Network Analytics Model Service for Customized NWDAF Subscription
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network standards, such as those defined by 3GPP, 3GPP2, ITU, and ETSI, are limiting in their configuration of Network Data Analytics Functions (NWDAF) and subscribed network devices, restricting their ability to manage network operations, policies, and configurations based on tailored management logic. Additionally, there is no specified procedure for creating and provisioning analytic models for NWDAF.
Innovation Solution
A network analytics model service is introduced, which includes an ingestion service for receiving customized models defined by a Network Data Analytics Function (NWDAF) model data specification (NMDS), and a provisioning service for providing these models to NWDAF and allowing subscription to customized analytics by other network devices. This service enables the creation, addition, and distribution of customized models, generating analytics information beyond current network standards.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If current network standards (3GPP, 3GPP2, ITU, ETSI) are used for configuring NWDAF and subscribed network devices, then network operation management is standardized and consistent, but the ability to manage network operations based on tailored management logic is restricted
Solution Approach 1:
The patent segments the network analytics functionality by introducing a separate analytics model service that operates independently from the standardized NWDAF configuration. This allows customized analytics models to be developed, trained, and deployed separately, then integrated with the standardized network operations management framework, thereby enabling tailored management logic without complicating the core standardized configuration.
Solution Approach 2:
The patent introduces an analytics model service as an intermediary layer between the standardized network standards and the customized analytics needs. This intermediary service handles the complexity of custom model development, training, and deployment, while presenting a standardized interface to the NWDAF and subscribed network devices, thus enabling adaptability without increasing configuration complexity.
2Adaptability or versatility
If customized analytics models are developed beyond current network standards, then analytics information generation becomes more versatile and tailored, but the procedure for creating and provisioning these models is not specified
Solution Approach 1:
The patent implements preliminary action by establishing a complete analytics model service framework in advance, including model development, training, validation, and deployment procedures. This pre-established framework provides a specified procedure for creating and provisioning customized analytics models, making the process systematic and manageable rather than ad-hoc.
Solution Approach 2:
The analytics model service enables self-service by allowing network operators to independently develop, train, and deploy their own customized analytics models using the provided framework. The service includes automated model training capabilities and provisioning mechanisms, reducing the need for manual intervention and complex coordination, thus improving ease of operation.
3Productivity
If a flexible approach to analytics generation is implemented, then network operations management becomes more versatile, but the system complexity increases due to custom model development and provisioning
Solution Approach 1:
The patent implements universality by designing the analytics model service to handle multiple functions within a single unified framework: model development, training, validation, deployment, and management. This multi-functional approach consolidates what would otherwise be separate complex systems into one cohesive service, enabling flexible analytics generation while controlling overall system complexity.
Solution Approach 2:
The patent incorporates feedback mechanisms in the analytics model service, including model performance monitoring, validation feedback loops, and automated retraining capabilities. These feedback mechanisms enable the system to self-optimize and adapt, improving network operations management efficiency while reducing the complexity of manual model management through automated closed-loop control.
Data Source
AI summary
A method, a device, and a non-transitory storage medium are described in which a network analytics model service is provided. The service may include ingesting customized models and non-customized models by a network analytics device. The ingestion procedure may include receiving the model and model specification data. The service may include provisioning the model and registering the model for discovery by other network devices. The service may include specifying customized identifiers to be used prospective subscribing network devices for the analytics information. The service may also facilitate the use of the analytics information by the subscribing network devices.