[standards-jig] Re:H.323 vs SIP for Jabber

Jean-Louis Seguineau/EXC/TEC jean-louis.seguineau at antepo.com
Fri Jan 24 21:26:45 UTC 2003


Hi,

Allow me to somehow temper the comparizon you have made between H323 and
SIP. They were intended to both be signaling protocols, but H323 has got
things a little bit mixed up as it also standardize the transport.

The H.323 standard allow transfer of multimedia streams over packet
networks. It is a descriptive standard and applies only to multimedia
communication, which include voice. It runs over a large number of packet
infrastructure which include among others IP, Frame Relay or ATM.
It was originally developed as an adaptation of H.320, which
addresses videoconferencing over ISDN and other circuit switched networks
and services. Since H.320 was ratified, H.323 has evolved beyond a logical
and necessary extension of the H.320 standard to include Corporate Intranets
and packet-switched networks generally. H.323 utilizes the Real-Time
Protocol (RTP/RTCP) from the IETF, along with internationally standardized
codecs. Since version 2, H.323 is also being used for video and other
communications, over the Internet.

In common with the other ITU multimedia teleconferencing standard, H.323
applies to multipoint and point-to-point sessions. H323 is only a
description standard that include a large number of other specialized
standards, such as

H.225.0      Specifies messages for call control including signaling,
registration
and admissions, and packetization/ synchronization of media streams
H.245  Specifies messages for opening and closing channels for media
streams,
and other commands, requests and indications.
H.450.x     Series of Suplimentary service recommendations. Defines
signalling and
procedures used to provide these telephony-like services
H.235     Defines the security framework used to provide authentication,
encryption and integrity to H.323 systems
H.332     Provides large scale, or loosely-coupled conferencing based upon
H.323.
H.261     Video codec for audiovisual services at P x 64 Kbps.
H.263     Specifies a new video codec for video over POTS.
G.711     Audio codec, 3.1 KHz at 48, 56, and 64 Kbps (normal telephony).
G.722     Audio Codec, 7 KHz at 48, 56, and 64 Kbps.
G.723     Audio Codec, for 5.3 and 6.3 Kbps modes
G.728     Audio Codec, 3.1 KHz at 16 Kbps.
G.729     Audio Codec, 8 kbps audio codec.

So comming back to your statement
> H.323
>  Pros:
>  Very Generalized And Fexible
>  Allows for Voice, Video and many other possible facets
>  Specifies Multipoint Control Unit protocols for Bridging or =
> Conferencing
>  Specifies interfacing with many other carrier methods outside of TCP/IP
>  Large Industry Support

H323 is as you said blotted, not flexible at all, as everybody can plug
anything (messy does not mean flexible :) In addition, its implementations
are often proprietary, and the interconection with different carriers
somewhat ectic. The industry support comes from the fact that many
manufacturer made the same mistake...

On the othe hand, SIP as it names implies was originaly intended to
establish communication session between to party. It does not provides any
transport mechanism for an undelying communication media. It just allow the
party to agree on a communication. But the actual media transport will need
to be carried out by a dedicated transport such as RTP, or any streaming
media protocol.
SIP is only handling the call establishement part in a VOIP call, the call
itself runs on a separate stream. In short, SIP never transport any VOIP by
itself but negotiate the establishement of a RTP session to carry VOIP.

This is where SIP is better that H323.

I would therefore rephrase a little your statement
> SIP
>  Pros:
>  Very Specific in that it was designed to directly address Voice over IP
>  Maximizes available bandwidth by keeping its messages lean and to the =
> point
>  Smaller System Footprint - More focused allowing for less code required =
> in its libraries
>  Light Weight - can process more call ser second than H.323
>  extensible
>
>  Cons:
>  Limited by its very nature to only voice communications
> =20

SIP is flexible as it can address many signaling needs beyond VOIP. It is
efficient, but it does not transport the voice that need to use another
streaming protocol (No VOIP packets are ever transported in SIP packet)
It has q wide acceptance in the industry, and there are a large number of
good qulity libraries around. But once again, there is no free VOIP
streaming libraries available.

But maybe we could come back to the underlying reason for the comparizon. It
is impossible to actually use Jabber XMPP to transport VOIP, due to several
performance issues that will arise(network latency, no guaranted quality of
service, etc...)
But it is possible to use XMPP as a way to negotiate a VOIP session between
two voice enabled client. It would be for example possible to use the
jabber:iq:oob with a specific syntax, then the client would have to start
some streaming protocol between themselves to actually carry the VOIP call.

