WO2024020747A1 - 一种模型的生成方法及装置 - Google Patents

一种模型的生成方法及装置 Download PDF

Info

Publication number
WO2024020747A1
WO2024020747A1 PCT/CN2022/107706 CN2022107706W WO2024020747A1 WO 2024020747 A1 WO2024020747 A1 WO 2024020747A1 CN 2022107706 W CN2022107706 W CN 2022107706W WO 2024020747 A1 WO2024020747 A1 WO 2024020747A1
Authority
WO
WIPO (PCT)
Prior art keywords
model
auxiliary information
network device
terminal device
calibrated
Prior art date
Application number
PCT/CN2022/107706
Other languages
English (en)
French (fr)
Inventor
牟勤
李松
Original Assignee
北京小米移动软件有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京小米移动软件有限公司 filed Critical 北京小米移动软件有限公司
Priority to PCT/CN2022/107706 priority Critical patent/WO2024020747A1/zh
Priority to CN202280003069.2A priority patent/CN117836785A/zh
Publication of WO2024020747A1 publication Critical patent/WO2024020747A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Definitions

  • the present disclosure relates to the field of communication technology, and in particular, to a model generation method and device.
  • AI Artificial intelligence
  • beam management requires not only beam configuration data from the network device side, but also feedback information from the terminal device side. Therefore, how to generate and arrange AI models in communication systems has become a current problem.
  • Embodiments of the present disclosure provide a model generation method and device, which can flexibly generate and configure models, further improving the performance and efficiency of the communication system.
  • embodiments of the present disclosure provide a method for generating a model.
  • the method is executed by a first device.
  • the method includes: obtaining a pre-trained model; receiving model calibration auxiliary information sent by the second device; and calibrating the auxiliary information based on the model. , calibrate the pre-trained model.
  • the first device after the first device obtains the pre-trained model and the model calibration auxiliary information sent by the second device, it can calibrate the pre-trained model based on the model calibration auxiliary information to obtain the model, and send the generated model to Secondary device.
  • the flexible generation and configuration of different calibrated models is achieved, further improving the performance and efficiency of the communication system.
  • embodiments of the present disclosure provide a method for generating a model.
  • the method is executed by a second device.
  • the method includes: sending model calibration assistance information to the first device; receiving a calibration based on the model sent by the first device. Auxiliary information Calibrated model after calibration.
  • the second device first sends the model calibration auxiliary information to the first device, and then receives the model obtained by the first device based on the model calibration auxiliary information.
  • the flexible generation and configuration of different calibrated models is achieved, further improving the performance and efficiency of the communication system.
  • an embodiment of the present disclosure provides a communication device.
  • the communication device is provided on the first device side.
  • the communication device includes:
  • Processing module used to obtain pre-trained models
  • a transceiver module configured to receive model calibration auxiliary information sent by the second device
  • the processing module is also configured to calibrate the pre-trained model based on the model calibration auxiliary information to generate a calibrated model
  • the transceiver module is also used to send the model to the second device.
  • an embodiment of the present disclosure provides a communication device.
  • the communication device is provided on the second device side.
  • the communication device includes:
  • a transceiver module configured to send model calibration auxiliary information to the first device
  • the transceiver module is also configured to receive the calibrated model sent by the first device and calibrated based on the model calibration auxiliary information.
  • an embodiment of the present disclosure provides a communication device.
  • the communication device includes a processor.
  • the processor calls a computer program in a memory, it executes the method described in the first aspect.
  • an embodiment of the present disclosure provides a communication device.
  • the communication device includes a processor.
  • the processor calls a computer program in a memory, it executes the method described in the second aspect.
  • an embodiment of the present disclosure provides a communication device.
  • the communication device includes a processor and a memory, and a computer program is stored in the memory; the processor executes the computer program stored in the memory, so that the communication device executes The method described in the first aspect above.
  • an embodiment of the present disclosure provides a communication device.
  • the communication device includes a processor and a memory, and a computer program is stored in the memory; the processor executes the computer program stored in the memory, so that the communication device executes The method described in the second aspect above.
  • an embodiment of the present disclosure provides a communication device.
  • the device includes a processor and an interface circuit.
  • the interface circuit is used to receive code instructions and transmit them to the processor.
  • the processor is used to run the code instructions to cause the The device performs the method described in the first aspect.
  • an embodiment of the present disclosure provides a communication device.
  • the device includes a processor and an interface circuit.
  • the interface circuit is used to receive code instructions and transmit them to the processor.
  • the processor is used to run the code instructions to cause the The device performs the method described in the second aspect above.
  • an embodiment of the present disclosure provides a communication system, which includes the communication device described in the third aspect and the communication device described in the fourth aspect, or the system includes the communication device described in the fifth aspect and The communication device according to the sixth aspect, or the system includes the communication device according to the seventh aspect and the communication device according to the eighth aspect, or the system includes the communication device according to the ninth aspect and the communication device according to the tenth aspect. the above-mentioned communication device.
  • embodiments of the present invention provide a computer-readable storage medium for storing instructions used by the above-mentioned terminal equipment. When the instructions are executed, the terminal equipment is caused to execute the above-mentioned first aspect. method.
  • embodiments of the present invention provide a readable storage medium for storing instructions used by the above-mentioned network device. When the instructions are executed, the network device is caused to perform the method described in the second aspect. .
  • the present disclosure also provides a computer program product including a computer program, which when run on a computer causes the computer to execute the method described in the first aspect.
  • the present disclosure also provides a computer program product including a computer program, which, when run on a computer, causes the computer to execute the method described in the second aspect.
  • the present disclosure provides a chip system.
  • the chip system includes at least one processor and an interface for supporting the first device to implement the functions involved in the first aspect, for example, determining or processing the functions involved in the above method. At least one of data and information.
  • the chip system further includes a memory, and the memory is used to store necessary computer programs and data for the terminal device.
  • the chip system may be composed of chips, or may include chips and other discrete devices.
  • the present disclosure provides a chip system, which includes at least one processor and an interface for supporting a second device to implement the functions involved in the second aspect, for example, determining or processing the functions involved in the above method. At least one of data and information.
  • the chip system further includes a memory, and the memory is used to store necessary computer programs and data for the network device.
  • the chip system may be composed of chips, or may include chips and other discrete devices.
  • the present disclosure provides a computer program that, when run on a computer, causes the computer to execute the method described in the first aspect.
  • the present disclosure provides a computer program that, when run on a computer, causes the computer to execute the method described in the second aspect.
  • Figure 1 is a schematic architectural diagram of a communication system provided by an embodiment of the present disclosure
  • Figure 2 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure
  • Figure 3 is a schematic flowchart of another model generation method provided by an embodiment of the present disclosure.
  • Figure 4 is a schematic flowchart of another model generation method provided by an embodiment of the present disclosure.
  • Figure 5 is a schematic flowchart of another model generation method provided by an embodiment of the present disclosure.
  • Figure 6 is a schematic flowchart of another model generation method provided by an embodiment of the present disclosure.
  • Figure 7 is a schematic flowchart of yet another model generation method provided by an embodiment of the present disclosure.
  • Figure 8 is a schematic flowchart of yet another model generation method provided by an embodiment of the present disclosure.
  • Figure 9 is a schematic flowchart of yet another model generation method provided by an embodiment of the present disclosure.
  • Figure 10 is an interactive schematic diagram of yet another model generation method provided by an embodiment of the present disclosure.
  • Figure 11 is an interactive schematic diagram of another model generation method provided by an embodiment of the present disclosure.
  • Figure 12 is an interactive schematic diagram of yet another model generation method provided by an embodiment of the present disclosure.
  • Figure 13 is a schematic structural diagram of a communication device provided by an embodiment of the present disclosure.
  • Figure 14 is a schematic structural diagram of another communication device provided by an embodiment of the present disclosure.
  • Figure 15 is a schematic structural diagram of a chip provided by an embodiment of the present disclosure.
  • FIG. 1 is a schematic architectural diagram of a communication system provided by an embodiment of the present disclosure.
  • the communication system may include but is not limited to a network device, a terminal device and a server.
  • the number and form of devices shown in Figure 1 are only for examples and do not constitute a limitation on the embodiments of the present disclosure. In actual applications, it may include two or Two or more network devices, two or more terminal devices.
  • the communication system shown in Figure 1 includes a network device 11, a terminal device 12 and a server 13 as an example.
  • LTE long term evolution
  • 5th generation fifth generation
  • 5G new radio (NR) system 5th generation new radio
  • the network device 11 in the embodiment of the present disclosure is an entity on the network side that is used to transmit or receive signals.
  • the network device 101 can be an evolved base station (evolved NodeB, eNB), a transmission point (transmission reception point, TRP), a next generation base station (next generation NodeB, gNB) in an NR system, or other base stations in future mobile communication systems. Or access nodes in wireless fidelity (WiFi) systems, etc.
  • the embodiments of the present disclosure do not limit the specific technologies and specific equipment forms used by network equipment.
  • the network equipment provided by the embodiments of the present disclosure may be composed of a centralized unit (CU) and a distributed unit (DU).
  • the CU may also be called a control unit (control unit).
  • the structure can separate the protocol layers of network equipment, such as base stations, and place some protocol layer functions under centralized control on the CU. The remaining part or all protocol layer functions are distributed in the DU, and the CU centrally controls the DU.
  • the terminal device 12 in the embodiment of the present disclosure is an entity on the user side for receiving or transmitting signals, such as a mobile phone.
  • Terminal equipment can also be called terminal equipment (terminal), user equipment (user equipment, UE), mobile station (mobile station, MS), mobile terminal equipment (mobile terminal, MT), etc.
  • the terminal device can be a car with communication functions, a smart car, a mobile phone, a wearable device, a tablet computer (Pad), a computer with wireless transceiver functions, a virtual reality (VR) terminal device, an augmented reality (augmented reality (AR) terminal equipment, wireless terminal equipment in industrial control, wireless terminal equipment in self-driving, wireless terminal equipment in remote medical surgery, smart grid ( Wireless terminal equipment in smart grid, wireless terminal equipment in transportation safety, wireless terminal equipment in smart city, wireless terminal equipment in smart home, etc.
  • the embodiments of the present disclosure do not limit the specific technology and specific equipment form used by the terminal equipment.
  • the server 13 is a device with data processing capabilities, which can be used to process and store data, and communicate with the terminal device 12 .
  • Figure 2 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure. The method is executed by the first device. As shown in Figure 2, the method may include but is not limited to the following steps:
  • Step 201 Obtain a pre-trained model.
  • the first device may be a network device, a terminal device, or a server, which is not limited in this disclosure.
  • the pre-trained model is an initial model generated by pre-training based on the training data by the first device.
  • the initial model generated by the second device may be pre-trained based on the training data received by the first device.
  • Step 202 Receive model calibration assistance information sent by the second device.
  • Step 203 Calibrate the pre-trained model based on the model calibration auxiliary information to generate a calibrated model.
  • the model calibration auxiliary information is auxiliary information determined by the second device side for calibrating the pre-training model, or auxiliary information received by the second device and sent by other devices for calibrating the pre-training model.
  • the pre-training model can be generated based on basic training data first, and then the model can be trained based on the training data (calibration auxiliary information) provided by each device. Perform calibration.
  • the second device can be a terminal device. That is, after the network device is trained to generate a pre-trained model, the terminal device can provide its own corresponding model calibration auxiliary information, for example, for In the beam management model, the terminal device can send feedback information of the beam determined by its measurement to the network device.
  • the model is used to encode and decode certain parameters, such as a model used to encode and decode channel state information (CSI) feedback information
  • the terminal device can feed back the CSI determined by its measurement. Information is sent to network devices.
  • CSI channel state information
  • the second device may be a network device.
