Mobile Localization for Indoor Iot Services: From Proof of Concept to Real-Word Experimentation

Abstract

The demand for indoor localization in Internet of Things (IoT) applications has prompted the need for improved positioning techniques. Existing methods face limitations due to Non-Line of Sight (NLOS) conditions. This paper presents an extended approach that introduces a novel indoor positioning system utilizing Mobile Access Points (MAP) instead of Static Access Points (SAP). An evaluation was conducted through simulations and real-world experiments in a 204x495 cm2 area using USRP platforms and Simulink MATLAB. Two client models were employed: one capturing distinct Received Signal Strength Indicator (RSSI) values from two access points with a novel frequency switch design and the other utilizing two USRPs to capture single RSSI values. The evaluation covered Line of Sight (LOS) and Non-Line of Sight (NLOS) conditions, employing metrics such as root mean square error (RMSE). The proposed system was compared to existing approaches, including the fingerprinting technique for signal strength analysis. Results demonstrated the system’s effectiveness, achieving an impressive RMSE of 0.26 m. This work contributes to bridging telecom and data processing elements, translating simulations into real-world experiences, and providing a comprehensive evaluation framework. The research addresses the need for enhanced indoor localization in IoT applications, advancing the state of the art in the field.

Keywords: RSS, NLOS, TOA, AOA, USRP, RMSE, Fingerprint localization method, Indoor positioning