Adaptive equalizer with complex signal regeneration and method of operation

• US 6,301,298 B1
• Filed: 11/03/1998
• Issued: 10/09/2001
• Est. Priority Date: 11/03/1998
• Status: Expired due to Term

• Current Assignee: Tektronix Incorporated (Fortive Corp.)
• Original Assignee: Tektronix Incorporated (Fortive Corp.)
• Inventors: Deshpande, Nikhil, Kuntz, Thomas L.
• Primary Examiner(s): Pham, Chi
• Assistant Examiner(s): Tran, Khai
• Application Number:
  US09/185,416
• Time in Patent Office: 1,071 Days
• Field of Search: 375/232, 375/229, 375/235, 375/233, 375/261, 375/270, 375/268, 375/320, 375/321, 375/350, 348/404, 348/472, 348/667, 708/322, 708/323
• US Class Current: 375/232
• CPC Class Codes: H04L 2025/03382 Single of vestigal sideband
• CPC Class Codes: H04L 2025/03477 not time-recursive
• CPC Class Codes: H04L 2025/03617 Time recursive algorithms H04L2025/03643 takes precedence
• CPC Class Codes: H04L 25/03038 with a non-recursive structure H04L25/03031 takes precedence
• CPC Class Codes: H04N 5/211 Ghost signal cancellation H04N23/81 takes precedence

Abstract

An adaptive equalizer for use in a digital transmission receiver includes Q-channel regeneration. A equalization filter produces complex valued equalized signal samples representative of a digital transmission signal. A slicer produces ideal real component values of the equalized signal samples. A Q-regeneration filter produces ideal imaginary component values of the equalized signal samples from the ideal real component values. The ideal real and imaginary component values are combined with the component values of the equalized signal samples to produce a complex valued error signal. The error signal is fed back to scale update values for updating coefficient values of the equalization filter. Delays in the equalizer provide storage to synchronize various equalizer signal and component values with the Q-regeneration filter output.

Claims

1. An adaptive equalizer with Q-channel regeneration comprising:

a filter system receiving complex valued unequalized signal samples and generating complex valued equalized signal samples;

a slicer receiving real component values of the complex valued equalized signal samples and generating ideal real component values; and

a complex valued error signal generator receiving the ideal real component values and the complex valued equalized signal samples and producing a complex valued error signal with real component values of the error signal being derived from the real component values of the complex valued equalized signal samples and the ideal real component values and imaginary component values of the complex valued error signal being derived from the imaginary component values of the complex valued equalized signal samples and ideal imaginary component values regenerated from the ideal real component values, with the complex valued error signal being combined with time aligned vectors of complex valued unequalized signal samples in the filter system for updating filter coefficient values of the filter system.

2. The adaptive equalizer as recited in claim 1 wherein the filter system further comprises:

a equalization filter having filter coefficient values updatable from initial values for generating the complex valued equalized signal samples from the received complex valued unequalized signal samples;

a delay receiving the complex valued unequalized signal samples and generating delayed vectors of complex valued unequalized signal samples time aligned with the complex valued error signal values;

a multiplier receiving the time aligned vectors of complex valued unequalized signal samples and the complex valued error signal for generating equalization filter coefficient correction values;

an adder receiving the coefficient correction values and current equalization filter coefficient values for generating updated equalization filter coefficient values.

3. The adaptive equalizer as recited in claim 1 wherein the complex valued error signal generation further comprises:

a regeneration filter receiving the ideal real component values and generating ideal imaginary component values;

a first delay receiving the ideal real component values and generating ideal real component values time aligned with the generated ideal imaginary component values;

a second delay receiving the complex valued equalized signal samples and generating complex valued equalized signal samples time aligned with the ideal real and imaginary component values; and

a combiner receiving the time aligned complex valued equalized signal samples and the ideal real and imaginary component values for generating the complex valued error signal from the difference between the time aligned complex valued equalized signal samples and the ideal real and imaginary component values.

