Monitoring method, device and system, NWDAF, SF and computer program product

Abstract

The invention relates to a monitoring method, device and system, an NWDAF, an SF and a computer program product, and relates to the technical field of mobile communication. The monitoring method is executed by a target network data analysis function NWDAF, and comprises the following steps: according to a model subscription request of a sensing function SF, determining and sending a sensing model to the SF; registration information of the SF is received, a feedback information subscription request is sent to the SF, and the registration information is used for indicating that a model used by the SF is a perception model; and monitoring the performance of the perception model according to feedback information of the SF.

Description

Technical Field

[0001] The present disclosure relates to the field of mobile communication technology, and in particular to a monitoring method, a monitoring device, a monitoring system, a NWDAF, a SF, a computer-readable storage medium, and a computer program product. Background Art

[0002] With the rapid development of wireless communication technology and the growing demand for high-quality data transmission, the integration of communication and perception technology is gradually becoming the focus of technological innovation. In the field of intelligent perception, the performance of intelligent perception is crucial to the accuracy of data transmission and service quality. Summary of the invention

[0003] The inventors of the present disclosure have discovered that the above-mentioned related technologies have the following problem: how to ensure the performance of intelligent perception.

[0004] In view of this, the present disclosure proposes a monitoring method, in which the target NWDAF determines and sends a perception model to the SF according to the model subscription request of the SF, and then receives the registration information of the SF to determine that the model currently used by the SF is the perception model. When it is determined that the model currently used by the SF is the perception model, a feedback information subscription request is sent to the SF, and finally the performance of the perception model is monitored through the feedback information of the SF. By sending the perception model to the SF, the SF can send feedback information to the target NWDAF according to the perception model. By sending the feedback information subscription request to the SF, the SF can send feedback information to the target NWDAF, so as to monitor the performance of the perception model according to the feedback information. By exchanging registration information and feedback information between the SF and the target NWDAF, close collaboration between the SF and the target NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception.

[0005] According to some embodiments of the first aspect of the present disclosure, a monitoring method is provided, which is executed by a target network data analysis function NWDAF, including: determining and sending a perception model to a perception function SF according to a model subscription request of the SF; receiving registration information of the SF, and sending a feedback information subscription request to the SF, wherein the registration information is used to indicate that the model used by the SF is the perception model; and monitoring the performance of the perception model according to the feedback information of the SF.

[0006] In some embodiments, the feedback information includes at least one of a perceptual inference result of a perceptual model provided by the SF and perceptual result feedback information provided by a perceptual service using device.

[0007] In some embodiments, the monitoring method further includes: if the performance of the perception model meets the performance requirements, not updating the perception model; if the performance of the perception model does not meet the performance requirements, updating the perception model.

[0008] In some embodiments, when the performance of the perception model does not meet the performance requirements, updating the perception model includes: re-collecting model training data and retraining the perception model; or replacing the perception model and retraining the replaced perception model.

[0009] In some embodiments, the monitoring method further comprises: sending result information to the SF, wherein the result information comprises a performance monitoring result of the perception model and / or an updated perception model.

[0010] In some embodiments, the monitoring method further includes: collecting perception monitoring data, wherein the perception monitoring data is used to monitor the performance of the perception model; and monitoring the performance of the perception model based on the feedback information and the perception monitoring data.

[0011] In some embodiments, the perception monitoring data includes at least one of historical performance of a perception model from an analytical data storage function ADRF, subscription data changes of a perception service usage device from a unified data management UDM, and measurement data from a user equipment UE or a base station.

[0012] In some embodiments, monitoring the performance of the perception model according to the feedback information includes: when the feedback information is feedback information sent by multiple SFs, monitoring the performance of the perception model according to the feedback information sent by multiple SFs.

[0013] In some embodiments, according to the model subscription request of the perception function SF, determining and sending the perception model to SF includes: collecting model training data according to the model subscription request; training the model corresponding to the model subscription request according to the model training data to determine the perception model; and sending the perception model to SF.

[0014] In some embodiments, the model subscription request includes business type, business requirements and model filtering information. Collecting model training data according to the model subscription request includes: receiving the model subscription request sent by SF; collecting model training data according to the business type, business requirements and model filtering information.

[0015] In some embodiments, the model subscription request further includes at least one of SF interoperability information and sensor node information.

[0016] According to some embodiments of the second aspect of the present disclosure, a monitoring method is provided, which is executed by a perception function SF, including: sending a model subscription request to a target NWDAF, and receiving a perception model sent by the target NWDAF; sending registration information to the target NWDAF, wherein the registration information is used to indicate that the model used by SF is a perception model; and sending feedback information to the target NWDAF upon receiving a feedback information subscription request sent by the target NWDAF.

[0017] In some embodiments, the feedback information includes at least one of a perceptual inference result of a perceptual model provided by the SF and perceptual result feedback information provided by a perceptual service using device.

[0018] In some embodiments, when the feedback information includes the perceptual reasoning results of the perceptual model provided by the SF, sending the feedback information to the target NWDAF includes: obtaining the perceptual reasoning results corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; and sending the perceptual reasoning results as feedback information to the target NWDAF.

[0019] In some embodiments, sending feedback information to the target NWDAF also includes: receiving perception node information sent by the perception service usage device; and collecting perception reasoning data based on the perception node information.

[0020] In some embodiments, when the feedback information includes the perceptual reasoning results of the perceptual model provided by the SF and the perceptual result feedback information provided by the perceptual service user device, sending the feedback information to the target NWDAF includes: obtaining the perceptual reasoning results corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; sending the perceptual reasoning results to the perceptual service user device, and receiving the perceptual result feedback information sent by the perceptual service user device; and sending the perceptual reasoning results and the perceptual result feedback information as feedback information to the target NWDAF.

[0021] In some embodiments, the monitoring method further includes: receiving result information sent by the target NWDAF, wherein the result information includes a performance monitoring result of the perception model and / or an updated perception model.

[0022] In some embodiments, the monitoring method further includes: determining a target NWDAF according to a perception service request sent by the perception service using device, wherein the perception service request includes at least one of a service type, a service requirement, and perception node information.