  • the calibrated model calibration auxiliary information sent by the network device can be the processing strategy or business configuration data adopted by the network device when processing the services processed by the model.
  • the second device may be a terminal device.
  • the model calibration auxiliary information sent by the terminal device to the server may include the second auxiliary information determined by the network device side, and may also include the first auxiliary information determined by the terminal device itself, etc.
  • Step 204 Send the calibrated model to the second device.
  • the first device after receiving the model calibration auxiliary information sent by the second device, the first device can calibrate the pre-trained model based on the model calibration auxiliary information to obtain a calibrated model. Afterwards, the first device can process the relevant data based on the calibrated model.
  • the second device can process relevant business data based on the calibrated model.
  • the first device after the first device obtains the pre-trained model and the model calibration auxiliary information sent by the second device, it can calibrate the pre-trained model based on the model calibration auxiliary information to obtain a calibrated model, and store the calibrated model The model is sent to the second device.
  • the flexible generation and configuration of models is achieved, improving the performance and efficiency of the communication system.
  • Figure 3 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure. The method is executed by a first device, where the first device is a network device and the second device is a terminal device. As shown in Figure 3, the method may include but is not limited to the following steps:
  • Step 301 Train to generate a pre-trained model.
  • the network device may first generate a pre-trained model based on basic training data training.
  • step 301 may refer to the detailed description of any embodiment of the present disclosure, and is not limited here.
  • Step 302 Receive the first auxiliary information determined by the terminal device and sent by the terminal device.
  • the first auxiliary information of the terminal device may include the information determined by the terminal device side for calibrating the pre-training model, so that the calibrated model not only contains the training data (second auxiliary information) of the network device side, but also Contains training data (first auxiliary information) on the terminal device side, etc.
  • the terminal device can report the first assistance required for model calibration to the network device as needed.
  • the first auxiliary information received by the network device may be reported by one terminal device after it is determined, or it may be that multiple terminal devices have reported their determined first auxiliary information, and then the network device will report the determined first auxiliary information by the multiple terminal devices. After the plurality of first auxiliary information are fused, the total first auxiliary information used for calibrating the pre-training model is determined, which is not limited by this disclosure.
  • Step 303 Calibrate the pre-trained model based on the first auxiliary information and the second auxiliary information determined by the network device to generate a calibrated model.
  • the network device can calibrate the pre-training model based on the first auxiliary information and its locally determined second assistance.
  • Step 304 Send the calibrated model to the terminal device.
  • the network device can first determine the service that the calibrated model is used to process, and send it to all users with the ability to process it. Terminal equipment with corresponding business capabilities.
  • the network device can send the calibrated model to multiple terminal devices with corresponding capabilities in the form of broadcast messages or system messages, which is not limited in this disclosure.
  • steps 303 and 304 reference can be made to the detailed description of any embodiment of the present disclosure, which is not limited here.
  • the network device after the network device generates the pre-training model and receives the first auxiliary information of the terminal device sent by the terminal device, it can perform the pre-training based on the first auxiliary information of the terminal device and the second auxiliary information on its own side.
  • the model is calibrated to obtain a model, and the calibrated model is sent to the terminal device.
  • Figure 4 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure. The method is executed by a first device, where the first device is a terminal device and the second device is a network device. As shown in Figure 4, the method may include but is not limited to the following steps:
  • Step 401 Receive the pre-trained model sent by the network device.
  • the network device after the network device generates a pre-trained model based on basic training data training, it can first send the pre-trained model to the terminal device, so that the terminal device calibrates the pre-trained model and generates a calibrated model.
  • Step 402 Receive the second auxiliary information determined by the network device and sent by the network device.
  • the second auxiliary information determined by the network device may include a service identifier, a processing policy or configuration data on the network device side, etc., which is not limited in this disclosure.
  • steps 401 and 402 can be executed at the same time.
  • the network device sends the pre-trained model and the second auxiliary information determined by the network device to the terminal device through a message; or, step 402 can also be executed first and then Step 401 is executed, and this disclosure does not limit this.
  • Step 403 Calibrate the pre-trained model based on the first auxiliary information of the terminal device and the second auxiliary information determined by the network device to generate a calibrated model.
  • Step 404 Send the calibrated model to the network device.
  • steps 403 and 404 For the specific implementation of steps 403 and 404, reference may be made to the detailed description of any embodiment of the present disclosure, which is not limited here.
  • the network device can then synchronize the model to other terminal devices that have business capabilities corresponding to the model, so that each terminal device and network device can Based on the calibrated model, relevant business data is processed.
  • the terminal device after receiving the pre-training model sent by the network device and the second auxiliary information determined by the network device, the terminal device can perform pre-training based on its own first auxiliary information and the second auxiliary information determined by the network device.
  • the model is updated to obtain its own model and reports the model to the network device.
  • Figure 5 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure. The method is executed by a first device, where the first device is a server and the second device is a terminal device. As shown in Figure 5, the method may include but is not limited to the following steps:
  • Step 501 Train to generate a pre-trained model.
  • the server can first train to generate a pre-trained model based on basic training data.
  • Step 502 Receive the first auxiliary information of the terminal device sent by the terminal device and the second auxiliary information determined by the network device.
  • the second auxiliary information determined by the network device may be sent by the network device to the terminal device, and may include a service processing policy on the network device side, etc., which is not limited in this disclosure.
  • step 501 and step 502 can be executed simultaneously. That is to say, during the process of training the pre-training model, the server can receive the first auxiliary information sent by the terminal device and the second auxiliary information determined by the network device.
  • step 402 may be performed first and then step 401 may be performed, and this disclosure does not limit this.
  • Step 503 Calibrate the pre-trained model based on the first auxiliary information and the second auxiliary information to generate a calibrated model.
  • Step 504 Send the calibrated model to the terminal device.
  • steps 503 and 504 For the specific implementation of steps 503 and 504, reference may be made to the detailed description of any embodiment of the present disclosure, which is not limited here.
  • the server can send the calibrated model to multiple terminal devices, so that each terminal device can process relevant business data based on the calibrated model.
  • the server after the server generates a pre-trained model through training and receives the second auxiliary information determined by the network device and the first auxiliary information determined by the terminal device sent by the terminal device, it can, based on the first auxiliary information and the second assistance, The pre-trained model is updated to obtain its own model, and the model is sent to the terminal device.
  • the pre-trained model is updated to obtain its own model, and the model is sent to the terminal device.
  • FIG. 6 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure.
  • the method is executed by a second device.
  • the method may include but is not limited to the following steps:
  • Step 601 Send model calibration assistance information to the first device.
  • the first device may be a network device, a terminal device, or a server, which is not limited in this disclosure.
  • the second device may be a terminal device. That is, after the network device is trained to generate a pre-trained model, the terminal device can send its own corresponding model calibration auxiliary information, such as the bandwidth, frequency band, etc. supported by the terminal device, to the network device.
  • the second device may be a network device.
  • the calibrated model calibration auxiliary information sent by the network device may be a processing strategy adopted by the network device when processing the service of the terminal device.
  • the second device may be a terminal device.
  • the model calibration auxiliary information sent by the terminal device to the server may include the model calibration auxiliary information on the network device side, and may also include model calibration auxiliary information in the terminal device, etc.
  • Step 602 Receive the calibrated model sent by the first device.
  • the first device can calibrate the pre-trained model based on the model calibration auxiliary information, thereby generating the calibrated model, and then sending the generated model to the terminal device.
  • the second device first sends the model calibration auxiliary information to the first device, and then receives the model obtained by the first device based on the model calibration auxiliary information.
  • FIG. 7 is a schematic flowchart of a method for generating a model provided by an embodiment of the present disclosure.
  • the method is executed by a second device, where the first device is a network device and the second device is a terminal device.
  • the method may include but is not limited to the following steps:
  • Step 701 Send the first auxiliary information determined by the terminal device to the network device.
  • the first auxiliary information of the terminal device may include feedback information related to the model on the terminal device side.
  • the first auxiliary information may be the feedback information of the beam determined by the terminal equipment measurement, etc.
  • the terminal device can report the first assistance required for model calibration to the network device as needed.
  • Step 702 Receive the calibrated model sent by the network device.
  • the network device can calibrate the pre-training model based on the first auxiliary information and the second assistance determined by the local network device to generate Calibrated model and send the model to the terminal device.
  • the terminal device first sends the first auxiliary information to the network device, and then receives the calibrated model calibrated by the network device based on the first auxiliary information and the second auxiliary information determined by the network device.
  • Figure 8 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure. The method is executed by a second device, where the first device is a terminal device and the second device is a network device. As shown in Figure 8, the method may include but is not limited to the following steps:
  • Step 801 Send the pre-trained model to the terminal device.
  • the network device after the network device is trained to generate a pre-trained model, it can first send the pre-trained model to the terminal device, so that the terminal device can calibrate the pre-trained model and generate its model.
  • the network device can send the pre-trained model to one terminal device, and the one terminal device completes the calibration of the pre-trained model; or, it can also send the pre-trained model to multiple terminal devices, and each terminal device can
  • the pre-trained models are calibrated independently, and this disclosure does not limit this.
  • Step 802 Send the second auxiliary information determined by the network device to the terminal device.
  • the second auxiliary information determined by the network device includes a service identifier, a service processing policy on the network device side or service configuration data, etc., which is not limited in this disclosure.
  • step 801 and step 802 can be performed at the same time.
  • the network device sends the pre-trained model and the second auxiliary information determined by the network device to the terminal device through a message; or, step 802 can also be performed first and then Step 801 is executed, and this disclosure does not limit this.
  • Step 803 Receive the calibrated model sent by the terminal device.
  • the terminal device can pre-train the pair based on its own first auxiliary information and the second auxiliary information determined by the network device.
  • the model is calibrated to generate a calibrated model, and the calibrated model is sent to the network device. In this way, the network device can perform data processing based on the calibrated model.
  • the model can also be synchronized to other terminal devices, so that each terminal device Relevant business data can be processed based on the calibrated model.
  • the network device sends the pre-trained model to multiple terminal devices, it can be agreed through the agreement that only one terminal device reports its calibrated model; or it can also report them all, but the network device only stores the received model. the first calibrated model, etc., this disclosure does not limit this.
  • the network device first sends the pre-trained model and the second auxiliary information determined by the network device to the terminal device, and then receives its model returned by the terminal device.
  • FIG. 9 is a schematic flowchart of a model generation method provided by an embodiment of the present disclosure. The method is executed by a second device, where the second device is a terminal device and the first device is a server. As shown in Figure 9, the method may include but is not limited to the following steps:
  • Step 901 Receive second auxiliary information determined by the network device and sent by the network device.
  • Step 902 Send the first auxiliary information and the second auxiliary information of the terminal device to the server.
  • Step 903 Receive the calibrated model sent by the server.
  • Step 904 Send the calibrated model to the network device.
  • the network device can synchronize the model to other terminal devices, so that each terminal device can process relevant business data based on the calibrated model.
