Goin' on a SIP trip
CED
June 2004
By Jeff Baumgartner, Editor
Cable is jumping on board with a technology that’s poised to play a key part in IP-based applications and services
cedmagazine.com
SIP (session initiation protocol), a technology considered more of a hobby in the late 1990s than the start of something great, is quickly coming into its own thanks to small but pioneering companies such as Vonage and its adoption by corporate behemoths like AT&T Corp.
Though SIP has traditionally been tied to voice services, it’s designed to do much more than that. Devices based on the technology are also being made to handle more advanced applications, including video telephony.
Network-agnostic, SIP, at its core, opens up communications between two intelligent end points. The end points also keep all of the communication “state” information, including all data about the call and everything that happens during a given session.
There are plenty of SIP devices out there today, many of which run on cable’s high-speed networks without the operator’s knowledge. But with so much innovation and general activity happening in the SIP arena, what has the cable industry been doing about it?
A lot, it turns out. But a quick look back also reveals that the industry looked into it early on, only to uncover many of the technology’s shortcomings. In fact, smart, SIP-like endpoints were considered during the formative days of PacketCable 1.x as operators and vendors began looking at the different protocols and approaches available.
As evaluations progressed in the late 1990s, two options were considered: a network-centric approach and distributed call signaling. At the time, AT&T Broadband was pushing for a distributed approach that leveraged smart, pre-SIP end points that didn’t particularly care much about the network on which they were riding.
Others argued that the distributed approach would enable the end points to establish calls without the operator’s knowledge, which presents a problem if the operator is thinking about replacing primary line telephony services.
The distributed model also didn’t enable the operator to prioritize and re-route calls should the network become congested or overloaded. On top of that, those with enough smarts could hack those end points or flush the buffers to get free service.
Because of those potential problems, the network-based call signaling (NCS) architecture eventually won out and formed what became PacketCable 1.x., a spec that enables MSOs to identify and authorize the clients on the network.
Though cable has somewhat reversed its original position on SIP and is looking into ways to support it, there’s a general misunderstanding of the technology and how it applies to PacketCable, says Steve Craddock, Comcast Cable’s senior vice president for new media development.
“There’s SIP the architecture, and there’s SIP the protocol,” he says. SIP the architecture, he explains, is what services like Vonage and AT&T’s CallVantage do today–connecting a SIP end point to a SIP end point.
“They will always be a best-effort service. They won’t have security like PacketCable will and QoS like PacketCable will,” Craddock says. “I’ll give Vonage a lot of credit. They’ve done a lot of great stuff on the front-end in terms of features and being able to get in and change your service and do the self-provisioning and so forth. But it’s stuff that any of us can and will do, as well.”
…
Net2Phone, meanwhile, has created an entire platform based on SIP called VoiceLine. “We certainly believe very strongly in the future of SIP, but the major deployments in 2004 will be PacketCable 1.x,” says Bryan Wiener, president of Net2Phone Global Services. “But we’ll be offering both SIP and PacketCable services on a single platform."
[See url for ntop graphic: Net2Phone created a SIP-based architecture that complements its
PacketCable strategy.
VIEW ENLARGED IMAGE]
….
“Will SIP obsolete PacketCable? I don’t know, which is why I think it’s important that we become knowledgeable on both technologies,” concludes Barber of Charter Communications. “There are certain things that PacketCable does that SIP may not be able to achieve in the short-term, but, thus far, SIP is proving itself as a very robust technology.”
(NTOP responded to a Charter RFP re cable telephony deployment a couple of months ago; we're waiting to see what comes of that) |