[0023] In some embodiments, determining the target NWDAF according to the perception service request sent by the perception service usage device includes: determining the target NWDAF according to the perception service request, the perception model training capability identifier of the NWDAF, and the model interoperability information of the NWDAF.

[0024] In some embodiments, the monitoring method further includes: determining whether to obtain a perceptual reasoning result corresponding to the perceptual service request through a perceptual model based on a perceptual service request sent by the perceptual service usage device; and when it is determined that the perceptual reasoning result is obtained through the perceptual model, performing an operation of determining the target NWDAF.

[0025] According to some embodiments of the third aspect of the present disclosure, there is provided an NWDAF, including: a determination unit, configured to determine and send a perception model to a perception function SF according to a model subscription request of the SF; a first sending unit, configured to receive registration information of the SF, and send a feedback information subscription request to the SF, wherein the registration information is used to indicate that the model used by the SF is the perception model; and a monitoring unit, configured to monitor the performance of the perception model according to the feedback information of the SF.

[0026] In some embodiments, the feedback information includes at least one of a perceptual inference result of a perceptual model provided by the SF and perceptual result feedback information provided by a perceptual service using device.

[0027] In some embodiments, the NWDAF further includes an updating unit configured to not update the perception model if the performance of the perception model meets the performance requirement; and to update the perception model if the performance of the perception model does not meet the performance requirement.

[0028] In some embodiments, the updating unit is further configured to re-collect model training data and re-train the perception model; or replace the perception model and re-train the replaced perception model.

[0029] In some embodiments, the first sending unit is further configured to send result information to the SF, wherein the result information includes a performance monitoring result of the perception model and / or an updated perception model.

[0030] In some embodiments, the monitoring unit is further configured to collect perception monitoring data, wherein the perception monitoring data is used to monitor the performance of the perception model; and the performance of the perception model is monitored based on the feedback information and the perception monitoring data.

[0031] In some embodiments, the perception monitoring data includes at least one of historical performance of a perception model from an analytical data storage function ADRF, subscription data changes of a perception service usage device from a unified data management UDM, and measurement data from a user equipment UE or a base station.

[0032] In some embodiments, the monitoring unit is further configured to monitor the performance of the perception model according to the feedback information sent by the multiple SFs when the feedback information is feedback information sent by the multiple SFs.

[0033] In some embodiments, the determination unit is further configured to collect model training data according to the model subscription request; train the model corresponding to the model subscription request according to the model training data to determine the perception model; and send the perception model to SF.

[0034] In some embodiments, the model subscription request includes a business type, business requirements, and model filtering information, and the determination unit is further configured to receive the model subscription request sent by the SF; and collect model training data according to the business type, business requirements, and model filtering information.

[0035] In some embodiments, the model subscription request further includes at least one of SF interoperability information and sensor node information.

[0036] According to some embodiments of the fourth aspect of the present disclosure, a SF is provided, including: a second sending unit, configured to send a model subscription request to a target NWDAF, and receive a perception model sent by the target NWDAF; a third sending unit, configured to send registration information to the target NWDAF, wherein the registration information is used to indicate that a model used by the SF is a perception model; and a fourth sending unit, configured to send feedback information to the target NWDAF upon receiving a feedback information subscription request sent by the target NWDAF.

[0037] In some embodiments, the feedback information includes at least one of a perceptual inference result of a perceptual model provided by the SF and perceptual result feedback information provided by a perceptual service using device.

[0038] In some embodiments, the fourth sending unit is further configured to obtain a perceptual reasoning result corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; and send the perceptual reasoning result as feedback information to the target NWDAF.

[0039] In some embodiments, the fourth sending unit is further configured to receive perception node information sent by the perception service usage device; and collect perception reasoning data based on the perception node information.

[0040] In some embodiments, the fourth sending unit is also configured to obtain the perceptual reasoning result corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; send the perceptual reasoning result to the perceptual service user device, and receive the perceptual result feedback information sent by the perceptual service user device; and send the perceptual reasoning result and the perceptual result feedback information as feedback information to the target NWDAF.

[0041] In some embodiments, the SF further includes a receiving unit configured to receive result information sent by the target NWDAF, wherein the result information includes a performance monitoring result of the perception model and / or an updated perception model.

[0042] In some embodiments, the SF further includes a determination unit configured to determine a target NWDAF according to a perception service request sent by the perception service usage device, wherein the perception service request includes at least one of a service type, a service requirement, and perception node information.

[0043] In some embodiments, the determination unit is further configured to determine the target NWDAF based on the perception service request, the perception model training capability identifier of the NWDAF, and the model interoperability information of the NWDAF.

[0044] In some embodiments, the determination unit is also configured to determine whether to obtain the perceptual reasoning result corresponding to the perceptual service request through the perceptual model based on the perceptual service request sent by the perceptual service usage device; when it is determined that the perceptual reasoning result is obtained through the perceptual model, perform the operation of determining the target NWDAF.

[0045] According to some embodiments of the fifth aspect of the present disclosure, a monitoring device is provided, comprising: a memory and a processor coupled to the memory, wherein the processor is configured to execute the monitoring method in any one of the above embodiments based on instructions stored in the memory.

[0046] According to some embodiments of the sixth aspect of the present disclosure, a monitoring system is provided, comprising: the NWDAF in any of the above embodiments and the SF in any of the above embodiments.

[0047] In some embodiments, the monitoring system further includes: a perception service using device configured to send a perception service request to the SF and receive a perception reasoning result.

[0048] According to some embodiments of the seventh aspect of the present disclosure, a computer-readable storage medium is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the monitoring method in any of the above embodiments is implemented.

[0049] According to some embodiments of the eighth aspect of the present disclosure, a computer program product is provided, comprising computer instructions, which, when executed by a processor, implement the monitoring method in any one of the above embodiments.

[0050] In the above embodiment, the target NWDAF determines and sends the perception model to SF according to the model subscription request of SF, and then receives the registration information of SF to determine that the model currently used by SF is the perception model. When it is determined that the model currently used by SF is the perception model, a feedback information subscription request is sent to SF, and finally the performance of the perception model is monitored through the feedback information of SF. By sending the perception model to SF, SF can send feedback information to the target NWDAF according to the perception model. By sending the feedback information subscription request to SF, SF can send feedback information to the target NWDAF, so that the performance of the perception model can be monitored according to the feedback information. By exchanging registration information and feedback information between SF and the target NWDAF, close cooperation between SF and the target NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0052] The present disclosure can be more clearly understood from the following detailed description with reference to the accompanying drawings.

