Self-configuring indoor localization based on low-cost ultrasonic range sensors

Area of Science:

• Computer Science
• Robotics
• Sensor Networks

Background:

• Target tracking in smart environments is vital for applications like activity recognition.
• Accurate sensor location mapping is critical for reliable target tracking.
• Manual sensor configuration and immobility limitations hinder current systems.

Purpose of the Study:

• To develop a self-configuring, device-free localization protocol for ultrasonic sensor networks.
• To autonomously identify network topology and detect structural changes.
• To reduce hardware and deployment costs through automated processes.

Main Methods:

• Utilized genetic algorithms for autonomous network topology identification.
• Implemented a device-free localization protocol for ultrasonic range sensors.
• Developed a system for rapid map generation and network change detection.

Main Results:

• Achieved topology mapping error between 7.16–17.53 cm.
• Demonstrated mobile target tracking with an average error of 11.71–18.43 cm.
• Successfully detected displacements of 1.41–3.16 m in approximately 30 seconds.

Conclusions:

• The proposed protocol enables self-configuration and accurate localization in smart environments.
• It significantly reduces costs associated with sensor deployment and maintenance.
• The system offers real-time adaptability to network changes for robust target tracking.