Geometry-Induced Bias Reduction in Synthetic Aperture Sonar Along-Track Micronavigation

Focusing synthetic aperture sonar (SAS) images requires estimation of the sensor motion to sub-wavelength precision. This can be achieved by exploiting the coherence between signals received by redundant phase centers of different pings, using a method termed micronavigation. In the along-track direction, this involves interpolating sparse samples of the spatial coherence function to estimate the location of the peak coherence, and interpreting this peak location estimate to find the along-track inter-ping motion, in a process similar to that used in correlation velocity logs. However, these along-track micronavigation estimates are affected by several sources of bias, which reduces their usefulness as a navigation aid for uncrewed underwater vehicles (UUVs). This article considers the along-track micronavigation bias observed in the presence of nonideal geometries, specifically when relative pitch exists between the array and the seafloor and the inter-ping motion is not colinear with the array axis, which is especially relevant for repeat-pass operations, such as coherent change detection and repeat-pass bathymetry estimation. A model is introduced for the location of the peak coherence as a function of the inter-ping motion and the along-track seafloor slope relative to the array axis. A surge estimation algorithm for interferometric SAS systems is detailed, and the resulting bias reduction and low residual model error are demonstrated using both simulated SAS data and using experimental SAS data collected by the CMRE MUSCLE UUV. These bias reductions will improve the utility of along-track micronavigation estimates for UUV navigation and will improve repeat-pass micronavigation accuracy, potentially enabling high resolution 3-D imaging of the underwater environment.