Autonomous Motion Compensation for UWB SAR Data

An innovative data-derived motion compensation technique is proposed to provide the accurate positional information needed for high quality near-field UWB SAR image formation. The low-latency, computationally efficient technique does not require any special targets or a prior knowledge of the scene developing. Assuming nominal values for aircraft velocity and altitude, the algorithm derives accurate estimates of the deviations from the nominal trajectory in three dimensions based on processing received radar data from a general scene. Therefore, the algorithm estimates the (x, y, z) position of the radar antenna on each pulse and hence enables high-resolution image formation. Position estimates in the horizontal cross-track and vertical directions are derived based on the spatial correlation existing across the synthetic aperture measured over a number of disjoint range subintervals. Velocity estimates are also obtained from the data. Phase I will consist of algorithm development and performance assessment as a function of latency, swath length of data, antenna beamwidth, average aircraft velocity, dynamics of the aircraft and signal-to-clutter/noise ratio. Phase II will consist of coding and demonstrating an optimized version of the algorithm on a real-time image formation processor assuming 1.5 Khz PRF, 50-200m/s aircraft velocities, and 256K point records. BENEFITS: The proposed data-derived measurement technique will provide accurate positional information to enable high quality foliage/ground penetration radar image formation. In addition to concealed target detection, commercial applications include such services as bald earth mapping for road construction planning, forest characterization, finding near-surface mineral deposits, locating downed aircraft, and humanitarian unexploded ordnance cleanup.