Passive Multistatic Radar Imaging with Prior Information

Our work extends the framework of Generalized Wirtinger Flow (GWF) [19] to incorporate prior information regarding the signal of interest (which is scene reflectively in our case). For interferometric multi-static radar measurements, we consider a linear-deterministic-discrete lifted forward model. Several structures on the signal can prove beneficial in the recovery process. In particular, we leverage the sparse structure of the unknown signal and show through numerical simulations that sparsely constrained GWF converges at a faster rate than a recovery method that does not acknowledge sparse structure.