×

TMCnet
ITEXPO begins in:   New Coverage :  Asterisk  |  Fax Software  |  SIP Phones  |  Small Cells
 

Feature Article
February 2002


Placing Your Chips On SIP?
SIP And H.323 Are Battling For Supremacy... And The Winner Is?

BY SHAI MOHABAN

As voice over IP has matured into the next-generation telephony technology and begins the second phase of its evolution, service providers and carriers are continuing to migrate to packet-based networks. The foundation of this new packet infrastructure is the IP telephony signaling and call control function for which two standards are competing for dominance: H.323 and Session Initiation Protocol (SIP).

Although VoIP was born in the 1990�s based on H.323, and H.323 persists as the dominant standard, examples of SIP�s increasing momentum abound, as evidenced by Level 3�s decision to support SIP exclusively, and by Microsoft�s support of SIP in Windows XP. So which standard will emerge as the protocol of choice for VoIP? A comparison of the two technologies and a look at how standards adoption played out in wireless telephony will provide insight into SIP�s potential to succeed H.323 as the dominant standard for VoIP and its likely rate of adoption.

Weighing In: H.323 Vs. SIP
Both SIP and H.323 define mechanisms for call routing, call signaling, capabilities exchange, media control, and supplementary services. H.323 is the international standard and current market leader for IP telephony and is responsible for a majority of the billions of VoIP minutes transmitted around the world today (estimated at 10 billion minutes or 6 percent of all international call traffic in 2001, according to a recent report by TeleGeography). H.323 is often referred to as an �umbrella� specification; that is, it provides recommendations for other specifications regarding control (H.245), video codecs (H.261, 263), audio codecs (G.711, 723, 729, etc.), and multimedia protocols (T.120 series). An International Telecommunication Union (ITU) standard, H.323 is the established VoIP protocol and has been widely used because, until recently, it was the only available, comprehensive standard for VoIP communications.

SIP is a new Internet Engineering Task Force (IETF) signaling protocol that promises flexibility, scalability, and ease of implementation when building complex systems. It was developed for establishing real-time communications (e.g., voice calls, conferences, instant messaging, gaming, and more). SIP is a traditional text-based Internet protocol and, in fact, resembles HTTP and SMTP. SIP is independent of the packet layer and has been designed as a general-purpose protocol. Today, there is limited commercial VoIP SIP traffic, although that is destined to change over the next several years due to major endorsements by developers, manufacturers, and carriers.

When considering the benefits of H.323 in relation to SIP, it is important to compare the technologies in terms of functionality, quality of service (QoS), scalability, flexibility, interoperability, and ease of implementation (see chart on next page). As far as functionality and QoS support, H.323 and SIP are very similar. However, H.323 is currently slightly better in supporting supplementary services because H.323 is a broad package of protocols. But due to its extensibility and flexibility, as it matures, SIP will be able to easily exceed the functionality of H.323 and support services that would not be possible with a monolithic standard like H.323.

The SIP Advantage
SIP�s primary advantages are in its extensibility and flexibility for adding new features, and its relative ease of implementation and debugging. This is at least part of the reason that industry heavyweights Microsoft, Cisco, Level 3, WorldCom, AOL, and others are weighing in on the side of SIP. An example of an up-and-coming feature much more easily enabled by SIP than H.323 is personal mobility -- the ability to reach a called party under a single, location independent address even when the user changes devices. Extensions of SIP are now available to allow third-party signaling enabling applications like �click-to-dial services.� SIP can also be used for signaling Internet real-time fax delivery and many other applications like instant messaging, application sharing, and gaming.

The second key advantage of SIP, its ease of implementation, also significantly differentiates the standard from H.323. While VoIP, in general, offers clear cost advantages to carriers and enterprises, development and implementation costs make up a significant part of the VoIP cost model. SIP�s simplified coding, whereby SIP offers an HTTP-like protocol, can save time and money on VoIP deployments, not to mention enable faster time to market and to new services. Moreover, the relative monolithic design of H.323 makes component updates difficult and expensive and debugging binary extensions is clearly more difficult than debugging an HTTP-like code of SIP.