[0053] Figure 1 A flow chart showing some embodiments of the monitoring method of the present disclosure.

[0054] Figure 2 Flowcharts showing other embodiments of the monitoring method of the present disclosure.

[0055] Figure 3 Schematic diagram showing some embodiments of the NWDAF of the present disclosure.

[0056] Figure 4 Schematic diagram showing some embodiments of the SF of the present disclosure.

[0057] Figure 5 Schematic diagrams showing some embodiments of the monitoring device of the present disclosure.

[0058] Figure 6 Schematic diagrams showing some embodiments of the monitoring system of the present disclosure.

[0059] Figure 7 Schematic diagrams showing still some embodiments of the monitoring method of the present disclosure. DETAILED DESCRIPTION

[0060] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless otherwise specifically stated.

[0061] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0062] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0063] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0064] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0065] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0066] The integration of communication and perception technologies is gradually becoming the focus of technological innovation. The core of this integrated technology is to complete sensing and communication tasks simultaneously through wireless radio frequency (RF) signals, without adding additional hardware costs, to achieve real-time monitoring of the environment and rapid data transmission. This not only helps to optimize the configuration of network resources and improve network efficiency, but also brings users a richer and more accurate service experience, such as location-based customized recommendations and intelligent traffic management systems, which greatly enriches the user experience and provides a solid technical foundation for the development of intelligent applications.

[0067] With the continuous advancement of AI (Artificial Intelligence) and ML (Machine Learning) technologies, these two technologies play an increasingly important role in the field of intelligent perception. AI or ML models can intelligently perceive changes in the environment or specific targets by analyzing and processing massive amounts of data. In the research and development of synaesthesia network architecture, SF (Sensing Function) has been widely introduced as a key unit in the core network, responsible for collecting and processing environmental perception data. At the same time, NWDAF (Network Data Analytics Function) is a network data analysis function newly introduced by 3GPP (3rd Generation Partnership Project) in the 5G (5th Generation Mobile Networks) R15 version, playing the role of a big data analysis engine in the core network. NWDAF is responsible for data collection, model training, reasoning and judgment, intelligent prediction, and provides analysis results to support policy decision-making for networks, network management, and applications. With the continuous evolution of network technology, network intelligence has become an irreversible trend. By combining the data collected by SF with the analytical capabilities of NWDAF, not only can deeper data integration and intelligent analysis be achieved, but network performance can also be further optimized and the overall user experience can be improved.

[0068] In the field of intelligent perception, the performance stability of intelligent perception is crucial to ensure data accuracy and service quality. Due to the inconsistency of data quality and the limitations of model generalization ability, intelligent perception may encounter performance fluctuations in actual applications. These problems may arise from a variety of factors, including but not limited to noise in data collection, environmental changes, model obsolescence, or improper parameter settings. How to ensure the performance of intelligent perception is a problem that needs to be solved.

[0069] In response to the above problems, the present disclosure proposes a monitoring method, in which the target NWDAF determines and sends a perception model to the SF according to the model subscription request of the SF, and then receives the registration information of the SF to determine that the model currently used by the SF is the perception model. When it is determined that the model currently used by the SF is the perception model, a feedback information subscription request is sent to the SF, and finally the performance of the perception model is monitored through the feedback information of the SF. By sending the perception model to the SF, the SF can send feedback information to the target NWDAF according to the perception model. By sending the feedback information subscription request to the SF, the SF can send feedback information to the target NWDAF, so as to monitor the performance of the perception model according to the feedback information. By exchanging registration information and feedback information between the SF and the target NWDAF, close collaboration between the SF and the target NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception, as follows.

[0070] Figure 1 A flow chart showing some embodiments of the monitoring method of the present disclosure.

[0071] like Figure 1 As shown, the monitoring method includes steps 110 to 130, and the monitoring method is performed by the target NWDAF. The target NWDAF is mainly responsible for model training and performance monitoring of the perception model, and the SF is mainly responsible for processing the perception reasoning data and reasoning calculation of the perception reasoning results. Among them, the SF can be combined with the AnLF (Analytics Logical Function) responsible for model reasoning or embed the module responsible for reasoning calculation, so that the SF can infer and calculate the perception reasoning results based on the perception model trained by the target NWDAF and the perception reasoning data collected by itself, and the SF can send the perception reasoning results to the perception service user device and the target NWDAF in different forms, so that the target NWDAF can monitor the performance of the perception model, as follows.

[0072] In step 110, according to the model subscription request of the perception function SF, the perception model is determined and sent to the SF.

[0073] In some embodiments, the SF may send a model subscription request to the target NWDAF and receive the perception model sent by the target NWDAF. For example, the SF may send a model subscription request to the target NWDAF through the Nnwdaf_MLModelProvision_Subscribe service.

[0074] In some embodiments, how the SF determines the target NWDAF from one or more NWDAFs is as follows: the SF determines the target NWDAF according to the sensing service request sent by the sensing service user device, wherein the sensing service request includes at least one of the service type (e.g., dynamic map, vehicle tracking, etc.), service requirements (e.g., sensing resolution, sensing accuracy, duration, target area information, latency, etc.), and sensing node information (e.g., UE information). The sensing service user device may be, for example, a UE (User Equipment) or an AF (Application Function).

[0075] For example, the SF determines the target NWDAF based on the perception service request, the perception model training capability identifier of the NWDAF, and the model interoperability information of the NWDAF.

[0076] Specifically, SF can determine whether NWDAF supports the training of the perception model based on the perception model training capability identifier of NWDAF, and can determine whether NWDAF supports model sharing with SF based on the model interoperability information of NWDAF.

[0077] For example, SF can discover the target NWDAF with MTLF (Model Training Logical Function) that supports perception model subscription and sharing through NRF (Network Repository Function) (the target NWDAF includes the perception model training capability identifier, NWDAF model filtering information, and NWDAF model interoperability information when registering). Among them, the model filtering information of NWDAF can be used for model selection to obtain the model corresponding to the model subscription request.

