Abstract:
|
A new algorithm, the Sparsified Adaptive Cross
Approximation (SPACA) is presented for fast iterative solution
of the Method of Moments linear system. Like ordinary ACA, it
is a completely kernel-independent method, but it is faster and
yields a higher compression rate than ordinary ACA, without
compromising the accuracy. As an example, the RCS of a
perfectly conducting sphere is computed using up to 786,432
basis functions. It is shown that SPACA exhibits close to NlogN
complexity for this problem. |