Introduction to ADSL Modems

Prof. Brian L. Evans Dept. of Electrical and Comp. Eng. The University of Texas at Austin http://signal.ece.utexas.edu
Last modified August 27, 2005 UT graduate students: Mr. Zukang Shen, Mr. Daifeng Wang, Mr. Ian Wong UT Ph.D. graduates: Dr. Gner Arslan (Silicon Labs), Dr. Biao Lu (Schlumberger), Dr. Ming Ding (Bandspeed), Dr. Milos Milosevic (Schlumberger) UT senior design students: Wade Berglund, Jerel Canales, David J. Love, Ketan Mandke, Scott Margo, Esther Resendiz, Jeff Wu Other collaborators: Dr. Lloyd D. Clark (Schlumberger), Prof. C. Richard Johnson, Jr. (Cornell), Prof. Sayfe Kiaei (ASU), Prof. Rick Martin (AFIT), Prof. Marc Moonen (KU Leuven), Dr. Lucio F. C. Pessoa (Motorola), Dr. Arthur J. Redfern (Texas Instruments)

Outline
Broadband Access
Applications Digital Subscriber Line (DSL) Standards

ADSL Modulation Methods

ADSL Transceiver Block Diagram Quadrature Amplitude Modulation Multicarrier Modulation

ADSL Transceiver Design

Inter-symbol Interference Time-Domain Equalization Frequency-Domain Equalization

Conclusion
25-2

Applications of Broadband Access

Residential Application Downstream Upstream Willing to pay Demand rate (kb/s) rate (kb/s) Potential 384 9 High Medium Database Access 384 9 Low High On-line directory; yellow pages 1,500 1,500 High Medium Video Phone 1,500 64 Low Medium Home Shopping 1,500 1,500 Medium Medium Video Games 3,000 384 High Medium Internet 6,000 0 Low High Broadcast Video 24,000 0 High Medium High definition TV
Business Application Downstream Upstream Willing to pay Demand rate (kb/s) rate (kb/s) Potential 384 9 Medium High On-line directory; yellow pages 1,500 9 Medium Low Financial news 1,500 1,500 High Low Video phone 3,000 384 High High Internet 3,000 3,000 High Low Video conference 6,000 1,500 High Medium Remote office 10,000 10,000 Medium Medium LAN interconnection 45,000 45,000 High Low Supercomputing, CAD
25-3

Interne t

DSL Broadband Access

DSLAM
downstream

Central Office

ADSL modem
upstream

ADSL modem

Voice Switch LPF LPF

Customer Premises

PSTN
25-4

DSL Broadband Access Standards

xDSL ISDN T1 HDSL SHDSL Splitterless ADSL Full-Rate ADSL VDSL Meaning Integrated Services Digital Network T-Carrier One (requires two pairs) High-Speed Digital Subscriber Line (requires two pairs) Single Line HDSL Splitterless Asymmetric DSL (G.Lite) Asymmetric DSL (G.DMT) Very High-Speed Digital Subscriber Line (proposed) Applications Internet Access, Voice, Pair Gain (2 channels) 1.544 Mbps Symmetric Business, Internet Service 1.544 Mbps Symmetric Pair Gain (12 channels), Internet Access, T1/E1 replacement 1.544 Mbps Symmetric Same as HDSL except pair gain is 24 channels Up to 1.5 Mbps Downstream Internet Access, Video Up to 512 kbps Upstream Phone Up to 10 Mbps Downstream Internet Access, Video Up to 1 Mbps Upstream Conferencing, Remote LAN Access Up to 22 Mbps Downstream Internet Access, VideoUp to 3 Mbps Upstream on-demand, ATM, Up to 6 Mbps Symmetric Fiber to the Hood
25-5

Data Rate Mode 144 kbps Symmetric

Courtesy of Mr. Shawn McCaslin (National Instruments, Austin, TX)

Spectral Compatibility of xDSL

POTS ISDN ADSL - USA ADSL - Europe HDSL/SHDSL HomePNA
optional

1.1 MHz

VDSL - FDD 1M Frequency (Hz) 10M 100M 12 MHz Mixed

25-6

10k

100k

Upstream

Downstream

ADSL Modem
N/2 subchannels N real samples
Bits quadrature amplitude modulation (QAM) encoder mirror data and N-IFFT add cyclic prefix D/A + transmit filter

