RSSI-based localization in industrial environments: A Wi-Fi/BLE hybrid approach

Abstract

Providing location information by taking advantage of signal features from networking activity is a cost-effective approach to substituting conventional GNSS technologies in harsh and indoor industrial environments for Industry 5.0. This paper presents a novel hybrid indoor positioning method that harnesses the strengths of both Bluetooth Low Energy (BLE) and Wi-Fi communications to mitigate both weaknesses by taking advantage of their high availability and great potential for Location-Enabled IoT (LE-IoT). In the proposed method, at first, Wi-Fi technology is used to perform location estimation at the zone level. In the next step, our approach integrates a weighted aging forecasting technique (for predicting the RSSI of lost packets) with a moving average filter (for noise filtering). This method effectively mitigates environmental noise effects. In the last step, a zone-specific path loss modeling method is used, which is based on diverse environmental scenarios encountered in various industrial zones. For the evaluation of the proposed method, we implemented a real testbed inside a lab environment with different zones to show the effect of noise filtering and zone-level path loss modeling. The experiment results demonstrate that the proposed method can improve location estimation accuracy by 40 percent and 18 percent in comparison to the raw dataset and other methods, respectively.