Tracking Mobile Targets Using Wireless Sensor Networks

Abstract

Mobile target tracking needs a sensor network able to autonomously adapt to the requirements of the data fusion algorithms. We propose a design architecture for a tracking algorithm in which the sensed data is processed in an abstract space called Information Space and the communication between nodes is modeled as an abstract space called Network Design Space. The two abstract spaces are connected through an interaction interface called InfoNet, that seamlessly translates the messages between the two. The proposed architecture is evaluated using simulations in NS2. The sensory data is collected from a laboratory testbed where a mobile robot (a Segway) moves along various types of trajectories over a pressure sensor network.