**7.6. Exploitation of polarimetric diversity gain**

The exploitation of polarization diversity in radar application provides additional information compared to mono-polarized sensing techniques. This polarization diversity gain enhances the efficiency of object classification according to the information contained in the backscattered signal. Hence, additional characteristics of objects such as shape, details of surface structure and orientation are gathered which may remain invisible for mono-polarized systems. However, in the literature polarization characteristics are rarely considered and most approaches use mono-polarized EM-waves. By the exploitation of polarization diversity the performance can be increased significantly.

Unpolarized electromagnetic incident waves on an object are diffracted or scattered in all directions. The spatial distribution of scattered energy depends on the object geometry, material composition, the operating frequency and polarization of the incident wave.

### *7.6.1. Polarization diversity gain in short-range UWB radar object imaging*

Investigations and results shall be demonstrated by a complex edged 2D object with 6 corners. This object is measured on a circular track at a 1◦ grid which results in 360 measurements. The contour of the object can easily be recognized on the right side of Fig. 41. The sensors are the previously introduced crossed Vivaldi antennas embedded in PTFE which allow full polarimetric measurements.

The imaging algorithm used in this work is Kirchhoff Migration (KM). KM relies on some form of coherent summation, which means that a pixel of the radar image is produced by integrating the phase-shifted radar data of the field amplitude measurements at each antenna position. KM image spots of high intensity correspond to the scattering centers of the object. The image contrast is higher with increasing number of recorded impulse responses at different antenna positions. Here, the object is of 1 m height with about 1 m distance to the object in a bi-static configuration with 0.25 m distance between the transmitter and receiver. Actually, the object is a column with no variation in the height. Thus, it has vertical predominant directions causing stronger reflection in co-polarization or VV, respectively (notation: the first index indicates the polarization of the transmitter, the second index the one of the receiver).

40 Will-be-set-by-IN-TECH

−0.1 −0.05 <sup>0</sup> 0.05 0.1 0.5

X [m]

**Figure 40.** Photograph of an 3D test object on the left. Extracted 3D Radar image with Fuzzy imaging in

The exploitation of polarization diversity in radar application provides additional information compared to mono-polarized sensing techniques. This polarization diversity gain enhances the efficiency of object classification according to the information contained in the backscattered signal. Hence, additional characteristics of objects such as shape, details of surface structure and orientation are gathered which may remain invisible for mono-polarized systems. However, in the literature polarization characteristics are rarely considered and most approaches use mono-polarized EM-waves. By the exploitation of polarization diversity the

Unpolarized electromagnetic incident waves on an object are diffracted or scattered in all directions. The spatial distribution of scattered energy depends on the object geometry,

Investigations and results shall be demonstrated by a complex edged 2D object with 6 corners. This object is measured on a circular track at a 1◦ grid which results in 360 measurements. The contour of the object can easily be recognized on the right side of Fig. 41. The sensors are the previously introduced crossed Vivaldi antennas embedded in PTFE which allow full

The imaging algorithm used in this work is Kirchhoff Migration (KM). KM relies on some form of coherent summation, which means that a pixel of the radar image is produced by integrating the phase-shifted radar data of the field amplitude measurements at each antenna position. KM image spots of high intensity correspond to the scattering centers of the object. The image contrast is higher with increasing number of recorded impulse responses at different antenna positions. Here, the object is of 1 m height with about 1 m distance to the object in a bi-static configuration with 0.25 m distance between the transmitter and receiver. Actually, the object is a column with no variation in the height. Thus, it has vertical predominant directions causing stronger reflection in co-polarization or VV, respectively

material composition, the operating frequency and polarization of the incident wave.

*7.6.1. Polarization diversity gain in short-range UWB radar object imaging*

−0.2

0.4

Y [m] X [m]

0.6

0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Z [m]

0

0.2

0.55 0.6 0.65 0.7

Y [m]

**7.6. Exploitation of polarimetric diversity gain**

performance can be increased significantly.

polarimetric measurements.

the middle and top view on the right.

**Figure 41.** Radargram of the object under test with cross-polarization and 45◦ rotated antennas on the left and the extracted KM image from this radar data.

However, this object has more or less parts which imitate the scattering and reflection characteristic of flat plates or 0◦- dihedrals. In the field of polarization research it is well known that dihedrals have strong polarizing effects. For example, a 0◦-dihedral (the angle between the fold line of the dihedral and the vertical axis) has only co-polarized components, whereas a 45◦-dihedral has only cross-polarized components. Therefore, dihedrals are especially suitable for calibration in polarimetric measurements.

In order to exploit polarimetric diversity gain a 45◦ shift is missing in the radar link [52]. As mentioned before, this would actually depolarize the wave. However, by rotating both antennas by 45◦, the scattering characteristic of the object edges are comparable to 45◦-dihedrals. Hence, in Fig. 42 both sensors are positioned diagonally. Using such a rotated configuration a cross-polarized measurement was performed with cross polarization.

**Figure 42.** Radargram of the object under test with cross-polarization and 45◦ rotated antennas on the left and the extracted KM image from this radar data.

#### 42 Will-be-set-by-IN-TECH 220 Ultra-Wideband Radio Technologies for Communications, Localization and Sensor Applications Cooperative Localization and Object Recognition in Autonomous UWB Sensor Networks <sup>43</sup>

In Fig. 42, it can clearly be seen that only the corners are focused in the resulting image but not the flat structures. The reason for this effect is that a flat plate does not depolarize but a 45◦-dihedral does.

Thus, the detection capability of a UWB radar system can be improved by exploiting polarization diversity. Under certain circumstances, the radar images detect object features which would have remained invisible in mono polarized radar systems. Supplementary information about the contour, orientation and dimensions of the object can thus be obtained, which upgrades super-resolution UWB radar significantly.
