Finite difference time domain tests of random media propagation theory

Abstract

The recent extension of the finite difference time domain (FDTD) method to frequency‐dependent media [Nickisch and Franke, 1992] allows FDTD to now be applied to ionospheric propagation. The FDTD method solves the Maxwell equations directly in the time domain by temporal integration. No approximations beyond that of finite differencing are necessary, although the direct enforcement of certain approximations is possible. By doing so the method can be used to explore the effect of many standard propagation approximations. We consider certain approximations typically applied to the problem of propagation in randomly structured ionization, such as the neglect of polarization coupling and the assumption of small‐angle scattering.

<title>Ionospheric distortion mitigation techniques for over-the-horizon radar</title>

“…2 The 'forward direction' may vary over large distances due to the smoothly inhomogeneous background of the medium. equations limited only by discretization error, shows that the parabolic approximation holds up remarkably well at HF frequencies [1].…”

“…A thicker random layer can be modeled by a succession of phase screens, provided that the overall angular scatter is not so great as to violate the parabolic assumption. k n u r r (1) which is derived from Maxwell's equations and is exact for scalar fields. The electric field u( ) r is scalar, since we are neglecting polarization, and n( ) r is the random index of refraction proportional to the electron density of the inhomogeneities of the medium.…”

“…al. using the Finite-Difference Time-Domain technique, which is a direct solution to Maxwell's 1 We should expect the distortion to vary with range and azimuth as well as time, since different range-azimuth cells correspond to rays refracted from different parts of the ionosphere. However, there may sometimes be some spatial correlation due to the elongated structure that the plasma takes along magnetic field lines.…”

Finite difference time domain tests of random media propagation theory: 2. Signal correlation length and channel bandwidth

“…In our previous work, high-resolution FDTD computations were used to study the effect of applying this formalism outside the regime of strict validity [Nickisch and Franke, 1996]. This work was limited to the two-position mutual coherence function (MCF), which defines the signal correlation length.…”

“…0048-6604/01/1999RS002416511.00 used to •tudy the adequacy of theoretical random media propagation formulations for both forward scatter [Nickisch and Franke, 1996] and backscatter [Nickisch and Franke, 1998] problems. In the present work, we extend our earlier analysis of forward scatter spatial decorrelation to include frequency decorrelation.…”