00110

S/P

P/S

TRANSMITTER
channel

RECEIVER N/2 subchannels

QAM demod decoder

N real samples
N-FFT and remove mirrored data remove S/P cyclic prefix
time domain equalizer (FIR filter)

P/S

invert channel =
frequency domain equalizer

receive filter + A/D

25-7

Bit Manipulations
Serial-to-parallel converter
110 00110
S/P

Parallel-to-serial converter
110
S/P

00110

00
Bits Words

00
Words Bits

Example of one input bit stream and two output words

Example of two input words and one output bit stream

25-8

Amplitude Modulation by Cosine Function

Multiplication in time is convolution in Fourier domain y ( t ) = f ( t ) cos( 0t )
Y ( ) = 1 F ( ) ( ( + 0 ) + ( 0 ) ) 2

Sifting property of the Dirac delta functional x( t ) ( t ) = ( ) x( t ) d = x( t )

x( t ) ( t t0 ) = ( t0 ) x( t ) d = x( t t0 )

Fourier transform property for modulation by a 1 1 cosine Y ( ) = F ( + ) + F ( )

2
0

25-9

Amplitude Modulation by Cosine Function

Example: y(t) = f(t) cos( 0 t)
f(t) is an ideal lowpass signal Assume 1 << 0 Y( ) is real-valued if F( ) is real-valued
-
1

F( ) 1

1 1 Y ( ) = F ( + 0 ) + F ( 0 ) 2 2
F( + 0) Y( ) - 0 - 0 + 1 0 01 0+
0

F( 0)

Demodulation is modulation then lowpass filtering Similar derivation for modulation with sin( 0t)
25-10

Amplitude Modulation by Sine Function

Multiplication in time is convolution in Fourier domain y ( t ) = f ( t ) sin ( 0t )
1 Y ( ) = F ( ) j ( ( + 0 ) ( 0 ) ) 2

Sifting property of the Dirac delta functional x( t ) ( t ) = ( ) x( t ) d = x( t )

x( t ) ( t t0 ) = ( t0 ) x( t ) d = x( t t0 )

Fourier transform property for modulation by a j j sine Y ( ) = F ( + 0 ) F ( 0 )

2 2
25-11

Amplitude Modulation by Sine Function

Example: y(t) = f(t) sin( 0 t)
f(t) is an ideal lowpass signal Assume 1 << 0 Y( ) is imaginary-valued if - 1 F( ) is real-valued j j Y ( ) = F ( + 0 ) F ( 0 ) 2 2
F( + 0) - 0 - 0 + 1 0 j Y( )
j

F( ) 1

F( 0) 0 00+ 1

-j

-j

Demodulation is modulation then lowpass filtering

25-12

Quadrature Amplitude Modulation (QAM)

Q
Xi

Modulator
I
I Q Lowpass filter Lowpass filter -

cos(2 fct )
TX Bandpass

Bits

00110

Constellation encoder

sin(2 fct ) channel

magnitude

fc

frequency

One carrier Single signal, occupying the whole available bandwidth The symbol rate is the bandwidth of the signal being centered on carrier frequency
25-13

Multicarrier Modulation
Divide broadband channel into narrowband subchannels Discrete Multitone (DMT) modulation
Based on fast Fourier transform (related to Fourier series) Standardized for ADSL every subchannel Proposed for VDSL behaves like QAM
magnitude
channel carrier subchannel

Subchannels are 4.3 kHz wide in ADSL

frequency
25-14

Multicarrier Modulation by Inverse FFT

e j 2 f1t
X1
g(t) x Q

Xi

e
X1

j 2

1 n N

e
X2
g(t)

j 2 f 2 t

j 2

2 n N

Discrete time

X2

e
X N /2
g(t)

j 2 f N / 2 t

e
X N /2

j 2

N /2 n N

g(t) : pulse shaping filter

Xi : ith symbol from encoder

25-15

Multicarrier Modulation in ADSL

Q 00101
QAM

Xi
I X0

N/2 subchannels (carriers)

X1 X2 XN/2 XN/2-1 * X2* X1* N-point Inverse Fast Fourier Transform (IFFT)

x0 x1 x2 N time samples xN-1

25-16

Multicarrier Modulation in ADSL

Inverse FFT
CP N
ADSL

v samples

N samples

downstream upstream 4 32 64 512

CP

s y m b o l (i) copy

CP

s y m b o l ( i+1) copy

D/A + transmit filter

CP: Cyclic Prefix

25-17

Multicarrier Demodulation in ADSL

S/P

~ X0
N/2 subchannels (carriers)