[0078] By adding the model training capability identifier, NWDAF model filtering information, and NWDAF model interoperability information during the target NWDAF registration process, the collaboration between SF and the target NWDAF, the subscription of perception models, and the sharing of perception models are ensured to proceed smoothly. Among them, the NWDAF model interoperability information includes the target NWDAF manufacturer identifier and the format and operating environment of the models it supports.

[0079] In some embodiments, before determining the target NWDAF, SF may also determine whether to obtain the perceptual reasoning result corresponding to the perceptual service request through the perceptual model based on the perceptual service request sent by the perceptual service usage device; if it is determined that the perceptual reasoning result is obtained through the perceptual model, the operation of determining the target NWDAF is performed.

[0080] For example, SF can decide whether to use traditional algorithms or perception model-based methods to calculate perception reasoning results based on built-in rules, which are defined by the manufacturer. The reference basis for judgment includes the performance requirements of perception reasoning results, the perceived environment (such as indoors and outdoors), and the perception scene (such as environmental intrusion monitoring and moving object tracking).

[0081] In some embodiments, step 110 includes steps (1) to (3), as follows.

[0082] In step (1), model training data is collected based on the model subscription request.

[0083] In step (2), the model corresponding to the model subscription request is trained according to the model training data to determine the perception model.

[0084] In step (3), the perception model is sent to the SF. For example, the target NWDAF may send the perception model or the information of the perception model to the SF via the Nnwdaf_MLModelProvision_Notify service.

[0085] In some embodiments, the model subscription request includes business type, business requirements and model filtering information (including slice information, target area, etc.). According to the model subscription request, collecting model training data can, for example, first receive the model subscription request sent by SF, and then collect model training data according to the business type, business requirements and model filtering information (model filtering information includes slice information).

[0086] In some embodiments, the model subscription request further includes at least one of SF interoperability information (the SF interoperability information includes the vendor identifier of the SF and the format and operating environment of the model supported by the SF) and sensor node information.

[0087] Through the interoperability information of SF and the target NWDAF, it is determined whether model sharing can be achieved between SF and the target NWDAF, ensuring the smooth progress of model sharing.

[0088] In step 120, the registration information of the SF is received, and a feedback information subscription request is sent to the SF, wherein the registration information is used to indicate that the model used by the SF is a perception model.

[0089] Subscribe to SF's feedback information by sending a feedback information subscription request to SF, so as to facilitate subsequent real-time monitoring of the performance of the perception model.

[0090] By receiving the registration information of SF to trigger the sending of feedback information subscription request to SF, the timeliness of monitoring of the perception model is guaranteed.

[0091] In some embodiments, the SF sends registration information to the target NWDAF, where the registration information is used to indicate that the model used by the SF is a perception model. The SF also receives a feedback information subscription request sent by the target NWDAF. If the SF receives the feedback information subscription request sent by the target NWDAF, the SF sends feedback information to the target NWDAF.

[0092] The SF sends registration information to the target NWDAF to inform the target NWDAF that the model currently used by the SF is the perception model.

[0093] If the feedback information includes the perceptual reasoning results of the perceptual model provided by the SF (for example, it may be relevant accuracy information of the perceptual model), the SF will obtain the perceptual reasoning results corresponding to the perceptual business request corresponding to the perceptual business request based on the collected perceptual reasoning data and the perceptual model, and then send the perceptual reasoning results as feedback information to the target NWDAF.

[0094] SF can collect perception and reasoning data in real time, perform inference calculations on the perception and reasoning data according to the perception model, and obtain perception and reasoning results. SF helps to optimize the allocation of network resources and improve the overall efficiency of the network.

[0095] In some embodiments, how to collect perception reasoning data is as follows: first, receiving the perception node information sent by the perception service user device, and then collecting the perception reasoning data according to the perception node information. The perception node information includes the identification information of the perception node, network information location information, etc.

[0096] If the feedback information includes the perceptual reasoning results of the perceptual model provided by the SF and the perceptual result feedback information provided by the perceptual service user device, the SF will obtain the perceptual reasoning results corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; then send the perceptual reasoning results to the perceptual service user device, and receive the perceptual result feedback information sent by the perceptual service user device; finally, send the perceptual reasoning results and the perceptual result feedback information as feedback information to the target NWDAF.

[0097] For example, the feedback information can be the accuracy of the perception reasoning results (such as the accuracy and false alarm rate of target detection), or the perception delay, the environmental adaptability of the perception model, etc.

[0098] In step 130, the performance of the perception model is monitored according to the feedback information of the SF. For example, the performance of the perception model may be the accuracy and working efficiency of the perception model.

[0099] For example, the feedback information includes at least one of a perception inference result of a perception model provided by the SF and perception result feedback information provided by a perception service using device.

[0100] Through the information interaction between SF and target NWDAF (registration information, feedback information subscription request and feedback information interaction), the performance of intelligent perception is guaranteed, and the perception performance and network intelligence level are improved.

[0101] In some embodiments, if the feedback information is feedback information sent by multiple SFs, the performance of the perception model is monitored according to the feedback information sent by the multiple SFs.

[0102] By monitoring the performance of the perception model based on the feedback information of multiple SFs, the target NWDAF can monitor the performance of the perception model under different environments and different data.

[0103] In some embodiments, monitoring the performance of the perception model may also be performed based on feedback information and perception monitoring data, as follows.

[0104] First, the perception monitoring data is collected, where the perception monitoring data is used to monitor the performance of the perception model, and then the performance of the perception model is monitored based on the feedback information and the perception monitoring data. The perception monitoring data can come from UE, RAN (Radio Access Network), and AMF (Access and Mobility Management Function).

[0105] For example, the perception monitoring data includes at least one of the historical performance (e.g., historical accuracy) of the perception model from the analytical data storage function ADRF, subscription data changes of the perception service usage device from the unified data management UDM, and measurement data from the user equipment UE or the base station (e.g., Doppler frequency shift from the UE or the base station, TOA (Time of Arrival) of the communication, TDOA (Time Difference of Arrival), AOA (Angle of Arrival), etc.).

[0106] By monitoring the performance of the perception model based on feedback information and perception monitoring data, the accuracy of monitoring the performance of the perception model is improved.

