Numerical Green's function for radio frequency tomography with complex geometry

We describe the use of numerical Green's functions applied to Radio Frequency Tomography. RF Tomography relies on a linear operator in matrix form, which depends on a Green's function. The availability of the correct Green's function can be an important limitation when imaging in non-trivial scenarios. If the function cannot be found analytically, numerical methods can be used efficiently. We present simulation results that show the benefit of using a numerical Green's function by comparing it to the result that would be obtained with the analytical function for the homogeneous space.