  • the terminal device first receives the second auxiliary information determined by the network device sent by the network device, and then sends its own first auxiliary information and second auxiliary information to the server, and after receiving the calibrated information returned by the server After the model is created, the model can also be sent to network devices.
  • the model can also be sent to network devices.
  • FIG. 10 is an interactive schematic diagram of a model generation method provided by an embodiment of the present disclosure. As shown in Figure 10, the method includes:
  • Step 1001 Network device training generates a pre-trained model.
  • Step 1002 The terminal device sends the determined first auxiliary information to the network device.
  • Step 1003 The network device calibrates the pre-trained model based on the first auxiliary information and the second auxiliary information determined by the network device, and generates a calibrated model.
  • Step 1004 The network device sends the calibrated model to the terminal device.
  • FIG. 11 is an interactive schematic diagram of a model generation method provided by an embodiment of the present disclosure. As shown in Figure 11, the method includes:
  • Step 1101 Network device training generates a pre-trained model.
  • Step 1102 The network device sends the pre-trained model and the second auxiliary information determined by the network device to the terminal device.
  • Step 1103 The terminal device calibrates the pre-trained model based on the determined first auxiliary information and second auxiliary information, and generates a calibrated model.
  • Step 1104 The terminal device sends the calibrated model to the network device.
  • FIG. 12 is an interactive schematic diagram of a model generation method provided by an embodiment of the present disclosure. As shown in Figure 12, the method includes:
  • Step 1201 Server training generates a pre-trained model.
  • Step 1202 The network device sends the determined second auxiliary information to the terminal device.
  • Step 1203 The terminal device sends the determined first auxiliary information and second auxiliary information to the server.
  • Step 1204 The server calibrates the pre-trained model based on the first auxiliary information and the second auxiliary information, and generates a calibrated model.
  • Step 1205 the server sends the calibrated model to the terminal device.
  • Step 1206 The terminal device sends the calibrated model to the network device.
  • FIG. 13 is a schematic structural diagram of a communication device provided by an embodiment of the present disclosure.
  • the communication device 1300 shown in FIG. 13 may include a processing module 1301 and a transceiver module 1302.
  • the transceiving module 1302 may include a sending module and/or a receiving module.
  • the sending module is used to implement the sending function
  • the receiving module is used to implement the receiving function.
  • the transceiving module 1302 may implement the sending function and/or the receiving function.
  • the communication device 1300 may be a first device, a device in the first device, or a device that can be used in conjunction with the first device.
  • the communication device 1300 is on the first device side, where:
  • Transceiver module 1302 configured to receive model calibration auxiliary information sent by the second device
  • the processing module 1301 is also used to calibrate the pre-trained model based on the model calibration auxiliary information to generate a calibrated model
  • the transceiver module 1302 is also used to send the calibrated model to the second device.
  • the first device is a network device
  • the second device is the terminal device.
  • the transceiver module 1302 is further configured to receive the first auxiliary information sent by the terminal device and determined by the terminal device.
  • the first device is the terminal device
  • the second device is a network device
  • the processing module 1301 is further configured to receive the pre-training model sent by the network device.
  • the transceiving module 1302 is also configured to receive the second auxiliary information sent by the network device and determined by the network device.
  • the first device is a server
  • the second device is the terminal device
  • the transceiver module 1302 is also used to:
  • the first device after the first device obtains the pre-trained model and the model calibration auxiliary information sent by the second device, it can calibrate the pre-trained model based on the model calibration auxiliary information to obtain the model, and send the generated model to Secondary device.
  • the first device obtains the pre-trained model and the model calibration auxiliary information sent by the second device
  • it can calibrate the pre-trained model based on the model calibration auxiliary information to obtain the model, and send the generated model to Secondary device.
  • flexible generation and configuration of the model is achieved, further improving the performance and efficiency of the communication system.
  • the communication device 1300 may be a second device, a device in the second device, or a device that can be used in conjunction with the second device.
  • Transceiver module 1302 used to send model calibration auxiliary information to the first device
  • the transceiver module 1302 is also configured to receive the model calibrated based on the model calibration auxiliary information sent by the first device.
  • the first device is a network device
  • the second device is the terminal device
  • the transceiver module 1302 is also used to:
  • the first device is the terminal device
  • the second device is a network device.
  • the transceiver module 1302 is further configured to: send the second auxiliary information determined by the network device to the terminal device. .
  • the transceiving module 1302 is also used to send the pre-trained model to the terminal device.
  • the first device is a server
  • the second device is the terminal device
  • the transceiver module 1302 is also used to:
  • the transceiver module 1302 is also used for:
  • the transceiver module 1302 is also used for:
  • the second device first sends the model calibration auxiliary information to the first device, and then receives the model obtained by the first device based on the model calibration auxiliary information.
  • FIG. 14 is a schematic structural diagram of another communication device provided by an embodiment of the present disclosure.
  • the communication device 1400 may be a first device, a second device, a chip, a chip system, a processor, etc. that supports the first device to implement the above method, or a chip, a chip system, or a processor that supports the second device to implement the above method.
  • Chip system, or processor, etc. The device can be used to implement the method described in the above method embodiment. For details, please refer to the description in the above method embodiment.
  • Communication device 1400 may include one or more processors 1401.
  • the processor 1401 may be a general-purpose processor or a special-purpose processor, or the like.
  • it can be a baseband processor or a central processing unit.
  • the baseband processor can be used to process communication protocols and communication data.
  • the central processor can be used to control communication devices (such as base stations, baseband chips, terminal equipment, terminal equipment chips, DU or CU, etc.) and execute computer programs. , processing data for computer programs.
  • the communication device 1400 may also include one or more memories 1402, on which a computer program 1404 may be stored.
  • the processor 1401 executes the computer program 1404, so that the communication device 1400 performs the steps described in the above method embodiments. method.
  • the memory 1402 may also store data.
  • the communication device 1400 and the memory 1402 can be provided separately or integrated together.
  • the communication device 1400 may also include a transceiver 1405 and an antenna 1406.
  • the transceiver 1405 may be called a transceiver unit, a transceiver, a transceiver circuit, etc., and is used to implement transceiver functions.
  • the transceiver 1405 may include a receiver and a transmitter.
  • the receiver may be called a receiver or a receiving circuit, etc., used to implement the receiving function;
  • the transmitter may be called a transmitter, a transmitting circuit, etc., used to implement the transmitting function.
  • the communication device 1400 may also include one or more interface circuits 1407.
  • the interface circuit 1407 is used to receive code instructions and transmit them to the processor 1401 .
  • the processor 1401 executes the code instructions to cause the communication device 1400 to perform the method described in the above method embodiment.
  • the communication device 1400 is the first device: the processor 1401 is used to perform steps 201 and 203 in Figure 2, step 301, step 303 in Figure 3, etc., and the transceiver 1405 is used to perform steps 202 and 204 in Figure 2 , or step 302, step 304 and so on in Figure 3.
  • the communication device 1400 is a second device: the transceiver 1405 is used to perform steps 601 and 602 in Figure 6, or to perform steps 701, 702, etc. in Figure 7.
  • the processor 1401 may include a transceiver for implementing receiving and transmitting functions.
  • the transceiver may be a transceiver circuit, an interface, or an interface circuit.
  • the transceiver circuits, interfaces or interface circuits used to implement the receiving and transmitting functions can be separate or integrated together.
  • the above-mentioned transceiver circuit, interface or interface circuit can be used for reading and writing codes/data, or the above-mentioned transceiver circuit, interface or interface circuit can be used for signal transmission or transfer.
  • the processor 1401 may store a computer program 1403, and the computer program 1403 runs on the processor 1401, causing the communication device 1400 to perform the method described in the above method embodiment.
  • the computer program 603 may be solidified in the processor 1401, in which case the processor 1401 may be implemented by hardware.
  • the communication device 1400 may include a circuit, which may implement the functions of sending or receiving or communicating in the foregoing method embodiments.
  • the processors and transceivers described in this disclosure may be implemented on integrated circuits (ICs), analog ICs, radio frequency integrated circuits (RFICs), mixed signal ICs, application specific integrated circuits (ASICs), printed circuit boards ( printed circuit board (PCB), electronic equipment, etc.
  • the processor and transceiver can also be manufactured using various IC process technologies, such as complementary metal oxide semiconductor (CMOS), n-type metal oxide-semiconductor (NMOS), P-type Metal oxide semiconductor (positive channel metal oxide semiconductor, PMOS), bipolar junction transistor (BJT), bipolar CMOS (BiCMOS), silicon germanium (SiGe), gallium arsenide (GaAs), etc.
  • CMOS complementary metal oxide semiconductor
  • NMOS n-type metal oxide-semiconductor
  • PMOS P-type Metal oxide semiconductor
  • BJT bipolar junction transistor
  • BiCMOS bipolar CMOS
  • SiGe silicon germanium
  • GaAs gallium arsenide
  • the communication device described in the above embodiments may be a network device or an intelligent relay, but the scope of the communication device described in the present disclosure is not limited thereto, and the structure of the communication device may not be limited by FIG. 14 .
  • the communication device may be a stand-alone device or may be part of a larger device.
  • the communication device may be:
  • the IC collection may also include storage components for storing data and computer programs;
  • the communication device may be a chip or a chip system
  • the schematic structural diagram of the chip shown in FIG. 15 refer to the schematic structural diagram of the chip shown in FIG. 15 .
  • the chip shown in Figure 15 includes a processor 1501 and an interface 1503.
  • the number of processors 1501 may be one or more, and the number of interfaces 1503 may be multiple.
  • Interface 1503 is used to execute step 202, step 204, etc. in Figure 2.
  • Interface 1503 is used to execute steps 601, 602, etc. in Figure 6.
  • the chip also includes a memory 1503, which is used to store necessary computer programs and data.
  • the present disclosure also provides a readable storage medium on which instructions are stored, and when the instructions are executed by a computer, the functions of any of the above method embodiments are implemented.
  • the present disclosure also provides a computer program product, which, when executed by a computer, implements the functions of any of the above method embodiments.
  • the above embodiments it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof.
  • software it may be implemented in whole or in part in the form of a computer program product.
  • the computer program product includes one or more computer programs.
  • the computer program When the computer program is loaded and executed on a computer, the processes or functions described in accordance with the embodiments of the present disclosure are generated in whole or in part.
  • the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.
  • the computer program may be stored in or transferred from one computer-readable storage medium to another, for example, the computer program may be transferred from a website, computer, server, or data center Transmission to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) means.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more available media integrated.
  • the usable media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., high-density digital video discs (DVD)), or semiconductor media (e.g., solid state disks, SSD)) etc.
  • magnetic media e.g., floppy disks, hard disks, magnetic tapes
  • optical media e.g., high-density digital video discs (DVD)
  • DVD digital video discs
  • semiconductor media e.g., solid state disks, SSD
  • At least one in the present disclosure can also be described as one or more, and the plurality can be two, three, four or more, and the present disclosure is not limited.
  • the technical feature is distinguished by “first”, “second”, “third”, “A”, “B”, “C” and “D” etc.
  • the technical features described in “first”, “second”, “third”, “A”, “B”, “C” and “D” are in no particular order or order.
  • the corresponding relationships shown in each table in this disclosure can be configured or predefined.
  • the values of the information in each table are only examples and can be configured as other values, which is not limited by this disclosure.
  • the corresponding relationships shown in some rows may not be configured.
  • appropriate deformation adjustments can be made based on the above table, such as splitting, merging, etc.
  • the title of the above tables indicates that the name of the first auxiliary can also be other names that can be understood by the communication device, and the value or expression of the first auxiliary can also be other values or expressions that can be understood by the communication device.
  • other data structures can also be used, such as arrays, queues, containers, stacks, linear lists, pointers, linked lists, trees, graphs, structures, classes, heaps, hash tables or hash tables. wait.
  • Predefinition in this disclosure may be understood as definition, pre-definition, storage, pre-storage, pre-negotiation, pre-configuration, solidification, or pre-burning.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Artificial Intelligence (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本公开实施例公开了一种模型生成方法及装置，可应用于通信技术领域，其中，由第一设备执行的方法包括:获取预训练模型(201)；接收第二设备发送的模型校准辅助信息(202);基于模型校准辅助信息，对预训练模型进行校准，以生成校准后的模型(203)；将校准后模型，发送给第二设备(204)。由此，通过第一设备与第二设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。