[0107] In some embodiments, if the performance of the perception model meets the performance requirement, the perception model is not updated; if the performance of the perception model does not meet the performance requirement, the perception model is updated.

[0108] For example, updating the perception model may include re-collecting model training data and retraining the perception model, or may include replacing the perception model and retraining the replaced perception model.

[0109] In some embodiments, the target NWDAF sends result information to the SF, wherein the result information includes the performance monitoring result of the perception model and / or the updated perception model.

[0110] If the target NWDAF updates the perception model by monitoring the performance of the perception model, the updated perception model needs to be sent to the SF through result information to optimize the performance of the intelligent perception.

[0111] By monitoring the performance of the perception model and providing feedback (the feedback mechanism refers to sending result information to SF), it is possible to ensure that the perception model maintains optimal performance in the intelligent perception business and promptly adapts to environmental changes and the evolution of data characteristics, thereby providing more accurate and reliable intelligent perception services.

[0112] In the above embodiment, the target NWDAF determines and sends the perception model to SF according to the model subscription request of SF, and then receives the registration information of SF to determine that the model currently used by SF is the perception model. When it is determined that the model currently used by SF is the perception model, a feedback information subscription request is sent to SF, and finally the performance of the perception model is monitored through the feedback information of SF. By sending the perception model to SF, SF can send feedback information to the target NWDAF according to the perception model. By sending the feedback information subscription request to SF, SF can send feedback information to the target NWDAF, so that the performance of the perception model can be monitored according to the feedback information. By exchanging registration information and feedback information between SF and the target NWDAF, close cooperation between SF and the target NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception.

[0113] Figure 2 Flowcharts showing other embodiments of the monitoring method of the present disclosure.

[0114] like Figure 2 As shown, the monitoring method includes steps 210 to 230, and the monitoring method is executed by SF.

[0115] In step 210, a model subscription request is sent to the target NWDAF, and the perception model sent by the target NWDAF is received.

[0116] In some embodiments, the SF further determines whether to obtain the perceptual inference result corresponding to the perceptual service request through the perceptual model according to the perceptual service request sent by the perceptual service user device; if it is determined that the perceptual inference result is obtained through the perceptual model, the operation of determining the target NWDAF is performed. In some embodiments, the SF further determines the target NWDAF according to the perceptual service request sent by the perceptual service user device, wherein the perceptual service request includes at least one of the service type, service requirements, and perceptual node information.

[0117] In some embodiments, how the SF determines the target NWDAF is specifically as follows: the target NWDAF is determined according to the perception service request, the perception model training capability identifier of the NWDAF, and the model interoperability information of the NWDAF.

[0118] In step 220, registration information is sent to the target NWDAF, wherein the registration information is used to indicate that the model used by the SF is a perception model.

[0119] In step 230, when a feedback information subscription request sent by the target NWDAF is received, feedback information is sent to the target NWDAF.

[0120] In some embodiments, if the feedback information includes the perceptual reasoning results of the perceptual model provided by the SF, sending the feedback information to the target NWDAF specifically includes: obtaining the perceptual reasoning results corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; and sending the perceptual reasoning results as feedback information to the target NWDAF.

[0121] Regarding how to collect perception reasoning data, the details are as follows: receiving perception node information sent by the perception service usage device; and collecting perception reasoning data based on the perception node information.

[0122] In some embodiments, if the feedback information includes the perceptual reasoning results of the perceptual model provided by the SF and the perceptual result feedback information provided by the perceptual service user device, sending the feedback information to the target NWDAF specifically includes: obtaining the perceptual reasoning results corresponding to the perceptual service request based on the collected perceptual reasoning data and perceptual model; sending the perceptual reasoning results to the perceptual service user device, and receiving the perceptual result feedback information sent by the perceptual service user device; sending the perceptual reasoning results and the perceptual result feedback information as feedback information to the target NWDAF.

[0123] In some embodiments, the SF also receives result information sent by the target NWDAF, wherein the result information includes performance monitoring results of the perception model and / or an updated perception model.

[0124] In the above embodiment, SF sends a model subscription request to the target NWDAF, receives the perception model sent by the target NWDAF, and then sends registration information to the target NWDAF, wherein the registration information is used to indicate that the model used by SF is the perception model. Finally, after receiving the feedback information subscription request sent by the target NWDAF, SF will send feedback information to the target NWDAF. By sending a model subscription request to the target NWDAF, the perception model is obtained, so that SF can subsequently obtain the perception reasoning results based on the perception model and perception reasoning data. By sending registration information to the target NWDAF, the target NWDAF is notified that the model currently used by SF is the perception model, so that the target NWDAF can start monitoring the perception model. By sending feedback information to the target NWDAF, the target NWDAF can monitor the perception model based on the feedback information, thereby ensuring the performance of intelligent perception.

[0125] Figure 3 Schematic diagram showing some embodiments of the NWDAF of the present disclosure.

[0126] like Figure 3 As shown, the NWDAF 30 includes a determination unit 31 , a first sending unit 32 and a monitoring unit 33 .

[0127] The determination unit 31 is configured to determine and send the perception model to the perception function SF according to the model subscription request of the SF.

[0128] In some embodiments, the determination unit 31 is further configured to collect model training data according to the model subscription request; train the model corresponding to the model subscription request according to the model training data to determine the perception model; and send the perception model to SF.

[0129] In some embodiments, the model subscription request includes a service type, service requirements and model filtering information, and the determination unit 31 is further configured to receive the model subscription request sent by the SF; and collect model training data according to the service type, service requirements and model filtering information.

[0130] In some embodiments, the model subscription request further includes at least one of SF interoperability information and sensor node information.

[0131] The first sending unit 32 is configured to receive registration information of the SF, and send a feedback information subscription request to the SF, wherein the registration information is used to indicate that the model used by the SF is a perception model.

[0132] In some embodiments, the first sending unit 32 is further configured to send result information to the SF, wherein the result information includes a performance monitoring result of the perception model and / or an updated perception model.

[0133] The monitoring unit 33 is configured to monitor the performance of the perception model according to the feedback information of the SF.

[0134] In some embodiments, the feedback information includes at least one of a perceptual inference result of a perceptual model provided by the SF and perceptual result feedback information provided by a perceptual service using device.

