US5168459A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Adaptive filter using continuous cross-correlation
Assignee: Agilent Technologies Inc
•
Filing: 1992-05-12
•
Granted: 1992-12-01
•
Expires: 2012-05-12
Abstract
Abstract Coefficients of an adaptive filter are updated continuously during the filter's normal operation. This is achieved by superimposing on the filter input signal a known noise signal. At the filter's output, a counterpart of this known noise signal is subtracted and the resultant signal is cross correlated with past samples of the noise signal. The filter coefficients are then updated in response to these correlations so as to minimize correlation of the resultant signal with the noise signal.
Inventors
- 👤 Donald R. Hiller
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-24 22:25
CPC Classifications
H; H03; H03H; H03H21/00; H03H21/0012
Claims Analysis
Total Claims: 27
Claim 1
Independent Claim
108 words
A method of tailoring the characteristics of an adaptive filter so that electronic signals filtered thereby are transformed in a desired manner, the method comprising: applying to the adaptive filter a composite signal that includes a data signal and a known pseudo random signal; determining a cross correlation between a filtered signal resulting from said applying step and the known pseudo random signal; and varying coefficients of the adaptive filter based on said cross correlation; wherein the variation of the adaptive filer coefficients based on the cross correlation tailors the characteristics of the adaptive filter so that the data signal filtered thereby is transformed in a desired manner.
Claim 2
Independent Claim
108 words
A method of tailoring the characteristics of an adaptive filter so that electronic signals filtered thereby are transformed in a desired manner, the method comprising: applying to the adaptive filter a composite signal that includes a data signal and a known pseudo random signal; determining a cross correlation between a filtered signal resulting from said applying step and the known pseudo random signal; and varying coefficients of the adaptive filter based on said cross correlation; wherein the variation of the adaptive filer coefficients based on the cross correlation tailors the characteristics of the adaptive filter so that the data signal filtered thereby is transformed in a desired manner.
Claim 3
Independent Claim
104 words
A method of programming an adaptive filter having a plurality of filter coefficients so that electronic signals filtered thereby are transformed in a desired manner, the method comprising: applying to the filter an input data signal; superimposing on the input data signal a known noise signal; removing from a filtered output signal a counterpart to the known noise signal to thereby produce a resultant output signal; detecting artifacts of the known noise signal in the resulting signal; and adapting the filter coefficients to minimize said artifacts, said adapting step occurring continuously during normal operation of the filter rather than in a dedicated calibration cycle.
Claim 4
Independent Claim
104 words
A method of programming an adaptive filter having a plurality of filter coefficients so that electronic signals filtered thereby are transformed in a desired manner, the method comprising: applying to the filter an input data signal; superimposing on the input data signal a known noise signal; removing from a filtered output signal a counterpart to the known noise signal to thereby produce a resultant output signal; detecting artifacts of the known noise signal in the resulting signal; and adapting the filter coefficients to minimize said artifacts, said adapting step occurring continuously during normal operation of the filter rather than in a dedicated calibration cycle.
Claim 5
Dependent Claim
25 words
The method of claim 2 in which the detecting step includes cross correlating the resultant output signal with a counterpart to the known noise signal.
Claim 6
Independent Claim
98 words
A method of filtering with an adaptive filter so as to transform an electrical data signal in a desired manner, the method comprising: superimposing a noise signal on a data signal applied to an input of the adaptive filter; subtracting a counterpart to said noise signal from a signal output from the adaptive filter to produce a resultant signal; determining a correlation between the resultant signal and past samples of the noise signal; and updating coefficients of the adaptive filter in response to said correlation so as to minimize correlation of the resultant signal with the noise signal.
Claim 7
Independent Claim
98 words
A method of filtering with an adaptive filter so as to transform an electrical data signal in a desired manner, the method comprising: superimposing a noise signal on a data signal applied to an input of the adaptive filter; subtracting a counterpart to said noise signal from a signal output from the adaptive filter to produce a resultant signal; determining a correlation between the resultant signal and past samples of the noise signal; and updating coefficients of the adaptive filter in response to said correlation so as to minimize correlation of the resultant signal with the noise signal.
Claim 8
Independent Claim
80 words
A method of calibrating an anti-alias filter so as to reduce undesired responses thereof, the method comprising the steps: filtering an analog data signal with an analog anti-alias filter; converting the filtered signal into digital form; compensating an imperfect response of the analog anti-alias filter by passing the digital signal through a digital compensation filter; and programming the digital compensation filter while filtering the analog data signal, rather than while filtering a known calibration signal in a dedicated calibration cycle.
