January 2005

Voice over Broadband: The Real Revolution In Personal Communications

BY Ari Rabban, Vice President for Corporate Development and Marketing, VocalTec Communications

While Voice over IP (VoIP) technology has been deployed for several years in many carriers’ backbone networks, only in the past year has it captured widespread public attention. This is because carrier VoIP was completely transparent to the end user and did not change the way people used the phone. In the next few years, with the growth of voice over broadband (VoBB) services, this is likely to change. VoBB is creating a paradigm shift in the way people conduct their personal communications and represents the first true change in end user behavior since the inception of VoIP. Service providers can leverage this fundamental change to offer new services that will increase revenues and market share.

This article explains how the VoBB model differs from the Class5 softswitches that preceded it and, in essence, breaks the century-old Class5 centralized switching model. It also examines the key considerations service providers must take into account when deploying VoBB services.

Background
VoIP was launched almost 10 years ago with the introduction of consumer Internet telephony products based on the integration of phone capabilities into computers, enabling “free” PC-to-PC voice communications over the Internet. The enormous public interest and high expectations for consumer VoIP software never materialized then, mainly due to quality and bandwidth issues.

About one year later, the first telephony gateway appeared, enabling PC-to-phone and phone-to-phone calling. Increased gateway capacities, together with routing and billing capabilities, allowed for the first commercial VoIP network deployments, chiefly for prepaid calling card and international wholesale services. In these deployments, only the trunking ports were VoIP (the endpoints were PSTN). In 2003, VoIP networks carried approximately 18 percent of the world’s international traffic and are expected to account for 63 percent of this market by 2008 (Probe Research).

All of the above, while critical for the evolution and development of VoIP, did nothing to alter people’s communications behavior — in fact in the majority of cases people didn’t even realize their calls were traversing IP networks. They did enjoy much lower rates for international calls using their “black” phones, in part due to carriers’ use of VoIP.

Emergence Of The Softswitch
One of the industry’s more nebulous terms, “softswitch” is most commonly used to describe the concept of separating call control from media processing, rather than referring to a physical entity (although many vendors use the term this way). When the first “softswitches” emerged in the late 1990’s, there was a clear distinction between Class4 (Trunking) and Class5 (Access) softswitches. However, soon afterward, even this distinction started to blur. Instead of referring to the control layer, “softswitch” morphed into an all-encompassing term for next generation networks (NGNs) that included both call control and transport (gateway) layers.

There was also confusion between “pure” VoIP networks on the Class4 level (e.g., Deutsche Telekom’s carrier services division and ITXC) and the NGN (Class5) switches, some of which included both TDM and VoIP ports.
Furthermore, some NGN solutions, such as Santera and Taqua, did not include VoIP. Instead, they were based on a “compact” TDM switch receiving aggregated “NGN TDM” traffic. Even in those solutions that supported IP, the end user experience didn’t change since he/she was still using a “black phone” and receiving black phone services from the service provider. As these solutions only provided a subset of basic Class5 features, they were not really a replacement for the traditional Class5 TDM switch, rather provided carriers with a cheaper alternative in terms of opex and capex to the big Class 5 iron (e.g., Lucent 5ESS and Nortel DMS).

This is evidenced by the fact that the CLECs that bought these solutions mainly deployed them alongside their Class5 switches in new or smaller POPs, in essence settling for less functionality than provided by the traditional TDM switch. Very few of them deployed the NGN switches as a forklift replacement providing full Class5 switching functionality.

As noted, the key benefit of these NGN switches was substantial savings in both capex and opex. For example, a recent analysis demonstrated that using an IP/softswitch architecture to replace a TDM switch of equal capacity would shrink the space required in the central office from 28 racks to two. Reductions in air conditioning and power costs alone will allow carriers to recoup the cost of a new softswitch in only 18 months. In comparison, the cost of most circuit switches needed to be spread over more than 10 years.

In summary, the state of carrier VoIP by the end of 2002 could be characterized as follows: legacy TDM switch vendors begin to announce end-of-life, Class4 international long-distance is being replaced by VoIP trunking solutions, and Class5 NGN is producing a lot of “buzz” but no real forklift upgrades on the carrier side, or behavioral changes on the end user side, are happening.

Though this article addresses the service provider market, it would be remiss to ignore the enterprise market, which was well ahead of the carriers in terms of VoIP deployments. Enterprises deployed IP telephony (e.g., IP PBX) services, using IP phones on their desks and softphone features that they could activate from their PCs (voice mail/e-mail convergence, unified messaging, drag and drop conferencing, etc.).

