Experimental Investigation on Weather Changes Influences on Wireless Localization System

Wireless Sensor Networks (WSNs) are widely explored because of their low cost, increasing capability of nodes, energy efficiency, accuracy and real time. A major issue is localization because it is based on the use of a number of sensor nodes deployed at unknown positions. Moreover, the need for more accurate and reliable localization system is increasing especially for certain applications, such as object tracking, surveillance, and disasters prediction. The reliability of the localization process should be investigated and external factors need to be considered in order to increase the accuracy of the localization. In this work, a localization system based on ultra-wide band technology is presented. The ranging system employs the two-way ranging method, which is based on the time of Arrival (ToA) technique. The DecaWave ranging system is, therefore, chosen for its high accuracy, which is about ±10 cm. To evaluate the proposed localization system, outdoor experiments were carried out, where the weather changes are considered. In this paper, the influences of weather changes on distance measurement are highlighted and a polynomial regression model for distance measurement prediction is provided with R-squared value of 78%. The regression model is designed to characterize the distance measurement variation in relevance to weather changes to enhance the localization system accuracy.