Algorithm Definition for the VLSI Design Implementation of the Electromagnetic Radiation Integral
Abstract
A classic method used to solve for the electric (E) and magnetic (H) fields produced by a radiating source involves first solving for the vector potential (A) through the evaluation of the radiation integral. This study was undertaken to define a numerical algorithm that solves for A, so that a reduction in computation time can be realized through VLSI design implementation. Fast Fourier Transforms, Newton-Cotes numerical integration, and variable parameterization, were applied to the problem with unsuccessful results. A numerical summation approximation proved to have all the desired qualities for the algorithm. These qualities are, ability to handle arbitrary antennas, accurate results, and all digital mathematics algorithms exist. Verification of the numerical solution was accomplished by comparing the results it produced for a dipole antenna with a triangular shaped current density against the far-field approximation, and an incomplete cylindrical functions solution developed by Lee. Solution for E and H involves taking derivatives of A. Complications that arise from numerical techniques dictate that numerical differentiation should be performed before numerical integration. Follow on work on this topic will therefore have to focus on this detail before the actual VLSI chip architecture definition stage can be addressed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1985
- Accession Number
- ADA163825
Entities
People
- Lawrence E. Jones
Organizations
- Air Force Institute of Technology