부산대학교 도서관

A key supporting technology for future naval capabilities (FNCs) in long-range search and survey will be synthetic aperture sonar (SAS). The few existing, R&D, SAS systems tend to have relatively narrow bandwidth and beamwidth. In order to attain FNC goals of high resolution at long ranges, system designs will be driven toward both larger bandwidth and wider beam-width. In this study we emulate broadband and wide-beam data in order to analyze their implications for SAS processing. We specifically break the processing into the three steps: motion estimation, motion compensation, and image formation. The modeled data include beam patterns that are wide and frequency dependent as expected for widebeam/broad-band systems. For such systems, the interplay between beam-width and bandwidth will cause the frequency content of the received waveforms to depend on both the range and the bearing of the target. This complicates motion estimates that are based on correlating signals from sequential pings. Because data are linear combinations of signals from potentially very different bearings, the motion compensation step is more complicated as well. Both the motion estimation and motion compensation steps are more naturally addressed after the data are transformed to range-bearing space.