Comments on assessment of polarization dependence of body shadow effect on dosimetry measurements in 2.4GHz band

Radiofrequency (RF) exposure measurements using personal dosimeters or exposimeters are influenced by the presence of the body, which reflects, diffracts, and absorbs the same RF electromagnetic fields (EMFs) that one wishes to measure [Bolte et al., 2011]. Two of these effects are a reduction in registered electric field (E-Field) strength and the influence of polarization. The latter is caused by differences in propagation around the human body, according to De Miguel-Bilbao et al. [2017]. In their research, they find that “the attenuation due to the body-shadow effect is greater when the antenna is vertically polarized” [De Miguel-Bilbao et al., 2017]. This would be mainly attributed to “a greater effective area where incident waves are scattered” [De Miguel-Bilbao et al., 2017]. In our opinion, the propagation and absorption mechanisms are more complex, in particular the polarization dependence, than what De Miguel-Bilbao et al. [2017] bring forward in the discussion section of their paper. Moreover, as we will show in this manuscript, the received polarization is not controlled or constant in the experiments used in De Miguel-Bilbao et al. [2017]. When RF EMFs are incident on a subject, a part of the incident RF EMFs will be absorbed, another part will be reflected away from the measurement devices, and a third part can propagate towards the exposimeters. This propagation has several components: propagating modes around the human body, specular components, including the line-of-sight component between the source exposimeter, and reflections from the environment, and the diffuse multipath component (DMC). This DMC is the part of the incident power density that cannot be attributed to any incident angle [Bamba et al., 2015] (in contrast to the specular components). First, the loss of EM power due to absorption in the human body is quantified using the whole-body averaged specific absorption rate (SARwb). Existing literature on the absorption of RF frequencies above >2GHz is not conclusive on polarization dependence of the SARwb. Some studies find a higher absorption for horizontally polarized (H-polarized) incident plane waves [Hirata et al., 2009; Uusitupa et al., 2010; Bamba et al., 2015] at frequencies higher than 2GHz. Others find only slightly higher absorption values for vertically polarized (V-polarized) incident plane waves at frequencies higher than 2GHz [K€uhn et al., 2009; Bakker et al., 2010]. Second, propagation around the human body is indeed polarization-dependent. However, there is again no clear consensus in literature that propagation of EMF waves around the human body would result in relatively more path loss for V-polarized plane waves at 2435MHz in comparison to H-polarized EMFs. Alves et al. [2011] found a higher path loss for E-fields polarized parallel to the main axis of a lossy cylinder in comparison to E-fields polarized orthogonal to the surface of the same cylinder at 2.4GHz, for EMFs that are emitted in very close proximity to the cylinder. Syed and Volakis [1991] and Mavridis et al. [2014] indicate that a difference in path loss between both polarizations depends on the size and dielectric properties of the phantom (cylinder), w.r.t. the used wavelength. Moreover, it is not clear how the two considered polarizations H and V would convert into parallel and orthogonally propagating modes on the body as both incident polarizations will have compo-