CN-122029793-A - Performance monitoring method and device

Abstract

The embodiment of the application provides a performance monitoring method and device. The method comprises the steps that the terminal equipment receives a first configuration sent by the network equipment, monitors the performance of an AI/ML model of an enabled or activated function according to the first configuration, wherein the function comprises a feature or a feature group enabled by the configuration indicated by the capability of the terminal equipment, and determines whether the function fails or not according to the monitoring.

Inventors

• WANG GUOTONG
• WANG XIN
• SUN GANG
• ZHANG QUN
• JIN LIQIANG
• SHAN YUJIA
• GENG TINGTING

Assignees

• 壹泛尼提株式会社

Dates

Publication Date

20260512

Application Date

20230927

Claims (20)

1. A performance monitoring apparatus comprising: A receiving unit that receives a first configuration transmitted by a network device; A monitoring unit for monitoring the performance of the AI/ML model of the enabled or activated functions according to the first configuration, the functions comprising features or groups of features enabled by the configuration indicated by the terminal device capabilities, and And a processing unit for determining whether the function fails according to the monitoring.
2. The apparatus of claim 1, wherein the AI/ML model is located on a terminal device side and the first configuration comprises at least one of a performance metric, a parameter controlling the monitoring, and a reference signal configuration.
3. The apparatus of claim 1, wherein the receiving unit further receives a second configuration sent by the network device, the apparatus further comprising: And the sending unit is used for sending the AI/ML performance information corresponding to the function to the network equipment according to the second configuration.
4. The apparatus of claim 3, wherein the second configuration comprises at least one of a reporting configuration, a manner of transmitting uplink resources of the AI/ML capability information, and reporting the AI/ML capability information, the manner comprising periodic reporting, semi-persistent reporting, or aperiodic reporting, and the AI/ML capability information comprises at least one of a failure of function indication, an input data drift, an output data drift.
5. The apparatus of claim 1, wherein the function is one or more, and the terminal device performs AI/ML performance monitoring separately for each function.
6. The apparatus of claim 1, wherein the receiving unit receives a reference signal for AI/ML performance monitoring sent by the network device; The processing unit determines whether the function fails according to the reference signal.
7. The apparatus of claim 6, wherein the apparatus further comprises: A transmitting unit that transmits instruction information to the network device in case of failure of the function, and The receiving unit also receives response information fed back by the network equipment according to the indication information, wherein the response information comprises at least one of function activation information, function deactivation information, function rollback information, function switching information or reconfiguration information.
8. The apparatus of claim 1, wherein the processing unit determines whether the function failed based on a performance metric; Wherein the failure instances are counted if the performance metric is below a threshold value, and the functional failure is determined if a plurality of consecutive failure instances are detected within a time window.
9. The apparatus of claim 6, wherein the reference signal is a periodic signal for AI/ML performance monitoring for one function.
10. The apparatus of claim 9, wherein the reference signal for AI/ML performance monitoring for one function is configured by radio resource control or the reference signal for AI/ML performance monitoring for one function is determined by the terminal device.
11. The apparatus of claim 7, wherein the indication information comprises a failure indication of a function.
12. The apparatus of claim 11, wherein the failure indication of the one function is transmitted via at least one of dedicated random access channel resources, uplink control channel resources, MAC CE, two-step random access, radio resource control message.
13. The apparatus of claim 11, wherein the response information is sent for the one function and via at least one of downlink control information, MAC CE, radio resource control message.
14. The apparatus of claim 6, wherein the reference signal is a periodic signal for AI/ML performance monitoring for a plurality of functions.
15. The apparatus of claim 14, wherein the reference signal for AI/ML performance monitoring for a plurality of functions is configured by radio resource control or is determined by the terminal device.
16. The apparatus of claim 7, wherein the indication information comprises a failure indication of one or more functions.
17. The apparatus of claim 16, wherein the failure indication of the one or more functions is transmitted via at least one of an uplink control channel resource, a MAC CE, a two-step random access, a radio resource control message.
18. The apparatus of claim 16, wherein the response information is sent for the one or more functions and via at least one of downlink control information, MAC CE, radio resource control message.
19. A performance monitoring apparatus comprising: a transmitting unit that transmits the first configuration to the terminal device; Wherein the terminal device monitors the performance of the AI/ML model of the enabled or activated functionality according to the first configuration, the functionality comprising a feature or set of features enabled by the configuration indicated by the terminal device capabilities, and determining if the functionality fails according to the monitoring.
20. A communication system, comprising: a network device that transmits a first configuration to a terminal device; The terminal device monitors the performance of an AI/ML model of an enabled or activated function according to the first configuration, the function comprising a feature or set of features enabled by a configuration indicated by terminal device capabilities, and determines whether the function failed according to the monitoring.

