent by the second device;

[0231] Sending the machine learning model parameters to a target functional unit in the first device.

[0232] Optionally, the processor is further configured to call the program instructions stored in the memory, and execute according to the obtained program:

[0233] The target functional unit is controlled to configure and generate a machine learning model according to the machine learning model parameters.

[0234] Optionally, the machine learning model parameters include one or a combination of the following parameters:

[0235] target functional unit identification;

[0236] machine learning model type;

[0237] Machine learning model configuration parameters.

[0238] Optionally, the machine learning model parameter is a list of functional units and machine learning model types that the second device supports to apply the machine learning model according to the first device, in the first device determined according to the deployment state in the machine learning model library. Machine learning model parameters corresponding to the target functional unit of the machine learning model to be deployed.

[0239] Optionally, the functional unit list, machine learning model type, and/or machine learning model parameters are carried by control signaling in the mobile communication system.

[0240] Optionally, the first device is a terminal device in a mobile communication system, and the second device is a base station device in a mobile communication system; or the first device is a base station distributed unit in a mobile communication system, so The second device is a centralized unit of a base station in a mobile communication system.

[0241] The transceiver 610 is used for receiving and transmitting data under the control of the processor 600 .

[0242] Among them, in Image 6 The bus architecture may include any number of interconnected buses and bridges, in particular one or more processors represented by processor 600 and various circuits of memory represented by memory 620 linked together. The bus architecture may also link together various other circuits, such as peripherals, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides the interface. Transceiver 610 may be a number of elements, including a transmitter and a receiver, that provide a means for communicating with various other devices over a transmission medium. For different user equipments, the user interface 630 may also be an interface capable of externally connecting the required equipment, and the connected equipment includes but is not limited to a keypad, a display, a speaker, a microphone, a joystick, and the like.

[0243] The processor 600 is responsible for managing the bus architecture and general processing, and the memory 620 may store data used by the processor 600 in performing operations.

[0244] Optionally, the processor 600 may be a CPU (central processor), an ASIC (Application Specific Integrated Circuit, application specific integrated circuit), an FPGA (Field-Programmable Gate Array, field programmable gate array) or a CPLD (Complex Programmable Logic Device, complex programmable logic devices).

[0245] On the second device side, see Figure 7 , a device for transmitting machine learning model parameters in a mobile communication system provided by an embodiment of the present application includes:

[0246] a memory 520 for storing program instructions;

[0247] The processor 500 is configured to call the program instructions stored in the memory, and execute according to the obtained program:

[0248] Receive the list of functional units that support the application of the machine learning model and the type of the machine learning model reported by the first device;

[0249] Determine the target functional unit of the machine learning model to be deployed in the first device and its corresponding function unit according to the list of functional units supporting the applied machine learning model and the deployment status of the machine learning model type in the machine learning model library reported by the first device machine learning model parameters;

[0250] Sending the machine learning model parameters to the first device.

[0251] Optionally, the machine learning model parameters include one or a combination of the following parameters:

[0252] target functional unit identification;

[0253] machine learning model type;

[0254] Machine learning model configuration parameters.

[0255] Optionally, the functional unit list, machine learning model type, and/or machine learning model parameters are carried by control signaling in the mobile communication system.

[0256] Optionally, the first device is a terminal device in a mobile communication system, and the second device is a base station device in a mobile communication system; or the first device is a base station distributed unit in a mobile communication system, so The second device is a centralized unit of a base station in a mobile communication system.

[0257] The transceiver 510 is used for receiving and transmitting data under the control of the processor 500 .

[0258] Among them, in Figure 7 The bus architecture may include any number of interconnected buses and bridges, in particular one or more processors represented by processor 500 and various circuits of memory represented by memory 520 linked together. The bus architecture may also link together various other circuits, such as peripherals, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides the interface. Transceiver 510 may be multiple elements, ie, including a transmitter and a receiver, providing a means for communicating with various other devices over a transmission medium. The processor 500 is responsible for managing the bus architecture and general processing, and the memory 520 may store data used by the processor 500 in performing operations.

[0259] The processor 500 may be a central processor (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or a complex programmable logic device (Complex Programmable Logic Device, CPLD).

[0260] On the first device side, see Figure 8 , another device for transmitting machine learning model parameters in a mobile communication system provided by the embodiment of the present application includes:

[0261] A reporting unit 11, configured to report to the second device the list of functional units and the type of the machine learning model supported by the first device to apply the machine learning model;

[0262] a receiving unit 12, configured to receive the machine learning model parameters sent by the second device;

[0263] The sending unit 13 is configured to send the machine learning model parameters to the target functional unit in the first device.

[0264] Optionally, the reporting unit 11 uses control signaling in the mobile communication system to report the list of functional units and the type of the machine learning model;

[0265] The receiving unit 12 uses the control signaling in the mobile communication system to receive the parameters of the machine learning model.

[0266] Optionally, the first device is a terminal device in a mobile communication system, and the second device is a base station device in a mobile communication system; or the first device is a base station distributed unit in a mobile communication system, so The second device is a centralized unit of a base station in a mobile communication system.

