DSP Applications and Student Projects - Digital Signal Processing and Applications with the C6713 and C6416 DSK - page 397

is used. For BPSK, an input value is segmented into sixteen 1-bit components; for
QPSK, it is fractioned into 8 dibits. This is achieved by masking the input with the
appropriate values, 0x0001, and 0x0003, respectively. In order to obtain the next
segment to be processed, the previous input data is shifted once for BPSK or twice
for QPSK.
Following the extraction of segments, values are assigned to sinusoids with cor-
responding phases. In BPSK, there are only two phases: 0° and 180° for bits 0 and
1, respectively. However, for QPSK, we need four phases (0°, 90°, 180°, and 270°)
corresponding to 00, 01, 11, and 10. This mapping is used in accordance with gray
encoding. This minimizes the error caused by interference during the transmission
of the signal by maximizing the distance between symbols with the most different
bits on the constellation diagram. Each input sample is represented with 16 bits.
Every sampled data contains 16 segments for BPSK, and 8 segments for QPSK.
Since each symbol is transmitted by a sinusoid generated digitally by four points,
an input sample is acquired every 64 and 32 output samples for BPSK and QPSK,
respectively.
At the PSK receiver, each sinusoid is mapped into the corresponding symbols
composed of 1 bit for BPSK or 2 bits for QPSK. The extracted symbols are then
aligned in the newly constructed 16-bit value by appropriate left shifts. The sample
is then sent to the codec, and the original waveform is regenerated.
Implementation Results
The necessary files are in the folder modulation _ schemes . The C source file
modulation_scheme.c contains all the schemes for both modulation and
demodulation, and a gel file to select the specific case. The 10 cases implement the
4-, 8-, and 16-PAM, BPSK, and QPSK for both modulation and demodulation. For
example, the slider in positions 1 and 2 implements the four-PAM scheme for
modulation and demodulation, respectively.
PAM
Three PAM modulation and demodulation schemes are implemented, based on a
lookup table and level assignment. The demodulation process is designed on the
same DSK, with the output of the modulator fed into the input of the demodula-
tor. The modulation output for each PAM scheme is obtained using a 1.3-kHz
sinusoid as input, with the output to a scope. For the four-PAM scheme, the output
is shown in Figure 10.37 a . The four levels are labeled to indicate the modulation
process. The 2's complement format of the codec reverses the negative and positive
values. For example, -0x8000 is shown as the most positive value. Figure 10.37 b
shows the modulation levels for the eight-PAM output with the same sinusoidal
input. Figure 10.37 c shows the output of the 16-PAM modulator, where 12 of the 16
levels are present. This describes the effect of increasing the number of levels. The
spacing between levels is smaller than in the other two PAM schemes. The higher
the number of levels, the harder it is to distinguish and demodulate the signal.