Description

一种模型的生成方法及装置 技术领域
本公开涉及通信技术领域，尤其涉及一种模型的生成方法及装置。
背景技术
人工智能(artificial intelligence，AI)已经逐步应用于通信系统中，用于辅助提升通讯算法的效率。通信系统中AI模型的训练可能需要各方的参与才能最大程度提高模型的准确性。比如波束管理，不仅需要网络设备侧的波束配置数据，也需要终端设备侧的反馈信息。所以在通信系统中如何生成和布置AI模型，成为当前面临的问题。
发明内容
本公开实施例提供一种模型的生成方法及装置，可以灵活生成和配置模型，进一步提升了通信系统的性能和效率。
第一方面，本公开实施例提供一种模型的生成方法，该方法由第一设备执行，方法包括：取预训练模型；接收第二设备发送的模型校准辅助信息；基于所述模型校准辅助信息，对所述预训练模型进行校准。
本公开中，第一设备在获取到预训练模型及第二设备发送的模型校准辅助信息后，即可基于模型校准辅助信息，对预训练模型进行校准以得到模型，并将生成的模型发送给第二设备。由此，实现了灵活生成和配置不同校准后的模型，进一步提升了通信系统的性能和效率。
第二方面，本公开实施例提供一种模型的生成方法，该方法由第二设备执行，方法包括：向第一设备发送模型校准辅助信息；接收所述第一设备发送的基于所述模型校准辅助信息校准后的校准后的模型。
本公开中，第二设备首先向第一设备发送模型校准辅助信息，之后，即可接收到第一设备基于模型校准辅助信息得到的模型。由此，实现了灵活生成和配置不同校准后的模型，进一步提升了通信系统的性能和效率。
第三方面，本公开实施例提供一种通信装置，所述通信装置设置在第一设备侧，所述通信装置包括：
处理模块，用于获取预训练模型；
收发模块，用于接收第二设备发送的模型校准辅助信息；
所述处理模块，还用于基于所述模型校准辅助信息，对所述预训练模型进行校准，以生成校准后模型；
所述收发模块，还用于将所述模型，发送给所述第二设备。
第四方面，本公开实施例提供一种通信装置，所述通信装置设置在第二设备侧，所述通信装置包括：
收发模块，用于向第一设备发送模型校准辅助信息；
所述收发模块，还用于接收所述第一设备发送的基于所述模型校准辅助信息校准后的校准后的模型。
第五方面，本公开实施例提供一种通信装置，该通信装置包括处理器，当该处理器调用存储器中的计算机程序时，执行上述第一方面所述的方法。
第六方面，本公开实施例提供一种通信装置，该通信装置包括处理器，当该处理器调用存储器中的计算机程序时，执行上述第二方面所述的方法。
第七方面，本公开实施例提供一种通信装置，该通信装置包括处理器和存储器，该存储器中存储有计算机程序；所述处理器执行该存储器所存储的计算机程序，以使该通信装置执行上述第一方面所述的方法。
第八方面，本公开实施例提供一种通信装置，该通信装置包括处理器和存储器，该存储器中存储有计算机程序；所述处理器执行该存储器所存储的计算机程序，以使该通信装置执行上述第二方面所述的方法。
第九方面，本公开实施例提供一种通信装置，该装置包括处理器和接口电路，该接口电路用于接收代码指令并传输至该处理器，该处理器用于运行所述代码指令以使该装置执行上述第一方面所述的方法。
第十方面，本公开实施例提供一种通信装置，该装置包括处理器和接口电路，该接口电路用于接收代码指令并传输至该处理器，该处理器用于运行所述代码指令以使该装置执行上述第二方面所述的方法。
第十一方面，本公开实施例提供一种通信系统，该系统包括第三方面所述的通信装置及第四方面所述的通信装置，或者，该系统包括第五方面所述的通信装置以及第六方面所述的通信装置，或者，该系统包括第七方面所述的通信装置以及第八方面所述的通信装置，或者，该系统包括第九方面所述的通信装置以及第十方面所述的通信装置。
第十二方面，本发明实施例提供一种计算机可读存储介质，用于储存为上述终端设备所用的指令，当所述指令被执行时，使所述终端设备执行上述第一方面所述的方法。
第十三方面，本发明实施例提供一种可读存储介质，用于储存为上述网络设备所用的指令，当所述指令被执行时，使所述网络设备执行上述第二方面所述的方法。
第十四方面，本公开还提供一种包括计算机程序的计算机程序产品，当其在计算机上运行时，使得计算机执行上述第一方面所述的方法。
第十五方面，本公开还提供一种包括计算机程序的计算机程序产品，当其在计算机上运行时，使得计算机执行上述第二方面所述的方法。
第十七方面，本公开提供一种芯片系统，该芯片系统包括至少一个处理器和接口，用于支持第一设备实现第一方面所涉及的功能，例如，确定或处理上述方法中所涉及的数据和信息中的至少一种。在一种可能的设计中，所述芯片系统还包括存储器，所述存储器，用于保存终端设备必要的计算机程序和数据。该芯片系统，可以由芯片构成，也可以包括芯片和其他分立器件。
第十八方面，本公开提供一种芯片系统，该芯片系统包括至少一个处理器和接口，用于支持第二设备实现第二方面所涉及的功能，例如，确定或处理上述方法中所涉及的数据和信息中的至少一种。在一种可能的设计中，所述芯片系统还包括存储器，所述存储器，用于保存网络设备必要的计算机程序和数据。该芯片系统，可以由芯片构成，也可以包括芯片和其他分立器件。
第十九方面，本公开提供一种计算机程序，当其在计算机上运行时，使得计算机执行上述第一方面所述的方法。
第二十方面，本公开提供一种计算机程序，当其在计算机上运行时，使得计算机执行上述第二方面所述的方法。
附图说明
为了更清楚地说明本公开实施例或背景技术中的技术方案，下面将对本公开实施例或背景技术中所需要使用的附图进行说明。
图1是本公开实施例提供的一种通信系统的架构示意图；
图2是本公开实施例提供的一种模型的生成方法的流程示意图；
图3是本公开实施例提供的另一种模型的生成方法的流程示意图；
图4是本公开实施例提供的另一种模型的生成方法的流程示意图；
图5是本公开实施例提供的另一种模型的生成方法的流程示意图；
图6是本公开实施例提供的另一种模型的生成方法的流程示意图；
图7是本公开实施例提供的又一种模型的生成方法的流程示意图；
图8是本公开实施例提供的又一种模型的生成方法的流程示意图；
图9是本公开实施例提供的又一种模型的生成方法的流程示意图；
图10是本公开实施例提供的再一种模型的生成方法的交互示意图；
图11是本公开实施例提供的另一种模型的生成方法的交互示意图；
图12是本公开实施例提供的又一种模型的生成方法的交互示意图；
图13是本公开实施例提供的一种通信装置的结构示意图；
图14是本公开实施例提供的另一种通信装置的结构示意图；
图15是本公开实施例提供的一种芯片的结构示意图。
具体实施方式
为了更好的理解本公开实施例公开的一种模型的生成方法，下面首先对本公开实施例适用的通信系统进行描述。
请参见图1，图1为本公开实施例提供的一种通信系统的架构示意图。该通信系统可包括但不限于一个网络设备、一个终端设备和一个服务器，图1所示的设备数量和形态仅用于举例并不构成对本公开实施例的限定，实际应用中可以包括两个或两个以上的网络设备，两个或两个以上的终端设备。图1所示的通信系统以包括一个网络设备11、一个终端设备12和一个服务器13为例。
需要说明的是，本公开实施例的技术方案可以应用于各种通信系统。例如：长期演进(long term evolution，LTE)系统、第五代(5th generation，5G)移动通信系统、5G新空口(new radio，NR)系统，或者其他未来的新型移动通信系统等。
本公开实施例中的网络设备11是网络侧的一种用于发射或接收信号的实体。例如，网络设备101可以为演进型基站(evolved NodeB，eNB)、传输点(transmission reception point，TRP)、NR系统中的下一代基站(next generation NodeB，gNB)、其他未来移动通信系统中的基站或无线保真(wireless fidelity，WiFi)系统中的接入节点等。本公开的实施例对网络设备所采用的具体技术和具体设备形态不做限定。本公开实施例提供的网络设备可以是由集中单元(central unit，CU)与分布式单元(distributed unit，DU)组成的，其中，CU也可以称为控制单元(control unit)，采用CU-DU的结构可以将网络设备，例如基站的协议层拆分开，部分协议层的功能放在CU集中控制，剩下部分或全部协议层的功能分布在DU中，由CU集中控制DU。
本公开实施例中的终端设备12是用户侧的一种用于接收或发射信号的实体，如手机。终端设备也可以称为终端设备(terminal)、用户设备(user equipment，UE)、移动台(mobile station，MS)、移动终端设备(mobile terminal，MT)等。终端设备可以是具备通信功能的汽车、智能汽车、手机(mobile phone)、穿戴式设备、平板电脑(Pad)、带无线收发功能的电脑、虚拟现实(virtual reality，VR)终端设备、增强现实(augmented reality，AR)终端设备、工业控制(industrial control)中的无线终端设备、无人驾驶(self-driving)中的无线终端设备、远程手术(remote medical surgery)中的无线终端设备、智能电网(smart grid)中的无线终端设备、运输安全(transportation safety)中的无线终端设备、智慧城市(smart city)中的无线终端设备、智慧家庭(smart home)中的无线终端设备等等。本公开的实施例对终端设备所采用的具体技术和具体设备形态不做限定。
服务器13为具有数据处理能力的设备，其可用于进行数据处理和存储，并与终端设备12进行通信。
