Method and apparatus for time-domain equalization in discrete multitone transceivers

• US 6,678,318 B1
• Filed: 01/11/2000
• Issued: 01/13/2004
• Est. Priority Date: 01/11/2000
• Status: Expired due to Term

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Agere Systems Incorporated (Broadcom, Inc.)
• Inventors: Lai, Yhean-Sen
• Primary Examiner(s): Corrielus, Jean B.
• Application Number:
  US09/481,027
• Time in Patent Office: 1,463 Days
• Field of Search: 375/229, 375/232, 375/233, 375/230, 375/234, 375/346, 375/350, 708/300, 708/322, 708/323
• US Class Current: 375/232
• CPC Class Codes: H04L 2025/03414 Multicarrier
• CPC Class Codes: H04L 2025/03477 not time-recursive
• CPC Class Codes: H04L 2025/03605 Block algorithms
• CPC Class Codes: H04L 25/0216 with estimation of channel length
• CPC Class Codes: H04L 25/03038 with a non-recursive structure H04L25/03031 takes precedence

Abstract

The invention is a method and apparatus for determining the coefficients to be used by a time domain equalization circuit in a receiver in a communication system using discrete multitone (DMT) technology. The method and apparatus is particularly efficient in that the order of the channel response equation after channel shortening is less than or equal to the allotted cyclic prefix length utilized in the applicable communication protocol being used. Further, the invention includes an efficient method and apparatus for determining symbol synchronization for the time domain equalization.

Claims

1. A method of generating coefficients for performing time domain equalization on a digital multitone signal comprising a plurality of symbols received from a transmitter over a communication channel, said method comprising the steps of:

(1) synchronizing said received digital multitone signal to a corresponding signal that was transmitted from said transmitter prior to inclusion of any channel response;

(2) determining time domain equalization coefficients, a_opt, for a TEQ filter via

 

(R_xy^TR_xx<sup>−

1</sup>R_xy−

R_yy)a_opt=r₂−

(R_xy^TR_xx<sup>−

1</sup>r₁)

or

2. The method as set forth in claim 1 wherein step (2) comprises solving for a_opt using one of

3. The method as set forth in claim 2 wherein step (2) comprises the steps of:

(2.1) calculating each of V^T and D_inv once;

(2.2) storing V^T and D_inv in a memory; and

(2.3) retrieving V^T and D_inv when needed for determining a_opt.

4. The method as set forth in claim 2 wherein step (2) further comprises:

(2.4) calculating VR_xy once and using that calculated value for each occurrence of VR_xy in the equation.

5. The method as set forth in claim 2 wherein step (2) comprises:

(2.5) solving the equation for a_opt by employing a Cholesky decomposition technique.

6. The method as set forth in claim 1 wherein said method is used in connection with a digital multitone receive architecture that utilizes a cyclic prefix of length L and wherein step (2) provides a coefficient wherein M≦

L, where M is the impulse response length of a shortening channel utilized in said method.

7. The method as set forth in claim 6 wherein L=M.

8. The method as set forth in claim 7 wherein step (2) comprises assuming that an impulse response of a channel of said received digital multitone signal has a transfer function of:

$H<br><br>\left(z\right)=\frac{B<br><br>\left(z\right)}{A<br><br>\left(z\right)}=\frac{b_0+b_1<br><br>z^{-1}+b_2<br><br>z^{-2}+\mathrm{\dots <br><br>}+b_{{\left(L-1\right)}^{z^{-\left(L-1\right)}}}}{1+a_1<br><br>z^{-1}+a_2<br><br>z^{-2}+\mathrm{\dots <br><br>}+a_{{\left(K-1\right)}^{z^{-\left(K-1\right)}}}}.$

9. The method as set forth in claim 6 wherein said time domain equalization is performed in connection with an asynchronous digital subscriber line.

10. The method as set forth in claim 9 wherein said time domain equalization is used in connection with a communications system in accordance with ITU-T G.992.2 specifications.

11. The method of claim 1 wherein step (1) comprises:

(1.1) cross correlating the received signal with the transmitted signal to generate an impulse response of the channel;

(1.2) determining a maximal energy window of said impulse response and a range of said channel response;

(1.3) determining a noise floor of said channel impulse response; and

(1.4) determining a start point of said channel impulse response using said noise floor.

12. The method as set forth in claim 11 wherein step (1.2) comprises passing said impulse response through an energy detector with a window of size L samples.

13. The method as set forth in claim 12 wherein step (2) is performed starting with a sample of said received signal that is L samples before said start point.

14. The method as set forth in claim 11 further comprising the steps of:

(1.5) converting said received signal to frequency domain before step (1.1); and

(1.6) converting said channel impulse response from frequency domain to time domain after step (1.1).

15. A method of recovering a signal that was received after being transmitted in digital multitone with a cyclic prefix, said method comprising the steps of:

(1) time domain equalizing said signal;

(2) removing said cyclic prefix from said signal;

(3) frequency domain equalizing said signal;

wherein step (1) comprises convolving said signal with 1+a(n) wherein a(n) is generated via

16. The method as set forth in claim 15 wherein step (1) comprises solving for a_opt using one of

17. The method as set forth in claim 16 wherein step (1) further comprises:

solving one of said equations for a_opt by employing a Cholesky decomposition technique.

