Hybrid indoor localization using multiple Radio Interfaces

Abstract

This paper presents a hybrid approach for real-time indoor localization without interaction with end users or network operators. The proposed solution uses signal strength information from multiple Radio Interfaces (RIs) to estimate locations of target devices. This solution benefits from different characteristics of radio signals at each RI to overcome challenges such as multipath propagation. The proposed algorithms may combine signal information prior to or after the localization process. The system operates blindly without a priori knowledge of the environment layout or target devices' radio settings. Results of real indoor experiments with both WiFi and GSM show an improvement by the proposed hybrid solution with a median error of 1.6m compared to 2.3m for WiFi and 3.0m for GSM.