4. The adaptive equalizer as recited in claim 3 wherein the regeneration filter is a FIR filter.

5. The adaptive equalizer as recited in claim 4 wherein the FIR filter has filter coefficients that produce imaginary component output values of a complex raised cosine filter.

6. The adaptive equalizer as recited in claim 1 further comprising:

an offset filter receiving the real component values of the error signal and generating an output signal representative of a DC offset in the complex valued unequalized signal samples; and

a combiner receiving the output of the offset filter and real component values of the complex valued unequalized signal samples and generating real component values of the complex valued unequalized signal samples with reduced DC offset from the difference between the complex valued unequalized signal samples and the output of the offset filter.

7. The adaptive equalizer as recited in claim 3 wherein the complex valued error signal generation further comprises a multiplier receiving the complex valued error signal and a variable scaler value for scaling the complex valued error signal.

8. The adaptive equalizer as recited in claim 3 further comprising an equalizer convergence filter receiving the real component values of the error signal and generating an output signal indicative of equalizer convergence.

9. An adaptive equalizer with Q-channel regeneration comprising:

a filter system receiving complex valued unequalized signal samples and generating complex valued equalized signal samples;

a slicer receiving real component values of the complex valued equalized signal samples and generating ideal real component values;

a regeneration filter receiving the ideal real component values and generating ideal imaginary component values;

a first delay receiving the ideal real component values and generating ideal real component values time aligned with the generated ideal imaginary component values;

a second delay receiving the complex valued equalized signal samples and generating complex valued equalized signal samples time aligned with the ideal real and imaginary component values; and

a combiner receiving the time aligned complex valued equalized signal samples and the ideal real and imaginary component values for generating a complex valued error signal from the difference between the time aligned complex valued equalized signal samples and the ideal real and imaginary component values, with the complex valued error signal being combined with time aligned vectors of complex valued unequalized signal samples in the filter system for updating filter coefficient values of the filter system.

10. The adaptive equalizer as recited in claim 9 wherein the filter system further comprises:

a equalization filter receiving filter coefficient values updatable from initial values and generating the complex valued equalized signal samples from the received complex valued unequalized signal samples;

a delay receiving the complex valued unequalized signal samples and generating delayed vectors of complex valued unequalized signal samples time aligned with the complex valued error signal values;

a multiplier receiving the time aligned vectors of complex valued unequalized signal samples and the complex valued error signal for generating equalization filter coefficient correction values;

an adder receiving the coefficient correction values and current equalization filter coefficient values for generating updated equalization filter coefficient values.

11. The adaptive equalizer as recited in claim 9 wherein the regeneration filter is a FIR filter.

12. The adaptive equalizer as recited in claim 11 wherein the FIR filter has filter coefficients that produce imaginary component output values of a complex raised cosine filter.

13. The adaptive equalizer as recited in claim 9 further comprising:

an offset filter receiving the real component values of the error signal and generating an output signal representative of a DC offset in the complex valued unequalized signal samples; and

a combiner receiving the output of the offset filter and real component values of the complex valued unequalized signal samples and generating real component values of the complex valued unequalized signal samples with reduced DC offset from the difference between the complex valued unequalized signal samples and the output of the offset filter.

14. The adaptive equalizer as recited in claim 10 wherein the complex valued error signal generation further comprises a multiplier circuit receiving the complex valued error signal and a scaler value for scaling the complex valued error signal.

15. The adaptive equalizer as recited in claim 14 further comprising an equalizer convergence filter receiving the real component values of the error signal and generating an output signal indicative of equalizer convergence.

16. A method of equalizing complex valued unequalized signal samples comprising the steps of:

a) filtering the complex valued unequalized signal samples using a equalization filter having filter coefficient values updatable from initial values to produce complex valued equalized signal samples;

b) slicing real component values of the complex valued equalized signal samples to produce ideal real component values;

c) generating a complex valued reference signal using the ideal real component values as an input to an imaginary component value regeneration filter and time aligning the ideal real component value with the ideal imaginary component value;

d) generating a complex valued error signal by determining the difference between time aligned complex valued equalized signal samples and the reference signal; and

e) updating the filter coefficients of the equalization filter by combining the complex valued error signal with time aligned vectors of complex valued unequalized signal samples to produce filter coefficient update values that are combined with the current filter coefficients.

