Multi-Level Approximations for Fast and Accurate Antenna Noise Temperature Calculation of Dual-Reflector Antennas

Antenna noise temperature calculation for a reflector antenna system can be a computationally expensive operation if the structure is electrically large. Several approximate methods have been suggested to speed up the calculation. These rely on removing one or more of the reflecting surfaces from the calculation domain and compensating for the effect through geometric optics arguments. This article presents a unified multilevel approximation formalism, which allows the user the freedom to tradeoff the simulation time, implementation complexity, and result accuracy as required for a specific application. As part of the formalism, a strategy to reliably remove both reflectors in a dual-reflector system from the calculation domain and rely on ray tracing to correct the background temperature is developed. It is shown that this new method reduces calculation time by several orders of magnitude while maintaining accuracy in the final antenna temperature result of the order of 1%. It is fast and accurate enough to be used in numerical optimization loops. The method requires the implementation of a ray tracing algorithm, and thus, may present some unnecessary complexity if only rough temperature estimates are required. To this end, a simplified method is also presented requiring no ray tracing and only simple postprocessing of readily available feed pattern data. Though not as accurate as of the ray tracing-based method, temperatures may be estimated with around 10% accuracy over bandwidths well in excess of a decade.