Spatial and frequency measurement optimization in Subsurface Imaging

This paper deals with microwave subsurface imaging for a multi-monostatic/multi-frequency configuration. The focus is on devising a suitable data sampling scheme that requires as low as possible data but preserves the achievable performance. To this end, we introduce two sampling schemes. The first one is based on analytical arguments subtended by the recently introduced warping approach. The second sampling method relies on a sensor selection procedure called maximal projection onto the minimum eigenspace. The two approaches are compared for different configuration parameters, and both show a dramatic data reduction as compared to sampling schemes commonly employed in literature