UAV Waypoint Opportunistic Navigation in GNSS-Denied Environments

Navigation of an unmanned aerial vehicle (UAV) to reach a desired waypoint in global navigation satellite system (GNSS)-denied environments is considered. The UAV is assumed to have an unknown initial state and the environment is assumed to possess multiple terrestrial signals of opportunity (SOPs) transmitters with unknown states and one anchor SOP whose states are known. The UAV makes pseudorange measurements to all SOPs to estimate its own states simultaneously with the states of the unknown SOPs. The problem is formulated as a multi-objective motion planning (MOMP) strategy, which guarantees the UAV gets to within a user-specified distance of the waypoint with a user-specified confidence. The MOMP strategy balances two objectives: (i) navigating to the waypoint and (ii) reducing UAVs position estimate uncertainty. It is demonstrated that in such an environment, formulating the waypoint navigation problem in a so-called naive fashion by heading directly to the waypoint results in failing to reach the waypoint due to poor estimability of the environment. In contrast, the MOMP strategy guarantees (in a probabilistic sense) reaching the waypoint. Simulation and experimental results are presented showing that the MOMP strategy achieves the desired objective while the naive strategy fails to do so.