[0135] In some embodiments, the monitoring unit 33 is further configured to collect perception monitoring data, wherein the perception monitoring data is used to monitor the performance of the perception model; and the performance of the perception model is monitored based on the feedback information and the perception monitoring data.

[0136] In some embodiments, the perception monitoring data includes at least one of historical performance of a perception model from an analytical data storage function ADRF, subscription data changes of a perception service usage device from a unified data management UDM, and measurement data from a user equipment UE or a base station.

[0137] In some embodiments, the monitoring unit 33 is further configured to monitor the performance of the perception model according to the feedback information sent by the multiple SFs when the feedback information is feedback information sent by the multiple SFs.

[0138] In some embodiments, NWDAF 30 further includes an updating unit configured to not update the perception model if the performance of the perception model meets the performance requirement; and to update the perception model if the performance of the perception model does not meet the performance requirement.

[0139] In some embodiments, the updating unit is further configured to re-collect model training data and re-train the perception model; or replace the perception model and re-train the replaced perception model.

[0140] In the above embodiment, NWDAF determines and sends the perception model to SF according to the model subscription request of SF, and then receives the registration information of SF to determine that the model currently used by SF is the perception model. When it is determined that the model currently used by SF is the perception model, NWDAF sends a feedback information subscription request to SF, and finally monitors the performance of the perception model through the feedback information of SF. By sending the perception model to SF, SF can send feedback information to NWDAF according to the perception model. By sending the feedback information subscription request to SF, SF can send feedback information to NWDAF, so as to monitor the performance of the perception model according to the feedback information. By exchanging registration information and feedback information between SF and NWDAF, close cooperation between SF and NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception.

[0141] Figure 4 Schematic diagram showing some embodiments of the SF of the present disclosure.

[0142] like Figure 4 As shown, SF40 includes a second sending unit 41 , a third sending unit 42 and a fourth sending unit 43 .

[0143] The second sending unit 41 is configured to send a model subscription request to the target NWDAF, and receive the perception model sent by the target NWDAF.

[0144] In some embodiments, SF40 further includes a determination unit configured to determine a target NWDAF according to a perception service request sent by the perception service usage device, wherein the perception service request includes at least one of a service type, a service requirement, and perception node information.

[0145] In some embodiments, the determination unit is further configured to determine the target NWDAF based on the perception service request, the perception model training capability identifier of the NWDAF, and the model interoperability information of the NWDAF.

[0146] In some embodiments, the determination unit is also configured to determine whether to obtain the perceptual reasoning result corresponding to the perceptual service request through the perceptual model based on the perceptual service request sent by the perceptual service usage device; when it is determined that the perceptual reasoning result is obtained through the perceptual model, perform the operation of determining the target NWDAF.

[0147] The third sending unit 42 is configured to send registration information to the target NWDAF, wherein the registration information is used to indicate that the model used by the SF is a perception model.

[0148] The fourth sending unit 43 is configured to send feedback information to the target NWDAF upon receiving a feedback information subscription request sent by the target NWDAF.

[0149] In some embodiments, the feedback information includes at least one of a perception inference result of a perception model provided by SF40 and perception result feedback information provided by a perception service using device.

[0150] In some embodiments, the fourth sending unit 43 is further configured to obtain a perception reasoning result corresponding to the perception service request based on the collected perception reasoning data and the perception model; and send the perception reasoning result as feedback information to the target NWDAF.

[0151] In some embodiments, the fourth sending unit 43 is further configured to receive the perception node information sent by the perception service using device; and collect the perception reasoning data according to the perception node information.

[0152] In some embodiments, the fourth sending unit 43 is also configured to obtain the perceptual reasoning result corresponding to the perceptual service request based on the collected perceptual reasoning data and the perceptual model; send the perceptual reasoning result to the perceptual service user device, and receive the perceptual result feedback information sent by the perceptual service user device; and send the perceptual reasoning result and the perceptual result feedback information as feedback information to the target NWDAF.

[0153] In some embodiments, SF40 further includes a receiving unit configured to receive result information sent by the target NWDAF, wherein the result information includes a performance monitoring result of the perception model and / or an updated perception model.

[0154] In the above embodiment, SF sends a model subscription request to the target NWDAF, receives the perception model sent by the target NWDAF, and then sends registration information to the target NWDAF, wherein the registration information is used to indicate that the model used by SF is the perception model. Finally, after receiving the feedback information subscription request sent by the target NWDAF, SF will send feedback information to the target NWDAF. By sending a model subscription request to the target NWDAF, the perception model is obtained, so that SF can subsequently obtain the perception reasoning results based on the perception model and perception reasoning data. By sending registration information to the target NWDAF, the target NWDAF is notified that the model currently used by SF is the perception model, so that the target NWDAF can start monitoring the perception model. By sending feedback information to the target NWDAF, the target NWDAF can monitor the perception model based on the feedback information, thereby ensuring the performance of intelligent perception.

[0155] Figure 5 Schematic diagrams showing some embodiments of the monitoring device of the present disclosure.

[0156] like Figure 5 As shown, the monitoring device 50 of this embodiment includes: a memory 51 and a processor 52 coupled to the memory 51 , and the processor 52 is configured to execute the monitoring method in any one of the aforementioned embodiments based on instructions stored in the memory 51 .

[0157] The memory 51 may include, for example, a system memory, a fixed non-volatile storage medium, etc. The system memory may store, for example, an operating system, an application program, a boot loader, and other programs.

[0158] The monitoring device 50 may also include an input / output interface 53, a network interface 54, a storage interface 55, etc. These interfaces 53, 54, 55, the memory 51, and the processor 52 may be connected, for example, via a bus 56. The input / output interface 53 provides a connection interface for input / output devices such as a display, a mouse, a keyboard, a touch screen, a microphone, and a speaker. The network interface 54 provides a connection interface for various networked devices. The storage interface 55 provides a connection interface for external storage devices such as an SD card and a USB flash drive.

