281
Mobile Communication Fields in Biological Systems
Obviously, the possibility to estimate the exact error bound of the results or to obtain
results with a desired accuracy would be extremely desirable, thus eliminating the need
for further (experimental or numerical) validation. However, the exact estimation of an
error for complicated EM problems, such as the interaction of near-field sources with
complex dielectric structures, is not yet possible. While international standards exist
for the evaluation of uncertainty in the maximum local SAR values for compliance tests
of cellular phones (IEEE 2003; IEC 2005), procedures to evaluate the uncertainty of
numerical dosimetry have not been established.
5.3 Experimental Dosimetry
Experimental dosimetry has been used for over 25 years to obtain SAR distributions for far-field and near-field actual exposure conditions (Balzano, Garay, and
Manning 1995; Balzano, Garay, and Steel 1978; Cleveland and Athey 1989; Gandhi
and Chen 1995; Johnson and Guy 1972; Schmid, Egger, and Kuster 1996; Stuchly and
Stuchly 1995). It enables accurate assessment of the induced field strengths under realistic conditions and thus, is widely applied for validation of the simulation results as well
as for compliance testing purposes. The weaknesses of experimental dosimetry include
restriction to liquid tissue-simulating media, limited spatial resolution of larger than
1 mm 3 , and limited applicability inside small substructures since the dimensions of the
measurement instrumentation must be considerably smaller than that of the substructure under study (Cavagnaro and Pisa 1996).
5.3.1 Human Phantoms
It is very difficult to measure the internal electric and magnetic field strength or temperature elevation in the actual human body exposed to RF EM fields emitted by mobile
communication devices by using noninvasive methods. Therefore, a phantom, a surrogate of the human body, which has electrical properties equivalent to those of the human
body, is used for experimental dosimetry.
Several kinds of materials have been developed to realize the tissue electrical properties. Their references may be found in international standards on RF dosimetry
(IEC 2005; IEEE 2002). Liquids and gels placed inside an enclosing shell that gives the
phantom the shape of the body to be modeled (most commonly made of plexiglass) have
been developed for RF dosimetry (Johnson and Guy 1972; Hartsgrove, Kraszewski, and
Surowiec 1987). These materials are easy to prepare and their electrical properties are
easily adjustable. Fluids rather than gels are preferred to facilitate rapid movement of
the probe tip sensor placed inside the phantom. However, it is worth noting that the
electrical properties of liquids and gels are not stable enough due to water evaporation.
Furthermore, although dry phantoms with fine stability have also been developed, they
are generally expensive and require complex and skilled procedures (Kobayashi et al.
1993; Nikawa, Chino, and Kikuchi 1996).
In the simplest phantoms, homogeneous liquids and gels are used. For example,
the experimental verification of compliance of mobile communication devices with
the basic limits (ICNIRP 1998; IEEE 2005) is usually performed by using plastic shell