AI-Assisted Base Station Positioning With LMF Model Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 5G mobile communication technologies face challenges in accurately locating user equipment (UE) due to the need for AI model training and updating on both the LMF device and base station sides, which is not adequately addressed by current methods.
Innovation Solution
A positioning method involving assistance information exchange between the base station and LMF device to train, adjust, or update AI models for accurate UE location, utilizing deployment scenario, LOS/NLOS probabilities, and equipment information to enhance AI positioning accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If AI model training and updating are performed on both LMF device and base station sides, then positioning accuracy is improved, but device complexity and information management burden increase
Solution Approach 1:
The patent extracts the AI model training and updating operations from the base station side and concentrates them on the LMF device side. The base station only needs to obtain assistance information from the LMF device and perform straightforward positioning calculations, while the LMF device handles the complex AI model training, updates, and management. This extraction of complex functions to a dedicated management entity resolves the technical contradiction by maintaining positioning accuracy through centralized AI processing while significantly reducing base station complexity.
2Measurement precision
If assistance information is exchanged between base station and LMF device, then AI positioning accuracy is improved, but channel occupancy and communication overhead increase
Solution Approach 1:
The patent applies preliminary action by having the LMF device pre-train AI models and prepare assistance information before actual positioning operations. The assistance information containing pre-processed AI model parameters, deployment scenario configurations, and equipment characteristics is prepared in advance and stored on the LMF device. During positioning operations, the base station simply retrieves this pre-prepared information rather than exchanging large volumes of raw data, thus improving positioning accuracy while minimizing channel occupancy.
3Measurement precision
If AI models are aligned with specific deployment scenarios and UE characteristics, then positioning accuracy is improved, but information processing requirements and system complexity increase
Solution Approach 1:
The patent implements local quality by creating specialized AI models tailored to specific deployment scenarios (e.g., urban canyon, rural area, indoor) and UE characteristics (e.g., device type, antenna configuration). The LMF device maintains multiple scenario-specific AI models and selects the appropriate model based on the current deployment environment and UE properties. This approach improves positioning accuracy by using locally optimized models while the LMF device's centralized management prevents information loss by systematically organizing and tracking which models apply to which scenarios.
Data Source
AI summary
A positioning method, performed by a base station, includes: sending assistance information, wherein the assistance information is configured for assisting a location management function (IMF) device in performing artificial intelligence (AI) positioning on a user equipment (UE).


