Integration of a 4x8 Antenna Array with a Reconfigurable 2-bit Phase
Shifter using RF MEMS Switches on Multilayer Organic Substrates
David J. Chung, Dimitrios E. Anagnostou, G. Ponchak, Manos Tentzeris, and
John Papapolymerou
Georgia Institute of Technology, School of Electrical and Computer Engineering,
85 Fifth Street, Technology Square Research Building,
Atlanta, GA 30308
Introduction
As water is a crucial factor for sustaining life, the study of its distribution around
the globe has also been a great importance. Satellites have been deployed with
antennas operating at 14 GHz to observe the rainfall. In the past, high-gain
parabolic reflectors have been used frequently for space missions with a cost of
low efficiency, large size, heavy weight, and difficulty in deploying. In order to
overcome these troubles, microstrip arrays antenna have been developed. Though
they have higher insertion loss, meaning less efficiency, they will be significantly
lighter, smaller, and simpler [1], [2]. Recently, NASA has been collaborating with
Georgia Institute of Technology to develop advanced, low cost, light weight, and
large surface area antennas that can be deployed in space. Previous works show
multi-functional small arrays [3], [4], but the size has been limited, and the
possibility of expanding the structure is small.
In this paper, Liquid Crystal Polymer (LCP) is used as the substrate, which
enables the fabrication of a flexible, low cost, and light weight antenna. Using
LCP, an attempt to expand from the previous works has been accomplished by
constructing a 4x8 antenna array made of two 4x4 sections being fed with a
reconfigurable 2-bit phase shifter. Each reconfigurable phase shifter is enabled by
eight RF MEMS switches and the two 2-bit phase shifter together, allow a 6°
phase shift in the final design.
Antenna Design
The unit design is based on a 4x4 antenna array. We can see in Fig. 1, the S11
response of a 4x4 array. The resonance is at a frequency about 13.9 GHz shown
by the line corresponding to thick foam and the measurements match the
simulation results. The thick foam (~2 inches) was placed on the patches to
facilitate the measurements when positioning the probes, and did not interfere
with the free-space results as the relative permittivity is very close to 1. In order to
verify that the resonance indeed represents radiation that takes place and does not
originate from anywhere along the feed network or the slots, the patch layer was
attached to a thick metal plane. With the metal plane on top of the patches, the
reflection coefficient was measured (Fig. 1). The results showed a cancellation of
the occurring resonance which verifies that the radiating elements are the metal
93 1-4244-0878-4/07/$20.00 ©2007 IEEE