May 2002

VoIP: In Through Data's Backdoor

BY STEVE GLADSTONE

In the November issue, I suggested that Rich Tehrani and I were getting indications that the industry was about to �get turned on its ear.� Well, it seems to be happening a little slower than we had hoped, but things continue to move quickly behind the scenes. Just last week I met with a large global player in the telecom equipment manufacturer space who named VoIP as one of only three key technology areas in which they continue to make any significant R&D investment. And we continue to see the bulk of our test gear being purchased by labs testing packet telephony network equipment. The NextGen network is happening, it is only a question of timing.

Carrier-class VoIP naysayers rightly point to the fact that state-of-the-art VoIP today is good only for the simplest telephony functions, i.e., making phone calls. Therefore, the technology is only suited to simple applications such as modem-offload, toll-bypass applications, etc. But, with long-distance call minutes becoming a commodity, that isn�t where the money is. So there is the holdup. CLECs and alternative carriers whose greenfield infrastructure buildup would justify the latest technology are evaporating and cash-strapped, established carriers can�t justify trading out the functioning Class 4 infrastructure for the marginal cost savings of packet-based long haul gear.

In �Switching the Switches,� a research paper published last September by UBS Warburg, Nikos Theodosopoulos and Stephen Chin did a nice job of arguing that the industry�s �Holy Grail� (in their words) is replacing the Class 5 Switches. Inherently there are many more such devices, therefore a much bigger replacement market. The problem is that the bounty of that market comes only with an order of magnitude � more functional complexity.

Class 5 switches, by definition, support a large set of features (4000+), like call forwarding, transfer, and hold, that have a complex set of interactions. Without a subset of those features that you commonly use, you don�t have a Class 5 switch. Today, none of the next generation switches have all these features, and developing and proving a softswitch architecture with Class 5 functionality requires substantial time and investment. On the other hand, I can tell you that a number of players in this space have product rapidly maturing and we are already testing a number of them.

It�s clear why network equipment manufacturers would seek to capture a piece of the Class 5 replacement market, but again what motivates carriers to swap out equipment which, although it�s been around for a while, is still working fine? There was the promise of cheaper switches, but Lucent and Nortel et al, have covered the price. But this can�t continue.

It really gets back to the prior premise that call minutes have turned into a commodity � so enhanced services are where the profit lies for carriers. The softswitch architecture, which uses VoIP to carry calls, opens the door to a whole new world of revenue generating services. It also opens the market to third parties and to the carriers themselves to develop those services and to thereby develop some much-needed differentiation. That is what excites the carriers and provides the light at the end of their tunnel.

The carriers are positioning themselves for the great promise of the new technology and know they need to go there eventually, but for now are only moving forward fast enough to stay in the game. It reminds me of the early laps of the 1,500-meter short track skating in the Olympics. The competitors all cruise along cautiously, not falling too far behind, not charging forward. But a few laps before the finish, someone is going to break ranks and go for the gold, and the others will have to follow; some will have fallen too far back to have any hope.

Well-positioned equipment manufacturers will enjoy a great boom when that sprint comes, but for now, they are sitting on the sidelines. And unlike the Olympics, no one knows how long this race is.

To shift analogies, the Warburg report uses a Trojan Horse to describe the approach that manufacturers are taking to maintain their positioning. They are selling equipment to meet current needs and capabilities � Internet offload and toll bypass � that can eventually be upgraded with more capability.

Cisco, for example, is the market leader in equipment sales for service providers that offer both voice and data networking services. What you may not be aware of is the more or less subtle change that has occurred in the data product content at Cisco: More than half of its service provider products and all of its AVVID (Architecture for Voice, Video and Integrated Data) for enterprises are VoIP-capable.

The hidden attack in the Cisco approach deals on a more subtle but direct level with the VoIP requirements of service providers. Cisco in mid-2001 won a very competitive benchmark test held by independent testing firm Mier Communication, demonstrating that the Cisco 12000 Internet Router delivers the lowest latency and highest levels of quality of service � a requirement for delay-sensitive applications such as voice and video. Service providers are responding. Major service providers including Qwest, AOL, France Telecom, Global Crossing, Velocita, China Unicom, and China Telecom, have announced that they have selected the Cisco 12000 Internet Router as the platform of choice for delivering both a data and VoIP networking backbone.

So on the service provider front, Cisco is winning over converts by providing data networking products for the next-generation service provider�s backbone that take into account today, as well as future VoIP performance issues such as latency and delay. Add to this the fact that Cisco is aggressively addressing IVR like Web data content delivery inside the service provider�s edge device of choice, the VoiceXML enabled Cisco Access Server 5300 Voice Gateway. The sum of Cisco�s game is to simply invade on all fronts and conquer.

What does this mean to more traditional players like Lucent, Siemens, or Nortel? They either need to look like a credible data networking player with reliable voice capabilities or rely on their existing switch and Class 5 feature business to hold off the onslaught of the data networking Huns. Next-generation service providers are looking for the best of both worlds: Lower operating costs, a smaller footprint, etc. for VoIP and services that are on equal footing with their existing PSTN counterparts.

There are two categories of traditional players, a dual player like Nortel Networks, which provides a combination of reliable voice networking and data networking products or the more traditional player like Lucent with a strong concentration on tried-and-true switching products that can be deployed today. This confronts the service provider with a basic choice: Either give the business to a data networking vendor that has no track record in services beyond Class 5, but knows the backbone, or try the newer products from a proven legacy vendor. These points are not lost on the incumbents; part of our business is providing test capability to both the traditional players who have focused lots of effort on becoming �like� the data networking competition, and the direct competition, the new next generation players. And, they are successfully making some very good systems.

Fortunately, for test guys like us, despite the economic malaise, the technology continues to advance. Upstream, carrier labs have all kinds of cool stuff that they are trying to make play together so they can be poised for eventual rapid deployment. Similarly, equipment suppliers need to provide small quantities of advancing technology to those labs, unprofitable though it may be, in anticipation of deployment. With minimal deployment going on, the lab is the only place to build confidence in new technology.

It�s no news that there are some strong macro forces keeping carriers from charging into next generation networks. But clearly the benefits are attractive enough to keep pulling them conservatively forward. As the economic and market shackles begin to release, that attraction will turn into critical need. Winning equipment manufacturers will be the ones that are in the accounts and have the technology. However, it�s still the case that carriers will need to see the new capabilities proved out in their labs. And, there is some wild stuff going on�

Wait till you see what comes next!

Steve Gladstone is the general manager of the Communications-Infrastructure Test Group of Empirix and is an acknowledged expert in testing computer telephony systems and networks. Steve founded Hammer Technologies, now a business unit of Empirix, Inc., whose products are in use today by developers of computer telephony, advanced switching, and enhanced services systems, most of the RBOCs and long-distance carriers, and at call centers of numerous Fortune 100 companies. Visit them online at www.empirix.com.

[ Return To The May 2002 Table Of Contents ]