Robust Sensor Fusion for Indoor Wireless Localization

TL;DR

This paper introduces a sensor fusion framework combining RSS, inertial sensors, and cooperative mobile nodes to improve indoor wireless localization accuracy in challenging environments.

Contribution

It proposes a novel cooperative sensor fusion approach using EKF that enhances indoor localization accuracy under severe multipath and NLOS conditions.

Findings

Significantly improved localization accuracy with cooperative mobile nodes.

Effective use of inertial sensors and RSS in a unified framework.

Theoretical and simulation results validate the approach's robustness.

Abstract

Location knowledge in indoor environment using Indoor Positioning Systems (IPS) has become very useful and popular in recent years. Indoor wireless localization suffers from severe multi-path fading and non-line-of-sight conditions. This paper presents a novel indoor localization framework based on sensor fusion of Zigbee Wireless Sensor Networks (WSN) using Received Signal Strength (RSS). The unknown position is equipped with two or more mobile nodes. The range between two mobile nodes is fixed as priori. The attitude (roll, pitch, and yaw) of the mobile node are measured by inertial sensors (ISs). Then the angle and the range between any two nodes can be obtained, and thus the path between the two nodes can be modeled as a curve. Through an efficient cooperation between two or more mobile nodes, this framework effectively exploits the RSS techniques. This constraint help improve the positioning accuracy. Theoretical analysis on localization distortion and Monte Carlo simulations shows that the proposed cooperative strategy of multiple nodes with extended Kalman filter (EKF) achieves significantly higher positioning accuracy than the existing systems, especially in heavily obstructed scenarios.