The SAR was simulated using a homogeneous spherical phantom of radius 75 mm. It has been established that the use of a homogeneous spherical phantom gives worse-case simulation of the peak 10 g SAR compared to an anatomically accurate heterogeneous phantom head. Computations times are also much less for the former.
The phantom material parameters chosen were εÏ’=41, conductivity σ=1.65 S/m and density ρ=1030 kg/m3 which resemble human tissue. The antenna radiating power external to the antenna was normalized at 125 mW. In all the simulations in this paper the sphere surface was placed 15 mm distant from the antenna surface of interest. Two antenna orientations with respect to the phantom were investigated: i) the monopole or dipole axis is parallel to the sphere surface and the dipole-like broadside beam is directed at the latter, ii) the monopole or dipole axis is perpendicular to the sphere surface and the dipole-like far-field pattern null is directed at the latter resulting in a significant SAR reduction. This is illustrated in Fig. 3 for a 30 X 30 X 30 mm material cube.
Fig. 3. Simulated radiation performance of 30 X 30 X 30 mm rectangular material coated antenna with central dipole excitation showing variation with relative permittivity; dipole radius = 0.5 mm; tanδε = tanδμ = tand. (a) SAR for perpendicular and parallel dipole orientation and (b) efficiency η and bandwidth BW.
The excitation dipole was varied in height from 9 to 20 mm to maintain tuning at 1800 MHz while μ’ and ε’ were also varied; the material losses were constant at tanδε,μ ~ 0.01 and 0.03. Maximum η, BW and SAR occur in the vicinity of μ |
