Balancing resolution improvement and artifacts in back-propagation reconstructions for a host medium having point-like inhomogeneities

In this contribution the inverse source problem for a multipath environment is addressed, with multipath being created by a grid of point scatterers appositely deployed in the scene. By adopting a weighted back-propagation imaging algorithm, we study how resolution improvement and artefacts depend on the grid layout. Successively, back-propagation is compared with a truncated singular value decomposition (TSVD) inverse filtering. It is shown that TSVD in principle allows for a larger resolution improvement and is less prone to artefacts. However, TSVD is more computationally demanding and when noise corrupts data, back-propagation definitely performs better.