18. The method as set forth in claim 15 wherein said cyclic prefix is of a length, L, and wherein step (1) shortens a channel of said signal by M, where M≦

L.

19. The method as set forth in claim 18 wherein L=M.

20. The method as set forth in claim 19 wherein step (1) further comprises assuming that an impulse response of a channel of said signal has a transfer function of:

$H<br><br>\left(z\right)=\frac{B<br><br>\left(z\right)}{A<br><br>\left(z\right)}=\frac{b_0+b_1<br><br>z^{-1}+b_2<br><br>z^{-2}+\mathrm{\dots <br><br>}+b_{{\left(L-1\right)}^{z^{-\left(L-1\right)}}}}{1+a_1<br><br>z^{-1}+a_2<br><br>z^{-2}+\mathrm{\dots <br><br>}+a_{{\left(K-1\right)}^{z^{-\left(K-1\right)}}}}.$

21. The method as set forth in claim 18 wherein said signal is transmitted via an asynchronous digital subscriber line.

22. The method as set forth in claim 21 wherein said method is used in connection with a communications system in accordance with ITU-T specification G.992.2.

23. The method as set forth in claim 15 further comprising the steps of:

(4) determining a starting point for calculating a_opt.

24. The method of claim 23 wherein step (4) comprises:

(4.1) cross correlating the received signal with the transmitted signal to generate an impulse response of the channel;

(4.2) determining a maximal energy window of said impulse response and a range of said channel response;

(4.3) determining a noise floor of said channel impulse response; and

(4.4) determining a start point of said channel impulse response using said noise floor.

25. The method as set forth in claim 24 wherein step (4.2) comprises passing said impulse response through an energy detector with a window of size L samples.

26. The method as set forth in claim 25 wherein step (1) is performed starting with a sample of said received signal that is L samples before said start point.

27. The method as set forth in claim 24 further comprising the steps of:

(4.5) converting said signal to frequency domain before step (4.1); and

(4.6) converting said channel impulse response from frequency domain to time domain after step (4.1).

28. A receiver for a digital multitone modem comprising:

means for synchronizing a received digital multitone signal to a corresponding signal that was transmitted from a transmitter prior to inclusion of any channel response;

means for determining, for said received signal, time domain equalization coefficients, a_opt, for a TEQ filter via

 

(R_xy^TR_xx<sup>−

1</sup>R_xy−

R_yy)a_opt=r₂−

(R_xy^TR_xx<sup>−

1</sup>r₁)

or

29. The receiver as set forth in claim 28 wherein said means for determining time domain equalization coefficients comprises solving for a_opt using one of

30. The receiver as set forth in claim 29 wherein said means for determining time domain equalization coefficients solves the equation for a_opt by employing a Cholesky decomposition technique.

31. The receiver as set forth in claim 28 wherein said receiver is used in connection with a digital multitone receive architecture that utilizes a cyclic prefix of length L and wherein said means for determining time domain equalization coefficients provides a coefficient wherein M≦

L, where M is the impulse response length of a shortening channel utilized in said method.

32. The receiver as set forth in claim 31 wherein L=M.

33. The receiver as set forth in claim 32 wherein said means for determining time domain equalization coefficients assumes that an impulse response of a channel of said received digital multitone signal has a transfer function of:

$H<br><br>\left(z\right)=\frac{B<br><br>\left(z\right)}{A<br><br>\left(z\right)}=\frac{b_0+b_1<br><br>z^{-1}+b_2<br><br>z^{-2}+\mathrm{\dots <br><br>}+b_{{\left(L-1\right)}^{z^{-\left(L-1\right)}}}}{1+a_1<br><br>z^{-1}+a_2<br><br>z^{-2}+\mathrm{\dots <br><br>}+a_{{\left(K-1\right)}^{z^{-\left(K-1\right)}}}}.$

34. The receiver as set forth in claim 31 wherein said receiver is used in connection with an asynchronous digital subscriber line.

35. The receiver as set forth in claim 34 wherein said receiver is used in connection with a communications system in accordance with ITU-T specification G.992.2.

36. The receiver as set forth claim 28 wherein said means for synchronizing comprises:

means for cross correlating the received signal with the transmitted signal to generate an impulse response of the channel;

means for determining a maximal energy window of said impulse response and a range of said channel response;

means for determining a noise floor of said channel impulse response; and

means for determining a start point of said channel impulse response using said noise floor.

37. The receiver as set forth in claim 36 wherein said means for determining a maximal energy window comprises means for passing said impulse response through an energy detector with a window of size L samples.

38. The receiver as set forth in claim 36 wherein said means for synchronizing comprises:

means for converting said received signal to frequency domain before cross correlating the received signal with the transmitted signal; and

means for converting said channel impulse response from frequency domain to time domain after cross correlating the received signal with the transmitted signal.