1: \begin{abstract}

2:         Radial basis function generated finite difference (RBF-FD) methods  for PDEs require a set of interpolation points which

3:         conform to the computational domain $\Omega$. One of the requirements leading to approximation robustness is to place

4:         the interpolation points with a locally uniform distance around the boundary of $\Omega$. However generating

5:         interpolation points with such properties is a cumbersome problem.

6:         Instead, the interpolation points can be extended

7:         over the boundary and as such completely decoupled from the shape of $\Omega$.  In this paper we present a modification to

8:         the least-squares RBF-FD method which allows the interpolation points to be placed in a box that encapsulates $\Omega$.

9:         This way, the node placement over a complex domain in 2D and 3D is greatly simplified.

10:         Numerical experiments on solving an elliptic model PDE over complex 2D geometries show that our approach is robust.

11:         Furthermore it performs better in terms of the approximation error and the runtime vs. error compared with the classic RBF-FD methods.

12:         It is also possible to use our approach in 3D, which we indicate

13:         by providing convergence results of a solution over a thoracic diaphragm.

14:     \end{abstract}

15: