Indoor Navigation for Autonomous Flight

Navigation for the autonomous flight of Unmanned Aerial Vehicles (UAVs) in an indoor space has attracted much attention recently. One of the main goals of an indoor navigation system is developing an alternative method to obtain position information that can replace or complement the global positioning system. While much research has focused on vision-based indoor navigation systems, this paper aims to develop a Received Signal Strength (RSS)-based Navigation System (RNS), which is a more cost e ective alternative. Then, the position and attitude of a UAV can be computed by the fusion of RSS measurements and measurements from the onboard inertial measurement unit. In order to improve the estimation, we rst consider a mathematical model of the RNS and formulate optimization problems to compute the parameter values which minimize the RSS measurement error. Using the optimal parameters, an autonomous fl ight system is developed whose estimator and controller components are designed to work well with the RNS. Simulations and experiments using a quadrotor demonstrate the feasibility and performance of the proposed RNS for UAVs operating in indoor environments.