A Range-Only Acoustic Localization Technique for Marine Robotics Applications

We present a parametric technique for simultaneous outlier rejection and position estimation from a set of range measurements. This approach utilizes an adapted version of the Random Sample Consensus (RANSAC) algorithm, which has found extensive use in image matching applications. Our implementation has been demonstrated to function in real time on an Autonomous Underwater Vehicle (AUV) for the purposes of localizing a stationary acoustic beacon and aiding navigation.