Resultat från bistatiskt SAR experiment på VHFbandet 2007

Authors:

  • Lars M H Ulander
  • Björn Flood
  • Per-Olov Frölind
  • Tommy Jonsson
  • Anders Gustavsson
  • Gunnar Stenström

Publish date: 2008-01-07

Report number: FOI-R--2389--SE

Pages: 41

Written in: Swedish

Keywords:

  • Bistatic
  • SAR
  • VHF-band
  • clutter suppression
  • forest
  • urban environment

Abstract

The report describes the first results from a bistatic SAR experiment in the VHF band (28-73 MHz) which was conducted in September 2007. The objective was to investigate the hypothesis that the signal-to-clutter ratio, for a vehicle in a background of forested terrain or urban environment, increases compared to the quasi-monostatic case as the bistatic elevation angle increases. The hypothesis is based on the assumption that the radar signal is dominated by a corner reflection between horizontal ground and a vertical smooth structure (e.g. ground-tree stem or ground-wall), which is large compared to the wavelength. The radar cross section is hence large in the backscattering direction but decreases significantly when the bistatic elevation angle changes. The radar cross section for a vehicle, on the other hand, remains relatively unchanged as the bistatic elevation angle changes since its vertical extent is small compared to the wavelength. The experiment was carried out in Switzerland by FOI in collaboration with armasuisse and University of Zürich. The CARABAS-II system acted as airborne transmitter and monostatic receiver, whereas the LORA system was configured in VHF receive-only mode and acted as ground-based bistatic receiver on the mountain Niesen in Berner Oberland. Measurements were made for six geometries, of which three were linear tracks with bistatic elevation angles 4°, 10° and 20°, and the remaining three were circular tracks centred on the aim point. Several problems were encountered during the campaign, of which the most serious was caused by interference from TV broadcasting which saturated two of the 10-MHz bands in the LORA receiver. The collected bistatic radar data therefore have a reduced sensitivity to weak ground returns. Data from the linear tracks have been processed and analysed. Both mono- and bistatic SAR images have been generated. The monostatic CARABAS-II images for three different incidence angles (51°, 61°, and 67°) show the forested and urban areas with similar character. The bistatic SAR images, on the other hand, show that the areas change character with the bistatic elevation angle. The image with a bistatic elevation angle of 4° (quasi-monostatic case) is similar to the monostatic images, whereas the others (bistatic elevation angle 10° and 20°) indicate a decreasing clutter level. Measurements in the latter images show that the signal-to-clutter ratio for a truck in a forest background increases by up to 10 dB compared with the quasi-monostatic case. This value would most likely be even larger with improved sensitivity for weak ground returns. The corresponding increase in signal-to-clutter ratio for a truck vehicle in an urban background is 2-8 dB, with the higher value obtained for multi-level buildings. We conclude that the signal-to-clutter ratio, and hence the detection performance, for a vehicle in a forest or urban environment may increase significantly in VHF-band SAR imagery by choosing a suitable bistatic geometry.