Abstract:

Target tracking is the most typical and challenging application of wireless sensor networks (WSN). In order to prove and test algorithm performance and reduce the existing gap between the established theoretical results and practical problems, the real test bed should be built. The WSNs target tracking platform contains the hardware and software. The hardware consists of ultrasonic sensors, passive infrared (PIR) and Micaz nodes, while the software includes the graphical user interface (GUI) and node embedded program. Aiming at the limited node resources and working in a dynamic environment, a target tracking collaborative scheduling algorithm based on minimum mean square estimation error is proposed and successfully realized. The experiments show that moving targets can be efficiently tracked within this testbed. With the purpose of saving energy consumption and enhancing real time tracking performance, a double wake up/sleep mechanism is adopted to extend network lifetime.