The contenders? Cable vs. telco, of course. Their most effective weapons? Cable modems against digital subscriber line (DSL). And the money? Well, most oddsmakers are calling for a draw.

Broadband access is emerging as the hot market for the coming year, with nothing less than the future of communications as we know it at stake. So far, cable modem deployment has the edge over DSL. But that lead might not hold for very long.

According to market research firm DataQuest, the installed cable modem base will jump from 1.7 million units in 1999 to 5.3 million in 2003. DSL devices will reach 1 million this year, but will shoot to 9.8 million by 2003.

Other firms’ numbers vary, of course, but nearly all analysts are certain that the cable modem lead will give way in the next decade. And already, telephone companies are lining up their resources for major pushes into the market this year, and they have the technology, marketing skills and customer base to make things mighty difficult for cable operators.

But if one thing has emerged from conversations with DSL and cable modem experts, it is this: Whoever gets to the market first will win it.

"It’s better to be first than to be better," says Terry Shaw, project director for network systems at CableLabs. "You have to get systems rolled out in a timely manner."

"The first person into the home will have the customer," echoes Jeffrey Waldhuper, director of technology and engineering at Bell Atlantic. "Once a customer makes an investment, they won’t want to go through the burden of changing it."

There are a lot of ways to gauge competitiveness in the broadband race. Let’s start with technology.

Nuts-and-bolts comparison

Both systems will offer roughly the same throughput, about 1.5 Mbps. Yes, cable modems have been clocked at up to 30 Mbps (downstream), but actual speeds will depend on the type of modulation you use and the number of users allocated to a given downstream data carrier. Remember, you have a shared network, so the more people on at a given time, the less bandwidth per user.

Telcos, on the other hand, offer a direct line to the consumer. The drawback is that DSL is limited to roughly three-mile runs. Right now, that means nearly all DSL customers have to be located fairly close to a central office that houses DSL gear. The industry has only recently developed DSL access multiplexer (DSLAM) gear that is rugged enough for the field. But expect a pretty heavy rollout from here on in.

"A lot of telcos have installed digital loop carrier (DLC) systems," says Kevin Kahn, director of communication architecture for Intel Architecture Labs. "Instead of running a line to each house, they run a high-speed line, put a cabinet in the field and run lines from there."

Essentially, telcos have developed their own answer to the neighborhood node.

Ultimately, data throughput from the customers’ perspective still may be limited by the pipe that cable and telcos use for their connections to the Internet. For instance, either provider may have very fast data transmission between the customer and the headend or central office, but if each were to use, say, a single T1 to the Internet, that would be a real bottleneck.

Most customers, however, don’t really care about network architectures. They want speed and simplicity. With data rates roughly equal, ease of installation will be a priority. Telcos may have stumbled in this area with integrated services digital network (ISDN) technology, which proved difficult to install and unreliable in practice. Soured ISDN customers might be unwilling to trust the telco again, but the industry is taking steps to ensure a smoother DSL setup.

Some telcos offer DSL self-installation kits; where service is available, customers can set up their own modems. Cable subs often need a technician at the home to turn on the service, although self-provisioning modems are one solution if an active cable outlet is present near the computer. As well, a DSL customer can move the modem to any phone jack in the home, while cable customers are limited to coax connections. Another DSL advantage: Customers choose Internet service providers (ISPs), while cable modem users sign on to the cable preferred provider.

Price points

What about price? Both technologies likely will have competitive rates: $200 to $300 for the modems and about $40 per month for basic-level service. The trick for cable operators is to beat telcos on the cost-per-customer side when it comes to network upgrades.

"Network costs are affordable," says Bell Atlantic’s Waldhuper, "a couple of hundred dollars per customer."

Marketing challenge

On the marketing side, most experts agree that telcos have a leg up in the business community, while cable has hooks into the residential space. But don’t expect telcos to sit still. Bell Atlantic, for one, has unveiled a major effort to double its DSL deployment this year, running 17 million lines throughout its East Coast territory.

"Cable companies have more to lose because they are ahead in the residential market," says Laurie Falconer, a DSL analyst with TeleChoice, a market research firm in Dallas. "(For DSL), it’s make it or break it in the consumer market. If DSL providers don’t get their foot in the door in 1999, it could be trouble for the industry on the residential side."

On the flip side, cable has nowhere to go but up in the business market, but it could be tough to overcome the inevitable press and advertising attacks on cable modems, particularly cable’s network reliability and the security of a shared network. The latter issue could have an impact on business leaders who would liken it to sharing a single local area network (LAN) with other businesses.

To reassure its customers that their service will not go down, cable operators will have to install rock-solid backup power and devise foolproof ways to prevent customers from peeking in on others’ data streams. It wouldn’t hurt to tout DSL’s distance limitations and the fact that no matter how you shake it, telco service still will come in on twisted copper for all but the largest corporations, which can afford direct fiber.

As for cable’s shared network, you might want to remind customers that DSL offers a private line only back to the telco’s central office. Once the data hits the Net at large, it’s as unsecure as all the other data flying around cyberspace.

Regulatory obstacles

Telcos also have a cloud hanging over their heads called the federal government. The industry faces a higher level of regulation than cable, and that could interfere with rollout plans.

"Cable has the freedom to build business cases in a more flexible manner," says Intel’s Kahn. "Telco regulation means they can’t move the cost around very much. But the indications are that the federal government will reduce the regulations on telcos."

Cable operators also have to keep in mind that they will not necessarily be competing against one major Bell company. Smaller, regional local exchange carriers (LECs) are jumping into the DSL market as well. A case in point is the recent agreement between West Coast providers Covad and GST Telecommunications to deliver integrated voice and data over DSL to the business community. Of course, if the new LECs take customers away from cable, they’ll take them from the major Bells, too.

What to do?

So what should the cable strategy be in the coming years? The answer is deploy, deploy, deploy. Right now, it’s crucial to get your service in front of the people. The No. 1 complaint on the Internet today is slow download speeds. Whoever meets that need first should receive undying gratitude.

What is G.lite?

G.lite is the new consumer digital subscriber line (DSL) standard recently adopted by the International Telecommunications Union. Although a voluntary standard, expect most vendors and telcos to conform to it for their residential systems.

G.lite provides a data rate of 1.5 Mbps downstream and about 500 kbps upstream. That’s much lower than DSL’s ultimate capability, but it allows telcos to push the signal farther out into the network. The chief benefit of G.lite is that it allows consumers to install their own modems, just as they do now.

"They’re hailing G.lite as the great savior," says CableLabs’ Terry Shaw. "You can look at it as a ‘dumbed-down’ DSL. They took the capability out to increase robustness."

Jeffrey Waldhuper, director of technology and engineering at Bell Atlantic, says the G.lite effort emerged in the mid-90s when the telco industry soured on video-on-demand (VOD) and thus had no real reason to deliver a higher data solution.

"The G.lite objective was to lower the cost of DSL to that of 56k modems," he says.

Does G.lite pose a particular threat to cable modem rollout? Not really. DOCSIS standards should provide the same amount of interoperability, and 1.5 Mbps is within easy reach for cable modems.

Art Cole is a contributing editor to "Communications Technology."