Instrumentation & Measurement Magazine 25-3 - 39

Fig. 3. The induced micro-Doppler signals from a plastic bottle exposed to
four different acoustic frequencies.
An acoustically driven target vibration is expected to proFig.
2. The laboratory setup for measurements of induced micro-Doppler
signals. The system comprises radar (signal generator, antennas and receiver)
and acoustic (loudspeaker) systems. The triangular transmit and receive
antennas are visible on the desk. The speaker is off to the left side, and the
target is a water bottle at the bottom of the image.
radar cross section and acoustic coupling. Measurements on
the metallic plate were performed to place practical limits on
the ranges of the potential acoustic and radar frequencies. This
initial experiment applied acoustic frequencies from 35 Hz to
150 Hz. For each acoustic frequency, micro-Doppler measurements
were collected at a number of discrete radar frequencies
spanning 750 MHz to 3.3 GHz. A preliminary analysis showed
a clear preference for detections at acoustic frequencies of
100 Hz or less and radar frequencies above 2.25 GHz. From
these observations, the measurements of the representative
targets were conducted at lower acoustic frequencies and a radar
frequency of 3.0 GHz.
To minimize acoustic reflections and radar clutter during
the data collection of the representative targets, the instrument
setup was moved to an outdoor field adjacent to the C-CORE
offices in St. John's, Newfoundland and Labrador, Canada.
Measurements were performed at fixed acoustic frequencies
which ranged from 10 Hz to 100 Hz in order to survey potential
resonant frequencies of the targeted objects. A 200 Hz band pass
filter, centered on the 1 kHz radar IF, was applied to the data.
This filter bandwidth was selected to be wide enough to prevent
attenuation of reasonably large target vibrations. Since the
radar and acoustic signal frequencies were constant in each test,
a frequency demodulation was applied to the filtered data and
the power spectral density (PSD) was then calculated.
May 2022
duce a sinusoidal micro-Doppler signal with an amplitude
proportional to the amplitude of the target vibration and an
oscillation frequency equal to the acoustic frequency. This
signal is detected by excess power in the PSD at the acoustic
frequency and is observed at a subset of acoustic frequencies
for all targets. For example, Fig. 3 shows a clear micro-Doppler
signature for the plastic bottled filled with water when
exposed to acoustic frequencies of 70 Hz, 80 Hz, and 90 Hz and
no detection when exposed to an acoustic frequency of 40 Hz.
For this target, no micro-Doppler signal was detected when
the acoustic frequency was set lower than 60 Hz, suggesting
that the resonances in vibro-acoustic coupling are strongest
between 60 Hz and 100 Hz. It is expected that the resonances
in vibro-acoustic coupling depend on the target material, composition,
and size.
The top of Fig. 3 plots the results when exposed to frequencies
in the high end of the acoustic range. The data show a peak
signal at the acoustic frequency used to drive the vibration of
the target. The bottom of Fig. 3 plots the result when applying
a 40 Hz acoustic frequency for target excitation. In this case, no
micro-Doppler signal is detected. For visual clarity, the data in
the top figure are normalized to their largest signal in the first
100 Hz. The data in the bottom figure is normalized to match
the background level of the other data sets.
Data were also recorded at two different loudspeaker volumes:
79 dB and 80 dB. For a given volume, the background
level is generally consistent across the full range of acoustic frequencies.
However, as the volume level increases by 1 dB, the
overall background level increases by a median factor of 2.3.
Fig. 4 shows the median background level measured across a
range of acoustic frequencies. The origin of this relationship
between speaker volume and background level is unknown
but is potentially due to a wideband acoustic noise level that
increases with frequency.
Finally, the noise in the data is shown to average down
as 1/ number of samples . This basic assessment indicates
IEEE Instrumentation & Measurement Magazine
39
Instrumentation & Measurement Magazine 25-3

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 25-3