~ X N 21 ~ XN 2 ~* X N 21 ~ X 1*

N-point Fast Fourier Transform (FFT)

~ x0 ~ x1 ~ x

N time samples

~ x N 1
25-18

Inter-symbol Interference (ISI)

2.1 1.7 111 1 1 .7 .4 1 .1 .7

Ideal channel
Impulse response is an impulse Frequency response is flat

*
-1

=
Channel impulse response Received signal

Non-ideal channel causes ISI

Channel memory Magnitude and phase variation

Threshold at zero
11 1 1 1

Received symbol is weighted sum of neighboring symbols

Weights are determined by channel impulse response
25-19

Detected signal

Channel Impulse Response

25-20

Channel Impulse Response

25-21

Cyclic Prefix Helps in Fighting ISI

Provide guard time between successive symbols
No ISI if channel length is shorter than +1 samples

Choose guard time samples to be a copy of the beginning of the symbol cyclic prefix
Cyclic prefix converts linear convolution into circular convolution Need circular convolution so that

Then division by the FFT(channel) can undo channel distortion

CP s y m b o l (i) copy CP s y m b o l ( i+1) copy
25-22

v samples

symbol channel FFT(symbol) x FFT(channel)

N samples

Cyclic Prefix Helps in Fighting ISI

Repeated symbol

cyclic prefix

* =
equal
to be removed

25-23

Combat ISI with Time-Domain Equalizer

Channel length is usually longer than cyclic prefix Use finite impulse response (FIR) filter called a time-domain equalizer to shorten channel impulse response to be no longer than cyclic prefix length
channel

Shortened channel

25-24

Convolution Review
Discrete-time convolution
y[k ] =
m=

Continuous-time convolution
y( t ) =

h[m] x[k m]

h ( ) x ( t ) d

For every k, we compute a new summation

x[k] h[k]
Represented by its impulse response

For every value of t, we compute a new integral

y[k] x(t) h(t)
Represented by its impulse response
25-25

y(t)

Finite Impulse Response (FIR) Filter

Assuming that h[k] is causal and has finite duration from k = 0, , N-1
y[k ] = h[m] x[k m]
m =0 N 1

Block diagram of an implementation (called a finite impulse response filter)

x[k]
h[0]

z-1
h[1]

z-1
h[2]

z-1
h[N-1]

y[k]
25-26

Example Time-Domain Equalizer

nk xk h z-
+

yk

w b

rk
- +

ek

zk

Minimize mean squared error E{ek2} where ek=bk- - hk*wk

Chose length of bk to shorten length of hk*wk

Disadvantages
Does not consider channel capacity Deep notches in equalizer frequency response
25-27

Frequency Domain Equalizer in ADSL

Problem: FFT coefficients (constellation points) have been distorted by the channel. Solution: Use Frequency-domain Equalizer (FEQ) to invert the channel. Implementation: N/2 single-tap filters with complex coefficients.

25-28

Frequency Domain Equalizer in ADSL

Y0 Y1

~ Yi = ci X i
YN/2-1

FEQ

~ X0 ~ X1 ~ X N 21

N/2 subchannels (carriers)

~ Xi
Q

FEQ Yi

0101

QAM Yi decoder

25-29

ADSL Modem
N/2 subchannels N real samples
Bits quadrature amplitude modulation (QAM) encoder mirror data and N-IFFT add cyclic prefix D/A + transmit filter

00110

S/P

P/S

TRANSMITTER
channel

RECEIVER N/2 subchannels

QAM demod decoder

N real samples
N-FFT and remove mirrored data remove S/P cyclic prefix
time domain equalizer (FIR filter)

P/S

invert channel =
frequency domain equalizer

receive filter + A/D

25-30

Crosstalk and Near-End Echo

TX
f

TX
f

H RX TX
f

cable FEXT

Near-End echo RX

NEXT TX H RX cable H Near-End echo RX

f
25-31

ADSL vs. FEXT, NEXT, Near-end Echo

ADSL with Freq. Division Multiplexing - FDM
Near-End Echo filtered out Self-NEXT (NEXT from another ADSL) mostly filtered out FEXT and NEXT (from another type of DSL) are problems

ADSL with overlapped spectrum (Echo Cancelled)

Near-End Echo Eliminated using an echo canceller FEXT, NEXT and self-NEXT are a problem Larger Spectrum available for downstream higher data rate
FDM EC

US

DS f

US

DS f
25-32