Related Experiment Video
Updated: Mar 23, 2026

Trajectory Data Analyses for Pedestrian Space-time Activity Study
Published on: February 25, 2013
Scalable Indoor Localization via Mobile Crowdsourcing and Gaussian Process
Qiang Chang1, Qun Li2, Zesen Shi3
1College of Information Systems and Management, National University of Defense Technology, Changsha 410073, China. changqiang@nudt.edu.cn.
This study introduces a new indoor positioning algorithm that creates a virtual signal database, improving accuracy by 25.5% and enabling localization in unsurveyed areas. The method enhances Received Signal Strength Indication (RSSI) fingerprinting efficiency.
Area of Science:
- * Computer Science
- * Electrical Engineering
- * Geomatics Engineering
Background:
- * Indoor localization relies heavily on Received Signal Strength Indication (RSSI) fingerprinting, which requires dense signal databases for accuracy.
- * Building and maintaining these databases is resource-intensive and impractical for many areas, limiting current algorithms.
- * Existing methods struggle in areas lacking calibration data, necessitating more efficient database creation and update strategies.
Purpose of the Study:
- * To develop a scalable indoor positioning algorithm effective in both surveyed and unsurveyed regions.
- * To address the limitations of traditional RSSI fingerprinting by reducing reliance on dense, manually created signal databases.
- * To enhance indoor localization accuracy and coverage through novel virtual database construction and estimation techniques.
Main Methods:
- * Proposed the Minimum Inverse Distance (MID) algorithm to construct a virtual database with uniformly distributed virtual Reference Points (RPs).
- * Employed a Local Gaussian Process (LGP) to estimate RSSI values for virtual RPs using crowdsourced training data.
- * Enhanced a Bayesian algorithm for user location estimation utilizing the generated virtual database.
Main Results:
- * Achieved a 25.5% improvement in positioning accuracy within surveyed areas compared to existing methods.
- * Demonstrated an average positioning error below 2.2 meters for 80% of test cases.
- * Successfully enabled indoor localization in neighboring unsurveyed areas, expanding the applicability of the algorithm.
Conclusions:
- * The proposed scalable algorithm significantly enhances indoor positioning accuracy and efficiency.
- * It overcomes the limitations of traditional fingerprinting by creating virtual databases, reducing manual effort.
- * The method offers practical solutions for indoor localization in diverse environments, including those without prior calibration data.
Related Concept Videos
Field Application of Global Positioning System
Types of Global Positioning System Surveys
Design Example: Identifying the Locations of Monuments in the Field Using Global Positioning System Device
Errors in Global Positioning System
Introduction to Global Positioning System
Distance Measurements by Taping
