Information Processing Device for Fixed Access Point Position Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for estimating a mobile terminal's position using radio waves from radio access points are compromised by movable access points, which introduce noise and reduce accuracy due to varying positions, especially when generating and using machine learning models.
Innovation Solution
An information processing device that acquires and filters measurement data to identify fixed radio access points by setting a threshold for time, position, and reception intensity differences, determining whether a target access point is fixed and useful for position estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data from all radio access points (including movable APs) is used for model generation and position estimation, then the quantity of available data increases, but the accuracy of position estimation deteriorates due to noise from movable APs
Solution Approach 1:
The patent extracts and separates data from fixed radio access points from data from movable radio access points. The determination unit identifies whether each radio access point is fixed or movable, and the extraction unit selectively extracts data only from fixed APs for model generation and position estimation, eliminating the noise introduced by movable APs while preserving valuable data from stable sources
Solution Approach 2:
The patent applies different quality criteria to different data sources. Fixed radio access points are treated as reliable data sources with high quality, while movable radio access points are treated as noisy data sources with low quality. The system selectively processes data based on the local quality characteristic of each AP type, using fixed AP data for accurate position estimation while excluding movable AP data
2Adaptability or versatility
If data from movable radio access points is included in learning data, then the model can be trained with more diverse data, but the model accuracy deteriorates due to position variations of movable APs
Solution Approach 1:
The determination unit extracts and identifies movable radio access points from the dataset, and the extraction unit removes data from these movable APs before feeding data to the model generation unit. This ensures the model is trained exclusively on data from fixed APs with stable positions, maintaining high accuracy while still providing sufficient training data from reliable sources
Data Source
AI summary
A server performs: acquiring measurement data of a mobile terminal; extracting fixed AP measurement data of a specific fixed AP; extracting measurement data including a measurement time of which a difference from the measurement time in the fixed AP measurement data is equal to or less than a threshold value, position information of which a difference from a position of the specific fixed AP is equal to or less than a threshold value, and a reception intensity of which a difference from the reception intensity in the fixed AP measurement data is equal to or less than a threshold value as target measurement data and to set the radio access point corresponding to the target measurement data as a target AP; and determining whether the target AP corresponds to a fixed AP on the basis of the measurement data corresponding to the target AP and the fixed AP measurement data.