Description

Performance monitoring method and device Technical Field The embodiment of the application relates to the technical field of communication. Background In NR Rel-18, AI/ML over the air was studied. AI/ML can be used for CSI feedback enhancement, beam management, positioning enhancement. CSI feedback enhancements may include CSI prediction, CSI compression, beam management may include spatial beam prediction, temporal beam prediction, positioning enhancements may include direct positioning, AI/ML assisted positioning. In some sub-use cases, a bilateral model may be used, i.e. the AI/ML model may be on the terminal device side and on the network device side, e.g. CSI compression may be used as a representative use case of the bilateral model. In other sub-examples, a single-sided model, i.e., the AI/ML model, may be used on the terminal device side or on the network device side. It should be noted that the foregoing description of the background art is only for the purpose of providing a clear and complete description of the technical solution of the present application and is presented for the convenience of understanding by those skilled in the art. The above-described solutions are not considered to be known to the person skilled in the art simply because they are set forth in the background of the application section. Disclosure of Invention The inventors have found that for the AI/ML model at the terminal device side, the performance should be monitored (monitor) by the terminal device. But there is no clear and accurate solution to the monitoring in particular. In view of at least one of the foregoing problems, embodiments of the present application provide a performance monitoring method and apparatus. According to an aspect of an embodiment of the present application, there is provided a performance monitoring method, including: The terminal equipment receives a first configuration sent by the network equipment; Monitoring the performance of an AI/ML model of an enabled or activated function according to the first configuration, the function comprising a feature or set of features enabled by a configuration indicated by terminal device capabilities, and And determining whether the function fails according to the monitoring. According to another aspect of an embodiment of the present application, there is provided a performance monitoring apparatus including: A receiving unit that receives a first configuration transmitted by a network device; A monitoring unit for monitoring the performance of the AI/ML model of the enabled or activated functions according to the first configuration, the functions comprising features or groups of features enabled by the configuration indicated by the terminal device capabilities, and And a processing unit for determining whether the function fails according to the monitoring. According to another aspect of an embodiment of the present application, there is provided a performance monitoring method, including: The network equipment sends a first configuration to the terminal equipment; Wherein the terminal device monitors the performance of the AI/ML model of the enabled or activated functionality according to the first configuration, the functionality comprising a feature or set of features enabled by the configuration indicated by the terminal device capabilities, and determining if the functionality fails according to the monitoring. According to another aspect of an embodiment of the present application, there is provided a performance monitoring apparatus including: a transmitting unit that transmits the first configuration to the terminal device; Wherein the terminal device monitors the performance of the AI/ML model of the enabled or activated functionality according to the first configuration, the functionality comprising a feature or set of features enabled by the configuration indicated by the terminal device capabilities, and determining if the functionality fails according to the monitoring. According to another aspect of an embodiment of the present application, there is provided a communication system including: a network device that transmits a first configuration to a terminal device; The terminal device monitors the performance of an AI/ML model of an enabled or activated function according to the first configuration, the function comprising a feature or set of features enabled by a configuration indicated by terminal device capabilities, and determines whether the function failed according to the monitoring. The method and the device have the advantages that the AI/ML model at the terminal equipment side can be accurately monitored, so that the accuracy and the reliability of the AI/ML are improved. Specific embodiments of the application are disclosed in detail below with reference to the following description and drawings, indicating the manner in which the principles of the application may be employed. It should be understood that the embodiments of