Efficient Macromodel for Interconnects Excited by Incident Fields

The analysis of interconnects in the presence of incident electromagnetic fields is an important part of the design process. Such an analysis is, however, very CPU expensive due to the large size of interconnects networks. Krylov based reduction methods were proposed in order to address the complexity of such simulations, however, they result in macromodels which are not optimal and still contain many redundant poles. In this paper a two level reduction method is proposed for interconnect affected by incident electromagnetic fields. The macro-models obtained using the proposed approach is less than half the size of those generated by previous techniques. The first level of reduction is a conventional Krylov subspace based reduction, while the second level is based on singular value decomposition. Furthermore, the proposed method is a projection method based on congruence transformation and is therefore passive by construction. Numerical examples are shown in order to illustrate the accuracy and efficiency of the proposed technique