Multiple-input multiple-output (MIMO) radar systems have been shown to have significant performance improvements
over their single-input multiple-output (SIMO) counterparts. For transmit and receive elements
that are collocated, the waveform diversity afforded by this radar is exploited for performance improvements.
These improvements include but are not limited to improved target detection, improved parameter identifiability and better resolvability. In this paper, we present the Synchronous Impulse Reconstruction Radar (SIRE)
Ultra-wideband (UWB) radar designed by the Army Research Lab (ARL) for landmine and improvised explosive
device (IED) detection as a 2 by 16 MIMO radar (with collocated antennas). Its improvement over its SIMO
counterpart in terms of beampattern/cross range resolution are discussed and demonstrated using simulated
data herein. The limitations of this radar for Radio Frequency Interference (RFI) suppression are also discussed
in this paper. A relaxation method (RELAX) combined with averaging of multiple realizations of the measured
data is presented for RFI suppression; results show no noticeable target signature distortion after suppression.
In this paper, the back-projection (delay and sum) data independent method is used for generating SAR images.
A side-lobe minimization technique called recursive side-lobe minimization (RSM) is also discussed for reducing
side-lobes in this data independent approach. We introduce a data-dependent sparsity based spectral estimation
technique called Sparse Learning via Iterative Minimization (SLIM) as well as a data-dependent CLEAN
approach for generating SAR images for the SIRE radar. These data-adaptive techniques show improvement in
side-lobe reduction and resolution for simulated data for the SIRE radar.