Tips for Using RL-GO

Several tips are presented to reduce runtime and memory consumption when using hybrid methods involving the ray launching geometrical optics (RL-GO).

  • Reduce the number of interactions

    The MoM / RL-GO is an optical technique that is based on the principle of ray-launching. From the MoM elements or sources, rays are launched in all directions. Where a ray hits geometry with the RL-GO setting, a Huygens source is created.

    The RL-GO requires that Huygens sources are spaced a minimum distance apart to maintain accuracy. If there is more than one interaction specified (the default is three), additional rays are launched from the previous set of Huygens sources, resulting in more Huygens sources. Once all the interactions are computed, integration over all the Huygens sources is done to obtain the solution.

  • Decouple the MoM and RL-GO parts if the coupling is negligible
    If the RL-GO part of the model does not significantly influence the MoM part of the model, then the MoM part can be decoupled from the RL-GO part to reduce computational resources.
  • Reduce the number of sources
    Each MoM unknown or source element acts as a source for the RL-GO region. Rays are launched from each MoM element or source point. Using an equivalent source (for example, radiation pattern point source, spherical expansion mode source, Hertzian dipole) reduces computational resources for the MoM / RL-GO as there is only a single source for the geometry part treated with the RL-GO.