We exploit the sparseness of induced currents on the surfaces of obscured perfectly conducting (PEC) scatterers in developing a single-frequency imaging algorithm in 2-D space. We model the induced currents on the targets by a set of unknown filament currents to which we apply the surface-equivalence theorem. We use$l_{1}$regularization to promote the solutions with a small number of equivalent sources placed in the vicinity of scatterer surfaces. The restored equivalent sources accurately determine the illuminated portions of the scatterer contours. We compare the performance of the algorithm with that of linear sampling (LSM) and MUSIC. We verify the proposed algorithm on the equivalent 3-D experimental model.

Evaluation of the time-domain response of multiconductor transmission lines is of great importance in the analysis of the crosstalk in fast digital circuit interconnections, as well as in the analysis of power lines. Several techniques for the computation of the line response, starting from the known circuit-theory parameters, are presented and evaluated. These methods are: time-stepping solution of the telegrapher equations, modal analysis in the time domain, model analysis in the frequency domain, and a convolution technique which uses line Green's functions. The last method can treat the most general case of lossy transmission lines with nonlinear terminal networks. Numerical and experimental results are presented to illustrate these techniques and to give insight into the crosstalk problems in fast digital circuits.