Ad hoc networks for localization and control

Abstract

We consider a team of mobile robots equipped with sensors and wireless network cards and the task of navigating to a desired location in a formation. We develop a set of algorithms for (a) discovery; (b) cooperative localization; and (c) cooperative control. Discovery involves the use of sensory information to organize the robots into a mobile network allowing each robot to establish its neighbors and, when necessary, one or more leaders. Cooperative control is the task of achieving a desired goal position and orientation and desired formation shape and maintaining it. Cooperative localization allows each robot to estimate its relative position and orientation with respect to its neighbors and hence the formation shape. We show numerical results and simulations for a team of nonholonomic, wheeled robots with omnidirectional cameras sharing a wireless communication network.