Computing radiation and scattering patterns using model-based parameter estimation

Model-based parameter estimation (MBPE) is a procedure by which data obtained from some experimental or analytical source such as Maxwell´s equations, which constitutes a generating model (GM), is approximated by a reduced order, simpler description called a fitting model (FM). Another EM observable that is suited to exponetial-series FMs is a radiation or scattering pattern since the far-field is given by an integration using a Green´s function. For radiation problems, obtaining the far-field pattern is not normally the most time-consuming step in the computation, unless, as in physical optics, where it comprises the only step. For scattering problems,the situation can be much different since the induced current an the object must be much be obtained for each angle of incidence. For these latter problems, it can be advantageous to minimize the number of samples that are needed to determine the pattern as a way of significantly reducing the overall computation time. This is the goal of an MBPE approach called WASPE (windowed adaptive sampling pattern estimation), which is described.