Chicago Tribune






















 

MARKETS

DOW:
-225.510598.4

NASDAQ:
-65.02821.15

S&P 500:
-27.951307.58
GET QUOTE:



 

WEB SPECIALS

GETTING IN THE FLOW OF DSL 
ALTHOUGH DIGITAL SUBSCRIBER LINE TECHNOLOGY ISN'T NEW, THE DEMAND FOR IT IS: THE INTERNET IS DRIVING NEW INTEREST IN THE SERVICE THAT SOME SAY WILL DELIVER VOICE AND DATA EFFICIENTLY AND CHEAPLY. 

By Jon Van
Tribune Staff Writer 
September 20, 1999 

DSL is the latest high-tech buzzword, intoned on radio commercials and splashed on the sides of CTA buses.

 But the technology that boosts data flow over pairs of traditional copper phone lines isn't very new at all, tracing its history over the course of 30 years' work at Bell Laboratories. In fact, this is the second time telephone companies have made a run at selling the technology to the public.

In the early 1980s, just before the breakup of the Bell system, the phone company's marketing people created a campaign for packet-switched service that would provide voice, data and video over a single copper line. They used cherries, pears and apples as symbols for each service flowing along in a big fruit salad of communications.

 Nothing much came of the marketing effort because demand was low for data transmission and nearly non-existent for video back in those days, so the technology was shelved.

 But it wasn't forgotten.

 Bell engineers continued to tweak and develop packet-switch technology, and they trotted it out again in the early '90s, when some regional Bell companies flirted briefly with the notion of video-on-demand that would enable customers to watch movies at home delivered over their phone system's copper lines.

 When interest in video-on-demand petered out, the engineers continued development work on DSL, or digital subscriber line, technology on the general theory that something that could magnify the efficiency of the existing copper phone lines so dramatically would some day find acceptance.

 "DSL failed in the early days because it is tough for us to force the customers to do something different, like video-on-demand," said Michael Meyers, director of marketing development at Lucent Technologies Inc., the current incarnation of Bell Labs.

 "But now the Internet is the driver. People want faster access to the Internet, and DSL can provide that. When you also have government deregulation of the phone business and the rise of competing phone companies looking for a technological edge, everything is coming together to really promote DSL."

 And now that there is demand, the engineers are eager to provide products to meet it. Lucent and others have hardware and software for phone companies that enable them to digitize and send voice, video and data in packets over copper lines in a profusion of pears, cherries and apples.

 This is in contrast to traditional circuit-switched architecture that provides one dedicated pair of copper wires for each voice conversation.

 Phone companies long ago began to digitize signals sent over great distances, but they have kept to the circuit technology when carrying conversations to and from neighborhoods. What DSL does is to digitize and packetize the signals in and out of the customer's premises, said Tony Grewe, a Lucent director of strategy and business development.

 As virtually all local phone companies embrace DSL technology to carry data at high speeds and to provide several voice channels over a single pair of copper wires, network costs will drop dramatically and the entire telecommunications industry is in for several years of radical upheaval.

 In Illinois, for instance, experts predict the current complex local phone service billing formula that uses time and distance of calls to calculate charges will be swept away by flat-rate deals.

 But as useful as DSL may be, it still has many limitations.

 For one thing, it works really well for customers who are very close to the phone company's central switching office, but not so well for customers who are a few miles away.

 In some localities, DSL technology may not be practical for 25 percent of the customers. But one reason for that is that today's circuit-switched network was never designed with DSL in mind. As DSL becomes more widely deployed, it's likely that service providers will find ways to reach far-flung customers with it.

 Engineers say it's still early days in advancing DSL technology, and the more experience they get in using it, the better its performance will become.

 In fact, to address the various trade-offs that engineers must make when designing a DSL system, they've come up with a whole family of DSL, about eight different flavors of the technology at last count.

 If the technology were still named for fruits, it would be a regular farmers' market by now.

 One significant distinction in DSL is whether a service is symmetric or not. Asymmetric DSL refers to a system that enables the customer to send data upstream to the central office at a much slower rate than he obtains data flowing downstream.

 ADSL is designed for customers at home who mostly want access to the Internet, which finds them sending brief messages upstream to access gushers of data coming back to their home computers.

 The flip side of ADSL is symmetric DSL, or SDSL, which is better suited to people working at home who exchange large data files with their colleagues in the office, often sending out as much data as comes back to them.

 Another flavor is VDSL, a very high bit-rate flavor intended to work at quite short distances that is geared toward video streaming. This service works well in areas where optical fiber, the granddaddy of big capacity pipes, is nearby and the copper need only carry signals a fraction of a mile from the fiber end point.

 A home consumer product called DSL Lite was developed to make the service easier to install with minimum expense. A limitation to the Lite version is some loss of transmission speed for transporting data.

 Even though DSL technology is more than two decades old, it still has much development ahead of it, said Amra Tareen, Lucent director of Internet product development. Growing popularity of DSL assures that the technology will just get faster and cheaper, she said.

 "This isn't an exact science," said Tareen. "Chip sets will improve speeds as we understand the performance parameters better. The same thing that happened with dial-up modem technology will happen with DSL."

 The demand that's causing phone companies to deploy DSL is also causing them to install more optical fiber connecting the network's backbone closer to the neighborhoods, said Lucent's Meyers. The various DSL flavors can complement fiber nicely as phone service providers bring more network capacity, or bandwidth, closer to their customers.

 But despite the excitement it is now generating, DSL's debut also brings a host of problems that promise a bumpy road for phone service providers and their customers.

 One potential problem is that the various flavors of DSL can interfere with each other when they run on copper wires that are wrapped together in bundles.

 Judith Hellerstein, author of a DSL technology report for the Chicago-based International Engineering Consortium, said that "we will have problems with interference and crosstalk stemming from wrapping SDSL with ADSL lines. It's not a problem now because these technologies aren't mass-deployed, but once they are, it will cause interference and static for voice calls."

 Besides technical issues that underlie DSL deployment, cultural and political factors also dog the industry's adoption of the new technology, Hellerstein said. For example, there are industry standards for some flavors of DSL but not for others, and one reason for that is timing.

 "The competitors who are deploying SDSL now are working on `Internet time' that says you want to get the first mover advantage," Hellerstein said. "If you wait, you lose. Well, Internet time is lightning-fast compared to telecommunications utility time, which governs how standards are set.

 "It can take years to hold meetings, measure performance and gain consensus. It took two years for the ADSL working group to come up with ADSL Lite, and that was considered fast."

 For economic reasons, the large incumbent local phone companies like Ameritech Corp. are deploying ADSL and marketing it mostly to residential customers. They do this because of competition from cable TV high-speed Internet connections and because they already offer expensive high-speed data connections called T1 lines to many business customers.

 Competitive phone companies like NorthPoint Communications Inc., based in San Francisco, usually offer SDSL because their customers are mostly small businesses that cannot afford T1 lines but are happy to get high-speed data services for $100 or $200 a month instead of $500 or $1,000 that T1 connections fetch.

 The standards groups are often dominated by the large incumbent carriers who favor ADSL, said Hellerstein, so there is fear that in the future there will be industry pressure to alter the unstandardized SDSL connections to alleviate interference problems that arise.

 "Even though there's an industry consensus that DSL is the future, and it's in everyone's interest to cooperate on setting standards and avoiding interference, there's no way to act quickly," she said.

 "All they can do is work out the fixes now and apply them to subsequent product offerings."
 




Chicago
Tribune