I would like to know which numerical technique is best for the analysis of patch antenna. Method of moments, finite element method, finite difference method, FDTD etc. are some of the most commonly used numerical technique in EM, now particularly for microstrip patch antenna which numerical technique will be the best and why ?
Thank You
Dhanu Chettri
@Dhanu
A normal microstrip patch antenna can be solved by any of the above methods, and since the structure is a simplistic one, you cannot leverage the benefits of one over another to a great extent. Each of the above methods that you pointed out have their own sweet-spots and limitations, depending on their applications.
Typically Method of Moments (MoM) is well-suited for radiation and scattering problems, antennas with conductor only geometries, and also for electrically large scatterers and radiators, where there are several fast algorithms to deal with those problems. This is because MoM only discretizes (meshes) the source, and not the solution domain and hence doesn't need a big radiation box in the solution domain to satisfy the Sommerfield Radiation condition at far field. Therefore it solves the problem in less number of mesh elements. Limitations: The matrix of the system is dense and hence provides computational challenge, unless fast algorithms are used.
FEM on the other hand is extremely well suited for dealing with dielectrics, especially when there is inhomogeneity. This is because FEM has volume based tetrahedral meshes, unlike traditional MoM which uses triangular surface-based meshes, and hence can capture the dielectric variations. Although, MoM has well defined surface and volume-based dielectric formulations in place, yet FEM wins over owing to sparse matrix in its formulation. Limitations- not suited for electrically large geometries as the number of unknowns (mesh elements) will be huge as both source and solution domain including the radiation box is discretized.
The two methods MoM and FEM are frequency-based methods, and hence are more suited when number of frequency points are less, or in other words narrow-band applications. For wideband and UWB antennas and systems, FDTD is an ideal choice as the problem translates to fewer time samples in time domain and hence FDTD wins over others.
i enclose for you a very useful paper that compares between the different Antenna solvers.
Good luck.
Prof. S. EL-Rabaie
- Badges
- Science method