Emulation of Wireless Sensor Networks for Object Tracking

Wireless sensor networks (WSN) represent a new technology that could be adopted to achieve ubiquitous computing and embedded Internet. One example of WSN applications is surveillance and monitoring of a deployment area. The goal of our project is the development of effective WSN for detecting and tracking human and/or vehicle presence. In this paper, our focus is two-folded. We first examine the issues of tracking objects using wireless sensor networks, including a brief introduction to object tracking, a survey of work related to object tracking using WSN, and the challenges of designing an object tracking sensor network using the emulation approach (in contrast to the simulation approach). The second focus of this paper is the design of a sample wireless sensor network application capable of detecting human presence. In the sample application, motes and sensor nodes capture data from the surrounding environment and send the data to the base station. A Java application running at the base station analyzes the data to determine whether a human is present. In addition to discussing our experience of developing the project, we examine the challenges of emulating a wireless sensor network, difficulties encountered during the project, and the algorithm implemented in the application. We also propose the design of a multi-node WSN for detecting/tracking humans. Projects are currently underway to implement and evaluate the design.