Claim 9
Independent Claim
80 words
A method of calibrating an anti-alias filter so as to reduce undesired responses thereof, the method comprising the steps: filtering an analog data signal with an analog anti-alias filter; converting the filtered signal into digital form; compensating an imperfect response of the analog anti-alias filter by passing the digital signal through a digital compensation filter; and programming the digital compensation filter while filtering the analog data signal, rather than while filtering a known calibration signal in a dedicated calibration cycle.
Claim 10
Independent Claim
199 words
A method of filtering an electronic signal to transform it in a desired manner, the method comprising the steps: applying an input signal to series coupled first and second filter stages, the first filter state having a response that roughly approximates a desired filter response, the second filter stage having a response that can be programmed by varying coefficients associated therewith; applying a noise signal to a third filter stage to yield a filtered noise signal ,the third filter stage having a response that more nearly approximates the desired filter response than does the first filter stage; applying a first counterpart to said noise signal to the cascaded first and second filter stages and producing an output signal corresponding thereto; subtracting the filtered noise signal from said output signal to yield a resultant signal; determining a correlation between said resultant signal and a second counterpart to said noise signal to produce accumulation signals; and accumulating the accumulation signals to yield the coefficients of the second filter stage; wherein the response of the second filter stage is thereby programmed so that the net response of the cascaded first and second filter stages mimics the response of the third filter stage.
Claim 11
Independent Claim
199 words
A method of filtering an electronic signal to transform it in a desired manner, the method comprising the steps: applying an input signal to series coupled first and second filter stages, the first filter state having a response that roughly approximates a desired filter response, the second filter stage having a response that can be programmed by varying coefficients associated therewith; applying a noise signal to a third filter stage to yield a filtered noise signal ,the third filter stage having a response that more nearly approximates the desired filter response than does the first filter stage; applying a first counterpart to said noise signal to the cascaded first and second filter stages and producing an output signal corresponding thereto; subtracting the filtered noise signal from said output signal to yield a resultant signal; determining a correlation between said resultant signal and a second counterpart to said noise signal to produce accumulation signals; and accumulating the accumulation signals to yield the coefficients of the second filter stage; wherein the response of the second filter stage is thereby programmed so that the net response of the cascaded first and second filter stages mimics the response of the third filter stage.
Claim 12
Dependent Claim
31 words
The method of claim 6 which further includes passing a data signal through the cascaded first and second filter stages simultaneously with passing the first counterpart of the noise signal therethrough.
Claim 13
Dependent Claim
238 words
In a filtering method that includes the steps of: filtering an analog input signal with an analog filter to pass a desired spectrum, said analog filter having an imperfect response across the desired spectrum; converting the filtered analog signal to digital form once each clock period; and filtering the digital signal with an N-tap finite impulse response filter to produce an output signal compensated for the imperfect response of the analog filter, the filtering step including filtering in accordance with N stored filter coefficients; an improved method for updating the filter coefficients of the N-tap finite impulse response filter, the method comprising the steps: providing a digital noise signal that alternates between first and second values; summing an analog counterpart of the noise signal with the analog input signal prior to analog filtering; coupling a first digital counterpart of the noise signal to an N-tap shift register, said register having N outputs through which the first counterpart signal is shifted as the shift register is clocked once each clock period; subtracting a second digital counterpart of the noise signal from the output signal of the N-tap finite impulse response filter to produce a digital difference signal; scaling the digital difference signal; processing the scaled digital difference signal and the N outputs of the N-tap shift register to produce N correction factors corresponding to the N filter coefficients; and updating the filter coefficients with the correction factors corresponding thereto.
Claim 14
Dependent Claim
238 words
In a filtering method that includes the steps of: filtering an analog input signal with an analog filter to pass a desired spectrum, said analog filter having an imperfect response across the desired spectrum; converting the filtered analog signal to digital form once each clock period; and filtering the digital signal with an N-tap finite impulse response filter to produce an output signal compensated for the imperfect response of the analog filter, the filtering step including filtering in accordance with N stored filter coefficients; an improved method for updating the filter coefficients of the N-tap finite impulse response filter, the method comprising the steps: providing a digital noise signal that alternates between first and second values; summing an analog counterpart of the noise signal with the analog input signal prior to analog filtering; coupling a first digital counterpart of the noise signal to an N-tap shift register, said register having N outputs through which the first counterpart signal is shifted as the shift register is clocked once each clock period; subtracting a second digital counterpart of the noise signal from the output signal of the N-tap finite impulse response filter to produce a digital difference signal; scaling the digital difference signal; processing the scaled digital difference signal and the N outputs of the N-tap shift register to produce N correction factors corresponding to the N filter coefficients; and updating the filter coefficients with the correction factors corresponding thereto.
