Super-Resolution Algorithm for UWB Snow Depth Measurement From Sparse Narrow Bands

This paper presents a super-resolution algorithm for reconstructing full band radar returns from sparse, narrow subband radar data (representing only <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2-6 \%$</tex> of the total bandwidth). An all-pole model and Root-MUSIC are used in band parameter estimation; the proposed method integrates linear and nonlinear least-squares techniques to obtain optimized estimation and full band reconstruction. Additionally, a weighted least squares fitting approach is applied for effective subband stitching. Testing on both synthesized target data using the actual radar and delay line setup and data from snow radar survey flight, demonstrates the algorithm's capability, yielding echograms that clearly delineate snow top and bottom layers. Considering the bandwidth limitation due to FCC regulations, these results underscore the algorithm's potential to enhance UWB snow radar's resolution from airborne platforms. Moreover, the algorithm bears great potential for spaceborne low-power UWB microwave radar deployments for snow thickness measurements with fine resolution, given the small, allowed space transmission bandwidth.