Tried & True? Not for VOIP

As carriers migrate their TDM voice networks to VOIP, they may be tempted to take the course of perceived lowest risk by using their legacy equipment vendors to supply new gear. They shouldn't: Giving in to that temptation could have dire long-term consequences.

That stay-the-course vendor strategy is enticing on several levels. First, carriers don't have to add new vendors and, probably more importantly, new network management systems to their network. Most incumbent Class 5 TDM switch vendors can also update existing equipment piecemeal by adding new VOIP cards or chassis into existing racks, which means lines can be switched over from TDM to VOIP gradually. Of course, the incumbent TDM vendors also have an unparalleled knowledge of the existing network.

But carriers that deploy next-gen equipment based solely on past equipment buys could be making a serious and costly mistake.

The future network cost and service revenue implications of picking a mediocre Class 5 migration vendor are enormous. Incumbent network operators of all sizes must be prepared to bite the bullet, cap their legacy systems, and migrate to a multisource paradigm with new vendors, if existing suppliers aren't up to the task.

This is one of the key findings of VOIP: A Comprehensive Competitive Analysis of Softswitches, one of two new VOIP equipment reports now available from Heavy Reading. Along with its companion report – VOIP: A Comprehensive Competitive Analysis of Media Gateways – it delivers the most complete study of VOIP carrier equipment ever undertaken.

There are many shortcomings that can relegate a Class 5 migration vendor or its products to mediocrity. This is a market with high entry barriers, and some of these are product features and vendor characteristics that I call "table stakes": If you don't have them, you're not a player.

The obvious table stakes for softswitches include the usual suspects for carrier-grade products: NEBS or equivalent compliance and "five-nines" availability for the combination of all products and networks in the system design. To these is added new requirements for access by law enforcement officials, such as those enumerated in CALEA (the Communications Assistance for Law Enforcement Act).

The major intangible requirements for a vendor looking to be the prime contractor for a large telco Class 5 migration from TDM to VOIP softswitching include a sound knowledge of telephony, preferably built up over decades of experience. Obviously, incumbent vendors have this experience, but with the layoffs and downsizing of the major vendors in the last three years, coupled with the other companies that were used as OEM suppliers of telephony hardware and software, it's not as rare a commodity as might be expected.

The key migration vendor also needs to be financially stable, so that the VOIP products that are new today can still be supported and maintained for years to come. Again, most incumbents meet this requirement.

For incumbents, the devil of VOIP mediocrity is in the product details. The combination of softswitches and media gateways must provide and support the services that carriers want to offer to their customers. Most ILECs now deploy more than 300 CLASS (Custom Local Area Signaling Service) services, such as call waiting and call forwarding, and more than 200 Centrex services, such as hosted PBX. They want to continue these services, but they also need to be looking for innovative new services that will increase revenue. Our competitive analysis of softswitch products shows that some legacy vendors aren't doing a particularly good job on the innovation front.

Operators should insist on multisourced and interoperating products, so that Vendor A's softswitch works with Vendor B's media gateways, and so that between them they can still support legacy and new services. Other key details include line density, call-processing capability, scaleability, protocol support, and codec support. This is where some incumbent vendors fall down.

In analyzing vendor VOIP strategies, I tried to make the comparisons as realistic as possible, using a simple network architecture but stretching the requirements and nailing down many parameters to ensure a level playing field and no cheating. For distributed softswitches, I set the network target at a demanding 1 million access lines.

For media gateways there were other challenging metrics, with some of my favorite "per cubic foot" measures being used to avoid doubt about chassis and rack claims. For softswitches, these metrics include subscribers handled per cubic foot, maximum BHCA (busy-hour call attempts), calls set up per second, and number of simultaneous calls handled. For media gateways, key metrics include number of lines per cubic foot, number of T1/E1s, and simultaneous calls handled.

After performing all of these checks and comparisons, it may well turn out that your incumbent TDM switch vendor does offer a solid migration path from Class 5 to VOIP: There are many good vendors and products out there. However, any carrier that simply puts its future network into the hands of its legacy vendor without doing due diligence could end up paying for that mistake with its very existence.

— Graham Beniston, Analyst at Large, Heavy Reading

Be the first to post a comment regarding this story.
Sign In