To ensure that our antenna was functioning properly a metal plate was placed at 15, 20 and 25 centimeters from the antenna. Ten measurements were taken for each distance and the average result was used. This is done to minimize the level of error in the measurements from the network analyzer. The results were then weighted, conjugated and inverse Fourier transformed. The signals received were then subtracted by the signal received by just the antenna. This is done to eliminate any reflections due to the antenna. The result can be seen in Figure 8.

The figure shown below is a magnified version of the figure shown above. It can be seen that each spike is a different distance from the antenna. Note that the separation distances are equal as in the experiment, 5 centimeters. This shows that the antenna is recording the reflections appropriately, our algorithm to find the signals is correct and that the signals can be detected over the noise.

The next step of the experiment is to try and use the phantom to see if a signal can be detected. Measurements were taken for the PVC pipe and the PVC pipe with copper pipe together. The same process as before was applied to see if difference between the PVC pipe and the PVC pipe with copper pipe could be detected. The result shown below demonstrates that we cannot determine a difference between the two experiments.


However, our results returned an abnormal signal and due to lack of time, further investigation and more measurements were not possible. The only result that was returned as expected was the copper pipe at 10 centimeters. These results show that the copper pipe can be identified and further tests must be done to identify the pipe. The findings with the copper pipe are encouraging for our group because the copper pipe is a much smaller object then the metal plate. The next steps that should be done are to continue testing with the copper pipe. Once identification of the copper pipe can be found at different distances adding the PVC tube with the copper pipe at different distances from the antenna will be an important step. After this is achieved, the model to turn the copper pipe should be used to take readings. These readings will then be applied to Dr. Elise Fear's algorithms and the exact place of the copper pipe will be identified.