Low-complexity 3D target tracking in Wireless Aerial Sensor Networks

Abstract

With the proliferation of the application areas of the Wireless Sensor Networks (WSNs) and the technological advances in sensor electronics, target tracking has become one of the most important functions in WSNs. The overwhelming majority of moving target tracking studies in WSNs has assumed either one-dimensional space (lineland), or two-dimensional domain (flat-land). However, many existing, such as environmental monitoring of rugged terrain, structural health monitoring of buildings, underground and underwater WSN applications, and emerging applications, such as search-and-rescue via Micro Aerial Vehicles, reconnaissance by insect-sized flying robots, need target tracking in three-dimensional space. In this study, we assume a Wireless Aerial Sensor Network (WASN) with conventional low-power and low-complexity features and devise a low spatial and temporal complexity moving target tracking mechanism in 3D based on g-h filter that considers measurement errors. Our prefatory simulations show promising preliminary results for our approach in terms of very low inaccuracy rates between the true location of the target and the forecasts at the sink node as well as negligible estimate errors in terms of Euclidean distance.