Improving the accuracy of equivalence principle algorithm with mixed discretization scheme

For the poor precision problem of identity operator using conventional numerical discretization in equivalent principle algorithm (EPA), a mixed discretization scheme is proposed to improve the accuracy in this paper. In the mixed discretization scheme, the electric currents on equivalence surfaces are expanded by Rao-Wilton-Glisson (RWG) basis functions and magnetic currents are expanded by Buffa-Christiansen (BC) basis functions. Then the testing functions are \boldsymbol\hat n×BC and \boldsymbol\hat n×RWG respectively, which means that identity operators are tested on the dual of their range. The multilevel fast multipole algorithm (MLFMA) is also used to accelerate the matrix-vector multiplication in EPA. Finally, numerical results are shown to demonstrate the accuracy and efficiency of the proposed technique.