Browse Prior Art Database

Error Control for Voice/Data Packet Network

IP.com Disclosure Number: IPCOM000062621D
Original Publication Date: 1986-Dec-01
Included in the Prior Art Database: 2005-Mar-09
Document File: 5 page(s) / 31K

Publishing Venue

IBM

Related People

Ahmadi, H: AUTHOR [+2]

Abstract

In the recent past, we have witnessed a surge of interest in applying packet switching technology to the transmission of digitized voice traffic. Although this technology has been used primarily as an efficient means of carrying computer communication traffic, recent advances in high speed, high capacity switching fabrics and progress in digitization of analog waveform and speech processing have created great potential for the handling of both voice and data in a common integrated packet switched network. While voice and data packets can use network resources uniformly, they impose dissimilar demands and requirements on the network protocols. One such case is the error control procedure. Data applications require error-free transmission, and most existing standard protocols provide this capability.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 27% of the total text.

Page 1 of 5

Error Control for Voice/Data Packet Network

In the recent past, we have witnessed a surge of interest in applying packet switching technology to the transmission of digitized voice traffic. Although this technology has been used primarily as an efficient means of carrying computer communication traffic, recent advances in high speed, high capacity switching fabrics and progress in digitization of analog waveform and speech processing have created great potential for the handling of both voice and data in a common integrated packet switched network. While voice and data packets can use network resources uniformly, they impose dissimilar demands and requirements on the network protocols. One such case is the error control procedure. Data applications require error-free transmission, and most existing standard protocols provide this capability. In the case of voice packet however, one can argue that an error or even the loss of a packet will not significantly reduce intelligibility since the amount of real time speech included in one packet is not high. Therefore, in this case one could ignore an error or discard an erroneous packet to avoid additional retransmission delay and to, thereby, improve transmission efficiency. In what follows we propose several error control schemes in a mixed voice/data traffic environment. We consider these schemes for both hop-by-hop and end-to-end cases. A novel error control approach, which incorporates two CRCs (cyclic redundancy checks) and was first introduced by Conrads and Kermani [*], is extended to the cases considered here. It should be emphasized that all cases considered here can use either selective repeat or go-back N protocols. In this class of protocols error checking and buffering of transmitted packets for possible retransmissions are performed at each node along communication paths, but different actions are taken depending on the type (voice/data) of the packet. That is, data packet integrity is preserved through retransmissions. However, in the case of voice packets, one could ignore an error (referred to as keep scheme) or discard an erroneous packet (referred to as drop scheme) to avoid the additional delay due to retransmission and improve link transmission efficiency. Thus, the protocol would be that the receiver has to determine if the received packet is a voice or data packet, whether there is an error in the received packet or not. The problem, however, is that if the receiver detects an error, it cannot trust the received information and thus is not capable of finding out whether the received erroneous packet is a voice packet or a data packet. If the packet were a voice packet, the receiver could simply discard the packet or send it towards the destination without requesting a correct copy. However, if the packet were a data packet, the receiver must call for retransmission after discarding the data packet. To overcome the above problem, two methods are proposed so that vo...