17. The method of equalizing complex valued unequalized signal samples as recited in claim 16 wherein the complex valued reference signal generating step further comprises the step of delaying the ideal real component values an amount corresponding to the delay produced by the regeneration filter.

18. The method of equalizing complex valued unequalized signal samples as recited in claim 16 wherein the complex valued error signal generating step further comprises the step of delaying the complex valued equalized signal samples an amount corresponding to the delay produced by the regeneration filter.

19. The method of equalizing complex valued unequalized signal samples as recited in claim 16 wherein the updating step further comprises the step of delaying the vectors of complex valued unequalized signal samples an amount corresponding to the delay produced by the regeneration filter.

20. The method of equalizing complex valued unequalized signal samples as recited in claim 16 further comprising the step of:

a) generating a value from the real component values of the complex valued error signal representative of a DC offset in the complex valued unequalized signal samples; and

b) combining the offset value with the complex valued unequalized signal samples to reduce the DC offset in the complex valued unequalized signal samples prior to filtering.

21. The method of equalizing complex valued unequalized signal samples as recited in claim 16 wherein the complex valued error signal generating step further comprises the step of scaling the complex valued error signal with a variable convergence factor.

22. The method of equalizing complex valued unequalized signal samples as recited in claim 21 further comprising the steps of:

calculating running standard deviation values of the real component values of the complex valued error signal; and

generating an output signal indicative of equalizer convergence.

23. A machine readable medium having stored thereon a series of instructions which, when executed by a processor of a transmission system receiver, equalizes complex valued unequalized signal samples by causing the processor to:

filter the complex valued unequalized signal samples using a equalization filter having filter coefficients values updatable from initial values to produce complex valued equalized signal samples;

slice real component values of the complex valued equalized signal samples to produce ideal real component values;

generate a complex valued reference signal using the ideal real component values as an input to an imaginary component value regeneration filter and time aligning the ideal real component value with the ideal imaginary component value;

generate a complex valued error signal by determining the difference between time aligned complex valued equalized signal samples and the reference signal; and

update the filter coefficients of the equalization filter by combining the complex valued error signal with time aligned vectors of complex valued unequalized signal samples to produce filter coefficient update values that are combined with the current filter coefficients.

24. The machine readable medium as recited in claim 23 further comprising instructions which, when executed by the processor, causes the processor to delay the ideal real component values an amount corresponding to the delay produced by the regeneration filter when generating the complex valued reference signal.

25. The machine readable medium as recited in claim 23 further comprising instructions which, when executed by the processor, causes the processor to delay the complex valued equalized signal samples an amount corresponding to the delay produced by the regeneration filter when generating of the complex valued error signal includes.

26. The machine readable medium as recited in claim 23 further comprising instructions which, when executed by the processor, causes the processor to delay the vectors of complex valued unequalized signal samples an amount corresponding to the delay produced by the regeneration filter when updating the equalization filter coefficient values.

27. The machine readable medium as recited in claim 23 further comprising instructions which, when executed by the processor, causes the processor to generate an offset value from the real component values of the complex valued error signal representative of a DC offset in the complex valued unequalized signal samples, and combine the offset value with the complex valued unequalized signal samples to reduce the DC offset in the complex valued unequalized signal samples prior to filtering.

28. The machine readable medium as recited in claim 23 further comprising instructions which, when executed by the processor, causes the processor to scale the complex valued error signal with a variable convergence factor.

29. The machine readable medium as recited in claim 28 further comprising instructions which, when executed by the processor, causes the processor to calculate running standard deviation values of the real component values of the complex valued error signal, and generate an output signal indicative of equalizer convergence.