Multilevel surface decomposition algorithm for rapid evaluation of transient near-field to far-field transforms

Amir Shlivinski, Amir Boag

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

A fast multilevel algorithm with reduced memory requirements for the evaluation of transient near-field to far-field transforms is presented. The computational scheme is based on a hierarchical decomposition of an arbitrary shaped enclosing surface over which the near-fields of an antenna or a scatterer are given. For surface subdomains at the highest decomposition level, the angular-temporal far-field patterns are calculated directly from the known near fields over a sparse angular grid of directions and a short temporal duration. The multilevel computation comprises angular and temporal interpolations thus increasing angular resolution and temporal duration of radiation patterns while aggregating the subdomain contributions between successive decomposition levels. These steps are repeated until obtaining the transient far-field response of the whole enclosing surface. The computational complexity of the proposed algorithm is substantially lower than that of the direct evaluation. Reduction in memory requirements is obtained by formulating the algorithm as a marching-on-in-time windowed scheme. This approach allows for embedding of the accelerated transforms within existing near-field modeling tools.

Original languageEnglish
Pages (from-to)188-195
Number of pages8
JournalIEEE Transactions on Antennas and Propagation
Volume57
Issue number1
DOIs
StatePublished - 1 Apr 2009

Keywords

  • Antenna transient analysis
  • Computation time
  • Finite-difference (FD) methods
  • Finite-volume (FV) methods
  • Near-field far-field (NF-FF) transformation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Multilevel surface decomposition algorithm for rapid evaluation of transient near-field to far-field transforms'. Together they form a unique fingerprint.

Cite this