Voice over Packet Protocols
VoIP and VoATM (VoAAL1, VoAAL2)
(continued)

Other Bandwidth Considerations

VoAAL1 provides the closest emulation of the legacy TDM telephony transport systems. It also exhibits the strengths and weaknesses of a TDM system. VoAAL2 and VoIP are different. Both offer various improvements especially in the areas of bandwidth management and efficiency. The more generic and real longer-term improvement offered by VoIP and VoAAL2 is in terms of bandwidth management. In both cases bandwidth is no longer wasted when channels are not in use, contrary to TDM and VoAAL1. Remember that in the case of TDM trunks (for example T1/E1) or VoAAL1 when some channels are not used or active, the bandwidth is still reserved and wasted. This is no longer the case with VoAAL2 or VoIP.

The other bandwidth utilization improvement that VoAAL2 and VoIP bring is due to their support for low bit rate coders and silence suppression. Silence suppression is a technique that has true lasting potential to save half the bandwidth with virtually no cost or quality penalties if the background noise insertion schemes to replace the silence deliver high enough quality. Low bit rate coders can improve bandwidth utilization substantially, but the jury is still out as to whether these bandwidth savings are worth the extra delay incurred. In situations where bandwidth is extremely constrained or at a high premium (for example wireless) the use of low-bit rate coders is justified and indeed necessary. In other situations where bandwidth is readily available (for example wireline), the cost and especially the quality penalty (excessive delays) that low bit rate coders may introduce in telecom networks do not warrant their generalized and widespread use.

Looking Forward

Although VoAAL2 and VoIP are both superior to VoAAL1 in terms of bandwidth management and efficiency, VoAAL2 is likely to be constrained to areas of application where ATM is already deployed. An example of this can be seen in a VoDSL network, where DSLAMs were mostly designed around a core ATM technology.  Similarly, VoAAL2 has also been deployed in applications such as UMTS (3G cellular) radio access networks. In this particular application, VoAAL2 was sufficiently superior to VoIP at the time of specifications to justify its use, given its slightly higher bandwidth efficiency, QoS capabilities, and trunking capabilities.

VoIP may offer improvements on bandwidth efficiency and management, but the voice impairment issues discussed in this article - excessive delays - must be addressed with corrective and preventive actions. The use of generalized echo cancellation to compensate for the extra packetization and processing delays that create the echo problem is a particular corrective action that is needed in order to offer users the same voice quality that is available now in legacy telephony systems.

Limiting the use of low bit rate coders (G.723, G.728, G.729 and the like) is a preventative action necessary to avoid the generation of excessively long transmission delays (larger than 150-200 milliseconds) that create impairments for which there is no technical cure. This limited use of low bit rate coders will result in their deployment being generally constrained to network segments where bandwidth is scarce or at a premium (ex. airwave in the cellular market) or to situations where the end-to-end network topology is under stringent network engineering controls.

While VoAAL2 implementations currently outpace the shipments of VoIP implementations, this is not an indication of what will become the dominant protocol in the future. It is evident that over time the VoIP deployment rate will outpace and overtake its VoATM counterparts and will eventually dominate and own the entire VoP market space.

Octasic designs voice processors that are optimized to perform key functions such as echo cancellation, VQE and VoIP/VoATM packetization. For more information about Octasic, visit www.octasic.com

Michel Laurence co-founded and co-financed Octasic Inc. in 1998 and has served as Chief Executive Officer and Chairman of the Board of Directors since it's inception. In 1996, Mr. Laurence co-founded InnoMediaLogic (IML), which was later bought by NMS Communications. Mr Laurence has held a variety of professional positions, and also co-founded two other companies, IML Research in 1989 and Ovalsys International in 1991. Both Companies were involved in the development of intercept equipment for law enforcement agencies.

Doug Morrissey is Vice President and Chief Technology Officer at Octasic and has over 9 years of experience in the definition and marketing of semiconductor devices. Joining Octasic in 1999, Mr. Morrissey strategically focuses on issues with regard to the technical evolution of future Octasic products within the Voice over Packet market. Prior to joining Octasic, Mr. Morrissey worked as Marketing Manager for ATM and DSL products for Agere (formerly Lucent Technologies, Microelectronics Group). Previously to that, he was Senior Systems Architect at Unisys Corporation. 

Nigel Harney is Senior Product Marketing Manager, responsible for Octasic's packet processing products. He was previously with Ericsson, where he worked on wireline and wireless systems.

<< Previous page

Page 4 of 4