This paper presents considerations regarding the signal quality for free-space UWB measurements. Scattered pulses from various small and irregularly shaped objects are subjected to multivariate calibration in order to develop a statistical model for the moisture content. The accuracy of this model strongly depends on both short term and long term variations of the received signals. Amplitude noise and jitter are referred to as short term variations. Low noise and low jitter are crucial for the accuracy of the multivariate calibration methods. The dependence of the signal quality on the accuracy is investigated. Furthermore, the influence of long term variations, e.g. temperature differences, is considered.
展开▼