[0267] On the second device side, see Figure 9 , another device for transmitting machine learning model parameters in a mobile communication system provided by the embodiment of the present application includes:

[0268] A receiving unit 21, configured to receive a list of functional units supporting the applied machine learning model and the type of the machine learning model reported by the first device;

[0269] A determination unit 22, configured to determine the target of the machine learning model to be deployed in the first device according to the list of functional units that support the applied machine learning model and the deployment status of the machine learning model type in the machine learning model library reported by the first device Functional units and their corresponding machine learning model parameters;

[0270] The sending unit 23 is configured to send the parameters of the machine learning model to the first device.

[0271] Optionally, the sending unit 23 uses control signaling in the mobile communication system to send the parameters of the machine learning model;

[0272] The receiving unit 21 uses the control signaling in the mobile communication system to receive the functional unit list and the machine learning model type.

[0273] Optionally, the first device is a terminal device in a mobile communication system, and the second device is a base station device in a mobile communication system; or the first device is a base station distributed unit in a mobile communication system, so The second device is a centralized unit of a base station in a mobile communication system.

[0274] It should be noted that the division of units in the embodiments of the present application is illustrative, and is only a logical function division, and other division methods may be used in actual implementation. In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

[0275] The integrated unit, if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application can be embodied in the form of software products in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, and the computer software products are stored in a storage medium , including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (processor) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program codes .

[0276] An embodiment of the present application provides a computing device, and the computing device may specifically be a desktop computer, a portable computer, a smart phone, a tablet computer, a personal digital assistant (Personal Digital Assistant, PDA), and the like. The computing device may include a central processing unit (Central Processing Unit, CPU), a memory, an input/output device, etc., the input device may include a keyboard, a mouse, a touch screen, etc., and the output device may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), Cathode Ray Tube (CRT), etc.

[0277] The memory may include read only memory (ROM) and random access memory (RAM) and provide the processor with program instructions and data stored in the memory. In the embodiments of the present application, the memory may be used to store the program of any of the methods provided in the embodiments of the present application.

[0278] The processor invokes the program instructions stored in the memory, and the processor is configured to execute any one of the methods provided in the embodiments of the present application according to the obtained program instructions.

[0279] An embodiment of the present application provides a computer storage medium for storing computer program instructions used by the apparatus provided by the above embodiment of the present application, including a program for executing any of the methods provided by the above embodiment of the present application.

[0280] The computer storage medium can be any available medium or data storage device that can be accessed by a computer, including but not limited to magnetic storage (eg, floppy disk, hard disk, magnetic tape, magneto-optical disk (MO), etc.), optical storage (eg, CD, DVD, BD, HVD, etc.), and semiconductor memory (eg, ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state disk (SSD)), and the like.

[0281] The methods provided in the embodiments of the present application may be applied to terminal devices, and may also be applied to network devices.

[0282]The terminal equipment may also be referred to as user equipment (User Equipment, referred to as "UE"), mobile station (Mobile Station, referred to as "MS"), mobile terminal (Mobile Terminal), etc. Optionally, the terminal may be Have the ability to communicate with one or more core networks via a Radio Access Network (RAN), for example, a terminal may be a mobile phone (or a "cellular" phone), or a computer with a mobile nature, etc., For example, the terminal may also be a portable, pocket-sized, hand-held, computer-built, or vehicle-mounted mobile device.

[0283] A network device, which may be a base station (eg, an access point), refers to a device in an access network that communicates with wireless terminals over an air interface through one or more sectors. The base station may be used to convert received air frames to and from IP packets, acting as a router between the wireless terminal and the rest of the access network, which may include an Internet Protocol (IP) network. The base station may also coordinate attribute management of the air interface. For example, the base station may be a base station (BTS, BaseTransceiver Station) in GSM or CDMA, a base station (NodeB) in WCDMA, or an evolved base station (NodeB or eNB or e-NodeB, evolutional Node B) in LTE ), or it can also be a gNB in ​​the 5G system, etc. There is no limitation in this embodiment of the present application.

[0284] The processing flow of the above method can be implemented by a software program, and the software program can be stored in a storage medium, and when the stored software program is called, the above method steps are executed.

[0285] To sum up, the embodiment of the present application uses a standardized interface to transfer parameters of the machine learning model. Provides a method for transferring machine learning model parameters through high-level signaling, a method for expressing machine learning model parameters, and a learning model parameter transfer system, which facilitates operators to deploy/update machine learning models to solve complex problems in wireless mobile communication systems and improve network performance. , including controlling the performance of the terminal side. In scenarios where the machine learning model is relatively fixed, the parameters are updated frequently, and the processing delay is relatively sensitive, applying the method of parameter transfer can effectively reduce the transmission overhead.

[0286] As will be appreciated by those skilled in the art, the embodiments of the present application may be provided as a method, a system, or a computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media having computer-usable program code embodied therein, including but not limited to disk storage, optical storage, and the like.

[0287] The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce in the process of realization Figure 1 process or processes and/or blocks Figure 1 A means for the functions specified in a block or blocks.

[0288] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The device is implemented in the process Figure 1 process or processes and/or blocks Figure 1 the function specified in a box or boxes.

[0289] These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that Instructions are provided for implementing the process in Figure 1 process or processes and/or blocks Figure 1 The steps of the function specified in the box or boxes.

[0290] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.