Hope it helps clarify the subject

--jean-louis

----- Original Message -----
> Message: 2
> From: "Vapor" <vapor at 66oc.org>
> To: <standards-jig at jabber.org>
> Date: Thu, 23 Jan 2003 22:19:13 -0600
> Subject: [standards-jig] H.323 vs SIP for Jabber
> Reply-To: standards-jig at jabber.org
>
> This is a multi-part message in MIME format.
>
> ------=_NextPart_000_0059_01C2C32D.75C42A50
> Content-Type: text/plain;
> charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
>
> I have been doing some research on VoIP and as far as I can see, it all =
> seems to come down to H.323 vs SIP.  H.323 I believe has entered version =
> 4 if I am correct and I am not exactly familiar with the differences =
> between each version but as far as I can see SIP seems to fit with =
> Jabber more that H.323.
>
> Here is a brief Comparison I have made followed by a link to a pdf that =
> I found very useful in identifying the differences between the two.
>
> H.323
>  Pros:
>  Very Generalized And Fexible
>  Allows for Voice, Video and many other possible facets
>  Specifies Multipoint Control Unit protocols for Bridging or =
> Conferencing
>  Specifies interfacing with many other carrier methods outside of TCP/IP
>  Large Industry Support
>
>  Cons:
>  Bloated - Desires to Address Everything, including little obscure =
> protocols that may have no Relevance to Jabber
>  Noisy - Can cause lots of uncessesary traffic on a network
>  Large System Footprint - Because of its extreme generalization, it =
> requires rather large libraries to implement it properly.
>  Overly Complex
>  Requires gatekeepers to manage calls
>  Difficult to troubleshoot due to its complexity
>
>
> SIP
>  Pros:
>  Very Specific in that it was designed to directly address Voice over IP
>  Maximizes available bandwidth by keeping its messages lean and to the =
> point
>  Smaller System Footprint - More focused allowing for less code required =
> in its libraries
>  Light Weight - can process more call ser second than H.323
>  extensible
>
>  Cons:
>  Limited by its very nature to only voice communications
> =20
> http://www.sipcenter.com/files/Wind_River_SIP_H323.pdf
>
> I am still researching but so far it looks to me that if Jabber were to =
> adopt a single standard for clients to use for VoIP integration that SIP =
> should probably be it.  And with the development of XMPP over SIP, this =
> makes the choice even more logical. =20
>
> Again, I am not proposing that we should integrate SIP and VoIP into the =
> Jabber server, but if the Jabber Software Foundation were to adopt a =
> single VoIP protocol as its recommended standard for implementing voice, =
> that VoIP actually begin to creep its way into more clients than just =
> Enigma 3.
>
> I am still looking for some suitable open source implementations of SIP =
> servers to recommend to Jabber server administators that want to offer =
> SIP and I do thing there may need to be some way for the Jabber server =
> to inform that client where the SIP server is.  This may just need to be =
> a transport or component of some sort.
>
> Vapor
> ------=_NextPart_000_0059_01C2C32D.75C42A50
> Content-Type: text/html;
> charset="iso-8859-1"
> Content-Transfer-Encoding: quoted-printable
>
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
> <HTML><HEAD>
> <META http-equiv=3DContent-Type content=3D"text/html; =
> charset=3Diso-8859-1">
> <META content=3D"MSHTML 6.00.2800.1126" name=3DGENERATOR>
> <STYLE></STYLE>
> </HEAD>
> <BODY bgColor=3D#ffffff>
> <DIV><FONT face=3DArial size=3D2>I have been doing some research on VoIP =
> and as far=20
> as I can see, it all seems to come down to H.323 vs SIP.  H.323 I =
> believe=20
> has entered version 4 if I am correct and I am not exactly familiar with =
> the=20
> differences between each version but as far as I can see SIP seems to =
> fit with=20
> Jabber more that H.323.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT> </DIV>
> <DIV><FONT face=3DArial size=3D2>Here is a brief Comparison I have made =
> followed by=20
> a link to a pdf that I found very useful in identifying the differences =
> between=20
> the two.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT> </DIV>
> <DIV><FONT face=3DArial size=3D2>H.323<BR> Pros:<BR> Very =
> Generalized And=20
