The patent badge is an abbreviated version of the USPTO patent document. The patent badge does contain a link to the full patent document.

The patent badge is an abbreviated version of the USPTO patent document. The patent badge covers the following: Patent number, Date patent was issued, Date patent was filed, Title of the patent, Applicant, Inventor, Assignee, Attorney firm, Primary examiner, Assistant examiner, CPCs, and Abstract. The patent badge does contain a link to the full patent document (in Adobe Acrobat format, aka pdf). To download or print any patent click here.

Date of Patent:
Jun. 09, 1998

Filed:

Sep. 29, 1995
Applicant:
Inventors:

Shawn Robert McCaslin, Austin, TX (US);

Nariankadu Datatreya Hemkumar, Austin, TX (US);

Bheeshmar Redheendran, Austin, TX (US);

Assignee:

Crystal Semiconductor Corp., Austin, TX (US);

Attorneys:
Primary Examiner:
Assistant Examiner:
Int. Cl.
CPC ...
H04M / ;
U.S. Cl.
CPC ...
379389 ; 379390 ; 379409 ; 379410 ; 379411 ; 370286 ; 370291 ;
Abstract

A double-talk detector for an echo canceller includes power estimators (60) and (62) which are utilized to measure the ERLE value in a calculator (64). This ERLE is stored in a register (70) when it is the largest value generated. This register (70) is updated whenever a new and better ERLE occurs. A fraction of the value in register (70) is utilized as an input to a comparator (88), and then compared to the current ERIE value. If the current ERLE differs from the SERLE in register (70) an inhibit signal is generated for blocking the updates of an adaptive filter (40). The value stored in the register (70) is periodically decremented to reduce the value thereof. This decrement operation is performed in response to detection of an utterance from the far-end. A half-Duplex operation is provided with two attentuators (352) and (354) to provide a switching operation and allow only one side access to the communication path. Speech, or the occurrence of an utterance is monitored at each end by comparing the raw input power at each end to a scaled up background noise power for that end. Different scaling factors for speech detection, or utterance detection, are provided, there being a high scaling factor and a low scaling factor. The side with the initially detected utterance will have access to the system and have the lowest scaling factor associated therewith, the other having the highest scaling factor associated therewith.


Find Patent Forward Citations

Loading…