Digital Signal Processing Reference
In-Depth Information
FIGURE 10.64.
Block diagram of an acoustic signal tracker.
ing. Since one microphone is closer to the source than the other, the signal received
by the more distant microphone is delayed in time. This time shift corresponds to
the angle where the source is located and the relative distance between the micro-
phones and the source. The angle
c
arcsin(
a
/
b
)
,
where the distance
a
is the product
of the speed of sound and the time delay (phase/frequency).
Figure 10.64 shows a block diagram of the acoustic signal tracker. Two 128-point
arrays of data are obtained, cross-correlating the first signal with the second and
then the second signal with the first. The resulting cross-correlation data are decom-
posed into two halves, each transformed using a 128-point FFT. The resulting phase
is the phase difference of the two signals.
This project was implemented on the C30 [17] and can be transported to the
C6713 processor. To test this project, a speaker was positioned a few feet from the
two microphones, which are separated by 1 foot. The speaker receives a 1-kHz signal
from a function generator. A track of the source speaker is plotted over time on
the PC monitor. Plots of the cross-correlation and the magnitude of the cross-
correlation of the two microphone signals were also displayed on the PC monitor.
=
10.23.3 Neural Network for Signal Recognition
The goal of this project is to recognize a signal. The FFT of a signal becomes the
input to a neural network that is trained to recognize the signal using the back-
propagation learning rule.
Design and Implementation
The neural network consists of three layers with a total of 90 nodes: 64 input nodes
in the first layer, 24 nodes in the middle or hidden layer, and 2 output nodes in the
third layer. The 64 points as input to the neural network are obtained by retaining
Search WWH ::
Custom Search