可以理解的是，本公开实施例描述的通信系统是为了更加清楚的说明本公开实施例的技术方案，并不构成对于本公开实施例提供的技术方案的限定，本领域普通技术人员可知，随着系统架构的演变和新业务场景的出现，本公开实施例提供的技术方案对于类似的技术问题，同样适用。
请参见图2，图2是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第一设备执行。如图2所示，该方法可以包括但不限于如下步骤：
步骤201，获取预训练模型。
可选的，第一设备可以为网络设备，或者也可以为终端设备，或者还可以为服务器，本公开对此不做限定。
其中，预训练模型为第一设备基于训练数据，预先训练生成的初始模型。或者，也可以为第一设备接收的第二设备基于训练数据，预训练生成的初始模型。
步骤202，接收第二设备发送的模型校准辅助信息。
步骤203，基于模型校准辅助信息，对预训练模型进行校准，以生成校准后的模型。
其中，模型校准辅助信息为，第二设备侧确定的用于对预训练模型进行校准的辅助信息，或者为第二设备接收的其他设备发送的用于对预训练模型进行校准的辅助信息。
本公开中，为了保证最终得到的模型，考虑了各个设备侧的信息，可以首先基于基本的训练数据，训练生成预训练模型，之后再基于各设备提供的训练数据(校准辅助信息)，对模型进行校准。
可选的，若第一设备为网络设备，则第二设备可以为终端设备，即网络设备在训练生成预训练模型后，终端设备可以将其自身对应的模型校准辅助信息，比如，对于用于波束管理的模型，终端设备可以将其测量确定的波束的反馈信息等等发送给网络设备。或者，若模型为用于对某些参量做编解码处理的，比如用于对信道状态信息(channel state information，CSI)反馈信息进行编解码的模型，则终端设备可以将其测量确定的CSI反馈信息发送给网络设备。
可选的，若第一设备为终端设备，则第二设备可以为网络设备。此时，网络设备发送给的校准后的 模型校准辅助信息，可以为网络设备在处理该模型处理的业务时采用的处理策略、或业务配置数据等。
可选的，若第一设备为服务器，则第二设备可以为终端设备。此时，终端设备发送给服务器的模型校准辅助信息，可以包括网络设备侧确定的第二辅助信息，还可以包括该终端设备自身确定的第一辅助信息等。
步骤204，将校准后模型，发送给第二设备。
本公开中，第一设备在接收到第二设备发送的模型校准辅助信息后，即可基于该模型校准辅助信息，对预训练模型进行校准，以得到校准后的模型。之后，第一设备即可根据基于该校准后的模型，对相关的数据进行处理。
同时，第一设备将校准后的模型发送给第二设备后，第二设备即可基于该校准后的模型，对相关的的业务数据进行处理。
本公开中，第一设备在获取到预训练模型及第二设备发送的模型校准辅助信息后，即可基于模型校准辅助信息，对预训练模型进行校准以得到校准后的模型，并将校准后的模型发送给第二设备。由此，通过各设备的交互，实现了模型的灵活生成和配置模型，提升了通信系统的性能和效率。
请参见图3，图3是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第一设备执行，其中，第一设备为网络设备，第二设备为终端设备。如图3所示，该方法可以包括但不限于如下步骤：
步骤301，训练生成预训练模型。
本公开中，网络设备可以首先基于基础训练数据训练生成预训练模型。
其中，步骤301的具体实现可以参照本公开任一实施例的详细描述，此处不做限定。
步骤302，接收终端设备发送的终端设备确定的第一辅助信息。
可选的，终端设备的第一辅助信息，可以包括终端设备侧确定的用于对预训练模型进行校准，以使得校准后的模型即包含网络设备侧的训练数据(第二辅助信息)，又包含终端设备侧的训练数据(第一辅助信息)等等。
需要说明的是，用于处理不同业务的模型校准时所需要的第一辅助可能不同，终端设备可以根据需要，将模型校准所需的第一辅助上报给网络设备。
另外，网络设备收到的第一辅助信息，可能是一个终端设备确定后上报的，也可能是多个终端设备都上报了其确定的第一辅助信息，之后网络设备将多个终端设备上报的多个第一辅助信息融合后，确定用于对预训练模型进行校准时使用的总的第一辅助信息，本公开对此不做限定。
步骤303，基于第一辅助信息及所述网络设备确定的第二辅助信息，对预训练模型进行校准，以生成校准后的模型。
本公开中，网络设备在接收到终端设备上报的第一辅助信息后，即可基于该第一辅助信息及其本地确定的第二辅助，对预训练模型进行校准。
步骤304，将校准后的模型，发送给终端设备。
需要说明的是，由于不同的终端设备具有的业务能力不同，当网络设备在生成校准后的模型后，可以首先确定该校准后的模型用于处理的业务，并将其发送给所有具有处理其对应的业务能力的终端设备。可选的，网络设备可以通过广播消息或者系统消息的形式，将该校准后的模型发送给具有对应能力的多个终端设备，本公开对此不做限定。其中，步骤303及304的具体实现可以参照本公开任一实施例的详细描述，此处不做限定。
本公开中，网络设备在生成预训练模型，并接收到终端设备发送的终端设备的第一辅助信息后，即可基于终端设备的第一辅助信息及自身侧的第二辅助信息，对预训练模型进行校准以得到模型，并将校准后的模型发送给终端设备。由此，通过网络设备与终端设备的交互，实现了模型的灵活生成和配置，提升了通信系统的性能和效率。
请参见图4，图4是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第一设备执行，其中，第一设备为终端设备，第二设备为网络设备。如图4所示，该方法可以包括但不限于如下步骤：
步骤401，接收网络设备发送的预训练模型。
本公开中，网络设备在基于基础训练数据训练生成预训练模型后，可以先将预训练模型发送给终端设备，以由终端设备对该预训练模型进行校准，生成校准后的模型。
步骤402，接收网络设备发送的网络设备确定的第二辅助信息。
其中，网络设备确定的第二辅助信息中，可以包括业务标识、网络设备侧的处理策略或配置数据等等，本公开对此不做限定。
可选的，步骤401和步骤402可以同时执行，比如网络设备通过一条消息，将预训练模型及所述网络设备确定的第二辅助信息一起发送给终端设备；或者，也可以先执行步骤402再执行步骤401，本公开对此不做限定。
步骤403，基于终端设备的第一辅助信息及网络设备确定的第二辅助信息，对预训练模型进行校准，以生成校准后的模型。
步骤404，将校准后的模型，发送给网络设备。
其中，步骤403及404的具体实现可以参照本公开任一实施例的详细描述，此处不做限定。
本公开中，终端设备在将校准后的网络模型发送给网络设备后，网络设备可以再将该模型同步给其余具有该模型对应的业务能力的其他终端设备，从而使得各终端设备和网络设备可以基于该校准后的模型，对相关业务数据进行处理。
本公开中，终端设备在接收网络设备发送的预训练模型及所述网络设备确定的第二辅助信息后，即可以根据其自身的第一辅助信息及网络设备确定的第二辅助，对预训练模型进行更新，以得到其自身的模型，并把该模型上报给网络设备。由此，通过网络设备与终端设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图5，图5是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第一设备执行，其中，第一设备为服务器，第二设备为终端设备。如图5所示，该方法可以包括但不限于如下步骤：
步骤501，训练生成预训练模型。
本公开中，服务器可以首先基于基础的训练数据，训练生成预训练模型。
步骤502，接收终端设备发送的终端设备的第一辅助信息及网络设备确定的第二辅助信息。
其中，网络设备确定的第二辅助信息，可以为网络设备发送给终端设备的，其可以包括网络设备侧的业务处理策略等，本公开对此不做限定。
可选的，步骤501和步骤502可以同时执行，也就是说服务器在训练预训练模型的过程中，可以接收终端设备发送的第一辅助信息及所述网络设备确定的第二辅助信息。或者，也可以先执行步骤402再执行步骤401，本公开对此不做限定。
步骤503，基于第一辅助信息及第二辅助信息，对预训练模型进行校准，以生成校准后的模型。
步骤504，将校准后的模型，发送给终端设备。
其中，步骤503及504的具体实现可以参照本公开任一实施例的详细描述，此处不做限定。
需要说明的是，服务器可以将校准后的模型发送给多个终端设备，从而各个终端设备均可基于该校准后的模型，对相关业务数据进行处理。
本公开中，服务器在训练生成预训练模型，并接收到终端设备发送的网络设备确定的第二辅助信息及终端设备确定的第一辅助信息后，即可以根据第一辅助信息及第二辅助，对预训练模型进行更新，以得到其自身的模型，并把该模型发送给终端设备。由此，通过服务器、终端设备预网络设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图6，图6是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第二设备执行。如图6所示，该方法可以包括但不限于如下步骤：
步骤601，向第一设备发送模型校准辅助信息。
可选的，第一设备可以为网络设备，或者也可以为终端设备，或者还可以为服务器，本公开对此不做限定。
相应的，若第一设备为网络设备，则第二设备可以为终端设备。即网络设备在训练生成预训练模型后，终端设备可以将其自身对应的模型校准辅助信息，比如终端设备支持的带宽、频段等等发送给网络设备。
可选的，若第一设备为终端设备，则第二设备可以为网络设备。此时，网络设备发送给的校准后的模型校准辅助信息，可以为网络设备在处理该的终端设备的业务时采用的处理策略等。
