United Arab Emirates University

1
United Arab Emirates University College of
Engineering Department of electrical
engineering Graduation project 2
Equalization and Cross-Talk Cancellation
Techniques for Digital Subscriber lines (DSL)
Done by Maitha AbdulRhman AlTameeme
200212533Faten Fadel Alshamsi
200202901Mouza Mohammed Neaimy
200209004 Advisor Dr.Imad
Barhumi
2
Introduction

• The aim of this project is to study, simulate and
  implement the different equalization and cross
  talk cancellation techniques using the DSL
  modems. These techniques include time and
  frequency domain algorithms.

3
Single user scenario

• DSL standard.
• Bet Error Rate.
• Water Filling.
• Channel Capacity.

4
Water Filling
The solution for maximization of the bit rate is
called the water-filling, and that can be done
graphically by inverted the curve of channel
signal-to-noise ratios (SNR) to a constant line
by being filled with energy (water).
New Energy
Old Energy
5
Types of Capacity

• Uniform capacity is the maximum number of bits
  transmitted over a parallel set of sub-channels.
• The optimal capacity is the largest number of
  bits that can be transmitted over a parallel set
  of sub-channels

6
Guntt Table
7
Definition of Cross-Talk

• Due to the high frequencies employed in VDSL,
  signals from different telephone lines leak into
  one another through electro-magnetic coupling.
  This effect is referred to as 'cross-talk'.
• In telecommunication the cross-talk is often
  distinguishable as pieces of speech or signaling
  tones leaking from other people's connections.

8
The Cross Talk
CPE of user 1
CPE of user 2
CPE of user L
CO/ONU
9
Types of Cross Talk

• Near-End Cross-Talk (NEXT)

CPE of user 1
CPE of user 2
CPE of user L
CO/ONU
10
Types of Cross Talk

• Far End Cross-Talk (FEXT)

CPE of user 1
CPE of user 2
CPE of user L
CO/ONU
11
Basic multi-channel DSL
n
H
?
Y
Data X
12
Digital Communications System
Decoder
Message Send
Message Source
Encoder
Noise
Modulator
Channel
Demodulator
Receiver
Transmitter
Cross-talk
13
ADSL Transceiver Data Transmission
N/2 subchannels
N real samples
N-IFFT
S/P
Quadrature amplitude modulation (QAM) encoder
add cyclic prefix
P/S
D/A transmit filter
Bits
TRANSMITTER
channel
RECEIVER
N real samples
N/2 subchannels
receive filter A/D
P/S
QAM demod decoder
N-FFT
S/P
remove cyclic prefix
invert channel frequency domain equalizer
14
Zero-Forcing Equalizer

• The Zero-Forcing Equalizer applies the inverse of
  the channel to the received signal, to restore
  the signal before the channel.

15
Zero-Forcing technique
The inverse of the channel response
Channel frequency response
16

• Ideally, the combination of channel and equalizer
  gives a flat frequency response and linear phase.
  

Zero forcing equalizer Inverse of the channel
frequency response
17
In reality, zero-Forcing equalization does not
work in most applications, for the following
reasons

• Even though the channel impulse response has
  finite length, the impulse response of the
  equalizer needs to be infinitely long
• The channel may have zeroes in its frequency
  response that cannot be inverted
• At some frequencies, may be very small. To
  compensate, grows very large.
• As a consequence, any noise added after the
  channel gets boosted by a large factor and
  destroys the overall signal-to-noise ratio.

18

• So, It is not useful for practical applications ,
  but a study case that is useful to explain more
  realistic equalizer concepts.

19
Zero-Forcing technique
Transmitter
Xk, k0,,N-1
xcpk, k-L1,,N-1
Xk, k0,N-1
xk, k0,N-1
Coding baseband modulation
Serial to parallel converter
Cyclic prefix insertion
Parallel to serial converter
Symbols
ZF
IFFT modulator
20
MatLab Result
21
Minimum Mean Squared Error

• It describes the statistical estimator with the
  least possible mean squared error.
• The minimum mean square error (MMSE) is a
  solution to vector pre-coding for frequency
  multi-user scenarios with a multi-user
  transmitter.

22
MMSE technique
Transmitter
Xk, k0,,N-1
xcpk, k-L1,,N-1
Xk, k0,N-1
xk, k0,N-1
Coding baseband modulation
Serial to parallel converter
Cyclic prefix insertion
Parallel to serial converter
Symbols
MMSE
IFFT modulator
23
MatLab Result
24
SVD Technique

• Singular Value Decomposition (SVD).
• SVD is a very powerful set of techniques dealing
  with sets of equations or matrices.
• It is a vector base technique.

25
SVD Technique

• The objective of vectored transmission is to
  eliminate Far End Crosstalk (FEXT ) and Near End
  Crosstalk (NEXT) by treating the channel as a
  multi input / multi output (MIMO).

26
SVD Technique
Transmitter
Xk, k0,,N-1
xcpk, k-L1,,N-1
Xk, k0,N-1
xk, k0,N-1
Coding baseband modulation
Serial to parallel converter
Cyclic prefix insertion
Parallel to serial converter
Symbols
Pre- Coding
IFFT modulator
channel
Yk, k0,,N-1
ycpk, k-L1,,N-1
Yk, k0,N-1
yk, k0,N-1
Estimated
Decoding Demodulation
Cyclic prefix extraction
Serial to parallel converter
Parallel to serial converter
Post- Coding
FFT demodulator
Symbols
Receiver
27
MatLab Result
28
QR Technique

• QR is a technique for cross-talk recommendation
  which attempts to diagonalizable the cross-talk
  channel.
• Its a linear technique.

29
QR Technique
Transmitter
Xk, k0,,N-1
xcpk, k-L1,,N-1
Xk, k0,N-1
xk, k0,N-1
Symbols
Coding baseband modulation
Serial to parallel converter
Cyclic prefix insertion
Parallel to serial converter
IFFT modulator
channel
Yk, k0,,N-1
ycpk, k-L1,,N-1
Yk, k0,N-1
yk, k0,N-1
Estimated
Decoding Demodulation
Cyclic prefix extraction
Serial to parallel converter
Parallel to serial converter
QR
FFT demodulator
Symbols
Receiver
30
MatLab Result
31
Comparison
32
Sound System
Limiting the input
Record
33
Cost estimation

• Two data acquisition
• (4000Dhs,not exist)
• Tow Cables
• (80Dhs, exist)
• One sound card creative
• (425Dhs, exist)

34
Conclusion

• Reading and searching about the cross-talk was
  done by recognizing the types of cross-talk, the
  disadvantages and the different techniques to
  remove the cross-talk.
• The MatLab program was used to simulate different
  techniques in order to cancel the cross-talk such
  as the
• Zero-Forcing technique
• MMSE technique
• SVD technique
• QR technique
• The main objective for graduation project (II)
  has been achieved.

35
References

• www.cross-talk.org
• G. Ginis and J. Cioffi "A Multi-user Precoding
  Scheme achieving Cross-talk Cancellation with
  Application to DSL Systems"Proceedings of the
  34th Asilomar Conference, Pacific Grove, CA, ,
  October 2000pg. 1627-1631
• Wei Yu, J. Cioffi Multi-users Detection in Vector
  Multiple Access Channels using Generalized
  Decision Feedback Equalization"5th International
  Conf. on Signal Processing, World Computer
  Congress, Beijing, August 2000 Proves that a
  version of the DFE is optimal in that it achieves
  the theoretical capacity of the channel

36
Thanks for listening