Accurate Indoor-Positioning Model Based on People Effect and Ray-Tracing Propagation

Area of Science:

• Computer Science
• Electrical Engineering
• Signal Processing

Background:

• Wireless local area networks (WLAN) fingerprinting is a key technology for indoor positioning.
• Accuracy is limited by signal strength fluctuations caused by obstacles, particularly people.
• Existing models often neglect the impact of human presence and multipath effects.

Purpose of the Study:

• To develop an accurate indoor positioning model that incorporates people's presence effect (PPE) and multipath propagation.
• To enhance the reliability of WLAN-based indoor positioning systems.

Main Methods:

• Proposed an accurate indoor positioning (AIRY) model using ray-tracing.
• Developed an automatic radio map generation technique via ray tracing.
• Implemented a constant people's effect for received signal strength indicator (RSSI) adaptation.
• Utilized a K-nearest-neighbor (KNN) algorithm for position determination.

Main Results:

• Simulations were performed using MATLAB.
• The model was tested in a real-world environment (Level 3, Menara Razak, Universiti Teknologi Malaysia).
• Initial positioning accuracy was 2.04 m, significantly improved to 0.57 m after considering PPE and multipath effects.

Conclusions:

• The proposed AIRY model effectively addresses the limitations of traditional WLAN fingerprinting.
• Incorporating people's presence and multipath effects leads to substantial improvements in indoor positioning accuracy.
• This research offers a more robust solution for precise indoor navigation.