Residential VoIP Goes Mainstream
The fundamental change in VoIP usage and public perception came in mid-2003 with the emergence of Vonage. This is the first time residential VoIP went mainstream in the Western world. (Softbank’s Yahoo BB! in Japan was the first commercial, large scale, voice over broadband provider and now has over 4 million users). The Vonage model shows how small companies with limited facilities can offer residential services with PSTN quality at highly competitive costs.
The concept of residential VoIP is simple. Its basic requirements include a computer with a broadband connection (DSL, cable) and a compact end-user device (IAD, etc.) which plugs into a regular black phone.
This is all the user needs to enjoy “sticky” services, not available over the PSTN, and with the impact to really change the way people behave in terms of personal communications. Here are a few examples:

• Set up conference calls with a few mouse clicks using Web browser;
• “Find Me’ services so you never miss that important call;
• Set up your own ‘virtual’ phone numbers in different areas;
• Video calling;
• Do Not Disturb to avoid those annoying dinnertime telemarketers.

Most residential users still use VoIP as a second line. However, as users get comfortable with the quality and reliability and regulatory requirements concerning emergency services and lawful interception are relaxed, we can expect users to disconnect their PSTN line and use VoIP instead.

VoBB may still not be ready to serve as VoIP’s long-awaited “killer app.” However, combined with WiFi, and later with WiMAX and video capabilities (as they become available to the masses), it very well could be.

Regardless, what is important is that the NGN capabilities have finally emerged — and that people will change their behavior to take advantage of these capabilities because they let you do things not possible with a PSTN phone. Major carriers and broadband service providers are aware that a fundamental behavioral change is taking place. Moreover, the success of initial VoBB networks has led the carrier market to the conclusion that VoIP-based Class5 solutions are the preferred model.

Today price is the key factor; later when incumbents inevitably lower prices to be competitive with the VoBB upstarts, services will be the key differentiator. VoBB allows service providers to offer enhanced services out-of-region since the entire network is IP-based and the local loop is circumvented. Thus, service providers can realize the true vision of NGN, in which end-user devices are connected directly to IP networks and boundaries between access and long distance networks cease to exist (i.e., the entire voice route is implemented over an IP network).

This new model presents opportunities for new types of service providers (incumbents, CLECs, MSOs, virtual network operators, startups), wishing to offer services to residential and SOHO/ SME customers based on VoBB.

Big players will always have the advantage in the huge residential market due to mass, resources, marketing capabilities, cash, etc. Good examples of such are the AT&T CallVantage prime time commercials during the 2004 Olympics. Small players could, however, penetrate the residential market with the right marketing plan and execution, while the enterprise market, due to its higher margins, is also a lucrative target for smaller players.

What Equipment Do Service Providers Need?
The VoBB model calls for robust application servers that can support residential services with carrier-grade reliability and scalability, Web-based management, feature database, emergency 911 services, and lawful interception support. The equipment should be vendor neutral to work with any independent or integrated access device.

We are still a long way from the seemingly inevitable transition to a pure IP world (today less than three percent of calls stay within the VOBB network), which means that the vast majority of calls terminate on the PSTN. And, since there is no direct IP interconnect between disparate VOBB networks, e.g., Packet8 to SoftBank, there is a need for strong IP peering capabilities. This can be done using wholesale partners for global termination, while the service provider need only supply the application.

Carriers require a VoBB platform that is interoperable with various types of SIP endpoints (IADs, SIP phones, softphones, etc.) and can communicate with SIP-based trunking gateways with robust SS7 capabilities for interconnection with the PSTN. In addition, as more carriers deploy VoIP as their technology of choice for voice trunking and enterprise solutions, there is a growing need for direct interconnection between carriers’ VoIP (H.323 or SIP) networks. Carriers require IP peering solutions for secure, cost-effective traffic exchange across VoIP networks without the need for TDM conversion. This allows them to expand network reach and minimize PSTN termination costs.

These solutions should also be able to manage the routing of traffic to and from affiliates, using the most advantageous route by providing the carrier with different routing options and flexible routing management. Such capabilities minimize routing costs and optimize QoS.

Conclusion
As VoBB networks reach critical mass, carriers and service providers must accommodate this business model within their overall VoIP/NGN strategy. In terms of networks, service providers must weigh several options: from building their own independent network to outsourcing the entire network and limiting operations to marketing and management, with multiple variations between these two extremes. They must also find the right service focus, business model and technology partners. While even the very big players are not yet sure of their direction, it is clear that we are witnessing the beginning of the fundamental change that was promised when commercial VoIP products were first introduced to the world less than a decade ago.

Ari Rabban is vice president for corporate development and marketing at VocalTec Communications. For more information, please visit the company’s Web site at www.vocaltec.com.

[ Return To The January 2005 Of Contents ]