Automated Experimental Data Collection System for Radio

Tomographic Imaging
Jason David Bonior, Dr. Zhen Hu

Introduction System Control Experimental Results Special Apparatus

With the proliferation of wireless devices comes the All devices are controlled using LabVIEW and GPIB There are basic tests which can give us a sense of The need to minimize the amount of conductive
increase in large-scale wireless communication interfaces. The experiments discussed here are whether or not our data appears accurate. Ultimately material inside the chamber required the design and
networks around the world. Remote sensing can be 0 0 0
conducted within an 18 × 18 × 10 RF-anechoic however, the final test is to apply the data and observe fabrication of specialty non-conductive apparatus. To
embedded into such a network providing a wide chamber and all experimental apparatus are either its performance. We verify the correct operation of our hold our dipole antennas we designed a clip-on plastic
reaching sensing platform [1]. However, wireless non-metallic and/or shielded using radiation absorbing system by utilizing known, working algorithms to carry antenna mount. This part was designed using
communication devices are not specifically designed material. These precautions allow us to treat the data out inverse scattering using the Born-Iterative Method SolidWorks and fabricated using a 3D printer (3D
for, nor do they meet the high accuracy measurement collected as free space phenomena. and construct an image of the target [2]. Figure 4 Systems Projet SD 3500).
requirements of remote sensing. Innovative algorithms shows the position of the target in each of the two
are required to perform sensing with communication experiments discussed here.
nodes. One of several solutions offered is wireless
tomography. Wireless Tomography combines wireless The value of each pixel in the generated images
communication and RF Tomography providing a novel represents the dielectric constant of the material
approach to remote sensing [2]. Experimental located in that space. Referring to the color bar on the
verification of such techniques requires a robust right of Figs. 5(a) and 5(b) we can see that the deep
experimental system. Here we outline our new red located at the center of each target is
automated experimental data collection system built approximately equal to 3 which is what we would
for furthering our research in Wireless Tomography. expect for this material. We also notice that the
Figure 2: Front Panel of Control System Virtual Instrument surrounding area is approximately equal to 1 which is
(a) 3D Rendering of Antenna (b) Completed Antenna
to be expected for air [4]. Mount Mount
Experimental Setup Figure 6: Custom Antenna Mount for use in our RF-Anechoic
Chamber
We focus on elongated objects as our targets. The
receiving antenna circles about the imaging area from
Conclusions
60◦ to 300◦ with the transmitting antenna located at
0◦. We measured the accuracy of the turn table to be We have demonstrated our ability to collect reliable
± 0.57◦ when moving at least 1◦ in either direction. scattering data using the system outlined on this
Measurements are taken with the receiver at 121 poster. We will expand upon this work by
positions from 60◦ to 300◦ in 2◦ increments. A (a) Target Position A (b) Target Position B
experimentally verifying previous theoretical work
diagram of the experimental layout can be seen in Fig Figure 4: Layout of the Imaging Area
conducted by students in our lab in the area of phase
3. reconstruction.
3

Figure 1: Experimental Setup within RF-Anechoic Chamber −0.06 2.8 −0.06

3

2.8

240.0° −0.04

−0.02
2.6

2.4
−0.04

−0.02
2.6

2.4
References
2.2

System Description 2.2

x (m)

x (m)
2
0 0
2

0.02
1.8

1.6
0.02 1.8 [1] J. Wilson and N. Patwari, “Radio tomographic imaging with
1.6
0.04 1.4 0.04 wireless networks,” Mobile Computing, IEEE Transactions on,
Our experimental setup is similar to that presented Imaging
0.06
1.2

1
0.06
1.4

1.2
vol. 9, no. 5, pp. 621–632, 2010.
1

in [3]. The system consists of a surface-mount Area

1.22 m Receiving −0.06 −0.04 −0.02 0
y (m)
0.02 0.04 0.06 −0.06 −0.04 −0.02 0
y (m)
0.02 0.04 0.06

1.58 m Antenna [2] R. C. Qiu, M. C. Wicks, L. Li, Z. Hu, and S. J. Hou, “Wireless
turntable, a vector network analyzer (VNA), two horn 0.19 m
(a) Target Position A (b) Target Position B Tomography, Part I: A Novel Approach to Remote Sensing,”
antennas, and a PC to serve as system control. The Figure 5: Imaging Results Using Collected Data in 5th International Waveform Diversity & Design Conference,
VNA, an Agilent PNA-L N5230, is calibrated to Niagara Falls, Canada, August 2010.
1.58 m

remove the effects of the feed cables. This ensures If we look closely at our resulting images we notice [3] J.-M. Geffrin, P. Sabouroux, and C. Eyraud, “Free space
that, with the exception of the antennas, only what appears to be a slight offset in the position of experimental scattering database continuation: experimental
scattering due to the target is observed. The effect of set-up and measurement precision,” inverse Problems, vol. 21,
the target. Upon closer inspection of the target it was no. 6, p. S117, 2005.
the antennas used is accounted for in the Transmitting
Antenna found that it had a slight bend resulting in the
post-processing of the data. [4] J. F. Shackelford and W. Alexander, CRC materials science
Figure 3: Physical Layout of Experiment within RF-Anechoic cross-section being shifted approximately 15 mm from and engineering handbook. CRC press, 2001.
The location of the transmitting antenna is fixed so we Chamber the expected location. Such accuracy in the resulting
must rotate the imaging area an appropriate amount images further increases our confidence in the Acknowledgments
for each transmitter position. This emulates the The target chosen for our experiments is a 4-foot performance of this system.
(1.22 m) long by 3 inch (7.62 cm) diameter piece of This work is funded by National Science Foundation through three grants
relative rotation associated with multiple transmitter (ECCS-0901420, ECCS-0821658, and CNS-1247778), and Office of Naval
positions while still providing a consistent experimental Chlorinated Polyvinyl Chloride (CPVC). This material Research through two grants (N00010-10-1-0810 and N00014-11-1-0006).
setup. The complete experimental setup can be seen has an approximate dielectric constant of 3 [4]. This work is partly funded by an AFOSR subcontract though the prime
contractor (RNET Technologies, Inc.) on the CIRE (Center for Innovative
in Fig. 1. Radar Engineering) contract FA8650-10-D-1750 with AFRL/RY.

http://iweb.tntech.edu/rqiu