Related Experiment Video
Updated: Oct 10, 2025

Using a Real-Time Locating System to Measure Walking Activity Associated with Wandering Behaviors Among Institutionalized Older Adults
Published on: February 8, 2019
A Survey of Recent Indoor Localization Scenarios and Methodologies
Tian Yang1, Adnane Cabani1, Houcine Chafouk1
1Normandie Université, UNIROUEN, ESIGELEC, IRSEEM, 76000 Rouen, France.
This survey reviews indoor localization techniques, focusing on improving accuracy by processing raw data and employing machine learning and Bayesian filtering methods. It compares various algorithms for better scalability, stability, and reliability in diverse network conditions.
Area of Science:
- Computer Science
- Electrical Engineering
- Signal Processing
Background:
- Indoor localization is crucial, with trilateration as a classic geometric approach using Received Signal Strength Indication (RSSI) data for distance estimation.
- Raw RSSI and other measurement data (e.g., Time Difference of Arrival, Distance of Arrival, Round Trip Time) are susceptible to errors from multi-path effects, interference, and noise, necessitating advanced processing.
- Existing techniques are categorized based on network structures and channel conditions, specifically Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS) scenarios.
Purpose of the Study:
- To provide a comprehensive survey of existing indoor localization techniques.
- To analyze and compare various methods for improving localization accuracy and reducing system complexity.
- To evaluate performance features including scalability, stability, and reliability across different algorithms and application scenarios.
Main Methods:
- Exploration of RSSI-based fingerprinting techniques using supervised machine learning (Support Vector Machine, K-Nearest Neighbors, Neural Networks) in offline training.
- Application of unsupervised methods (Isolation Forest, K-Means, Expectation Maximization) for enhancing online localization accuracy.
- Introduction of Bayesian filtering methods, including Linear Kalman Filters (LKF) and nonlinear filters (Extended KF, Cubature KF, Unscented KF, Particle Filters) for dynamic models.
Main Results:
- Comparison of various indoor localization techniques, highlighting their strengths and weaknesses in different network and channel conditions (LOS/NLOS).
- Evaluation of machine learning and Bayesian filtering approaches for their effectiveness in mitigating errors and improving positioning precision.
- Assessment of algorithms based on key performance indicators: accuracy, scalability, stability, reliability, and computational complexity.
Conclusions:
- Advanced data processing and sophisticated algorithms like machine learning and Bayesian filtering are essential for accurate indoor localization.
- The choice of technique depends on specific application requirements, network structure, and channel characteristics.
- This survey offers a comparative perspective to guide the selection and development of practical indoor localization systems with improved performance.
Related Concept Videos
Types of Global Positioning System Surveys
Design Example: Identifying the Locations of Monuments in the Field Using Global Positioning System Device
Field Application of Global Positioning System
Methods of Obtaining Topography
Errors in Global Positioning System
Design Example: Measuring Distance Between Two Points with Obstructions

