Efficient monostatic anisotropic point scatterer model for high performance computing

Research output: Contribution to journalArticlepeer-review

Abstract

High performance computing (HPC) electromagnetic (EM) emulators are used to simulate real-time EM wave interactions between numerous radar targets. Radar Cross Section (RCS) data stores reflection profiles of radar targets as a table; however, the needed storage size increases with frequency sampling density, aspect angle sampling density, and number of target types. The large quantity of data often exceeds storage capability and limits manipulation and representation feasibility. The spherical harmonic based monostatic anisotropic point scatterer model is proposed for HPC EM simulations where scattering responses can be computed with finite impulse response (FIR) filters. An efficient algorithm constructing this model with large scale RCS data is discussed. The scatterer position and the reflection profile of each scatterer are solved using least squares methods and particle swarm optimization (PSO). In addition, the function evaluations in PSO are accelerated by taking advantage of the matrix structure, making the algorithm 22 times faster.

Original languageEnglish (US)
Pages (from-to)522-537
Number of pages16
JournalJournal of Electromagnetic Waves and Applications
Volume38
Issue number4
DOIs
StatePublished - 2024

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • General Physics and Astronomy
  • Electrical and Electronic Engineering

Cite this