Claim 15
Dependent Claim
26 words
The method of claim 8 in which the first digital counterpart of the noise signal is different than the second digital counterpart of the noise signal.
Claim 16
Dependent Claim
27 words
The method of claim 8 in which the processing step includes cross correlating the scaled digital difference signal with the first digital counterpart of the noise signal.
Claim 17
Dependent Claim
59 words
The method of claim 8 which further includes: subtracting the scaled difference signal from each of the filter coefficients for which the corresponding output of the N-bit shift register has the first value; and summing the scaled difference signal with each of the filter coefficients for which the corresponding output of the N-bit shift register has the second value.
Claim 18
Dependent Claim
19 words
The method of claim 8 in which the updating step is performed less frequently than once each clock period.
Claim 19
Dependent Claim
19 words
The method of claim 8 in which the processing step is performed less frequently than once each clock period.
Claim 20
Dependent Claim
19 words
The method of claim 8 in which the subtracting step is performed less frequently than once each clock period.
Claim 21
Dependent Claim
25 words
The method of claim 8 which further includes generating the second digital counterpart signal by filtering the digital noise signal with a desired filter shape.
Claim 22
Dependent Claim
121 words
The method of claim 8 which further includes: generating a first digital counterpart signal having at least three possible signal values, the sequence of values that define said signal being a function, at least in part, of preceding values of the digital noise signal; shifting the first digital counterpart signal through the shift register; subtracting the scaled difference signal from each of the filter coefficients for which the corresponding output of the shift register has the first value; summing the scaled difference signal with each of the filter coefficients for which the corresponding output of the shift register has the second value; and leaving unaltered the filter coefficients for which the corresponding output of the shift register has the third value.
Claim 23
Dependent Claim
153 words
An adaptive filer for tailoring phase and amplitude response of a circuit, the filter comprising: an N-tap finite impulse response filter whose filtering characteristics are determined by N filter coefficients stored in a memory, the filter being serially coupled with the circuit whose response is to be tailored, the serial combination having an input and an output; a noise source having an output; the output of the noise source being coupled to the input of said serial combination; the output of the noise source and the output of the serial combination being coupled to inputs of a difference circuit, said difference circuit having an output; a cross correlation circuit having a first input coupled to the output of the difference circuit and having a second input coupled to the output of the noise source, the cross correlation circuit having an output coupled to the memory in which the N filter coefficients are stored.
Claim 24
Dependent Claim
153 words
An adaptive filer for tailoring phase and amplitude response of a circuit, the filter comprising: an N-tap finite impulse response filter whose filtering characteristics are determined by N filter coefficients stored in a memory, the filter being serially coupled with the circuit whose response is to be tailored, the serial combination having an input and an output; a noise source having an output; the output of the noise source being coupled to the input of said serial combination; the output of the noise source and the output of the serial combination being coupled to inputs of a difference circuit, said difference circuit having an output; a cross correlation circuit having a first input coupled to the output of the difference circuit and having a second input coupled to the output of the noise source, the cross correlation circuit having an output coupled to the memory in which the N filter coefficients are stored.
Claim 25
Dependent Claim
42 words
The filter of claim 17 which further include a model filter that exhibits a characteristic to be mimicked by the serial combination, said model filter being serially interposed between the output of the noise source and the input of the difference circuit.
Claim 26
Dependent Claim
26 words
The filter of claim 17 which further includes a signal processor serially interposed between the noise source and the second input of the cross correlation circuit.
Claim 27
Dependent Claim
29 words
The filter of claim 17 which further includes a scaling circuit serially interposed between the output of the difference circuit and the first input of the cross correlation circuit.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1995-12-01 | 1996-06-01 | 1996-12-01 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 1999-12-01 | 2000-06-01 | 2000-12-01 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2003-12-01 | 2004-06-01 | 2004-12-01 | $7700.00 | Window Closed / Lapsed |