Title :
Automated synthesis of a printed dipole antenna with FD-TD and a novel artificial neural network
Author :
Delgado, Heriberto J. ; Thursby, Michael H.
Author_Institution :
Harris Corp., Melbourne, FL, USA
Abstract :
A novel artificial neural network (ANN) used in combination with the finite difference time domain method (FD-TD), is presented. Practical applications are illustrated by optimizing the voltage standing wave ratio (VSWR) of a printed dipole antenna. An artificial neural network (ANN) architecture, consisting of a hetero-associative memory utilizing a simple sum number representation, is employed. The number representation revealed insight into a new method of fault tolerant computing. Furthermore, a new randomization process allows synthesizing antenna geometrical parameters. Further work is needed to evaluate the potential of this new paradigm.
Keywords :
dipole antennas; fault tolerant computing; finite difference time-domain analysis; microstrip antennas; neural nets; ANN; ANN architecture; FDTD method; VSWR; artificial neural network; automated synthesis; fault tolerant computing; finite difference time domain method; hetero-associative memory; printed dipole antenna; randomization process; sum number representation; voltage standing wave ratio; Artificial neural networks; Dipole antennas; Feeds; Finite difference methods; Geometry; Impedance matching; Microstrip; Network synthesis; Time domain analysis; Training data;
Conference_Titel :
Antennas and Propagation Society International Symposium, 2004. IEEE
Print_ISBN :
0-7803-8302-8
DOI :
10.1109/APS.2004.1330327
