Using Wireless Sensors to Build Mobility Models for the Water Flow Environment

Abstract

Wireless sensor networks (WSNs) have been studied extensively in recent years. The tiny and inexpensive wireless sensors can be deployed close to the target objects or phenomena to provide in-situ and direct measurements and to transmit the collected data to the back-end sinks in real-time through wireless connections. In this thesis, we take advantages of this characteristic of WSNs and apply it to study the mobility models of objects in a water flow. We first modify off-the-shelf wireless sensor nodes to include GPS modules. The sensors are then put on a river to collect real trajectory data when the node drifts in the river. The raw data are used to derive the parameters for representative mobility models under study. Finally, the synthetic trajectories out of these mobility models are compared with real data to evaluate their ability of modeling moving objects in water flows. Our evaluations show that existing mobility models cannot model mobile objects in water flows very well, primarily due to a lack of topographic modeling. We then propose new mobility models that take into account of topography as well as obstacles in the water flow. The new models are evaluated again using the collected real data to show their ability to model moving objects in the water flows.