Subspace-based optimization method for reconstructing extended scatterers: transverse electric case.

The subspace-based optimization method is generalized to the transverse electric (TE) case for reconstructing the relative-permittivity profiles of extended scatterers. The essence of the subspace-based optimization method is that part of the contrast source is determined from the spectrum analysis without recourse to optimization, whereas the rest is determined by using an optimization procedure. We show that the dimension of the signal subspace behaves somewhat as a regularization parameter. Due to the vectorial property of the TE wave, the reconstructed profile in TE case is smoother than that in the transverse magnetic (TM) case.