How Will Verizon Go National?
Why Verizon? Several reasons. It's got this year's biggest RBOC budget (see Carrier Capex Cuts to Slow a Bit). It's got approval via Section 271 of the 1996 Telecommunications Act to offer long-distance services in nearly all states. And it's already got an impressive corporate services base to use as a launching pad for nationwide data, thanks to its geography and regulatory status.
"Verizon has coverage with its sales force of 45 percent of all corporate headquarters sites in the U.S.," says Tom Nolle, president of CIMI Group, a telecom consultancy. That's significantly more than other RBOCs can boast, he maintains.
But just what are Verizon's plans? And which equipment vendors will have a hand in making them happen?
Details are sketchy. It's no secret Verizon's been evaluating Multiprotocol Label Switching (MPLS) for nationwide services. Like other RBOCs, Verizon hopes MPLS can be used as a centralizing protocol to move Asynchronous Transfer Mode (ATM) and Frame Relay across LATA (local access and transport area) boundaries that define its regional network (see MPLS vs ATM? Vendors Weigh In).
But sources say the carrier's shifted gears in its plans to make this happen. Back in early 2002, Verizon talked about setting up multiple points of presence (POPs), just like an IXC, equipped with MPLS-capable multiservice switches that would bring Frame Relay and ATM services into an IP backbone.
Since November 2002, though, Verizon's taken a different tack. Instead of multiservice switches in multiple POPs, it describes linking LATAs with virtual circuits. This implies that Verizon wants to use ATM as the main technology for bringing legacy services to data customers.
MPLS would still be used in the network core -- perhaps to link multiple DWDM boxes. But the architecture would be based on ATM aggregation points in each region, linked by ATM's NNI (network to network inteface).
What happened? Why did Verizon change course, and what are the implications?
Sources say Verizon changed plans for its network based on a vendor selection process gone awry. Specifically, it had a startup vendor picked in the fall of 2002 to deliver MPLS-based multiservice switches in multiple POPs. Not wanting to risk going with an untried company, Verizon reportedly asked the startup to partner with a larger player in delivering its solution. When no partnership materialized among Verizon's leading vendors, which include Lucent Technologies Inc. (NYSE: LU) as the prime supplier of ATM switching, Verizon rejiggered its architecture.
Verizon's current strategy for its nationwide network was unveiled in a series of public presentations and a press release in November 2002 and dubbed Enterprise Advance (see Verizon Does Enterprise Data). A feature of the announcement was Verizon's plan to build its future on existing network infrastructure, calling a halt to new spending.
Sources say the spending slowdown was related to Verizon deciding to use existing Lucent ATM gear in its network. When the decision was made to forego using new multiservice gear, Verizon simply transferred some Lucent switches that had already been ordered into its long-distance project.
Verizon did not respond to inquiries regarding the RFPs mentioned in this story. But a spokesman said frame relay, ATM, and Sonet services in New York, New Jersey, Massachusetts, and Pennsylvania have been integrated and progress toward a nationwide backbone is going "swimmingly."
Who was the original vendor? No one seems ready to say. But of all the rumors swirling, one seems to stick the most: Digging by Light Reading has produced information that Laurel Networks Inc. was involved in tests at a series of carriers last year. Could Laurel have been the chosen solution?
Calls to execs at Laurel produced only a refusal to comment. Smoking gun -- or smoke screen?
It may be a moot point. Verizon seems to be on to other plans, ones that could supersede Enterprise Advance and ensure the carrier doesn't put itself at a disadvantage for future packet services by picking too much ATM. "They're looking at some partnerships and deals... I think they've got several new RFPs in the boiler," says Sam Greenholtz of Communications Industry Researchers Inc.
Still, scuttlebutt has it Verizon's looking seriously at ATM aggregation, and some say the chance that WaveSmith Networks Inc. might win a Verizon contract was behind last week's news that Ciena Corp. (Nasdaq: CIEN) plans to buy the startup (see Ciena Nabs WaveSmith).
Rumors persist that other vendors, including Équipe Communications Corp., which also has a Ciena relationship, may be in line for consideration. Verizon's also got IP deals with router vendors Cisco Systems Inc. (Nasdaq: CSCO) and Juniper Networks Inc. (Nasdaq: JNPR) that could affect future choices (see Marketing Mantra: Prove Your Worth and Juniper Boosts Spirits).
Ultimately, it looks as if Verizon's nationwide IP network will be the product of many RFPs. Some products may be introduced onto small portions of the network, such as virtual circuit aggregation, and move to bigger, more important functions, such as coordinating multiple service orders at Layer 2.
Others warn that all the planning and speculating may come to very little until the carrier starts serious deployment of its next-gen data network. "They haven't spent any money," says Frank Dzubeck, president of Communications Network Architects. "There are lots of projects on hold." There's no reason an RFP for nationwide services would fare any differently than dozens of other carrier RFPs that have amounted to nothing yet, he notes.
— Mary Jander, Senior Editor, Light Reading