62
IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 46, NO. 1, FEBRUARY 2004
Study of the Coupling Between Human Head and
Cellular Phone Helical Antennas
Stavros Koulouridis and Konstantina S. Nikita, Senior Member, IEEE
radiate in two modes depending on their dimensional charac-
Abstract--The interaction between normal-mode helical
antennas and human head models is analyzed, using both a
teristics [1]. For mobile communication handsets, helical an-
novel accurate semi-analytical method and finite-difference
tennas are constructed to radiate in normal mode, hence their
time-domain (FDTD) simulations. The semi-analytical method is
physical dimensions are very small compared to the radiation
based on the combination of Green's functions theory with the
wavelength.
method of moments (Green/MoM) and is able to model arbitrarily
The evaluation of the power absorbed by the user's head as
shaped wire antennas radiating in the close proximity of layered
lossy dielectric spheres representing simplified models of the
well as the antenna performance in the presence of the user's
human head. The purpose of the development of the Green/MoM
head represent key tasks for both design and compliance testing
technique is to provide a reliable tool for preliminary (worst
of cellular phones. These tasks can be efficiently addressed
case) estimation of human head exposure to the field generated
by means of numerical techniques, with the finite-difference
by different antenna configurations with emphasis on the helical
time-domain (FDTD) method [2], [3] dominating over the other
antenna, representing the most diffused antenna type used in
modern cellular handsets. Furthermore, the accurate semi-analyt-
numerical methods in recent years, due to its simplicity and its
ical character of the Green/MoM technique permits the accuracy
ability to treat highly nonhomogeneous structures. Although
assessment of purely numerical techniques, such as the FDTD,
FDTD is able to model anatomically detailed human head
which is currently the most widely used computational method
structures, significant difficulties are encountered in modeling
in mobile communication dosimetric problems, since it allows
antenna structures not conforming to the used grid. Thus, while
modeling of anatomically based head models. After appropriate
benchmarking, FDTD simulations are used to study the interac-
monopole and planar antennas can be easily implemented
tion between a heterogeneous anatomically correct model of the
within FDTD codes, modeling of a helix can become a rather
human head exposed to a normal-mode helix monopole operating
difficult task. In fact, rather large structures have been ana-
at 1710 MHz mounted on the top of a metal box representing
lyzed using a pure FDTD scheme [4], [5], while for smaller
a realistic mobile communication terminal. The study of both
structures a graded mesh [6], equivalent sources [7], [8], and a
canonical and realistic exposure problems includes computations
of specific absorption rates (SARs) inside the human head,
hybrid method of moments (MoM) /FDTD [9], [10] have been
total power absorbed by the head and assessment of antenna
proposed. Furthermore, even if intensive research work with
performance. Emphasis is placed on the comparative dosimetric
the FDTD method has led to high confidence in the obtained
assessment between adults and children head models.
results, the exact error estimation of FDTD simulations still
Index Terms--Biological effects of electromagnetic radiation,
remains a difficult task. It has been shown that uncertainties
helical antenna, human head of adults and children, mobile
are involved and discrepancies can be observed in the results
phone, semi-analytical and finite-difference time-domain (FDTD)
obtained by different research groups nominally using the
techniques.
same numerical method, even for well defined canonical cases
involving linear dipoles [11].
I. INTRODUCTION
Due to the above described problems in FDTD modeling
T
of the helical antenna and uncertainties encountered in FDTD
HE wide expansion of cellular phones around the world
modeling even for linear antenna structures, the availability of
is indisputable. The phone radiating element used to vary,
accurate methods able to treat canonical exposure problems
in previous years, among quarter-wavelength, half-wavelength,
becomes of primary importance. Such methods can provide an
three quarter-wavelength linear antennas and whip antennas.
efficient tool for benchmarking of purely numerical techniques
These antenna models have been extensively used to equip var-
(FDTD), which can then be used to analyze realistic exposure
ious cellular handsets. In the last few years, however, different
problems.
antenna designs have been proposed, motivated by the need to
reduce the size of handset devices while keeping radiation char-
In this paper, the electromagnetic dosimetry problem refer-
acteristics similar with the "older" antenna models. One of the
ring to the exposure of human head models to helical antennas is
most diffused new antenna types is the helical antenna. Helical
treated in detail. To this end, a semi-analytical technique is pro-
antennas have been used in the last 40 years and, in general,
posed, based on the use of the dyadic Green's function theory
[12], [13] and the MoM (Green/MoM), for studying the interac-
tion between a layered spherical head model and an arbitrarily
Manuscript received October 31, 2002; revised June 14, 2003. The work of
shaped wire antenna. The proposed method provides a means of
S. Koulouridis was supported by the Hellenic State Scholarships Foundation.
The authors are with the National Technical University of Athens, School
(worst case) preliminary estimation of the human head exposure
of Electrical and Computer Engineering, 15780 Athens, Greece (e-mail:
to the field generated by different wire antenna configurations
knikita@cc.ece.ntua.gr).
such as helical or linear dipoles. It is also employed as a reliable
Digital Object Identifier 10.1109/TEMC.2004.823612
0018-9375/04$20.00 © 2004 IEEE