US 6,785,384 B2 · Two-step algorithm for training an echo cancellation filter

Abstract

A two-step training method for the estimation filter in the echo cancellation (EC) path of the analog front-end (AFE) circuit for a modem, such as an asymmetric digital subscriber line (ADSL) modem. During the first step, the transmit (TX) path is turned off and a digital TX signal (e.g., a pseudo white gaussian noise signal) is applied to the EC path essentially bypassing the EC estimation filter. During the second step, the EC path is turned off and a digital TX signal (e.g., the same pseudo white gaussian noise signal) is applied to the TX path. Coefficients for the EC estimation filter are then generated based on the digital RX signals recorded during the two training steps. In particular, the recorded digital RX signals are transformed to the frequency domain, where one is divided into the other. The resulting digital signal is then transformed back to the time domain, where it is used to generate the coefficients for the EC estimation filter. For Digital Multi-Tone (DMT) applications, the two-step training method of the present invention should be able to be implemented using at most four DMT symbols.

Claims23

Dependent claims

In a circuit comprising (1) a transmit (TX) path configured to convert a digital TX signal into an analog TX signal;

(2) a receive (RX) path configured to convert an analog RX signal into a digital RX signal;
and (3) an echo cancellation (EC) path configured to generate an analog EC signal, based on the digital TX signal, to be subtracted from the analog RX signal prior to digitization, a method for training an EC estimation filter in the EC path, comprising the steps of:
(a) generating a first digital RX signal by applying a first digital TX signal to the circuit with the TX path turned off and with the EC and RX paths turned on;
(b) generating a second digital RX signal by applying a second digital TX signal to the circuit with the EC path turned off and with the TX and RX paths turned on;
and
(c) generating coefficients for the EC estimation filter based on the first and second digital RX signals.

3 more elements

In a circuit comprising (1) a transmit (TX) path configured to convert a digital TX signal into an analog TX signal;

(2) a receive (RX) path configured to convert an analog RX signal into a digital RX signal;
and (3) an echo cancellation (EC) path configured to generate an analog EC signal, based on the digital TX signal, to be subtracted from the analog RX signal prior to digitization, a method for training an EC estimation filter in the EC path, comprising the steps of:
(a) generating a first digital RX signal by applying a first digital TX signal to the circuit with the TX path turned off;
(b) generating a second digital RX signal by applying a second digital TX signal to the circuit with the EC path turned off, and
(c) generating coefficients for the EC estimation filter based on the first and second digital RX signals, wherein step (c) comprises the steps of:
(1) transforming the first and second digital RX signals to generate first and second digital RX signals in the frequency domain;
(2) applying a division operation between the second digital RX signal in the frequency domain and the first digital RX signal in the frequency domain to generate a third digital signal in the frequency domain;
(3) transforming the third digital signal in the frequency domain to generate a third digital signal in the time domain;
and
(4) generating the coefficients for the EC estimation filter from the third digital signal in the time domain.

7 more elements

Dependencies are inferred from the claim text.

Two-step algorithm for training an echo cancellation filter

Life of this patent

Two-step algorithm for training an echo cancellation filter

Life of this patent

Abstract

Claims23