可选的，若第一设备为服务器，则第二设备可以为终端设备。此时，终端设备发送给服务器的模型校准辅助信息，可以包括网络设备侧的该模型校准辅助信息，还可以包括该终端设备中的模型校准辅助信息等。
步骤602，接收第一设备发送的校准后的模型。
本公开中，第二设备将模型校准辅助信息发送给第一设备后，第一设备即可基于模型校准辅助信息对预训练模型进行校准，从而生成该校准后的模型，进而将生成的模型发送给终端设备。
本公开中，第二设备首先向第一设备发送模型校准辅助信息，之后，即可接收到第一设备基于模型校准辅助信息得到的模型。由此，通过第一设备与第二设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图7，图7是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第二设备执行，其中，第一设备为网络设备，第二设备为终端设备。如图7所示，该方法可以包括但不限于如下步骤：
步骤701，向网络设备发送终端设备确定的第一辅助信息。
其中，终端设备的第一辅助信息，可以包括终端设备侧的与该模型相关的反馈信息。比如，针对波数管理模型，第一辅助信息，可以为终端设备测量确定的波束的反馈信息等。
需要说明的是，用于处理不同业务的模型校准时所需要的第一辅助可能不同，终端设备可以根据需要，将模型校准所需的第一辅助上报给网络设备。
步骤702，接收网络设备发送的校准后的模型。
本公开中，网络设备在接收到终端设备上报的第一辅助信息后，即可基于该第一辅助信息及其本地的所述网络设备确定的第二辅助，对预训练模型进行校准，以生成校准后的模型，并将该模型发送给终端设备。
本公开中，终端设备首先向网络设备发送第一辅助信息，之后，即可接收到网络设备基于第一辅助信息及网络设备确定的第二辅助信息校准得到的校准后的模型。由此，通过网络设备与终端设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图8，图8是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第二设备执行，其中，第一设备为终端设备，第二设备为网络设备。如图8所示，该方法可以包括但不限于如下步骤：
步骤801，向终端设备发送预训练模型。
本公开中，网络设备在训练生成预训练模型后，可以先将预训练模型发送给终端设备，以由终端设备对该预训练模型进行校准，生成其的模型。
本公开中，网络设备可以将预训练模型发送给一个终端设备，由该一个终端设备完成对该预训练模型的校准；或者，也可以将预训练模型发送给多个终端设备，由各个终端设备分别对该预训练模型，各自独立进行校准，本公开对此不做限定。
步骤802，向终端设备发送网络设备确定的第二辅助信息。
其中，所述网络设备确定的第二辅助信息中包括业务标识、网络设备侧的业务处理策略或业务配置数据等等，本公开对此不做限定。
可选的，步骤801和步骤802可以同时执行，比如网络设备通过一条消息，将预训练模型及所述网络设备确定的第二辅助信息一起发送给终端设备；或者，也可以先执行步骤802再执行步骤801，本公开对此不做限定。
步骤803，接收终端设备发送的校准后的模型。
本公开中，网络设备在将预训练模型及所述网络设备确定的第二辅助信息发送给终端设备后，终端设备即可基于自身的第一辅助信息及网络设备确定的第二辅助对预训练模型进行校准，以生成校准后的模型，并把校准后的模型发送给网络设备。从而网络设备即可基于该校准后的模型，进行数据处理。
可选的，网络设备若将预训练模型发送给了一个终端设备，那么在接收到该一个终端设备返回的校准后的模型后，还可以将该模型同步给其他终端设备，从而使得各个终端设备均可基于该校准后的模型处理相关业务数据。
可选的，网络设备若将该预训练模型发送给了多个终端设备，那么可以通过协议约定，仅由一个终端设备上报其校准后的模型；或者也可以都上报，但是网络设备仅存储接收到的第一个校准后的模型，等等，本公开对此不做限定。
本公开中，网络设备首先将预训练模型及所述网络设备确定的第二辅助信息发送给终端设备，之后即可接收到终端设备返回的其的模型。由此，通过网络设备与终端设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图9，图9是本公开实施例提供的一种模型的生成方法的流程示意图，该方法由第二设备执行，其中，第二设备为终端设备，第一设备为服务器。如图9所示，该方法可以包括但不限于如下步骤：
步骤901，接收网络设备发送的网络设备确定的第二辅助信息。
步骤902，向服务器发送终端设备的第一辅助信息及第二辅助信息。
步骤903，接收服务器发送的校准后的模型。
步骤904，将校准后的模型，发送给网络设备。
本公开中，终端设备将校准后的模型发送给网络设备后，网络设备即可将该模型同步给其余终端设备，由此，使得各个终端设备均可基于该校准后的模型处理相关业务数据。
本公开中，终端设备首先接收网络设备发送的网络设备确定的第二辅助信息，之后即可将自身的第一辅助信息及第二辅助信息发送给服务器，并且在接收到服务器返回的校准后的模型后，还可以将该模型发送给网络设备。由此，通过服务器、网络设备与终端设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图10，图10是本公开实施例提供的一种模型的生成方法的交互示意图。如图10所示，该方法包括：
步骤1001，网络设备训练生成预训练模型。
步骤1002，终端设备向网络设备发送确定的第一辅助信息。
步骤1003，网络设备基于第一辅助信息及网络设备确定的第二辅助信息，对预训练模型进行校准，生成校准后的模型。
步骤1004，网络设备将校准后的模型发送给终端设备。
本公开中，通过终端设备及网络设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图11，图11是本公开实施例提供的一种模型的生成方法的交互示意图。如图11所示，该方法包括：
步骤1101，网络设备训练生成预训练模型。
步骤1102，网络设备将预训练模型及网络设备确定的第二辅助信息发送给终端设备。
步骤1103，终端设备基于确定的第一辅助信息及第二辅助信息，对预训练模型进行校准，生成校准后的模型。
步骤1104，终端设备将校准后的模型发送给网络设备。
本公开中，通过终端设备及网络设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图12，图12是本公开实施例提供的一种模型的生成方法的交互示意图。如图12所示，该方法包括：
步骤1201，服务器训练生成预训练模型。
步骤1202，网络设备将确定的第二辅助信息发送给终端设备。
步骤1203，终端设备将确定的第一辅助信息及第二辅助信息，发送给服务器。
步骤1204，服务器基于第一辅助信息及第二辅助信息，对预训练模型进行校准，生成校准后的模型。
步骤1205，服务器将校准后的模型发送给终端设备。
步骤1206，终端设备将校准后的模型发送给网络设备。
本公开中，通过服务器、终端设备及网络设备的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图13，为本公开实施例提供的一种通信装置的结构示意图。图13所示的通信装置1300可包括处理模块1301和收发模块1302。收发模块1302可包括发送模块和/或接收模块，发送模块用于实现发送功能，接收模块用于实现接收功能，收发模块1302可以实现发送功能和/或接收功能。
可以理解的是，通信装置1300可以是第一设备，也可以是第一设备中的装置，还可以是能够与第一设备匹配使用的装置。
通信装置1300在第一设备侧，其中：
处理模块1301，用于获取预训练模型；
收发模块1302，用于接收第二设备发送的模型校准辅助信息；
所述处理模块1301，还用于基于所述模型校准辅助信息，对所述预训练模型进行校准，以生成校准后模型；
所述收发模块1302，还用于将所述校准后模型，发送给所述第二设备。
可选的，第一设备为网络设备，第二设备为所述终端设备，所述收发模块1302，还用于接收所述终端设备发送的所述终端设备确定的第一辅助信息。
可选的，所述第一设备为所述终端设备，所述第二设备为网络设备，所述处理模块1301，还用于接收所述网络设备发送的所述预训练模型。
可选的，所述收发模块1302，还用于接收所述网络设备发送的所述网络设备确定的第二辅助信息。
可选的，第一设备为服务器，所述第二设备为所述的终端设备，所述收发模块1302，还用于：
接收所述终端设备发送的所述终端设备确定的第一辅助信息及所述网络设备确定的第二辅助信息。
本公开中，第一设备在获取到预训练模型及第二设备发送的模型校准辅助信息后，即可基于模型校准辅助信息，对预训练模型进行校准以得到模型，并将生成的模型发送给第二设备。由此，通过第一设备与第二设备间的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
可以理解的是，通信装置1300可以是第二设备，也可以是第二设备中的装置，还可以是能够与第二设备匹配使用的装置。
通信装置1300，在第二设备侧，其中：
收发模块1302，用于向第一设备发送模型校准辅助信息；
所述收发模块1302，还用于接收所述第一设备发送的基于所述模型校准辅助信息校准后的模型。
可选的，所述第一设备为网络设备，第二设备为所述终端设备，所述收发模块1302，还用于：
向所述网络设备发送所述终端设备确定的第一辅助信息。
可选的，所述第一设备为所述终端设备，所述第二设备为网络设备，所述收发模块1302，还用于：向所述终端设备发送所述网络设备确定的第二辅助信息。
可选的，收发模块1302，还用于向所述终端设备发送预训练模型。
