Blog Post

Here Comes Trouble: The Thin Edge of SIP

Vint Cerf’s Facebook profile includes a picture of him wearing his favorite t-shirt: it reads “IP on Everything.” Cerf co-authored the 1973 paper that led to TCP/IP being used as a means to interconnect previously incompatible computers and networks associated with the ARPANET. Increasingly, Session Initiation Protocol (SIP) is playing a similar role as the common denominator interconnecting diverse communication devices and networks. And although the protocol geeks either love or hate SIP, its rapid adoption makes it impossible to ignore.

Although Microsoft and Cisco offer competing visions of the future of communication, they both support SIP. Skype rose to fame via a proprietary protocol, but Skype utilizes SIP as the means to connect with the telephone network. Several dozen device manufacturers — from Nokia and Philips to Sony and Siemens — offer SIP-enabled devices, and virtually every other consumer electronics company on the planet plans to roll out SIP-enabled devices over the next 12 months. Ten million SIP-enabled phones have sold to enterprise customers. Avaya, Nortel and Siemens may argue over who has the best features, but they all support SIP.

The entry-level price for an SIP telephone fell to $40 in 2007 from $400 in 2002. Chip manufacturers like Texas Instruments and Broadcom already have third-generation functionality in the pipeline. Best Buy et al do not currently carry SIP phones, but web sites dedicated to SIP-enabled products (e.g. arrived in 2007. Hundreds of companies (e.g. Betamax Group) bridge SIP calls to the traditional telephone network. Fring provides free software that turns mobile handsets into SIP clients enabling voice and IM functionality via Wi-Fi and 2G or 3G data plans.

The patent woes of SIP-based Vonage seem to have squelched the stream of SIP VoIP startups for the time being. For some 20 years, the TCP-IP protocol that Vint co-created achieved very little in the way of public awareness until the arrival of Mark Andreessen’s web browser. Cheap telephone calls represent SIP’s thin edge, but SIP still needs its web browser moment.

Solutions exist for the early obstacles encountered by SIP, such as NAT and firewall traversal. Adobe’s plans for integrating SIP into Flash may go a long way toward unleashing more creativity. SIP continues to evolve with peer-to-peer SIP arriving to challenge client-server SIP during 2008. Yet we remain in the horseless carriage phase, in which everything gets framed in terms of the old model. SIP phones do little more than replicate the features and functions of traditional telephones.

In any case, to quote Victor Hugo, “Nothing is as powerful as an idea whose time has come.” In the 100 years between 1876 and the 1980s, the painfully slow pace of innovation associated with wired telephone monopolies meant that a mere 600 million people were able to use the telephone as a means of communication. Over the next 25 years, competition between cellular carriers increased the pace of innovation enough to allow the technology to reach two billion people. Now, an even faster pace of cost performance improvements positions an infocom ecosystem of SIP devices as the solution to bring communication to four billion people. The time has come for SIP.

21 Responses to “Here Comes Trouble: The Thin Edge of SIP”

  1. Thought-provoking post, Dan, though I can’t go so far as to predict that SIP, as we now know it, will be the basis for the advance to 4 billion users. I think we will see SIP continue to evolve, and the future service model for the protocol may look substantially different than our current model. I am particularly interested in the peer-to-peer concepts that are under development. More on my blog at

  2. SIP is (mis-)used by the telco companies to replace existing PSTN with VoIP services, no big innovation here. Exisiting SIP IM and presence implementations are mostly proprietary (e.g. MS OCS) and XMPP/Jabber seems to be the better choice. From the beginning, SIP was flawed by not solving the NAT/firewall traversal problem, and by being totaly unsecure. After 10 years of development, these issues are still not solved (e.g. ICE is still in draft and SIP/SRTP doesn’t work because key exchange isn’t standardized).

    SIP is a 2nd generation realtime communications (RTC) protocol, it is time to work on the next generation of RTC protocols that allow for real innovation and do solve the problems mentioned above. One such initiative was started by ITU-T SG 16 called Advanced Multimedia System (AMS). Have a look at for further information about this project.

  3. Vince,

    The headline and column identify a disconnect between the capabilities of SIP and the implementation delivered to end users. The present manifestation offers cheaper ways of delivering telephony, but the functionality does not qualify as revolutionary from an end user perspective. I am speculating there needs to be a “SIP equivalent of the web browser” in order for SIP to reach its full potential.

  4. I don’t see how your headline makes sense. The ascendance of SIP seems like a good thing… but I am biased. Folks that use Java for application development may want to check out JSR 289, which defines APIs for creating SIP “applications” in Java.

  5. The “Browser Moment” (for me) could not have happened without the pull of the web in the first place. I discovered Netscape only after first hearing about applications on the web like News, Chat and E-mail.

    • Santosh
  6. as maybe the largest sip network, we’ve found the peering and integration opportunities growing rapidly in the past 6 months. it’s not economical to embed any other protocol in your hardware or software app. just like other open codecs and protocols, there’s a tipping point at which open takes over closed.

    for example, all of nokia’s phone come with sip support. in 24 months, every mobile phone on the market will have sip. this is a boon for voice application developers.

    the one aspect of sip not mentioned here is open-ness. many networks are based on sip (mostly because of hardware support), but few are open.

  7. Sean, I agree with you that SIP is being used for a lot of stuff like IM, Conferencing and IMS applications like Presence, Push to Talk etc. However, all these applications have been done without SIP too, so don’t know if these applications will qualify as the “browser moment”.

    SIP is definitely a next generation protocol, and has a lot of market momentum, but there are lots of other realities in the world of communications, that did not apply when the www took over from ftp, telnet, Veronica, Archie etc.

  8. I’m a bit biased, but I believe SIP has already gone beyond just PSTN interconnect. At Microsoft, we have used SIP to enable instant messaging, presence, video conferencing, and even web conferencing. SIP is definitely the protocol to bet on as many companies including Microsoft have already done. Any protocol that stops at just replicating the PSTN is doomed… fortunately SIP is not that kind of protocol.

  9. “TCP-IP protocol [had] little in the way of public awareness until the arrival of [the] web browser…SIP still needs its web browser moment.”

    The difference is, TCP-IP was the protocol behind the browser from day one. There were no “legacy” browsers in 1990. SIP has a much tougher battle in that there are several billion non-SIP telephony devices out there.

    I like SIP, and it has a strong foothold. But we are still early enough in the adoption process that this whole thing could go another way. More here:

  10. Agree, so far SIP has not been used for much beyond what the PSTN. But then the PSTN (including cellular) has grown to >2Billion endpoints. And that is still growing, so don’t see how or why the “horseless” paradigm will change. Maybe with IMS…

  11. Verizon’s patent claims that lead to the Vonage infringement would not hold up in court against a defendant with the financial resources to adequately defend themselves.

    This was a calculated tactical move on Verizon’s part against a wounded enemy. In the long term (Long term meaning 2-5 years), I do not believe these will hold any water. You have not seen them pursuing it beyond Vonage, they have no interest in tangling with a worthy adversary.

    There is a wealth of substantially close prior art out there. The methods and technologies Verizon is laying claim to were in use well before their application.

  12. Great post! I like it very much. Go SIP!

    I love it to squeeze out my phone bill with SIP. Betamax already gave me so many freedays that I virtually don’t have to pay for phone calls anymore. Tringme and also the Voxalot click to call widget on my website let people from all over the world call me for free.

    With callback to my mobile phone and Betamax as SIP provider in the game I can call the world from my cell phone for prices cheaper than local calls with my own mobile network operator. That’s why I use Voxalot’s mobile callback even for local mobile calls.

    All thanks to SIP.