AI Model Prediction for Wireless Network Status Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems lack efficient methods for utilizing artificial intelligence (AI) to optimize the performance of radio access networks (RAN) in terms of energy saving, load balancing, traffic steering, and mobility optimization.
Innovation Solution
A method and apparatus that involve transmitting requests and information between nodes to utilize trained AI models for predicting wireless network status, including traffic load, reliability, latency, and link quality, with feedback mechanisms for model improvement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If AI technology is introduced to optimize RAN performance, then energy saving, load balancing, traffic steering, and mobility optimization are improved, but device complexity and implementation difficulty increase
Solution Approach 1:
The patent introduces a NWDAF (Network Data Analytics Function) as an intermediary component that centralizes AI model management and analytics processing. This mediator separates the complexity of AI operations from individual RAN nodes, allowing them to focus on core networking functions while benefiting from centralized intelligence for optimization.
Solution Approach 2:
The patent divides the AI-based optimization system into separate functional segments: model training is performed separately from model inference, with different network functions (AMF, SMF, RAN nodes) having specific roles. This segmentation allows each component to be optimized independently and simplifies the overall system architecture.
2Measurement precision
If AI models are trained and deployed for network status prediction, then prediction accuracy improves, but information processing complexity and data requirements increase
Solution Approach 1:
The patent extracts and specifies the exact information elements needed for AI model training and inference, such as subscriber location information, service area details, and network status parameters. By clearly defining the required information structure, the system reduces data processing complexity while maintaining prediction accuracy.
Solution Approach 2:
The patent transforms raw network data into standardized parameters and features that are suitable for AI processing. This parameter transformation involves selecting, transforming, and aggregating data in specific ways that optimize model performance while managing information processing requirements.
3Reliability
If feedback mechanisms are implemented for AI model improvement, then model accuracy over time improves, but system complexity and operational overhead increase
Solution Approach 1:
The patent implements feedback mechanisms where the NWDAF receives actual network status information and compares it with AI predictions. This feedback loop enables continuous model improvement by adjusting parameters based on actual performance, thereby increasing reliability while managing complexity through structured feedback processing.
Data Source
AI summary
The present application relates to a method and an apparatus for determining a prediction for a status of a wireless network. One embodiment of the present disclosure provides a method for determining a prediction for a status of a wireless network, comprising: transmitting a first request associated with the prediction to a first node and a second node; receiving information of a trained AI model from the second node; transmitting, to the first node, input for the trained AI model; and receiving the prediction from the first node, wherein the prediction is determined based on the trained AI model and the input.