Ioannis P. A. Papadopoulos,Sheehan Olver
Abstract:We develop a sparse spectral method for a class of fractional differential equations, posed on $\mathbb{R}$, in one dimension. These equations can include sqrt-Laplacian, Hilbert, derivative and identity terms. The numerical method utilizes a basis consisting of weighted Chebyshev polynomials of the second kind in conjunction with their Hilbert transforms. The former functions are supported on $[-1,1]$ whereas the latter have global support. The global approximation space can contain different affine transformations of the basis, mapping $[-1,1]$ to other intervals. Remarkably, not only are the induced linear systems sparse, but the operator decouples across the different affine transformations. Hence, the solve reduces to solving $K$ independent sparse linear systems of size $\mathcal{O}(n)\times \mathcal{O}(n)$, with $\mathcal{O}(n)$ nonzero entries, where $K$ is the number of different intervals and $n$ is the highest polynomial degree contained in the sum space. This results in an $\mathcal{O}(n)$ complexity solve. Applications to fractional heat and wave equations are considered.
What problem does this paper attempt to address?
### Problems the paper attempts to solve
The paper aims to develop a sparse - spectral method for solving a class of one - dimensional fractional - order differential equations defined on the real axis $\mathbb{R}$. These equations may contain terms such as the square - root Laplace operator $(-\Delta)^{1/2}$, the Hilbert transform $H$, the derivative $\frac{d}{dx}$, and the identity operator $I$. Specifically, the paper focuses on equations in the following form:
\[ L_{\lambda, \mu, \eta}[u] := (\lambda I + \mu H + \eta \frac{d}{dx}+(-\Delta)^{1/2})[u]=f, \]
where $\lambda, \mu, \eta\in\mathbb{R}$ are known constants, $H$ is the Hilbert transform, $(-\Delta)^{1/2}$ is the square - root Laplace operator, and the domain is the entire real axis $\mathbb{R}$. The goal of the paper is to find a solution $u$ that satisfies the following condition:
\[ \lim_{|x|\to\infty}u(x) = 0. \]
### Main contributions
1. **Sparse - spectral method**: The paper proposes a sparse - spectral method based on weighted Chebyshev polynomials and their Hilbert transforms. This method not only makes the matrix of the linear system sparse, but also decouples the operators in subspaces under different affine transformations, thereby greatly improving the solution efficiency.
2. **Efficient solution**: By decomposing the solution process into multiple independent sparse linear systems, each of size $O(n)\times O(n)$ with the number of non - zero elements being $O(n)$, an $O(n)$ - complexity solution is finally achieved.
3. **Scope of application**: This method is applicable not only to one - dimensional problems, but can also be extended to two - dimensional and three - dimensional problems, and can handle different powers of the fractional - order Laplace operator.
4. **Numerical stability**: The paper proves that the condition number of the linear system can be controlled within the $O(1)$ range by a diagonal pre - conditioner, independent of the truncation degree and interval selection, thus ensuring numerical stability.
5. **Theoretical analysis**: The paper analyzes in detail the relationships among the square - root Laplace operator, the Hilbert transform, and the derivative, and proves the properties of related operators in specific function spaces.
### Application examples
The paper considers the applications of the fractional - order heat equation and the fractional - order wave equation, demonstrating the effectiveness and efficiency of this method in practical problems.
### Conclusion
The paper proposes an efficient sparse - spectral method for solving a class of fractional - order differential equations containing the square - root Laplace operator and the Hilbert transform. This method not only has high solution efficiency, but also has good numerical stability and wide applicability.