Culture Clash: IMS vs. Peer to Peer for VoIP

After many years of promise, Voice over IP is now seeing strong growth in all sectors of the voice communications industry. Upstarts, incumbents, cable providers, and other telephony companies are all embracing VoIP. However, there are two very different ideas about how VoIP and its services will evolve. In the IP Multimedia Subsystem (News - Alert) (IMS), all-IP services are built using a framework that borrows from both the PSTN and IP worlds. By contrast, there is a growing peer-to-peer (P2P) VoIP movement that sees voice as just another IP application, with most of the intelligence provided at the endpoints. This article will compare the two models and consider the resulting impact on the ecosystems and go-to-market strategies for equipment vendors, carriers, and customers.

Back to the Future: IMS
During the last 20 years, it has become apparent to most telephone companies that the path to improved profitability lies not just in providing voice phone calls to consumers, but also various other services. One of the key rationales for the Intelligent Network (IN) was to provide a framework that would enable new services to be rapidly developed and open up such service development to a wide array of developers. The IN uses circuit-switched technology and has had some successes, but never really unleashed the large array of profitable services its creators had envisioned. In addition, even as IN and its embedded SS7 signaling were getting off the ground, the success of applications such as the World Wide Web and e-mail over IP networks caused network architects to reconsider how they should be building future communications networks.

There were lots of arguments in the telephony standards community about which way to go, but eventually, a consensus began to emerge in the wireless community, which was attempting to figure out how to build its next generation (3G) networks. This consensus centered on a few basic ideas:

1. The network should move from circuit-switched to IP technology;

2. The focus of the next generation architecture should be on enabling a rich variety of multimedia services; and

3. The primary protocol that would enable these services should be SIP.

With these directions in mind, the 3GPP (Third Generation Partnership Project) developed a series of specifications that collectively became known as the IP Multimedia Subsystem (IMS). In addition, a related consortium (3GPP2) that has focused on building architectures for CDMA wireless systems, also endorsed the ideas of the IMS.

One of the useful ideas of IMS is its layered architecture (Figure 1).

There are three key layers within IMS, dealing with Service, Control, and Connectivity. A key part of the vision is that the core of the architecture is built solely on IP protocols. However, since there is a need to connect to other networks as well, the Connectivity layer was developed in order to enable users on other networks to communicate with the users of an IMS-enabled network and take advantage of common services.

IMS was developed by the wireless community, which already had an existing second generation network architecture based on the use of SS7 and other PSTN technologies. Hence, a connectivity layer was a must for the third generation wireless network in order to be able to communicate with users of the existing wireless and landline networks. However, other network flavors, such as WiFi (News - Alert), MPLS, and WiMAX, were emerging, so the connectivity layer concept was extended to support virtually any network fabric to which a 3G network might connect.

Once the problem of connecting to a variety of networks was resolved via the connectivity layer, IMS became interesting to service providers outside of the original wireless target market. In particular, landline providers were also looking for a blueprint as they began to move their networks to VoIP. Hence, IMS emerged as the candidate service architecture for both wireless and wireline networks.

A key element of IMS is that it has adopted SIP as its primary protocol. This enables IMS to build upon the already strong momentum of SIP in the application provider community. In fact, IMS encourages further decomposition of services into a combination of software and hardware components. For example, an application server vendor may develop a prepaid calling card application using SIP and then communicate with a Media Server that also uses SIP in order to handle media operations such as collection of digits and verification of pin numbers. With IMS, the application vendor no longer needs to worry about generalized service issues, such as authorizing the use of the network and the generation of related accounting records. All of this can be done using standard elements of the IMS control layer. In addition, existing SIP applications, such as prepaid and unified messaging, can be ported onto the IMS architecture.

For all of these reasons, IMS is quite popular among incumbent vendors in the wireless and wireline communities, and has also gotten the attention of more specialized providers, such as cable MSOs, that offer voice. In addition, IMS is not just about voice, but also is intended to enable a new generation of media-rich services that support other media, such as text, video, and instant messaging. IMS has gotten quite a bit of mind share among providers of hardware and application software. These vendors are rapidly developing their IMS strategies and recruiting a roster of IMS partners. The next few years should be very interesting as both traditional and next generation equipment providers devise their IMS strategies and build alliances with the many companies that are investing in IMS related technology and applications.