> Fexible<BR> Allows for Voice, Video and many other possible=20
> facets<BR> Specifies Multipoint Control Unit protocols for Bridging =
> or=20
> Conferencing<BR> Specifies interfacing with many other carrier =
> methods=20
> outside of TCP/IP<BR> Large Industry Support</FONT></DIV>
> <DIV> </DIV>
> <DIV><FONT face=3DArial size=3D2> Cons:<BR> Bloated - Desires =
> to Address=20
> Everything, including little obscure protocols that may have no =
> Relevance to=20
> Jabber<BR> Noisy - Can cause lots of uncessesary traffic on a=20
> network<BR> Large System Footprint - Because of its extreme =
> generalization,=20
> it requires rather large libraries to implement it =
> properly.<BR> Overly=20
> Complex<BR> Requires gatekeepers to manage calls<BR> Difficult =
> to=20
> troubleshoot due to its complexity</FONT></DIV>
> <DIV> </DIV><FONT face=3DArial size=3D2>
> <DIV><BR>SIP</DIV>
> <DIV> Pros:<BR> Very Specific in that it was designed to =
> directly=20
> address Voice over IP<BR> Maximizes available bandwidth by keeping =
> its=20
> messages lean and to the point<BR> Smaller System Footprint - More =
> focused=20
> allowing for less code required in its libraries<BR> Light Weight - =
> can=20
> process more call ser second than H.323<BR> extensible</DIV>
> <DIV> </DIV>
> <DIV> Cons:<BR> Limited by its very nature to only voice=20
> communications<BR> </FONT></DIV>
> <DIV><FONT face=3DArial size=3D2><A=20
> href=3D"http://www.sipcenter.com/files/Wind_River_SIP_H323.pdf">http://ww=
> w.sipcenter.com/files/Wind_River_SIP_H323.pdf</A></FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT> </DIV>
> <DIV><FONT face=3DArial size=3D2>I am still researching but so far it =
> looks to me=20
> that if Jabber were to adopt a single standard for clients to use for =
> VoIP=20
> integration that SIP should probably be it.  And with the =
> development of=20
> XMPP over SIP, this makes the choice even more logical.  =
> </FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT> </DIV>
> <DIV><FONT face=3DArial size=3D2>Again, I am not proposing that we =
> should integrate=20
> SIP and VoIP into the Jabber server, but if the Jabber Software =
> Foundation were=20
> to adopt a single VoIP protocol as its recommended standard for =
> implementing=20
> voice, that VoIP actually begin to creep its way into more clients than =
> just=20
> Enigma 3.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT> </DIV>
> <DIV><FONT face=3DArial size=3D2>I am still looking for some suitable =
> open source=20
> implementations of SIP servers to recommend to Jabber server =
> administators that=20
> want to offer SIP and I do thing there may need to be some way for the =
> Jabber=20
> server to inform that client where the SIP server is.  This may =
> just need=20
> to be a transport or component of some sort.</FONT></DIV>
> <DIV><FONT face=3DArial size=3D2></FONT> </DIV>
> <DIV><FONT face=3DArial size=3D2>Vapor</FONT></DIV></BODY></HTML>
>
> ------=_NextPart_000_0059_01C2C32D.75C42A50--
>
>
>
> --__--__--
>
> Message: 3
> Date: Fri, 24 Jan 2003 08:33:36 +0100
> From: Jacek Konieczny <jajcus at bnet.pl>
> To: standards-jig at jabber.org
> Subject: Re: [standards-jig] H.323 vs SIP for Jabber
> Reply-To: standards-jig at jabber.org
>
> On Thu, Jan 23, 2003 at 10:19:13PM -0600, Vapor wrote:
> > SIP
> >  Pros:
> >  Very Specific in that it was designed to directly address Voice over IP
> >  Maximizes available bandwidth by keeping its messages lean and to the
point
> >  Smaller System Footprint - More focused allowing for less code required
in its
> > libraries
> >  Light Weight - can process more call ser second than H.323
> >  extensible
> >
> >  Cons:
> >  Limited by its very nature to only voice communications
> This is not true. SIP can be used to initialize any session (but mostly
> multimedia sessions), and even to carry any not-session-related messages
> (eg. SIMPLE which is IM over SIP). I think SIP is much more generic
> then H323. But the most important thing is that SIP is much simpler than
> H323.
>
> Greets,
> Jacek
>
>
> --__--__--
>
> _______________________________________________
> Standards-JIG mailing list
> Standards-JIG at jabber.org
> http://mailman.jabber.org/listinfo/standards-jig
>
>
> End of Standards-JIG Digest
>





More information about the Standards mailing list