Apparatus of a digital echo canceller and method therefor

• US 20020101983A1
• Filed: 03/19/2001
• Published: 08/01/2002
• Est. Priority Date: 12/08/2000
• Status: Active Grant

• Current Assignee: Topic Semiconductor Corporation
• Original Assignee: Topic Semiconductor Corporation
• Inventors: Lee, Claymens
• Granted Patent: US 6,934,385 B2
• Time in Patent Office: Days
• US Class Current: 379/406.01
• CPC Class Codes: H04B 3/23 using a replica of transmitted signal in the time domain, e.g. echo cancellers

Abstract

An apparatus of a digital echo canceller comprises a first-set delay circuits, a selector, a second-set delay circuits, a plurality of multipliers and an adder. The first-set delay circuits are arranged into groups, and each group has N delay circuits. An exhaustive search function is carried out to produce a plurality of energy sums. The selector will select the biggest energy sum to serve as a significant part and then transmit to the outputs of the second-set delay circuits, the multipliers and the adder. An estimated echo signal is produced to cancel the echo signal.

Claims

1. An apparatus of a digital echo canceller that is suitable to use on a full-duplex digital echo transceiver and is used to cancel a produced echo signal, the apparatus comprises:

a plurality of first-set delay circuits installed, wherein each delay circuit has an input and an output, and are all connected in series, and the first input receives an input signal that transmitted from the full-duplex digital transceiver, and the first-set delay circuits are arranged in groups and each group has N delay circuits;

a selector with an input and an output, wherein the input is based on an exhaustive search that chooses to connect to one the outputs of the first delay circuits;

a plurality of second-set delay circuits, wherein each circuit has an input and an output and are all connected in series, and the first input is connected to the output of the selector;

a plurality of multipliers wherein the number of the multipliers is the same as the number of the second-set delay circuits, and the multipliers are connected respectively to the outputs of the second-set delay circuits, and the transmitting signals from the second-set delay circuits are multiplied respectively to correlation coefficients;

an adder that adds the results from the multiplication operation together to produce an estimated echo signal, wherein the estimated echo signal cancels the echo signal.

2. The apparatus of claim 1, wherein the exhaustive search function depends on a largest energy sum that is produced from the outputs of each group of delay circuits to select an output to connect to the selector.

3. The apparatus of claim 2, wherein the energy sum is calculated from an equation

4. The apparatus of claim 3, wherein the value of N_s is set at 20.

5. A method of a digital canceller that is suitable for a full-duplex digital transceiver and is used to cancel an echo signal, with a first receiving end connected to a second receiving end by a cable, and the first receiving end transmits an input signal, the steps of the method comprise:

(a) calculating total number of searching S by using a N_is,max value and a Δ

N value, wherein N_is,max is the largest response number of the insignificant part of the response signal and Δ

N is an incremental number, N_is,max found first;

the value Δ

N is determined from the value of N_is,max, and once these two values are known, the total number of searching S can be calculated;

(b) using an equation N_is(s)=s*Δ

N to search for a starting point of the echo response, wherein Δ

N is assumed to be 10 and the value of s starts from 0, and once the starting point of the echo response N_is(s) is found, the response number of searching N_s can be determined;

the N_s value is used to calculate the energy sum of EC from an energy equation

$E<br><br>\left(s\right)=\underset{i=1}{\overset{N_s}{\sum <br><br>}}<br><br>\text{\hspace{1em}<br><br>}<br><br>{\mathrm{ci}}^2<br><br>\left(s\right),$

wherein E(s) is the energy sum of EC, and Ci(s) is the EC coefficient;

(c) calculating the value of S in step (a) and using the S value as a reference to determine the value of s in order to get s=S, the calculation process has to repeat step (b) until s=S and N_is(s)=N_is,max;

(d) calculating the largest energy sum during each searching is chosen for the significant part;

(e) multiplying the values of the significant part to corresponding correlation coefficients, and the results of this multiplication operation are then added together to give an estimated echo signal that will cancel the echo signal.

6. The method of claim 5, wherein the value of Δ

N is set at 10, and the value of N_s is set at 20.