Target Tracking with Cooperative Networked Robots

Abstract

The problem of mobile target tracking has received significant attention in the last few years. Many algorithms have been proposed for different variations and scenarios of this problem. In this paper, we consider a scenario where multiple mobile sensors (robots) need to track a single target by coordinating with each other, however, robots are restricted to remain in their mutually exclusive bounded region. We propose two coordination approaches, one having low tracking accuracy with low communication cost, another with high tracking accuracy with high communication cost. Implicit communication is used to achieve coordination by sharing target locations along with the current state of the Kalman filter. The simulation model includes motion planning and sensor noise in detecting the target. Three parameters, tracking accuracy, communication cost and degree of detection (number of times target goes detected) are evaluated to show the performance of both algorithms.