Patent · US Expired

Method of encoding a signal

US6546515B1 · kind B1 · utility

18Cited by

1References

2Claims

0Family size

Assignee

ALCATEL · FR

Inventors

Peter Vary · Herzogenaurach, DE
Stefan Heinen · Krefeld, DE
Stefan Bleck · Mettmann, DE

Key dates

Filing date	Feb 16, 2000
Grant date	Apr 8, 2003
Priority date	—
Expiry date	Feb 16, 2020

Classification

Technology area (CPC H)Electricity
CPC primaryH04L1/0052
WIPO fieldDigital communication
WIPO sectorElectrical engineering

Abstract

A method of encoding a signal, particularly a message signal for speech and/or image transmission, in which signal-matched encoding of the signal with M bits is performed using Q quantization levels, with Q<2M, and in which at least part of the remaining bit rate or the entire redundancy of M−Id(Q) bits is used to protect the signal against errors. The error protection uses a nonlinear code, for example. Furthermore, additional, signal-matched code redundancy can be added to an arbitrary source-coded signal. To decode the redundant signals in both cases, parameter estimation is performed at the receiving end using at least part of the redundancy.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.