prof. brian l. evans dept. of electrical and computer...
TRANSCRIPT
• Prof. Brian L. Evans• Dept. of Electrical and Computer Engineering• The University of Texas at Austin
EE345S Real-Time Digital Signal Processing Lab Spring 2006
Lecture 14
Digital Pulse Amplitude Modulation
14 - 2
Introduction• Modulate M = 2J discrete messages or J bits of
information into amplitude of signal• If amplitude mapping changes at symbol rate of
fsym, then bit rate is J fsym
• Conventional mapping of discretemessages to M uniformly space amplitudes
• Pulse amplitude modulated (PAM) signal)12( −= idai 2
, ... , 0 ..., , 12
MMi +−=
) ( )(*sym
kk Tktats −= �
∞
−∞=
δ No pulses overlap in time:requires infinite bandwidth
fsym = 1 / Tsym
14 - 3
Pulse Shaping• Infinite bandwidth cannot be sent in practice• Limit bandwidth by pulse shaping
filter with impulse response gT(t)
• L samples per symbol durationn is symbol index
m is sample index in a symbol: m = 0, 1, 2,…, L-1.
( )symTk
k Tktgatssym
)(* −= �∞
−∞=
��
���
� −+=��
���
� + �∞
−∞=symsymsymT
kksymsym kTT
Lm
nTgaTLm
nTssym
*
( ) ( ) ( )mknLgakLmnLgamLnssymsym T
kkT
kk +−=−+=+ ��
∞
−∞=
∞
−∞=
)( *
At each time t, k is indexed over number of
overlapping pulses
At indices n & m, k is indexed over number of overlapping pulses
14 - 4
Pulse Shaping Example
2-PAM with Raised Cosine Pulse Shaping
14 - 5
Pulse Shaping Block Diagram
• Upsampling by L denoted as LOutpus input sample followed by L-1 zerosUpsampling by converts symbol rate to sampling rate
• Pulse shaping (FIR) filter gTsym[m]Fills in zero values generated by upsamplerMultiplies by zero most of time (L-1 out of every L times)
D/A TransmitFilter
akgTsym[m]L
symbol rate
sampling rate
sampling rate
analog analog
14 - 6
Digital Interpolation Example
• Upsampling by 4 (denoted by 4)Output input sample followed by 3 zerosFour times the samples on output as inputIncreases sampling rate by factor of 4
• FIR filter performs interpolationLowpass filter with stopband frequency ωstopband ≤ π / 4For fsampling = 176.4 kHz, ω = π / 4 corresponds to 22.05 kHz
Digital 4x Oversampling Filter
16 bits44.1 kHz
28 bits176.4 kHz4 FIR Filter16 bits
176.4 kHz
1 2
Input to Upsampler by 4
n
0
n’
Output of Upsampler by 4
1 2 3 4 5 6 7 80
1 2
Output of FIR Filter
3 4 5 6 7 8n’
0
14 - 7
Pulse Shaping Filter Bank
• Simplify by avoiding multiplication by zeroSplit the long pulse shaping filter into L short polyphase filters
operating at symbol rategTsym,0[n]
gTsym,1[n]
gTsym,L-1[n]
akD/A Transmit
Filter
s(Ln)
s(Ln+1)
s(Ln+(L-1))
Filter Bank Implementation
D/A TransmitFilter
akgTsym[m]L
symbol rate
sampling rate
sampling rate
analog analog
14 - 8
Pulse Shaping Filter Bank Example• Pulse length 24 samples and L = 4 samples/symbol
• Derivation in Tretter's manual,
• Define mth polyphase filter
• Four six-tap polyphase filters (next slide)
��
���
� −+=��
���
� + �+
−=symsymsymT
n
nkksymsym kTT
Lm
nTgaTLm
nTs 3
2
* 1,...,1 ,0 −= Lm
��
���
� += symsymTmT TLm
nTgngsymsym
][,
[ ]kngaTLm
nTs mT
n
nkksymsym sym
−=��
���
� + �+
−=,
3
2
*
( ) ( )mknLgamnLs T
n
nkk +−=+ �
+
−=
)( 3
2
* Six pulses contribute to each output sample
1,...,1 ,0 −= Lm
14 - 9
Pulse Shaping Filter Bank Example
gTsym,0[n]24 samples
in pulse
x marks samples ofpolyphase
filter
4 samples per symbol
Polyphase filter 0 response is the first sample of the
pulse shape plus every fourth sample after that
Polyphase filter 0 has only one non-zero sample.