Revolution in Process � P2P for VoIP
Even as IMS was being addressed in the standards bodies, another set of companies took a completely different approach to rolling out multimedia communications services over IP. The best known of these companies is Skype (News - Alert). Over the last three years, Skype has created a mostly �free� IP voice service that now has over 100 million registered users.

Skype has used the concept of a P2P network as its starting point, where almost all of the sophistication of the application is built into the endpoints themselves (Figure 2). For example, a user can download the Skype communications application onto a PC and use a headset for speaking and listening. The Skype client also has built-in use of buddy lists and presence concepts, so that a Skype user can check to see if the person he wants to talk with is online before initiating a voice conversation. Skype also borrowed heavily from the original P2P services such as Kazaa and Napster in its marketing approaches, which are viral and rely more on social networking than on traditional advertising channels. On the technical side, these companies had also figured out how to cross the barriers erected by the NAT (Network Address Translation) devices used on most routers. This has been a crucial element in ensuring connectivity among P2P users, who use a wide range of commercial router and IP switching equipment.

Skype is not the only example of a P2P VoIP network, but it is the best known. Other advocates of P2P IP voice include the leading Instant Messaging (IM) companies: AOL, Yahoo!, and Microsoft (News - Alert). All of these companies have added voice capabilities into their IM clients, so users of these networks can communicate via text or voice messaging. One of the controversial aspects of Skype is that it makes use of proprietary communications protocols. By contrast, most of their competitors in the P2P space are using SIP as a standard protocol.

There is a standards movement afoot to further enhance SIP to make it more P2P friendly. A mailing list has been set up as a first step in creating a working group in the Internet Engineering Task Force to gain agreement on SIP specifications for enabling P2P network communication. In the meantime, Skype continues to move forward very aggressively with its own approach. In particular, Skype has courted the application developer community by creating an affiliate program and an application programming interface. This has already resulted in numerous software developers creating their own add-on services for Skype users and the development of a new generation of mobile phones that include support for Skype software clients.

Summary
There are two rather different movements emerging to shape the future of IP communications. Many players in the telecom industry have lined up in force behind IMS and are taking steps to develop this architecture over the next several years. There are already many related requests for proposals (RFPs) from operators and the early days of deployment have also begun, albeit sometimes referred to as �pre-IMS.� Industry groups, such as the Multi-Service Forum, are hosting interoperability events that are designed to foster cooperation between the companies that are building IMS-compliant products and services. Consumers will not see IMS directly, but will be able to gain the benefit of new services, such as video mail and online mobile gaming, as they roll out using the common IMS infrastructure. For application developers, IMS holds the promise that an application can be developed once and then used on several different networks. In addition, its built-in security features and �blue chip� heritage should enable new services based on IMS to be effectively marketed to enterprises that want to take advantage of a hosted service business model.

By contrast, P2P IP communications is a classic disruptive technology, which has a much different starting point than IMS and is, therefore, attractive to a different business ecosystem. P2P is fertile ground for application developers, particularly if they want to leverage the large network of Skype users and become a Skype affiliate. There will also be opportunities to develop new applications based on P2P SIP, but the market for these is at an earlier stage of development. To date, P2P has been mostly limited to PC-to-PC communications and has been primarily used by consumers rather than businesses. Given its roots in file sharing systems, peer-to-peer is perceived to be more of an �outlaw� technology than IMS, which is likely to hamper adoption of P2P in the enterprise.

In summary, the IMS and P2P have excellent momentum as infrastructures upon which to build new IP communications applications and services. IMS is shaping up as the preferred approach among many telecom industry players for deploying an IP service architecture that will enable voice and other multimedia services such IPTV (News - Alert) to run on multiple networks. Peer-to-peer technology for IP communications has shown tremendous growth to date, but has been primarily of interest to PC-based users. As a quickly growing, viral technology, it will continue to challenge the status quo in IP communications. It�s likely to be on the forefront as real-time communications evolve and continue to take advantage of the ever increasing power of the CPU chips that are used in computers and mobile communication devices. IT

James Rafferty is the senior product manager for VoIP at Cantata Technology (News - Alert). For more information, please visit www.cantata.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 [email protected] or by phone at 800-290-5460.