可选的，所述第一设备为服务器，所述第二设备为所述的终端设备，所述收发模块1302，还用于：
向所述服务器发送所述终端设备确定的第一辅助信息及所述网络设备确定的第二辅助信息。
可选的，收发模块1302，还用于：
接收网络设备发送的所述网络设备确定的第二辅助信息。
可选的，收发模块1302，还用于：
将所述校准后模型，发送给所述网络设备。
本公开中，第二设备首先向第一设备发送模型校准辅助信息，之后，即可接收到第一设备基于模型校准辅助信息得到的模型。由此，通过第一设备与第二设备间的交互，实现了模型的灵活生成和配置，进一步提升了通信系统的性能和效率。
请参见图14，图14是本公开实施例提供的另一种通信装置的结构示意图。通信装置1400可以是第一设备，也可以是第二设备，也可以是支持第一设备实现上述方法的芯片、芯片系统、或处理器等，还可以是支持第二设备实现上述方法的芯片、芯片系统、或处理器等。该装置可用于实现上述方法实施例中描述的方法，具体可以参见上述方法实施例中的说明。
通信装置1400可以包括一个或多个处理器1401。处理器1401可以是通用处理器或者专用处理器等。例如可以是基带处理器或中央处理器。基带处理器可以用于对通信协议以及通信数据进行处理，中央处理器可以用于对通信装置(如，基站、基带芯片，终端设备、终端设备芯片，DU或CU等)进行控制，执行计算机程序，处理计算机程序的数据。
可选的，通信装置1400中还可以包括一个或多个存储器1402，其上可以存有计算机程序1404，处理器1401执行所述计算机程序1404，以使得通信装置1400执行上述方法实施例中描述的方法。可选的，所述存储器1402中还可以存储有数据。通信装置1400和存储器1402可以单独设置，也可以集成在一起。
可选的，通信装置1400还可以包括收发器1405、天线1406。收发器1405可以称为收发单元、收发机、或收发电路等，用于实现收发功能。收发器1405可以包括接收器和发送器，接收器可以称为接收机或接收电路等，用于实现接收功能；发送器可以称为发送机或发送电路等，用于实现发送功能。
可选的，通信装置1400中还可以包括一个或多个接口电路1407。接口电路1407用于接收代码指令并传输至处理器1401。处理器1401运行所述代码指令以使通信装置1400执行上述方法实施例中描 述的方法。
通信装置1400为第一设备：处理器1401用于执行图2中的步骤201、步骤203，图3中的步骤301，步骤303等，收发器1405用于执行图2中的步骤202、步骤204，或者图3中的步骤302，步骤304等等。
通信装置1400为第二设备：收发器1405用于执行图6中的步骤601、步骤602，或者，执行图7中的步骤701、步骤702等。
在一种实现方式中，处理器1401中可以包括用于实现接收和发送功能的收发器。例如该收发器可以是收发电路，或者是接口，或者是接口电路。用于实现接收和发送功能的收发电路、接口或接口电路可以是分开的，也可以集成在一起。上述收发电路、接口或接口电路可以用于代码/数据的读写，或者，上述收发电路、接口或接口电路可以用于信号的传输或传递。
在一种实现方式中，处理器1401可以存有计算机程序1403，计算机程序1403在处理器1401上运行，可使得通信装置1400执行上述方法实施例中描述的方法。计算机程序603可能固化在处理器1401中，该种情况下，处理器1401可能由硬件实现。
在一种实现方式中，通信装置1400可以包括电路，所述电路可以实现前述方法实施例中发送或接收或者通信的功能。本公开中描述的处理器和收发器可实现在集成电路(integrated circuit，IC)、模拟IC、射频集成电路RFIC、混合信号IC、专用集成电路(application specific integrated circuit，ASIC)、印刷电路板(printed circuit board，PCB)、电子设备等上。该处理器和收发器也可以用各种IC工艺技术来制造，例如互补金属氧化物半导体(complementary metal oxide semiconductor，CMOS)、N型金属氧化物半导体(nMetal-oxide-semiconductor，NMOS)、P型金属氧化物半导体(positive channel metal oxide semiconductor，PMOS)、双极结型晶体管(bipolar junction transistor，BJT)、双极CMOS(BiCMOS)、硅锗(SiGe)、砷化镓(GaAs)等。
以上实施例描述中的通信装置可以是网络设备或者智能中继，但本公开中描述的通信装置的范围并不限于此，而且通信装置的结构可以不受图14的限制。通信装置可以是独立的设备或者可以是较大设备的一部分。例如所述通信装置可以是：
(1)独立的集成电路IC，或芯片，或，芯片系统或子系统；
(2)具有一个或多个IC的集合，可选的，该IC集合也可以包括用于存储数据，计算机程序的存储部件；
(3)ASIC，例如调制解调器(Modem)；
(4)可嵌入在其他设备内的模块；
(5)接收机、终端设备、智能终端设备、蜂窝电话、无线设备、手持机、移动单元、车载设备、网络设备、云设备、人工智能设备等等；
(6)其他等等。
对于通信装置可以是芯片或芯片系统的情况，可参见图15所示的芯片的结构示意图。图15所示的芯片包括处理器1501和接口1503。其中，处理器1501的数量可以是一个或多个，接口1503的数量可以是多个。
对于芯片用于实现本公开实施例中第一设备的功能的情况：
接口1503，用于执行图2中的步骤202、步骤204等。
对于芯片用于实现本公开实施例中第二设备的功能的情况：
接口1503，用于执行图6中的步骤601、步骤602等。
可选的，芯片还包括存储器1503，存储器1503用于存储必要的计算机程序和数据。
本领域技术人员还可以了解到本公开实施例列出的各种说明性逻辑块(illustrative logical block)和步骤(step)可以通过电子硬件、电脑软件，或两者的结合进行实现。这样的功能是通过硬件还是软件来实现取决于特定的应用和整个系统的设计要求。本领域技术人员可以对于每种特定的应用，可以使用各种方法实现所述的功能，但这种实现不应被理解为超出本公开实施例保护的范围。
本公开还提供一种可读存储介质，其上存储有指令，该指令被计算机执行时实现上述任一方法实施例的功能。
本公开还提供一种计算机程序产品，该计算机程序产品被计算机执行时实现上述任一方法实施例的功能。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实 现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机程序。在计算机上加载和执行所述计算机程序时，全部或部分地产生按照本公开实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机程序可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机程序可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(digital subscriber line，DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质(例如，软盘、硬盘、磁带)、光介质(例如，高密度数字视频光盘(digital video disc，DVD))、或者半导体介质(例如，固态硬盘(solid state disk，SSD))等。
本领域普通技术人员可以理解：本公开中涉及的第一、第二等各种数字编号仅为描述方便进行的区分，并不用来限制本公开实施例的范围，也表示先后顺序。
本公开中的至少一个还可以描述为一个或多个，多个可以是两个、三个、四个或者更多个，本公开不做限制。在本公开实施例中，对于一种技术特征，通过“第一”、“第二”、“第三”、“A”、“B”、“C”和“D”等区分该种技术特征中的技术特征，该“第一”、“第二”、“第三”、“A”、“B”、“C”和“D”描述的技术特征间无先后顺序或者大小顺序。
本公开中各表所示的对应关系可以被配置，也可以是预定义的。各表中的信息的取值仅仅是举例，可以配置为其他值，本公开并不限定。在配置信息与各第一辅助的对应关系时，并不一定要求必须配置各表中示意出的所有对应关系。例如，本公开中的表格中，某些行示出的对应关系也可以不配置。又例如，可以基于上述表格做适当的变形调整，例如，拆分，合并等等。上述各表中标题示出第一辅助的名称也可以采用通信装置可理解的其他名称，其第一辅助的取值或表示方式也可以通信装置可理解的其他取值或表示方式。上述各表在实现时，也可以采用其他的数据结构，例如可以采用数组、队列、容器、栈、线性表、指针、链表、树、图、结构体、类、堆、散列表或哈希表等。
本公开中的预定义可以理解为定义、预先定义、存储、预存储、预协商、预配置、固化、或预烧制。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本公开的范围。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
以上所述，仅为本公开的具体实施方式，但本公开的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本公开揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本公开的保护范围之内。因此，本公开的保护范围应以所述权利要求的保护范围为准。

