Browsing by Subject "Green's functions"

Now showing 1 - 3 of 3

Open Access
Efficient computation of nonparaxial surface fields excited on an electrically large circular cylinder with an impedance boundary condition
(Institute of Electrical and Electronics Engineers, 2006) Alisan, B.; Ertürk, V. B.; Altintas, A.
An alternative numerical approach is presented for the evaluation of the Fock-type integrals that exist in the uniform geometrical theory of diffraction (UTD)-based asymptotic solution for the nonparaxial surface fields excited by a magnetic or an electric source located on the surface of an electrically large circular cylinder with an impedance boundary condition (IBC). This alternative approach is based on performing numerical integration of the Fock-type integrals on a deformed path on which the integrands are nonoscillatory and rapidly decaying. Comparison of this approach with the previously developed one presented in [1], which is based on invoking the Cauchy's residue theorem by finding the pole singularities numerically, reveals that the alternative approach is considerably more efficient.
Open Access
Efficient methods for electromagnetic characterization of 2-D geometries in stratified media
(1997) Çalışkan, Fatma
Numerically efficient method of moments (MoM) algorithms are developed for and applied to 2-D geometries in multilayer media. These are, namely, the spatial-domain MoM in conjunction with the closed-from Green's functions, the spectral-domain MoM using the generalized pencil of functions (GPOF) algorithm and a FFT algorithm to evaluate the MoM matrix entries. These approaches are mainly to improve the computational efficiency of the evaluation of the MoM matrix entries. Among these, the spectral-domain MoM using the GPOF algorithm is the most efficient approach for printed multilayer geometries. The assessment of the efficiency of this method is performed on several problems, by comparing the matrix fill times for these three approaches. In addition a new iterative algorithm is developed to solve the MoM matrix equation, which is based on dividing a large object into subregions and solving the matrix equation on each subregion by considering the effects of other regions. This iterative algorithm is applied to some large geometries and is compared to a traditional LU decomposition algorithm for the assessment of its numerical efficiency. It is observed that the iterative algorithm is numerically more efficient as compared to the LU decomposition.
Open Access
A novel approach for the efficient computation of 1-D and 2-D summations
(Institute of Electrical and Electronics Engineers Inc., 2016) Karabulut, E. P.; Ertürk, V. B.; Alatan, L.; Karan, S.; Alisan, B.; Aksun, M. I.
A novel computational method is proposed to evaluate 1-D and 2-D summations and integrals which are relatively difficult to compute numerically. The method is based on applying a subspace algorithm to the samples of partial sums and approximating them in terms of complex exponentials. For a convergent summation, the residue of the exponential term with zero complex pole of this approximation corresponds to the result of the summation. Since the procedure requires the evaluation of relatively small number of terms, the computation time for the evaluation of the summation is reduced significantly. In addition, by using the proposed method, very accurate and convergent results are obtained for the summations which are not even absolutely convergent. The efficiency and accuracy of the method are verified by evaluating some challenging 1-D and 2-D summations and integrals. © 2016 IEEE.