VIEWS: 660 PAGES: 7

• pg 1
```									EE345S Real-Time Digital Signal Processing Lab   Fall 2009

Prof. Brian L. Evans
Dept. of Electrical and Computer Engineering
The University of Texas at Austin

Lecture 16
Outline
• Introduction

• Carrier detection

• Symbol clock recovery

• Demodulation

• Conclusion

16 - 2
Introduction
• Channel has linear distortion, additive noise, and
nonlinear distortion
linear time-invariant distortion in channel
Channel equalizer coefficients adapted during modem startup
At startup, transmitter sends known PN training sequence
Channel equalizer would be placed right after A/D converter
• Receiver uses two matched filters to maximize
peak pulse signal-to-noise ratio (see lecture 13)
Minimizes symbol errors for a single-carrier system
Lowpass filters because pulse shapes are lowpass      16 - 3
AGC                Carrier Detect
I(m)            I(n)

r0(t)              r1(t)   r(t)          r(m)                         X   LPF             L
A/D           Equalizer
Filter                                                                 Q(m)            Q(n)
X   LPF             L
Symbol
Clock               90o
Recovery
L samples/symbol
• Automatic gain control
Increase gain when received signal level is low
Lowpass filters are matched filters & extract baseband                       16 - 4
Carrier Detection
• If receiver is not currently receiving a signal, then
it listens for known training sequence
• Detect energy of received signal
p[m]  c p[m  1]  (1  c) r 2 [m]        Transfer function?
c is a constant where 0 < c < 1
• Check if received energy is larger than threshold
• If receiver is currently receiving signal, then it
detects when transmission has stopped
Check whether it is smaller than a smaller threshold
16 - 5
Symbol Clock Recovery
• Two single-pole bandpass filters in parallel
One tuned to upper Nyquist frequency u = c + 0.5 sym
Other tuned to lower Nyquist frequency l = c – 0.5 sym
Bandwidth is B/2 (100 Hz for 2400 baud modem)            Pole
• A recovery method                                     locations?
Multiply upper bandpass filter output with conjugate of lower
bandpass filter output and take the imaginary value
Sample at symbol rate to estimate timing error     See Reader
v[n]  sin( sym  )   sym  when  sym   1 handout M
Smooth timing error estimate to compute phase advancement
p[n]   p[n  1]   v[n]        Lowpass
16 - 6
IIR filter
• QAM transmit signal x(t )  a(t ) cos(c t )  b(t ) sin(c t )
• QAM demodulation by modulation then filtering
Construct in-phase i(t) and quadrature q(t) signals
Lowpass filter them to obtain baseband signals a(t) and b(t)
i (t )  2 x(t ) cos( ct )  2a (t ) cos 2 (c t )  2b(t ) sin(c t ) cos(c t )
 a(t )  a(t ) cos(2ct )  b(t ) sin(2ct )
baseband       high frequency component centered at 2 c
q(t )  2 x(t ) sin(c t )  2a(t ) cos(c t ) sin(c t )  2b(t ) sin 2 (c t )
 b(t )  a(t ) sin(2ct )  b(t ) cos(2ct )
baseband     high frequency component centered at 2 c
1                                                     1
cos 2   (1  cos 2 )     2 cos sin   sin 2     sin 2   (1  cos 2 )   16 - 7
2                                                     2

```
To top