**1. Introduction**

178 Advanced Topics in Measurements

Vallbo, Å.B. & Johansson, R.S. (1984). Properties of Cutaneous Mechanoreceptors in the

The error analysis has been investigated since 1975 since recent publications, as shown in references as (Newell et al., 1975) where the classical 18 terms for planar near field systems where established. It is also worth mentioning the more recent studies developed inside the actions under the "Antenna Measurement" activity of the "Antenna Centre of Excellence (ACE)" within the sixth framework research program of the European Union. In particular, that work pretended to establish common error calculation criteria in spherical near-field and far-field antenna measurement systems. The results of that research were summarized in an exhaustive deliverable (Alexandridis et al., 2007), which detailed the observations stated by several research institutions. In that deliverable it was agreed that the causes of uncertainties and errors in a spherical near-field antenna measurement are divided in six categories:


The second part of the chapter will deal with the description of four different classes of correction techniques for antenna measurements.:

Overview of Novel Post-Processing

measurement facility considered.

measurement were divided in six categories:

in a spherical coordinate system.

the amplitude and phase shift in rotary joints.

Techniques to Reduce Uncertainty in Antenna Measurements 181

widely explained. Actually, the above mentioned studies of Hansen in (Hansen, 1988) and Newell in (Newell, 1988), published in 1988 and concerning the uncertainties in antenna measurements, settle the basis to evaluate the inaccuracies that may affect the results achieved from the antenna measurements. However, as the technology has incredibly evolved since then, an update of the uncertainty analysis is considered appropriate and therefore, this section will try to cover the uncertainty study of measurements that nowadays could be carried in indoor facilities, summarizing the work developed in (Alexandridis et al., 2007), under the "Antenna Measurement" activity of the "Antenna Centre of Excellence (ACE)" within the sixth framework research program of the European Union, involving researchers from the National Center of Scientic Research "Demokritos" (NCSRD), the Technical University of Denmark (DTU), the Technical University of Madrid (UPM), SATIMO (Société d'Applications Technologiques de l'Imagerie Micro-Onde S.A.), Saabgroup, FTR&D (France Telecom Recherche & Développement) and IMST GmbH. In particular, that work pretended to establish common error calculation criteria in spherical near-field and far-field antenna measurement systems. Thus, the outcomes of this activity became important instruments to verify the measurement accuracies for each range and to investigate and evaluate possible improvements in measurement set-up and procedures.

It was seen in this research that these errors depend not only on the gain determination technique applied to establish the antenna gain, but also on the AUT that will be measured and on the measurement range employed. Hence, a generic evaluation cannot be carried out and in each case the study has to be particularized and adapted to the AUT and

Besides, it is noticeable that the procedure to evaluate the uncertainties has to be different in the near-field systems than in the far-field ranges. This is due to the fact that in the far-field systems the measurement is direct and thus, the uncertainty could also straightly be obtained. On the other hand, in near-field ranges the information and the uncertainties related to this data are processed and so the near-to-far-field transformation has an influence on the both magnitudes. The sources of uncertainty present in the near-field measurements are either systematic or random quantities. The systematic errors can be characterized using the mean. On the other hand, for the random sources of uncertainties, the mean, the variance (σ2) and the probability distribution are employed to carry out the description of the magnitude studied. The first step to accomplish a complete uncertainty study is to identify all the sources of inaccuracies (uncertainties or errors) affecting the measurements. In the exhaustive deliverable carried out under the "Antenna Measurement" activity of the ACE, the observations, stated by the participant research institutions, were detailed and the causes of uncertainties and errors in a spherical near-field antenna

1. Mechanical uncertainties and errors: this group includes the axes intersection, the axes orthogonality, the horizontal pointing, the probe vertical position, the probe pointing and the measurement distances. Fig. 1 illustrates how is defined a non-intersection error

2. Electrical uncertainties and errors: this class contains the amplitude and phase drift, the amplitude and phase noise, the leakage and crosstalk, the amplitude non-linearity and

3. Probe-related uncertainties and errors: this kind of uncertainties takes into account the channel balance amplitude and phase, the polarization amplitude and phase and the


This paper is divided as follows: section 2 explains the main uncertainty sources based on the works developed on the Antenna Centre of Excellence (Alexandridis, 2007). Sections 3 to 6 detail the different aspects mentioned before, including some practical examples.
