Localization Process for WSNs with Various Grid-Based Topology Using Artificial Neural Network

Abstract

Wireless Sensor Network (WSN) is a technology that can aid human life by providing ubiquitous communication, sensing, and computing capabilities. It allows people to be more able to interact with the environment. The environment contains many nodes to monitor and collect data. Localizing nodes distributed in different locations covering different regions is a challenge in WSN. Localization of accurate and low-cost sensors is an urgent need to deploy WSN in various applications. In this paper, we propose an artificial automatic neural network method for sensor node localization. The proposed method in WSN is implemented with network-based topology in different regions. To demonstrate the accuracy of the proposed method, we compared the estimated locations of the proposed feedforward neural network (FFNN) with the estimated locations of the deep feedforward neural network (DFF) and the weighted centroid localization (WCL) algorithm based on the strength of the received signal index. The proposed FFNN model outperformed alternative methods in terms of its lower average localization error which is 0.056m. Furthermore, it demonstrated its capability to predict sensor locations in wireless sensor networks (WSNs) across various grid-based topologies.