14 - 10
Pulse Shaping Filter Bank Example24 samples
in pulse
x marks samples ofpolyphase
filter
4 samples per symbol
Polyphase filter 1 response is the second sample of the
pulse shape plus every fourth sample after that
gTsym,1[n]
14 - 11
Pulse Shaping Filter Bank Example24 samples
in pulse
x marks samples ofpolyphase
filter
4 samples per symbol
Polyphase filter 2 response is the third sample of the
pulse shape plus every fourth sample after that
gTsym,2[n]
14 - 12
Pulse Shaping Filter Bank Example24 samples
in pulse
x marks samples ofpolyphase
filter
4 samples per symbol
Polyphase filter 3 response is the fourth sample of the
pulse shape plus every fourth sample after that
gTsym,3[n]
14 - 13
Intersymbol Interference• Eye diagram is empirical measure of quality of
received signal
• Intersymbol interference (ISI):
• Raised cosine filter has zeroISI when correctly sampled
� �∞
−∞=
∞+
≠−∞= �
��
�
�
���
�
� −+=−=
knk
k
symsymknsymsymk g
kTnTgaagTkTnganx
)0(
)( )0() ( )(
��∞
≠−∞=
∞
≠−∞=
−=−
−≤nkk
sym
nkk
symsym
g
kTgdM
g
kTnTgdMD
,, )0(
)( )1(
)0(
)( )1(
14 - 14
Symbol Clock Recovery• Transmitter and receiver normally have different
crystal oscillators• Critical for receiver to sample at correct time
instances to have max signal power and min ISI• Receiver should try to synchronize with
transmitter clock (symbol frequency and phase)First extract clock information from received signalThen either adjust analog-to-digital converter or interpolate
• Next slides develop adjustment to A/D converter• Also, see Handout M in the reader
Optional
14 - 15
Symbol Clock Recovery• g1(t) is impulse response of LTI composite channel
of pulse shaper, noise-free channel, receive filter)( )()()( 11
*sym
kk kTtgatgtstq −=∗= �
∞
−∞=
� �∞
−∞=
∞
−∞=
−−==k m
symsymmk mTtgkTtgaatqtp )( )( )()( 112
)(
)( )( } {)}({
21
2
11
symk
symk
symm
mk
kTtga
mTtgkTtgaaEtpE
−=
−−=
�
� �∞
−∞=
∞
−∞=
∞
−∞=
Optional
ReceiveB(ω)
Squarer BPFH(ω)
PLLx(t)
q(t) q2(t)
p(t)
z(t)
E{ak am} = a2 δδδδ[k-m]
g1(t) is deterministic
s*(t) is transmitted signal
Periodic with period Tsym
14 - 16
Symbol Clock Recovery• Fourier series representation of E{ p(t) }
• In terms of g1(t) and using Parseval’s relation
• Fourier series representation of E{ z(t) }
tkj
kk
symeptpE )}({ ω�
∞
−∞=
= �−= sym sym
T tkj
symk dtetpE
Tp
0
)}({1 ω
where
Optional
( ) ( ) ( )��∞
∞−
∞
∞−
− −== ωωωωπ
ω dkGGT
adtetg
Ta
p symsym
tjk
symk
sym
11
221
2
2
ReceiveB(ω)
Squarer BPFH(ω)
PLLx(t)
q(t) q2(t)
p(t)
z(t)
( ) ( ) ( ) ( )�∞
∞−
−== ωωωωπ
ωω dkGGTa
kHkHpz symsym
symsymkk 11
2
2
14 - 17
Symbol Clock Recovery• With G1(ωωωω) = X(ωωωω) B(ωωωω)
Choose B(ω) to pass ± ½ωsym � pk = 0 except k = -1, 0, 1
Choose H(ω) to pass ±ωsym � Zk = 0 except k = -1, 1
• B(ωωωω) is lowpass filter with ωωωωpassband = ½ωωωωsym
• H(ωωωω) is bandpass filter with center frequency ωωωωsym
Optional
( ){ } )cos(2 teeeZtzE symtjtj
k
tjkk
symsymsym ωωωω =+== −�
( ) ( ) ( ) ( )�∞
∞−
−== ωωωωπ
ωω dkGGT
akHkHpZ sym
symsymsymkk 11
2
2
ReceiveB(ω)
Squarer BPFH(ω)
PLLx(t)
q(t) q2(t)
p(t)
z(t)