A Wireless Sensor Network Based on Unmanned Boundary Sensing Technique for Minimizing Human Elephant Conflicts

Unmanned boundary sensing technique targets to preserve and protect both the elephant breed and the human inside their protected areas without crossing the boundary. This paper proposes a novel integrated wireless sensor network architecture for elephant tracking. For this purpose, the boundary sense utilizes the most advanced sensors including vibration sensors, acoustic sensors and camera sensors, as these three sensors mutually balance their demerits leading to a promising solution. The architecture to deploy and activate boundary sensing is presented in this paper. The system offers three variations of detections using vibration, sound and image, and assesses them on a massive dataset of real time recordings. Moreover, the system integrates these sensors and fulfills the requirements, which is not adequately addressed in the existing literature. Based on this context, the detector is further enhanced based on the distance metric, SHORT, MEDIUM and LONG, and determined through a fuzzy inference system. Finally, the overall performance evaluation of the proposed system achieves 97% detection rate and 0% false alert for elephant detection.