Convergence may be todays buzzword, but to deliver its full potential, it must go far beyond delivering silos of voice, data, and video applications over the same bit-pipe. Instead, new revenue-generating applications and services that blend triple-play media across both fixed and wireless networks must offer more value to end users than simply lower prices on POTS equivalents.
In recent years, the development and increasing deployment of the session initiation protocol (SIP) is helping to lay the track needed to bring this potential to life. And, for cable service providers, SIP interworking with the PacketCable network call signaling (NCS) protocol within the IP multimedia subsystem (IMS) architecture will open up major opportunities for multimedia applications and mobility services.
These new applications and services will blend media and offer users presence-awareness and location-based features across a unified network domain, a common interface across endpoint devices, and mobile transparency whether at home, in the office, or on the go. In turn, these applications and services will provide service providers with new revenue streams and greater competitiveness. They will also help attract and retain customers.
Looking back, CableLabs an interoperability testing body backed by a consortium of multiple system operators (MSOs) developed its PacketCable specs to define a complete architecture for providing primary line telephony services over a data over cable service interface specification (DOCSIS)-qualified infrastructure. The aim was to equal to or better the voice services provided by the public switched telephone network (PSTN). That goal has now been realized by the likes of Cablevision, Comcast, Time Warner Cable, and other MSOs as they begin to deliver high-quality VoIP telephony services on a large scale.
So whats next for cable VoIP and IP applications in general?
In short, the answer is multimedia and mobility. PacketCables longer-term goal has always been to provide subscribers with high-quality, secure, and authenticated multimedia services through PacketCable 1.5, also known as the PacketCable Multimedia (PCMM) specifications. These specifications have occurred coincidently with the maturation of SIP, a distributed signaling protocol between intelligent endpoints that was defined by the Internet Engineering Task Force (IETF) as early as 1999. Although it has been a buzzword of sorts in the past few years, SIPs ubiquity, if not its legitimacy, was assured when Microsoft included it in its 2002 release of Windows XP.
Today, SIP appears not only in PCs but also in phones, PDAs, set-top boxes, and even soft clients on PCs and PDAs that enable telephony, gaming, and other multimedia applications and services between and among SIP endpoints. As a distributed, application-layer signaling protocol, SIP is quickly finding its way into all sorts of communication devices, more and more of which will be connected to DOCSIS networks. This suggests the need for SIP to coexist with the NCS protocol that provides call control and routing within cable networks and through gateways outside to the PSTN.
This coexistence is inevitable. After all, despite SIPs continuing evolution and its many flavors today, nearly all endpoint devices currently in development have SIP in their roadmaps, except those POTS-only phones designed for lowest-cost production and pricing. In other words, SIP is coming and MSO networks need to be ready for it. On the one hand, this coexistence may require additional engineering at the network core, especially in the call-management server (CMS), to enable better SIPNCS interworking and prevent any undermining of QoS, reliability, security, and authentication that DOCSIS and PacketCable have defined.
On the other hand, enabling if not optimizing SIP interworking and compatibility across DOCSIS networks coupled with PCMM and PacketCable 2.0 implementations will help spawn a new generation of feature-rich applications that MSOs can use to generate incremental revenues, profits, and competitive differentiation. For that matter, SIP can also help MSOs bridge the mobility gap in order to better compete with telcos, which now possess a competitive advantage through wireless offerings.
Network Versus Endpoint Control
In the early days of PacketCable, operators defined a common architectural approach for their networks. They faced the same debate telco architects had argued for decades: Should intelligence (e.g., session control and call states) reside in the network or be distributed in smart endpoint devices? At the time, SIP had not yet been defined, but its concepts favored smart endpoints. The problem was that, in the view of those favoring an NCS approach, a distributed model would give operators less control over network traffic and routing. At the same time, they were concerned that calls could be made and services stolen without their knowledge.
The NCS approach won, of course, with the PacketCable specification ensuring QoS, security, and authentication over hybrid fiber/coaxial (HFC) networks. And similar to what SS7 signaling did to enhance POTS by enabling custom local-area signaling service (CLASS) features like caller ID, call return, and call blocking, NCS likewise provided not all CLASS features but those that have continually proven themselves most popular. It is important to note that, aside from these technical concerns, there were economic ones as well. NCS, for example, enabled MSOs to debut VoIP services much faster and more cost-effectively over MTAs/eMTAs compared to the wait and cost of developing and deploying smart endpoints. Analog POTS phones may not have multimedia bells and whistles, but they are everywhere, and they are cheap and cost-effective endpoints for PacketCable networks.
NCS Versus SIP: The Best Of Both Worlds?
While NCS and SIP each has its strengths, together they are highly complementary, made more so by PCMM and PacketCable 2.0 specifications (Table 1).
As protocols go, SIP is among the most straightforward. It is text-based, similar to the World Wide Webs hypertext transfer protocol (HTTP), and the signaling messages generated by its call-processing language (CPL) are akin to the hypertext markup language (HTML) used to create web pages. Structurally, session set-ups, connections, and terminations by endpoints are as straightforward as well, using syntax reflective of interpersonal communications among humans. (Hi, can you chat? Yes? Great. Shall we talk, use pictures, or movies, or all three? All three? Terrific, lets go. OK, all done. Goodbye.)
Currently Vonage and AT&Ts Call Advantage VoIP service offerings are good examples of a SIP-based approach to architecture. Both services use SIP to provide telephony services that ride over any public IP broadband network. The downside to these non-facilities-based SIP services, however, is that they are inherently best-effort, which means QoS, bandwidth, reliability, and security cannot be assured. The QoS mechanisms that manage jitter, delay, and packet loss are simply not available to the overlay SIP service provider.
Unleashing An Exciting Future
Although NCS and PacketCable 1.0 helped operators enter the VoIP business quickly and cost-effectively, voice telephony alone hardly expresses MSOs ultimate potential for new services. That is especially true given cables two-way digital plant and the amount of available programming and content poised for interactive, multimedia/modal applications and services.
The PCMM specification in PacketCable 1.5 and coming specifications of PacketCable 2.0 promise enormous strides in this direction. By adding SIP interworking, the potential for taking the subscriber experience to new levels becomes much greater because SIP enhances the ability of developers to create new applications and services that leverage such IP technologies as instant messaging, presence, IPTV, and video telephony.
Combining these and other technologies including URLs, Java applets, XML calls, and other objects and Internet protocols can result in an almost infinite variety of applications and services, bounded only by the imagination. It is possible to envision a future in which services are created dynamically, adapting to inferences drawn from users behavior patterns, with the killer app being a unique, temporal combination of applications and services that meets a single users need when, where, and however that need appears. These possibilities may include wireless applications too, bridging the mobility gap with fixed-mobile convergence via PacketCable 2.0 or an IMS architecture as specified by the 3GPP working group and being adopted by service providers worldwide (Figure 1).
The interworking of NCS and SIP is inevitable and available now. Fixed-mobile convergence along with the convergence of public and private networks can then enable these capabilities to interwork seamlessly over unified domains supporting SIP-based devices that help to unify the end users experience. The result? Anytime, anywhere information access, entertainment, and communications, together the Holy Grail of the digerati for years now. IT
Mike Clement is director, NGN Networks, Siemens Communications, Inc. For more information please visit the company online at networks.siemens.com.
If you are interested in purchasing reprints of this article (in either print or PDF format), please visit Reprint Management Services online at www.reprintbuyer.com or contact a representative via e-mail at firstname.lastname@example.org or by phone at 800-290-5460.