National Research Council Post doctoral Research Associate at NIST. He is now with the Signal and Information Sciences Laboratory University of Texas Applied Research Laboratory, Austin, TX USA
Auto-generated profile, based on publicly available information
Ryan J. Pirkl
National Research Council Post doctoral Research Associate at NIST. He is now with the Signal and Information Sciences Laboratory University of Texas Applied Research Laboratory, Austin, TX USA
We assess the potential error in measurements of the power transfer function corresponding to a reverberation chamber set-up when different antenna types are used for the reference and device-under-test measurements. We derive a mathematical description of the transfer function that accounts for differences in the amount of unstirred energy, represented by the spatially averaged K factor, arising from various antenna types. Our results show that loaded chamber configurations, combined with reference/device antenna pairs having significantly different radiation patterns, can result in statistically significant errors in the prediction of the transfer function. If it is possible to obtain an estimate of the K factor associated with the device antenna, this correction can improve the estimate of the transfer function that would be experienced by the device under test. Finally, we develop a method that could be used in standardized test methods to bound the uncertainty associated with the unknown K factor for common antenna types.
9 April 2014 in IEEE Transactions on Electromagnetic Compatibility
In this paper, we study statistical methods for estimating the Rician K-factor when this parameter is small. A fiducial approach for making statistical inference on the K-factor is discussed. The approach requires a Monte Carlo method to compute the uncertainty of a K-factor estimate. Simulation and measurement studies are carried out to evaluate the performance of the proposed procedure and compare it with some of the existing procedures. A procedure for assessing the adequacy of the estimation of the K-factor via quantile plots is presented.