[0159] In the above embodiment, the target NWDAF determines and sends the perception model to SF according to the model subscription request of SF, and then receives the registration information of SF to determine that the model currently used by SF is the perception model. When it is determined that the model currently used by SF is the perception model, a feedback information subscription request is sent to SF, and finally the performance of the perception model is monitored through the feedback information of SF. By sending the perception model to SF, SF can send feedback information to the target NWDAF according to the perception model. By sending the feedback information subscription request to SF, SF can send feedback information to the target NWDAF, so that the performance of the perception model can be monitored according to the feedback information. By exchanging registration information and feedback information between SF and the target NWDAF, close cooperation between SF and the target NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception.

[0160] Figure 6 Schematic diagrams showing some embodiments of the monitoring system of the present disclosure.

[0161] like Figure 6 As shown, the monitoring system 60 includes the NWDAF 61 in any of the above embodiments and the SF 62 in any of the above embodiments.

[0162] In some embodiments, the monitoring system 60 further includes: a perception service using device configured to send a perception service request to the SF and receive a perception reasoning result.

[0163] In some embodiments, the sensing service using device is further configured to send sensing result feedback information to the SF.

[0164] In the above embodiment, NWDAF determines and sends the perception model to SF according to the model subscription request of SF, and then receives the registration information of SF to determine that the model currently used by SF is the perception model. When it is determined that the model currently used by SF is the perception model, NWDAF sends a feedback information subscription request to SF, and finally monitors the performance of the perception model through the feedback information of SF. By sending the perception model to SF, SF can send feedback information to NWDAF according to the perception model. By sending the feedback information subscription request to SF, SF can send feedback information to NWDAF, so as to monitor the performance of the perception model according to the feedback information. By exchanging registration information and feedback information between SF and NWDAF, close cooperation between SF and NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception.

[0165] Figure 7 Schematic diagrams showing still some embodiments of the monitoring method of the present disclosure.

[0166] like Figure 7 As shown, the monitoring method includes steps 710 to 721.

[0167] In step 710, a sensing service using device (such as UE (User Equipment) or AF (Application Function)) sends a sensing service request (Sensing request) to the SF. The sensing service request includes service type (such as dynamic map, vehicle tracking, etc.), service requirements (such as sensing resolution, sensing accuracy, duration, target area information and latency, etc.) and sensing node information (such as UE information).

[0168] In step 711 , the SF sends a model subscription request to the target NWDAF (including the MTLF), that is, the SF subscribes to the perception model for AI / ML perception from the target NWDAF through the Nnwdaf_MLModelProvision_Subscribe service.

[0169] In some embodiments, when SF determines to use the perception model for perception, SF will discover the target NWDAF containing MTLF that supports subscription and sharing of the perception model through NRF (the target NWDAF includes the perception model training capability identifier, the model filtering information of the NWDAF, and the model interoperability information of the NWDAF when registering).

[0170] In step 712, the target NWDAF collects model training data, and trains the model corresponding to the model subscription request according to the model training data to obtain a perception model.

[0171] In step 713, after completing the model training, the target NWDAF sends the perception model (perception model information) subscribed by the SF to the SF through the Nnwdaf_MLModelProvision_Notify service.

[0172] In step 714, after receiving the perception model trained by the target NWDAF, the SF collects perception reasoning data, performs reasoning, and obtains the perception reasoning result.

[0173] In step 715, the SF sends the perception reasoning result to the perception service user device.

[0174] In step 716, the SF sends registration information to the target NWDAF, that is, the SF registers the model used by the SF with the target NWDAF, so that the target NWDAF can track the usage and performance of the perception model.

[0175] In step 717, the target NWDAF sends a feedback information subscription request to the SF, that is, the target NWDAF subscribes to the feedback information from the SF to monitor the real-time performance of the perception model.

[0176] In some embodiments, the feedback information includes at least one of a perceptual inference result of a perceptual model provided by the SF and perceptual result feedback information provided by a perceptual service using device.

[0177] In step 718, the SF sends the feedback information to the target NWDAF for the target NWDAF to monitor the performance of the perception model.

[0178] In step 719, the target NWDAF collects perception monitoring data (for example, the perception monitoring data may come from UE, RAN, AMF, etc.), wherein the perception monitoring data is used to monitor the performance of the perception model.

[0179] In step 720, the target NWDAF monitors the performance of the perception model based on the feedback information and the perception monitoring data, and determines whether to update the perception model.

[0180] In some embodiments, updating the perception model includes: re-collecting model training data and re-training the perception model; or replacing the perception model and re-training the replaced perception model.

[0181] In step 721, the target NWDAF sends result information to the SF, where the result information includes the performance monitoring result of the perception model and / or the updated perception model.

[0182] In the above embodiment, the target NWDAF determines and sends the perception model to SF according to the model subscription request of SF, and then receives the registration information of SF to determine that the model currently used by SF is the perception model. When it is determined that the model currently used by SF is the perception model, a feedback information subscription request is sent to SF, and finally the performance of the perception model is monitored through the feedback information of SF. By sending the perception model to SF, SF can send feedback information to the target NWDAF according to the perception model. By sending the feedback information subscription request to SF, SF can send feedback information to the target NWDAF, so that the performance of the perception model can be monitored according to the feedback information. By exchanging registration information and feedback information between SF and the target NWDAF, close cooperation between SF and the target NWDAF is achieved to continuously monitor the performance of the perception model and ensure the performance of intelligent perception.

[0183] In some embodiments, a computer program product is protected, including a computer program or instruction, which implements the above-mentioned monitoring method when executed by a processor. The computer program product includes a computer instruction carried on a computer-readable medium, and the computer instruction contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer instruction can be downloaded and installed from a network through a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by the CPU, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0184] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present disclosure may take the form of a computer program product implemented on one or more computer-usable non-transient storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0185] So far, the monitoring method, device, system, NWDAF, SF and computer program product of the present disclosure have been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed herein.

[0186] The method and system of the present disclosure may be implemented in many ways. For example, the method and system of the present disclosure may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present disclosure are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present disclosure may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present disclosure. Therefore, the present disclosure also covers a recording medium storing a program for executing the method according to the present disclosure.

[0187] Although some specific embodiments of the present disclosure have been described in detail by way of example, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. It should be understood by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A monitoring method, performed by a target network data analysis function NWDAF, comprising: Determine and send a perception model to the perception function SF according to a model subscription request of the SF; receiving registration information of the SF, and sending a feedback information subscription request to the SF, wherein the registration information is used to indicate that a model used by the SF is the perception model; The performance of the perception model is monitored according to the feedback information of the SF.