The Cellular Adoption Model -- Will History Repeat Itself?
Looking at another technology adoption model -- mobile phone standardization in the U.S. -- may provide insight into what lies ahead for VoIP standards adoption and evolution. Mobile phones began the transition from first-generation analog (AMPS) service to second-generation digital service in the early 1990�s. At that time, there were three primary competing technologies: TDMA, GSM, and CDMA. Both TDMA and GSM are time division multiple access technologies whereas CDMA is a code division multiple access technology. The CDMA proponents touted the technology as superior because of its ability to support a greater number of calls-per-unit of bandwidth. And giving credence to this claim is the fact that, a decade later, all third-generation mobile services are designed based on CDMA technology. However, despite CDMA�s technical superiority, it was difficult to distinguish the hype from the facts surrounding the various technologies. Consequently, the European community of nations decided to standardize on GSM. In the U.S., there was no standardization and instead of adopting CDMA or GSM, TDMA initially gained most of the market share due to its head start in the market.

By 1997 about 31 percent of world subscribers were GSM. So one would expect that the U.S. market would migrate to GSM in order to be compatible with much of the rest of the world�s digital mobile phones. However, by 2000 only 7 percent of U.S. subscribers were GSM. Industry observers predict that it will not be until 2007 that GSM covers almost one-third (31 percent) of U.S. mobile subscribers. And in that year, based on third-generation (3G) deployment, CDMA technology will have already become the market share leader in the U.S. with approximately a 43 percent share.

What can we extrapolate from these percentages? In the case of network deployments, even logical shifts in technology can take a long time. Once it became clear that most of the rest of the world was adopting GSM one would expect the U.S. to quickly follow, yet the reality is that the shift will take a decade. And in the case of CDMA, despite its technical advantages over TDMA and GSM, it will take almost a decade and a half for CDMA to overtake the early deployment advantages of the other digital technologies.

And The Winner Is?
The cellular adoption model clearly points to a couple of compelling conclusions. First, initial technologies die hard. So, despite SIP�s obvious advantages, to say that H.323 will be dead and buried in the next several years would be erroneous indeed. Second, both protocols, along with potentially one or two others, will have to co-exist and inter-work for quite some time. As with the cellular example, this could be quite a lengthy period of co-existence. However, it is also clear from the cellular model that a new protocol can clearly overtake the incumbent and inevitably wrestle market share away. Certainly, the benefits of SIP point to this in the case of VoIP. In fact, one can claim that SIP will have a greater rate of success in grabbing market share than was seen with GSM or CDMA because SIP has the technology advantages associated with CDMA as well as the market design win momentum that was associated with GSM.

H.323 has been the initial, standard protocol that has been instrumental in launching and building a strong foundation for the VoIP industry. But SIP is poised to take the torch and continue the momentum. SIP has all of the necessary ingredients to be the successor protocol and is gaining visible market momentum to prevail over H.323 in the near future (2-3 years). Most development and investment by the VoIP industry is SIP oriented. Microsoft, as mentioned earlier, could enable hundreds of millions of SIP users in the future, and with the adoption of SIP by 3GPP, the wireless world has now endorsed SIP as the signaling solution.

At the same time, it is equally apparent that as equipment manufacturers and service providers consider their strategies to purchase and manufacture VoIP products based on standards consistent with those adopted by customers and competitors, they will find themselves with no alternative but to support multiple protocols. As new missing pieces to the packet voice evolution emerge, they will need to be protocol-agnostic. While the ultimate winner in the standards contest will be SIP, it appears that H.323 support will continue to be part of the VoIP family for quite some time.

Shai Mohaban is co-founder and chief technology officer of Kagoor Networks. Kagoor Networks is the new leader in VoIP Traffic Management and is committed to helping carriers and service providers overcome the technical and business roadblocks that have tended to inhibit the widespread, commercial deployment of VoIP services. Please visit their Web site at www.kagoor.com.

[ Return To The February 2002 Table Of Contents ]



Today @ TMC
Upcoming Events
ITEXPO West 2012
October 2- 5, 2012
The Austin Convention Center
Austin, Texas
MSPWorld
The World's Premier Managed Services and Cloud Computing Event
Click for Dates and Locations
Mobility Tech Conference & Expo
October 3- 5, 2012
The Austin Convention Center
Austin, Texas
Cloud Communications Summit
October 3- 5, 2012
The Austin Convention Center
Austin, Texas