Time- and frequency-domain MIMO FLGPR imaging

Multiple-input multiple-output (MIMO) forward looking ground penetrating radar (FLGPR) systems have shown the capability to improve performance for landmine detection. By leveraging multiple collocated transmitters and receivers, the waveform diversity offered by MIMO FLGPR systems provides performance enhancement in target detection, parameter identifiability, and increased cross-range resolution. In this paper, we introduce a data-dependent image formation algorithm based on the sparse learning via iterative minimization (SLIM) approach for high-resolution imaging applied to both time- and frequency-domain. In addition, we compare time-domain SLIM (TD-SLIM) and frequency-domain SLIM (FD-SLIM) with data-independent methods such as delay-and-sum (DAS) and recursive sidelobe minimization (RSM), designed by the Army Research Laboratory (ARL). These algorithms are applied to data collected by the Synchronous Impulse Reconstruction (SIRE) Ultra-wideband (UWB) radar developed by ARL for landmine detection. When applied to data collected by this 2 by 16 MIMO radar system, both the TD-SLIM and FD-SLIM algorithms yield results with higher resolution than the two conventional methods applied to the data. FD-SLIM provides significantly reduced computational time compared to TD-SLIM, motivating its application for high-resolution MIMO FLGPR imaging.