 |
Archives Ouvertes : HAL - IMT |  |
Over the last 20 years, the wireless technology has known a large development in terms of devices and services, accompanied consequently by considerable doubts about the level of exposure to the electromagnetic fields radiated by these systems. An intermediate step in the assessment of the real exposure is to statistically model the input power delivered to these personal devices during an uplink transmission. In this spirit, we propose a systematic methodology for the characterization of the variability of the delivered input power for given propagation conditions. This methodology first addresses the influence of the personal devices relative positioning, with respect to the user's body on the antennas performance as well as on the exposure level. Second, it add-resses the sensitivity of the input power to the characteristics of the propagation channel. The exposure levels have been investigated from specific absorption rates computed using finite-difference time-domain simulations on a realistic numerical model of child bodies. The input power is directly related to the antenna radiation pattern and to the local propagation scenario, through effects such as body obstruction and multipath diversity. The statistical analysis shows that the delivered input power can be modeled by an inverse Gaussian or a log-normal distribution.
