Comparative Evaluation of Some Empirical Design Techniques for CAD Optimization of Wideband U-Slot Microstrip Antennas

Abstract

In this paper three different empirical
techniques, for design of wideband U-slot microstrip
patch antennas on infinite, grounded substrates (r
= 1.0, 4.0, and 6.15), has been studied extensively.
Methods I and III, developed by the present authors,
commonly utilize the property of dimensional invariance
in designing the radiating patch of the U-Slot microstrip
antenna. These two methods use empirical (quadratic)
curve-fit equations, and formulas for design of probe-fed
rectangular patches, respectively, to initiate the corre-
sponding probe-fed U-Slot patch designs. The second
(method II) approach, published earlier by different
authors, principally utilizes the distinction between
four resonant frequencies and calculates the various
dimensions associated with the U-Slot patch radiator.
The initial U-slot designs from the three empirical
techniques have been further optimized for enhanced
bandwidths via: (a) parametric simulation, and (b)
built-in global optimizers such as Powell and Genetic
algorithms in the commercially available microstrip
CAD software, IE3D. Extensive analysis, based on: (a)
comparison of the three empirical design algorithms
(methods I, II and III) , and, (b) impedance behavior
(and VSWR characteristics) of U-slot antennas on low,
medium and high permittivity substrates, show that
method III is generally superior to methods I and II.
Additionally, since the overall U-Slot patch dimensions
follow the relation W
L ≈ 1.385 (methods I and III), and,
W
L ≈ 2.0 (method II), U-Slot designs via methods I
and III have lower cross-polar levels in the principal
(φ = 0◦, 90◦) planes compared to those obtained via
method II