Resolution and Accuracy in Multiband Radar Sensors

This thesis deals with the jointly processing of radar signals stemming from
independent sensors observing the same scattering scenario at different
frequency
bands. The suitability of a simple point-scatterer model to describe
complex scattering mechanisms is verified over wide bandwidths. The minimum
error in the estimation of multiband radar signal parameters is obtained
using the Cramer-Rao theorem. The resolution problem (separation of targets)
is treated as a test of the hypotheses of one or two closely-spaced
scatterers
present in the radar signal. Accuracy and resolution limits for multiband
data sets are obtained for different distances between the frequency bands.
To process multiband radar signals, high-resolution spectral estimators are
adapted to cope with the multiband case. While the accuracy improves with
increasing band distance, no general improvement of the resolution
performance
is observed. Additionally, an approach to estimate the distance between
the frequency bands is developed. Monte Carlo analyses confirm the
fundamental behavior of the resolution and accuracy dependence on the band
gap.