Improving Navigation Through Cooperation and Path Planning

Abstract

The ability to reliably estimate own-states is very important for Unmanned Aerial Vehicles (UAVs) in executing their missions. Most current approaches for UAV state estimation rely on fusing inertial information from accelerometers and gyroscopes with absolute position information from a position sensor. Global Positioning System (GPS) is one of the most widely used position sensors. However, GPS signals are not reliable, and can be jammed by adversarial forces. Without the aid of an absolute position reference such as GPS the navigation solution of the system is going to drift with time. The problem of two autonomous vehicles traveling in a two dimensional environment from an initial location to a known goal location without any absolute position reference is considered. The effect of cooperation between the vehicles by considering the measurements such as relative range to help in improving the navigation state estimation and its effect on the observability of the system is discussed. The reduction in the navigation solution drift of the system, with cooperation between the agents, using measured relative information and its effect on the observability of the system while taking different paths is discussed. Simulations and theoretical results show that relative motion between the agents helps reduce the navigation drift of the agents when there is no absolute position reference.