Velocity estimation from a forward-looking bottom profiling sonar

Velocity estimation plays an important role in many underwater applications. This paper discusses different approaches towards velocity estimation from a forward-looking sonar system intended to profile the terrain ahead. Beginning with a signal model, the covariance functions for a range-spread target are developed and used to describe two common velocity estimation methods: the ping pair incoherent method, and the single ping covariance method. Simulated and experimental data are used to evaluate the performance of these estimators, which are seen to perform in a way that is consistent with theoretical expectations. A new velocity estimation method is introduced, which generates velocity estimates by processing the image that is formed by the forward-looking sonar system. The forward-scan image, which plots the intensity of the received signal versus range for all the pings in the data set, shows “intensity patterns” which change in range when the sonar system moves in the direction of the targets that create these signals. By estimating the change in range over the change in time of these intensity patterns, the velocity of the sonar platform is estimated.