Cell Tower Localization Approach for Vehicles Tracking

Abstract

Tracking systems are common tools for managing transportation, monitoring vehicles, and securing vehicles. However, current tracking systems depend on GPS technology, which necessitates hardware and communication with satellites, making power consumption higher and more costly. Besides, the functioning of GPS technology may be affected in regions with reduced visibility of satellites. For this reason, new localization systems have received much attention. The aim of this study is to design an IoT vehicle tracking system by using cell tower localization in lieu of the existing GPS-based tracking systems. The vehicle tracking system has been designed using the Raspberry Pi 3 model B, 4G LTE modem, Google Geolocation API, and Google Maps services. The cell tower IDs have been captured by using the LTE modem and transformed into geographical positions to track the vehicles in real time. For testing the performance, both the proposed approach and the GPS localization technique have been applied on the same path with ten predetermined points. Coordinates obtained in both localization methods were highly consistent within the test locations. An average positioning error was measured in meters by the Haversine distance between cell tower localization coordinates and GPS coordinates. As a result, an average positioning error of 1.48 m was calculated, with the minimal value being equal to 0.00 m and the maximum value being 4.98 m. The system successfully implemented vehicle monitoring and visualization of the vehicle location via a web interface. Experimental results show that the use of cell tower localization is a promising low-cost option for vehicle tracking in Internet of Things systems. While GPS localization is still the most popular method for positioning, it is evident that the suggested method produces equally accurate positioning results. In future studies, the use of hybrid localization methods will be examined.