Process for fast blind equalization of an adaptive equalizer

• US 5,517,213 A
• Filed: 09/29/1994
• Issued: 05/14/1996
• Est. Priority Date: 09/29/1994
• Status: Expired due to Term

• Current Assignee: Thomson Consumer Electronics Incorporated (Vantiva SA)
• Original Assignee: Thomson Consumer Electronics Incorporated (Vantiva SA)
• Inventors: Dsouza, Adolf, Bhatt, Bhavesh B.
• Primary Examiner(s): Chin, Stephn
• Assistant Examiner(s): GHEBRETINSAE, TEMESGHEN
• Application Number:
  US08/315,050
• Time in Patent Office: 593 Days
• Field of Search: 375/232, 375/235, 375/229, 333/28, 364/724.1, 364/724.19, 364/724.2
• US Class Current: 375/232
• CPC Class Codes: H04L 2025/03585 Modifying the length
• CPC Class Codes: H04L 2025/03662 to a fixed value
• CPC Class Codes: H04L 2025/03732 according to the convergence state
• CPC Class Codes: H04L 25/0305 using blind adaptation
• CPC Class Codes: H04N 5/211 Ghost signal cancellation H04N23/81 takes precedence

Abstract

A blind equalization process (200) for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, is disclosed which comprises the following steps. Blind equalization (204) is performed on a subset of the coefficients in the adaptive equalizer. Then, if the coefficients have converged (206), a decision directed adaptive equalization process (220) is initiated.

Claims

1. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer, said subset comprising less than all of said coefficients;

determining if the coefficients of the subset have converged; and

if the coefficients of the subset have converged, then initiating a decision directed adaptive equalization process.

2. The process of claim 1, wherein the decision directed adaptive equalization process comprises the step of executing a least means squares algorithm.

3. The process of claim 1, further including the step of initializing a predetermined plurality of the coefficients to predetermined fixed values before the step of performing blind equalization.

4. The process of claim 1, wherein the adaptive equalizer includes 128 taps.

5. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process;

wherein the step of performing blind equalization comprises the steps of;

receiving a signal representing a received symbol;

determining if the received signal is within a predetermined distance of an ideal signal representing a valid symbol;

incrementing a count if the received signal is within the predetermined distance of the ideal signal;

accumulating information from each symbol in response to the received signal; and

repeating the receiving, determining, incrementing and accumulating steps a fixed number of times.

6. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer, including the steps of

(a) receiving a signal representing a received symbol;

(b) determining if the received signal is within a predetermined distance of an ideal signal representing a valid symbol;

(c) incrementing a count if the received signal is within the predetermined distance of the ideal signal;

(d) accumulating information from each symbol in response to the received signal; and

(e) repeating the receiving, determining, incrementing and accumulating steps a fixed number of times;

adjusting the subset of coefficients in response to the accumulated information;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process.

7. The process of claim 6 further including, after the repeating step, the step of repeating the receiving, determining, incrementing, accumulating and repeating step a predetermined number of times.

8. The process of claim 7 wherein the predetermined number of times is 4.

9. The process of claim 6 wherein the adjusting step comprises the step of executing a block least mean square algorithm.

10. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer, including the steps of

(a) receiving a signal representing a received symbol;

(b) determining if the received signal is within a predetermined distance of an ideal representing a valid symbol;

(c) incrementing a count if the received signal is within the predetermined distance of the ideal signal;

(d) accumulating information from each symbol in response to the received signal; and

(e) repeating the receiving, determining, incrementing and accumulating steps a fixed number of times;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process;

wherein the receiving step comprises the step of receiving a pulse amplitude modulated signal modulated by multilevel digital symbols in the form of a constellation of valid symbols in which rows and columns are spaced apart.

11. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer, including the steps

(a) receiving a signal representing a received symbol

(b) determining if the received signal is within a predetermined distance of an ideal signal representing a valid symbol;

(c) incrementing a count if the received signal is within the predetermined distance of the ideal signal;

(d) accumulating information from each symbol in response to the received signal; and

(e) repeating the receiving, determining, incrementing and accumulating steps a fixed number of times;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process;

wherein the repeating step repeats the receiving, determining, incrementing and accumulating steps 512 times.

12. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer, including the steps of

(a) receiving a signal representing a received symbol;

(b) determining if the received signal is within a predetermined distance of an ideal signal representing a valid symbol;

(c) incrementing a count if the received signal is within the predetermined distance of the ideal signal;

(d) accumulating information from each symbol in response to the received signal; and

(e) repeating the receiving, determining, incrementing and accumulating steps a fixed number of times;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process;

wherein the step of determining if the coefficients have converged comprises the steps of;

comparing the count to a fixed predetermined threshold value; and

if the count is greater than the fixed predetermined threshold value, then determining that the coefficients have converged.

13. The process of claim 12 wherein the fixed predetermined threshold value is substantially 100.

14. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

initializing a predetermined plurality of the coefficients to predetermined fixed values;

performing blind equalization on a subset of the coefficients in the adaptive equalizer;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process;

wherein the adaptive equalizer includes 128 taps symmetrically arranged about a middle point, and the predetermined plurality of coefficients comprises coefficients corresponding to the eight middle taps.

15. The process of claim 14 wherein the predetermined fixed values are real values of substantially:

0.5 for the two middle taps;

0.25 for the next two surrounding taps;

0.125 for the next two surrounding taps; and

0.0625 for the next two surrounding taps.

16. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer;

determining if the coefficients have converged; and

if the coefficients have converged, then initiating a decision directed adaptive equalization process;

wherein

the adaptive equalizer includes taps symmetrically arranged around a center point; and

the step of performing blind equalization includes the step of selecting the subset of the coefficients corresponding to taps symmetrically arranged around the center point.

17. The process of claim 16 wherein the step of performing blind equalization includes the step of further selecting the subset of the coefficients corresponding to taps adjacent the center point.

18. The process of claim 16 wherein the adaptive equalizer has 128 taps and the subset of the coefficients correspond to the 32 middle taps.

19. A blind equalization process for an adaptive equalizer including a plurality of taps, each having a corresponding coefficient, comprising the steps of:

performing blind equalization on a subset of the coefficients in the adaptive equalizer;

determining if the coefficients have converged;

if the coefficients have converged, then initiating a decision directed adaptive equalization process; and

if the coefficients have not converged, then repeating the step of performing blind equalization using a larger subset of coefficients.

20. The process of claim 19, wherein:

a maximum amount of time is allocated for blind convergence;

the repeating step terminates after the expiration of the maximum amount of time; and

the blind equalization process restarts after the expiration of the maximum amount of time.

21. The process of claim 19, wherein:

the adaptive equalizer includes 128 taps;

the step of performing blind equalization includes the step of selecting 32 taps as the subset of coefficients; and

the repeating step successively uses subsets of 64, and 128 taps.