With its home-page download counter incrementing by the second, much like McDonald's used to boast on its signs about the number of burgers sold, there's little question that Skype is way ahead of the pack in the peer-to-peer VoIP arena. But backers of the SIP protocol are on the brink of releasing what VoIP pioneer Erik Lagerway believes is a vastly more useful and powerful open-source alternative.
While Luxembourg-based Skype's closed-source software currently holds a huge lead in peer-to-peer VoIP, enthusiastic backers of the open source SIP protocol believe that an under-development peer-to-peer SIP approach is poised to catch up with Skype and even surpass it.
“SIP is an open standard and that has a longer term impact than a closed network philosophy,” says Eric Lagerway, co-founder of the softphone company Xten (a company he left earlier this year) and founder and CEO of Vocalscape. “While a closed network can foster innovation it does nothing for the long-term for IP communications as a whole. In the long term you're better off with an open standard and that's what peer-to-peer SIP is.”
By relying on an open source development model similar to that used with the popular Asterisk PBX, Lagerway believes, the sheer amount of engineering and software talent applied to it is significantly greater than the effort going into furhter enhancements of Skype.
“Taking Skype past where it is today will be hard,” Lagerway says, “because there are only so many people working on it while all of the working groups of the IETF (Internet Engineering Task Force) are working on [some aspect of] SIP.”
The benefit of peer-to-peer SIP is simplicity. It eliminates the server from the network by distributing the SIP methodology throughout the network itself.
“Decentralizing allows us to take advantage of a network that can be put together almost instantaneously by way of auto account creation and integrated network address translation and firewall traversal,” he says. “That answers the question people are asking, 'Can peer-to-peer SIP do everything that a Skype-like network can do and still be open standard?' The answer is simply, 'yes.'”
The potential benefits of decentralization are far-reaching, Lagerway said, citing an interview he conducted with Cisco Systems Engineer Cullen Jennings on August 3, 2005 about the IETF's recent ad-hoc meeting peer-to-peer SIP (the full interview, in MP3 format, can be heard at Lagerway's blog, SipThat.com) for a number of examples.
Imagine, Lagerway asks, if we were UN workers “in the middle of the desert and all we had was basically electricity. How are we going to put together a SIP network? And how are we going to connect to Skype? That's just not going to happen,” he said. “With peer-to-peer SIP and some wireless access points, all of a sudden you have a method for IP communications that requires no servers.”
Or consider the possibilities when an event like the recent London terrorist bombing or the September 11 attacks shuts down parts of the communications infrastructure.
“We could put up a wireless access network using peer-to-peer SIP,” says Lagerway. “With auto-provisioning inherent in that model and auto account creation — all of a sudden you have your communications.”
Disaster relief teams aren't the only ones who stand to benefit from peer-to-peer SIP. VoIP consumers also stand to gain.
“In terms of consumer VoIP, you have a lot of these SIP networks evolving,” Lagerway explains. “You have Free World Dialup from Jeff Pulver's efforts and SIPphone from Jeff Robertson's efforts and as far as the eye can see you have these free VoIP networks popping up using SIP. But it costs money to run these networks — it's not free to them, it's free to you.”
Peer-to-peer SIP drives down the cost of operating provider networks because the basic infrastructure exists in the endpoints, and not in expensive centralized servers.
“You wouldn't have all these expensive servers,” says Lagerway. “The endpoint would be empowered to do it on its own. Simply from a cost perspective it would make a lot of sense.”
Security is one area that is more challenging with an open system, but Lagerway says that issue is being addressed.
“There's work going on in the IETF to secure SIP identities, to secure against phishing and spoofing. One standard that has a great deal of promise, which Cullen Jennings is responsible for bringing to the IETF, is the SIP Cert technology where certificates are dished out and have to be recognized at both endpoints.”
Lagerway points out that there are security standards that can be used by SIP — for example, TLS (Transport Layer Security) SRTP (Secure Real-time Transport Protocol and S/MIME (Secure / Multipurpose Internet Mail Extensions).
“I don't think it's going to take that long to incorporate that methodology into a peer-to-peer SIP model,” he says, “where instead of having central servers for dishing out certificates or SIP proxy we can use the methodology of the peer-to-peer model and apply some of these ideas to it.”
So how soon can we expect to see peer-to-peer SIP implementations?
Soon, Lagerway thinks.
“I don't think it's going to be long before you see a peer-to-peer SIP extension to some of the SIP stacks that are out there,” he says. One example Lagerway gives is SIPFoundry's reSIProcate SIP stack which is used by Xten in their SIP softphones.
“We could certainly see some changes to that stack and see it being offered as a peer-to-peer enabled SIP stack, in fairly short order,” he says. “I don't think it's going to take long. I'd say no more than a few months.”