Abstract
INTRODUCTION: This study proposes a framework for determining the calculation error in online SAR supervision introduced by directional couplers. MATERIALS AND METHODS: A mathematical framework is introduced showing how the error in the measured excitation vector compared to the actual excitation vector can be rewritten as a new set of virtual observation points (VOPs). By comparing the new set of VOPs to the original VOPs through an optimization, the maximum underestimation of SAR can be calculated. The framework is then applied to five different RF arrays. RESULTS: The results show that the error in SAR calculation is very dependent on the position of the reference plane of the directional coupler measurements and the S-parameters of the array. To have an error of less than 5%, directional couplers with a directivity better than 40 dB are necessary for the worst case of the investigated arrays. DISCUSSION: The framework presented in this paper provides an approach for calculating a safety factor to account for the inaccuracies introduced by directional coupler measurements in online SAR supervision. The framework can also be adapted to other types of measurements.