Generative Model Fingerprint Map Generation for Indoor Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint-based indoor positioning methods require significant time and resources to create fingerprint maps, which are labor-intensive and costly, especially in unexplored areas where satellite navigation signals are unavailable.
Innovation Solution
A method utilizing a generative model, specifically a GAN-based approach, to automatically generate fingerprint maps by extracting feature vectors from reference points and access points, and combining these to create a virtual fingerprint map for unexplored areas, reducing the need for manual data collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual data collection methods are used to create fingerprint maps, then positioning accuracy can be achieved, but the time and cost required increase significantly
Solution Approach 1:
The patent uses a generative model to create virtual fingerprint maps that copy the essential characteristics of real fingerprint maps without requiring actual physical data collection. The model learns from training data and generates synthetic fingerprint representations that maintain positioning accuracy while eliminating time-consuming manual surveys
Solution Approach 2:
The generative model is trained in advance on available fingerprint data to learn the characteristics of wireless signal propagation in specific environments. This preliminary training enables the model to quickly generate accurate virtual fingerprint maps for unexplored areas without requiring real-time data collection
2Measurement precision
If manual data collection methods are used to create fingerprint maps, then positioning accuracy can be achieved, but the cost increases significantly
Solution Approach 1:
The patent replaces expensive manual data collection processes with automated virtual map generation using generative models. By copying the essential signal characteristics through synthetic data generation, the system maintains positioning accuracy while eliminating costs associated with human surveyors, equipment deployment, and manual fieldwork
Solution Approach 2:
The system uses existing wireless network infrastructure and automatically generates virtual fingerprint maps without requiring external surveying teams. The generative model self-services by learning from available data and autonomously creating positioning maps, eliminating the need for specialized manual intervention
3Productivity
If generative models are used to generate virtual fingerprint maps, then time and cost are reduced, but the complexity of the system increases
Solution Approach 1:
The patent introduces a generative model as an intermediary between available training data and the need for positioning maps in unexplored areas. This intermediary component learns signal propagation characteristics and generates virtual fingerprints, simplifying the overall process while managing complexity through specialized AI architecture
4Measurement precision
If traditional fingerprint collection methods are used, then accurate positioning is possible, but the process becomes labor-intensive
Solution Approach 1:
The system automates the entire fingerprint map creation process using generative models that self-service by learning from training data and autonomously generating virtual fingerprints for positioning. This eliminates labor-intensive manual data collection while maintaining positioning accuracy through AI-driven virtual map generation
Data Source
AI summary
A method, a data processing apparatus, and a positioning apparatus for generating a fingerprint map for an unexplored area are proposed. The method includes receiving, by a data processing apparatus, locations of reference points and locations of a plurality of access points (APs) for a service area, extracting, by the data processing apparatus, feature vectors for each AP on a basis of the locations of the reference points and the locations of the APs, generating, by the data processing apparatus, a fingerprint map for each of the APs by inputting, as conditions, the feature vectors extracted for each corresponding AP into a previously trained generative model for each of the APs and inputting random noise, and generating, by the data processing apparatus, a final fingerprint map for the service area by combining the fingerprint maps generated for the APs.