Claims (16)

  1. 一种模型的生成方法，其特征在于，由第一设备执行，所述方法包括：
    获取预训练模型；
    接收第二设备发送的模型校准辅助信息；
    基于所述模型校准辅助信息，对所述预训练模型进行校准，以生成校准后的模型；
    将所述校准后的模型，发送给所述第二设备。
  2. 如权利要求1所述的方法，其特征在于，所述第一设备为网络设备，第二设备为终端设备，所述接收第二设备发送的模型校准辅助信息，包括：
    接收所述终端设备发送的所述终端设备确定的第一辅助信息。
  3. 如权利要求1所述的方法，其特征在于，所述第一设备为所述终端设备，所述第二设备为网络设备，所述获取预训练模型，包括：
    接收所述网络设备发送的所述预训练模型。
  4. 如权利要求3所述的方法，其特征在于，所述接收第二设备发送的模型校准辅助信息，包括：
    接收所述网络设备发送的所述网络设备确定的第二辅助信息。
  5. 如权利要求1所述的方法，其特征在于，第一设备为服务器，所述第二设备为所述终端设备，所述接收第二设备发送的模型校准辅助信息，包括：
    接收所述终端设备发送的所述终端设备确定的第一辅助信息及所述网络设备确定的第二辅助信息。
  6. 一种模型生成方法，其特征在于，由第二设备执行，所述方法包括：
    向第一设备发送模型校准辅助信息；
    接收所述第一设备发送的基于所述模型校准辅助信息校准后的校准后的模型。
  7. 如权利要求6所述的方法，其特征在于，所述第一设备为网络设备，第二设备为所述终端设备，所述向第一设备发送模型校准辅助信息，包括：
    向所述网络设备发送所述终端设备确定的第一辅助信息。
  8. 如权利要求6所述的方法，其特征在于，所述第一设备为所述终端设备，所述第二设备为网络设备，所述向第一设备发送模型校准辅助信息，包括：
    向所述终端设备发送所述网络设备确定的第二辅助信息。
  9. 如权利要求8所述的方法，其特征在于，还包括：
    向所述终端设备发送预训练模型。
  10. 如权利要求6所述的方法，其特征在于，所述第一设备为服务器，所述第二设备为所述的终端设备，所述向第一设备发送模型校准辅助信息，包括：
    向所述服务器发送所述终端设备确定的第一辅助信息及关联的所述网络设备确定的第二辅助信息。
  11. 如权利要求10所述的方法，其特征在于，还包括：
    接收网络设备发送的所述的所述网络设备确定的第二辅助信息。
  12. 如权利要求10所述的方法，其特征在于，还包括：
    将所述校准后模型，发送给所述网络设备。
  13. 一种通信装置，其特征在于，所述装置设置于第一设备，所述装置包括：
    处理模块，用于获取预训练模型；
    收发模块，用于接收第二设备发送的模型校准辅助信息；
    所述处理模块，还用于基于所述模型校准辅助信息，对所述预训练模型进行校准，以生成校准后模型；
    所述收发模块，还用于将所述校准后模型，发送给所述第二设备。
  14. 一种通信装置，其特征在于，所述装置设置于第二设备，所述装置包括：
    收发模块，用于向第一设备发送模型校准辅助信息；
    所述收发模块，还用于接收所述第一设备发送的基于所述模型校准辅助信息校准后的校准后的模型。
  15. 一种通信装置，其特征在于，所述装置包括处理器和存储器，所述存储器中存储有计算机程序，所述处理器执行所述存储器中存储的计算机程序，以使所述装置执行如权利要求1至5中任一项所述的方法，或者执行如权利要求6至12中任一项所述的方法。
  16. 一种计算机可读存储介质，用于存储有指令，当所述指令被执行时，使如权利要求1至5中任 一项所述的方法被实现，或者使如权利要求6至12中任一项所述的方法被实现。
PCT/CN2022/107706 2022-07-25 2022-07-25 一种模型的生成方法及装置 WO2024020747A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2022/107706 WO2024020747A1 (zh) 2022-07-25 2022-07-25 一种模型的生成方法及装置
CN202280003069.2A CN117836785A (zh) 2022-07-25 2022-07-25 一种模型的生成方法及装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2022/107706 WO2024020747A1 (zh) 2022-07-25 2022-07-25 一种模型的生成方法及装置

Publications (1)

Publication Number Publication Date
WO2024020747A1 true WO2024020747A1 (zh) 2024-02-01

Family

ID=89704898

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/107706 WO2024020747A1 (zh) 2022-07-25 2022-07-25 一种模型的生成方法及装置

Country Status (2)

Country Link
CN (1) CN117836785A (zh)
WO (1) WO2024020747A1 (zh)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200169311A1 (en) * 2017-08-30 2020-05-28 Telefonaktiebolaget Lm Ericsson (Publ) Wireless device for determining a best antenna beam and method thereof
WO2021088000A1 (zh) * 2019-11-08 2021-05-14 Oppo广东移动通信有限公司 通信方法及装置
WO2021213376A1 (zh) * 2020-04-22 2021-10-28 维沃移动通信有限公司 定位方法、通信设备和网络设备
WO2021258798A1 (zh) * 2020-06-22 2021-12-30 华为技术有限公司 一种确定波束对的方法及装置
WO2022041947A1 (zh) * 2020-08-24 2022-03-03 华为技术有限公司 一种更新机器学习模型的方法及通信装置
WO2022048546A1 (zh) * 2020-09-04 2022-03-10 华为技术有限公司 数据传输方法和装置
WO2022073615A1 (en) * 2020-10-08 2022-04-14 Nokia Technologies Oy Arrangement for removing transmitter power amplifier distortion at a receiver

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200169311A1 (en) * 2017-08-30 2020-05-28 Telefonaktiebolaget Lm Ericsson (Publ) Wireless device for determining a best antenna beam and method thereof
WO2021088000A1 (zh) * 2019-11-08 2021-05-14 Oppo广东移动通信有限公司 通信方法及装置
WO2021213376A1 (zh) * 2020-04-22 2021-10-28 维沃移动通信有限公司 定位方法、通信设备和网络设备
WO2021258798A1 (zh) * 2020-06-22 2021-12-30 华为技术有限公司 一种确定波束对的方法及装置
WO2022041947A1 (zh) * 2020-08-24 2022-03-03 华为技术有限公司 一种更新机器学习模型的方法及通信装置
WO2022048546A1 (zh) * 2020-09-04 2022-03-10 华为技术有限公司 数据传输方法和装置
WO2022073615A1 (en) * 2020-10-08 2022-04-14 Nokia Technologies Oy Arrangement for removing transmitter power amplifier distortion at a receiver

Also Published As

Publication number Publication date
CN117836785A (zh) 2024-04-05

Similar Documents

Publication Publication Date Title
WO2023010471A1 (zh) 一种传输配置指示tci状态配置的方法及其装置
WO2023206179A1 (zh) 一种确定传输配置指示状态的方法及装置
WO2023216203A1 (zh) 资源配置方法及装置
WO2024026799A1 (zh) 数据传输方法和装置
WO2024026795A1 (zh) 一种侧行链路sl定位参考信号prs的发送方法及装置
WO2024021128A1 (zh) 同步参考源的确定方法及装置
WO2023197187A1 (zh) 一种信道状态信息的处理方法及装置
WO2022236625A1 (zh) 一种定位方法及其装置
WO2023015424A1 (zh) 一种定位方法及其装置
WO2024020747A1 (zh) 一种模型的生成方法及装置
WO2023216165A1 (zh) 一种控制智能超表面ris发射参考信号的方法及装置
WO2023193271A1 (zh) 一种双连接中终端设备小区组的更新方法及装置
WO2023201755A1 (zh) 一种移动性管理的配置方法及装置
WO2024168917A1 (zh) 一种ai模型的注册方法/装置/设备及存储介质
WO2024082196A1 (zh) 一种基于ai模型的终端定位方法及装置
WO2023201502A1 (zh) 测量配置的处理方法和装置
WO2023184452A1 (zh) 终端设备使用的模型的确定方法和装置
WO2024164124A1 (zh) 一种反馈方法、装置、设备及存储介质
WO2023173257A1 (zh) 一种请求系统信息的方法及其装置
WO2024082195A1 (zh) 一种基于ai模型的终端定位方法及装置
WO2024016245A1 (zh) 一种信息指示方法、装置、设备及存储介质
WO2024164123A1 (zh) 一种反馈方法、装置、设备及存储介质
WO2023184437A1 (zh) 信道状态信息csi报告的获取方法、装置及存储介质
WO2023123476A1 (zh) 一种时域资源传输位置的确定方法和装置
WO2024036519A1 (zh) 一种侧行链路pdcp复用的激活方法及装置

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 202280003069.2

Country of ref document: CN

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22952214

Country of ref document: EP

Kind code of ref document: A1