FW: H.323/VoIP QoS

Malcolm McDonald malcolm at NETSPEAK.COM
Thu Mar 4 15:56:48 EST 1999


I am a bit confused by the charter of this proposed group, but in general, I
am very interested in having a place to discuss generic QoS issues related
to voice over a packet data network.

Are you planning to include the following topics in your proposed group

1. VoIP flow reserving resource (flow based RSVP interacting with QoS aware
switch/router , SLA based diffserv, SLA based MPLS) - will this group be
recommending the model, or will it discuss the pros/cons or
2. switch/router resource mgmt (MPLS, Constrained-Based MPLS, diffserv,
etc.) - ditto
3. Control of the resource allocation (GK/GW/H.323 endpoints) - problems
such as GK granting the bandwidth without having the real control on actual
usage (bandwidth mgmt with collaboration from many components.)
4. speech coding/quality - encourage single LBRV to reduce transcoding.
5. absolute delay control.
6. echo cancellation problem with the existence of large jitter.
7. transport efficiency - concerns about header overhead that triggers the
proposed solution hop2hop header compression (rfc-2508) and the recent work
in MEGACO regarding g.729/udp/rtp/ip/atm & g.729/aal2/atm interworking.


Wen-Ping Ying
AT&T Wireless Services
PO Box 97059
Redmond, WA 98073
Voice 425-702-2725
Fax   425-702-2568
wying at att.com
wen-ping.ying at attws.com

> -----Original Message-----
> From: Joon Maeng [SMTP:Joon_Maeng at VTEL.COM]
> Sent: Wednesday, March 03, 1999 8:40 AM
> Subject:      H.323/VoIP QoS
> Greetings!  My name is Joon Maeng.  I work for VTEL Corp in USA. I've
> talked to
> several companies while I was attending ITU and IETF meetings about
> establishing
> an ad-hoc group to address QoS (Quality of Service) issues in TRANSPORTING
> REAL-TIME TRAFFICS in the packet switched network, especially in the IP
> network.
> I was very encouraged by responses from those companies.  Most of them
> agreed
> that lack of QoS support for real-time traffics  was one of major barriers
> in
> deploying H.323/VoIP products and  the H.323 industry would benefit by
> having a
> QoS activity group.  I would like to ask if your company would be
> interested in
> the activities described in the following:
> Proposed name of the group: H.323/VoIP QoS Activity Group.
> Motivation:  ITU defined a good conference standard, H.323, for the packet
> network.  IETF is actively defining QoS standards for the IP network. The
> deployment of H.323/VoIP products has been slow partially due to lack of
> QoS
> supports in current IP network.  The H.323/VoIP industry shall work
> together to
> access the issues and provide solutions for the users.
> Proposed charter of the group:
> - Assess QoS issues in transporting real-time traffics (calls with
> long-duration
> compared to web-access) in the IP network,
> - Find the solutions to resolve issues,
> - Execute solutions via various processes such as building consensus on
> various
> QoS solutions among participants,
>   submitting recommendations to standardization bodies if needed, and
> facilitating inter-operablity testing between QoS standards and
> application
>   standards,
> - and education,  promotion, etc.
> I would expect the final charter and the structure of the group will be
> determined by the participants and their interests.    I would say that
> everything is open at this time except that the industry shall look at the
> QoS
> issues seriously for real-time traffics.  I'd like to clarify that the
> proposed
> group is not a group to standardize the QoS.  It is a group or a forum
> that is
> interested in helping and promoting the real-time QoS for H.323/VoIP
> industry
> through various methods (to be determined).   I plan to set up an email
> reflector for the initial discussions among participants and decide the
> following steps based on the discussions. Please email me (joon_maeng@
> vtel.com)
> if you and your company would be interested in joining the initial talks
> via
> email reflector.   Your comments and suggestions will be greatly
> appreciated.
> Regards,
> Joon

More information about the sg16-avd mailing list