Digital Signal Processing Reference
In-Depth Information
and the leftmost bit in the path history of this state is output into a variable
dec
_
output
. For every other time instant afterward, this process is repeated and the
leftmost bit of the selected
path
_
history
variable is output to
dec
_
output
. On com-
pleting the decoding algorithm,
dec
_
output
contains the desired decoder output.
A variable named
output
_
table
, lists the output symbols for every input at a
particular state, as shown in the following table:
Output Symbols If:
Current State
Input
=
0
Input
=
1
00
000
111
01
011
100
10
001
110
11
010
101
The soft decision Viterbi algorithm functions in a similar fashion, except that the
metric is computed in a different way. The metric is specified using the Euclidean
distance between the signal points in a signal constellation. In the soft decision algo-
rithm, the output of the encoder is sent in the form of BPSK-modulated symbols,
that is, 0 is sent as
1. Before this distance is found, BPSK
modulation is performed on the possible channel symbol triad. Assume that a
channel symbol triad containing {
a
1,
a
2,
a
3} is received, and the expected input
channel symbol triad is 001. After BPSK modulation, it can be written as {
b
1,
b
2,
b
3}, where
b
1
-
1 and 1 is sent as
+
=-
1,
b
2
=-
1, and
b
3
=+
1. Then, the distance between these two
channel symbols is found using
(
)
+
(
)
+
(
)
distance =
abs b
11
-
a
abs b
2 2
-
a
abs b
33
-
a
10.16 SPEECH SYNTHESIS USING LINEAR PREDICTION OF
SPEECH SIGNALS
Speech synthesis is based on the reproduction of human intelligible speech through
artificial means [42-45]. Examples of speech synthesis technology include
text-to-
speech
systems. The creation of synthetic speech covers a range of processes; and
even though they are often lumped under the general term
text-to-speech
, a lot of
work has been done to generate speech from sequences of speech sounds. This
would be a speech-sound (phoneme) to audio waveform synthesis, rather than going
from text to phonemes (speech sounds) and then to sound. One of the first practi-
cal applications of speech synthesis was a speaking clock. It used optical storage for
phrases and words (noun, verb, etc.), concatenated to form complete sentences. This
led to a series of innovative products such as vocoders, speech toys, and so on.
Search WWH ::
Custom Search