DSL

Startup Promises New Carrier Class DSL Hardware

Out of Canada, which (in DSL terms) is more advanced than the U.S., comes a startup with a nifty little DSLAM that, the company says, goes way beyond next generation DSL to deliver a network designed for services.

by Alex Goldman ISP-Planet Managing Editor [October 4, 2004]

The latest standards battle in DSL is between ADSL2+, which can deliver higher speeds now, and VDSL, which should be able to deliver even higher speeds later. However, every time we hear about a massive deployment of fiber, it's in the future. It's not about something that's already been done.

Executives at Ottawa, Canada-based Critical Telecom, a carrier-class hardware startup with 40 employees, say that all the incentives are now in place for a true fiber buildout.

The fiber buildout will not be fiber to the home, but a gradual upgrade of the copper infrastructure. Ken Davison, the company's vice president of sales and marketing, says, "it will be a gradual migration of copper to fiber."

But even Davison is not 100 percent certain that the ILECs will go with fiber. "This is an opportunity for the industry," he says. "The industry needs to make a choice. We hope the industry takes advantage of this opportunity."

Hardened to a wet, wet world
The opportunity that Critical Telecom is announcing today is the hardy, small GEmini Ethernet Remote DSLAM (GEmini for short) built to be deployed outside the CO. Critical is making several technology bets with this product release: that Ethernet will replace ATM in carrier networks, that ADSL2+ will beat out VDSL because ADSL2+ is available now and VDSL is not, and that the ILECs will opt for a hybrid fiber-copper network architecture over alternatives such as pure fiber, pure copper, or fiber to the wireless AP.

Critical has been working closely with major Canadian ISP Telus, which is also an investor. Davison reports that Telus needed a DSLAM that could be deployed remotely.

Mark Labbe, Critical CTO, says that deploying hardware that was designed for the controlled CO environment in remote terminals can be expensive. "Traditional outside plant design does not have warm dry places. You need to design the product for the locations that are there today."

Davison adds, "the challenge is upgrading cost effectively. You need to avoid civil engineering costs such as air conditioning, battery backup, and power delivery problems."

The unit is powered over the copper infrastructure using power over Ethernet. This is a very good idea. Wireless equipment designers tell us that the antennas fail less often than the power units attached to them. Placing the power unit in the controlled environment of the CO removes this problem.

The product spec sheet reports that the device operates in a temperature range from -40°C to +60°C. You don't usually expect a spec sheet to be entertaining, but this entry was striking:

Humidity: Module meets IEC 60529 (IP67, submerged in 1 meter [ed. note: =3 feet] of water for 30 minutes)

The device can be placed in a cabinet, or on a telephone pole in a weatherproof metal box like those built for antennas.

VLANs within VLANs
Beyond better remote terminals is the promise of better services. Critical says that its new DSLAM supports services in the following ways:

• Allows virtualization through VLAN support (including PVC to VLAN mapping and what the company called "stacked VLAN transport" which we understand to mean VLANs within VLANs)
• Supports DHCP relay
• Supports IGMP at the DSLAM
• Enables remote management features such as OAM F5 (Operation, Administration, and Maintenance) and remote diagnostics

The idea is to push to the edge of the network the intelligence required to support services such as IPTV and self-provisioning of new services. The idea is that a customer should be able to order a pay per view video or even just better bandwidth for a download or less jitter for a single VoIP call.

While the RBOCs are the eventual targets for this device, Davison says that the decision cycle is faster with RLECs, and RLECs also have smaller, faster deployments. He notes that RLECs that already provide cable service of some kind should be especially eager to purchase this device.

Joe Laszlo, Jupiter Research senior analyst, agrees. He says that it is not yet clear what the RBOCs will do, and it is possible that each RBOC will embrace a different network architecture, from pure copper to pure fiber to any of many potential hybrid architectures. While the RBOCs are taking time to make these technology decisions, he suspects RLECs will be more focused on rapidly delivering advanced services over the copper network.

—End