Terrain-aided navigation using the Viterbi algorithm

Without some form of correction, an inertial navigation system’s (INS’s) positional error grows without bound over time. Several techniques have been used to reduce the INS’s positional error. One of these techniques, terrainaided navigation (TAN), has been widely researched over the last several decades with the research focusing on two methods: Sandia inertial terrain-aided navigation (SITAN), which is based on an extended Kalman filter (EKF), and terrain contour matching, which is based on correlation techniques. These methods are applied most successfully to fixed-wing aircraft flying over specific types of terrain; they perform poorly for low-velocity, highly maneuverable aircraft such as helicopters, especially when flying over flat or very rough terrain. In this paper, we introduce VATAN, a new TAN method based on the Viterbi algorithm (VA), and compare its performance to an implementation of SITAN based on a single EKF. Simulation results show that the VA in VATAN overcomes divergence problems associated with the EKF in SITAN and provides position estimates with smaller averagesquared error. These results show that VATAN has the potential to provide good performance for low-velocity aircraft flying over all types of terrain.