2. The monitoring method according to claim 1, wherein: The feedback information includes at least one of a perception inference result of the perception model provided by the SF and perception result feedback information provided by a perception service using device.

3. The monitoring method according to claim 1, further comprising: When the performance of the perception model meets the performance requirement, the perception model is not updated; When the performance of the perception model does not meet the performance requirement, the perception model is updated.

4. The monitoring method according to claim 3, wherein: When the performance of the perception model does not meet the performance requirement, updating the perception model includes: Re-collect model training data and re-train the perception model; or The perception model is replaced and the replaced perception model is retrained.

5. The monitoring method according to claim 4, further comprising: Sending result information to the SF, wherein the result information includes the performance monitoring result of the perception model and / or the updated perception model.

6. The monitoring method according to any one of claims 1 to 5, further comprising: Collecting perception monitoring data, wherein the perception monitoring data is used to monitor the performance of the perception model; The performance of the perception model is monitored according to the feedback information and the perception monitoring data.

7. The monitoring method according to claim 6, wherein: The perception monitoring data includes at least one of historical performance of the perception model from an analytical data storage function ADRF, subscription data changes of perception service usage devices from a unified data management UDM, and measurement data from a user equipment UE or a base station.

8. The monitoring method according to any one of claims 1 to 5, wherein: The monitoring of the performance of the perception model according to the feedback information includes: In a case where the feedback information is feedback information sent by multiple SFs, the performance of the perception model is monitored according to the feedback information sent by the multiple SFs.

9. The monitoring method according to any one of claims 1 to 5, wherein: The determining and sending the perception model to the perception function SF according to the model subscription request of the SF comprises: Collecting model training data according to the model subscription request; According to the model training data, a model corresponding to the model subscription request is trained to determine the perception model; The perceptual model is sent to the SF.

10. The monitoring method according to claim 9, wherein: The model subscription request includes a business type, business requirements, and model filtering information. The collecting of model training data according to the model subscription request includes: Receiving a model subscription request sent by the SF; Model training data is collected according to the business type, the business requirements and the model filtering information.

11. The monitoring method according to claim 10, wherein: The model subscription request also includes at least one of SF interoperability information and perception node information.

12. A monitoring method, performed by a sensing function SF, comprising: Send a model subscription request to a target NWDAF, and receive a perception model sent by the target NWDAF; Sending registration information to the target NWDAF, wherein the registration information is used to indicate that the model used by the SF is the perception model; In case of receiving the feedback information subscription request sent by the target NWDAF, feedback information is sent to the target NWDAF.

13. The monitoring method according to claim 12, wherein: In a case where the feedback information includes a perceptual inference result of the perceptual model provided by the SF, the sending the feedback information to the target NWDAF includes: Acquire a perception reasoning result corresponding to the perception service request according to the collected perception reasoning data and the perception model; The perception reasoning result is sent to the target NWDAF as the feedback information.

14. The monitoring method according to claim 13, wherein: The sending feedback information to the target NWDAF further includes: Receiving the sensing node information sent by the sensing service user device; The perception reasoning data is collected according to the perception node information.

15. The monitoring method according to claim 12, wherein: In a case where the feedback information includes the perception inference result of the perception model provided by the SF and the perception result feedback information provided by the perception service using device, the sending the feedback information to the target NWDAF includes: Acquire a perception reasoning result corresponding to the perception service request according to the collected perception reasoning data and the perception model; Sending the perception reasoning result to the perception service usage device, and receiving the perception result feedback information sent by the perception service usage device; The perception reasoning result and the perception result feedback information are sent to the target NWDAF as the feedback information.

16. The monitoring method according to any one of claims 12 to 15, further comprising: Receive result information sent by the target NWDAF, wherein the result information includes a performance monitoring result of the perception model and / or an updated perception model.

17. The monitoring method according to any one of claims 12 to 15, further comprising: The target NWDAF is determined according to a sensing service request sent by a sensing service using device, wherein the sensing service request includes at least one of a service type, a service requirement, and sensing node information.

18. The monitoring method according to claim 17, wherein: The determining of the target NWDAF according to the perception service request sent by the perception service using device includes: The target NWDAF is determined according to the perception service request, the perception model training capability identifier of the NWDAF, and the model interoperability information of the NWDAF.

19. The monitoring method according to any one of claims 12 to 15, further comprising: Determining, according to the perception service request sent by the perception service user device, whether to obtain a perception reasoning result corresponding to the perception service request through a perception model; In the case where it is determined that the perceptual inference result is obtained through the perceptual model, the operation of determining the target NWDAF is performed.

20. A network data analysis function NWDAF, comprising: a determining unit configured to determine and send a perception model to the perception function SF according to a model subscription request of the SF; a first sending unit, configured to receive registration information of the SF, and send a feedback information subscription request to the SF, wherein the registration information is used to indicate that a model used by the SF is the perception model; The monitoring unit is configured to monitor the performance of the perception model according to the feedback information of the SF.

21. A sensing function SF, comprising: A second sending unit is configured to send a model subscription request to a target NWDAF, and receive a perception model sent by the target NWDAF; A third sending unit is configured to send registration information to the target NWDAF, wherein the registration information is used to indicate that the model used by the SF is the perception model; The fourth sending unit is configured to send feedback information to the target NWDAF when receiving the feedback information subscription request sent by the target NWDAF.

22. A monitoring device comprising: Memory; and A processor coupled to the memory, the processor being configured to execute the monitoring method according to any one of claims 1 to 19 based on instructions stored in the memory.

23. A monitoring system comprising: The NWDAF as claimed in claim 20; The SF as claimed in claim 21.

24. The monitoring system of claim 23, further comprising: The perception service using device is configured to send the perception service request to the SF and receive the perception reasoning result.

25. A computer-readable storage medium having computer instructions stored thereon, wherein the instructions, when executed by a processor, implement the monitoring method according to any one of claims 1 to 19.

26. A computer program product, comprising computer instructions, wherein when the computer instructions are executed by a processor, the monitoring method according to any one of claims 1 to 19 is implemented.

Images