Abstract:In this paper we are concerned with the plane wave method for the discretization of time-harmonic Maxwell's equations in three dimensions. As pointed out in [6], it is difficult to derive a satisfactory L2 error estimate of the standard plane wave approximation of the time-harmonic Maxwell's equations. We propose a variant of the plane wave least squares (PWLS) method and show that the new plane wave approximations possess the desired L2 error estimate. Moreover, the numerical results indicate that the new approximations have sightly smaller L2 errors than the standard plane wave approximations. More importantly, the results are derived for more general models in layered media.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: For the three - dimensional time - harmonic Maxwell equations, it is difficult to obtain a satisfactory \(L^2\) error estimate using the standard plane - wave approximation method. Specifically:
- **Problem Background**: In the discretization process of the plane - wave method for the time - harmonic Maxwell equations, although it can provide high precision, it is theoretically difficult to derive a satisfactory \(L^2\) error estimate. This problem is particularly prominent in layered media because the material coefficients are no longer constant.
- **Specific Problem**: Although the existing plane - wave least - squares method (PWLS) performs well in some cases, it still lacks an effective error estimate for more general models (such as layered media).
To solve the above problems, the author proposes an improved variant of the plane - wave least - squares method (PWLS) and proves that this method can obtain the required \(L^2\) error estimate. In addition, numerical results show that the approximate solution generated by the new method is more accurate than the existing PWLS method.
### Formula Summary
1. **First - order form of Maxwell equations**:
\[
\begin{cases}
\nabla\times\mathbf{E}-i\omega\mu\mathbf{H} = 0, \\
\nabla\times\mathbf{H}+i\omega\varepsilon\mathbf{E} = 0, \\
\nabla\cdot(\varepsilon\mathbf{E}) = 0, \\
\nabla\cdot(\mu\mathbf{H}) = 0,
\end{cases}
\]
where \(\omega>0\) represents the time frequency, and \(\varepsilon\) and \(\mu\) represent the permittivity and permeability respectively.
2. **Boundary Conditions**:
\[
-\mathbf{E}\times\mathbf{n}+\sigma(\mathbf{H}\times\mathbf{n})\times\mathbf{n}=g\quad\text{on}\quad\gamma = \partial\Omega.
\]
3. **New Objective Functional**:
\[
J(F)=\sum_{k = 1}^N\delta\int_{\gamma_k}\left|-\mathbf{F}_k\times\mathbf{n}_k+\Phi(\mathbf{F}_k)\times\mathbf{n}_k - g\right|^2ds+\sum_{l < j}\left(\alpha\int_{\Gamma_{lj}}\left|[\![\mathbf{F}\times\mathbf{n}]\!]\right|^2ds+\beta\int_{\Gamma_{lj}}\left|[\![\Psi(\mathbf{F})\times\mathbf{n}]\!]\right|^2ds+\theta\int_{\Gamma_{lj}}\left|[\![\varepsilon\mathbf{F}\cdot\mathbf{n}]\!]\right|^2ds\right)
\]
where \(\delta,\alpha,\beta,\theta\) are positive parameters, \(\Phi(\mathbf{F}_k)=\frac{\sigma}{i\omega\mu}((\nabla\times\mathbf{F}_k)\times\mathbf{n}_k)\), \(\Psi(\ma