Optimizing iterative physical optics by using an ACA compression on interaction matrices

This paper presents a numerical compression of Iterative Physical Optics (IPO), known as an efficient asymptotic method for computation of Radar Cross Section (RCS) of complex targets including cavities. The compression is applied to interaction matrices of IPO between coupled surfaces. It is made by using the algebraic Adaptive Cross Approximation (ACA), which is known to provide a good compression and acceleration when used with dense matrices involved in rigorous numerical methods. Results show that this method highly compresses the IPO matrices, without damaging the RCS prediction by IPO.