Browse Prior Art Database

The Augmentation of Differential Values in Case of Intentional Loss

IP.com Disclosure Number: IPCOM000008506D
Original Publication Date: 2002-Jun-18
Included in the Prior Art Database: 2002-Jun-18
Document File: 3 page(s) / 159K

Publishing Venue

Motorola

Related People

Baruh Hason: AUTHOR

Abstract

A method to avoid the error propagation in case of intentional loss is described. By intentional loss, the replacement of the original data by some other data is meant.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 50% of the total text.

The Augmentation of Differential Values in Case of Intentional Loss

by Baruh Hason

Abstract

A method to avoid the error propagation in case of intentional loss is described.  By intentional loss, the replacement of the original data by some other data is meant.

Introduction

In digital wireless communication, the voice is compressed in units of packets and transmitted using the transmission services provided by the digital wireless communication system. The receiver decompresses the received digital information and presents it to the loudspeaker after proper time alignment. Any loss of information packet causes gaps in the received audio.  If the compressed audio carries differential information, any loss of packet causes cumulative error, which propagates to the coming packets. Exact correction is achieved only after the reception of absolute, non-deferential, values for these parameters. There are different methods for mitigating the audio gaps caused due to packet loss. Each one of the mitigation methods degrades the audio quality when compared to the audio stream without loss. The use of any one of these mitigation methods is inevitable when the packet loss is unpredictable by the transmitter.

If the loss of the packet is not due to the communication channel errors but the transmitter initiates it, the error recovery method can be designed to share the contribution between both the transmitter and the receiver. Consequently the error is limited to the time gap created by the loss and it does not propagate to the coming packets.

When the recovery method for intentional packet loss is applied at the transmitter together with complimentary action at the receiver, the audio quality degradation will be not only lower but also limited in time.

Problem To Be Solved

When the receiver detects erroneously received data and decides that this data shouldn’t be used, it uses different mitigation methods to fill in the gap created by the information loss. The mitigation methods either repeat the last information of the instead of the lost one or try to predict the actual value by interpolation/extrapolation methods. When the transmitted data is differential, since the data replacement in the receiver is not exact, the error propagates to the continuation of the data. A method to limit the error propagation is inserting absolute data from time to time instead of the differential data. In any case a loss of a single packet causes errors either up to the end of transmission or until the reception of absolute information. The method described below avoids the error propagation in case of intentional loss.

Proposed Solution to the Problem

The known technologies do not take into consideration that the loss of the differential information may be intentional at the transmitter side, and therefore the available solutions treat the intentional losses as if they are unpredictable losses.

            When the transmitter initiates the loss of original data,...