|Standards War Looms Over DSL For Voice|
By Craig Matsumoto, EE Times
Apr 13, 2001 (8:15 AM)
DENVER—A year-old proposal for carrying voice over DSL is gaining attention as a service for home users, but debates about the technology's purpose and business prospects are stirring up questions as to whether it should become a standard. Behind the arguments, channelized voice-over-DSL (CVoDSL) offers some insight into the state of the DSL business: Its very existence is a product of the industry's renewed emphasis on residential rather than corporate users, and the friction surrounding it reflects vendors' wariness during financial bad times.
Efforts to standardize the proposal, although in the very early stages, are generating their share of heat, as gauged by the reactions of vendors attending this week's DSLCon show, held here.
Championed by DSL intellectual-property firm Aware Inc., Bedford, Mass., CVoDSL would add one to four lines of voice traffic alongside a DSL connection, reserving 64 kbit/s of bandwidth for each voice line, against 80 kbit/s allocated under VoDSL (or 32 and 40 kbit/s, respectively, for compressed voice).
The key is that rather than carry voice as an ATM service, as standard VoDSL does, CVoDSL would carry it the way the traditional phone network does, as time-division-multiplexed traffic.
The number of phone lines is small because the technology is targeting home users, not businesses. And the use of TDM voice lines is geared toward next-generation digital loop carriers and DSL access multiplexers (DSLAMs), which are expected to handle traffic types beyond pure asynchronous transfer mode, said David Benini, DSL marketing manager at Aware.
Backers claim CVoDSL simplifies DSL voice traffic. DSL lines currently carry all traffic as ATM, including voice, and this means the voice traffic has to be separated using a gateway inside the carrier's point of presence or central office.
Those gateways are “really designed for legacy DSLAMs” that expect all incoming traffic to be ATM, Benini said. Newer systems targeting the service-provider end of DSL will accept multiple traffic types, so it makes sense not to convert voice into ATM at the customer premises, he said.
One factor making CVoDSL viable is the increased emphasis on bringing DSL to home users. This contrasts with older assumptions that small businesses would be the main customers for the 16 voice ports available in VoDSL.
That concept arose at a point when DSL was expected to have a large business presence, Benini said. In fact, entire carriers sprang up on the premise of selling DSL to small and medium-size corporations, and many of them are failing as residences begin making up the bulk of DSL sales.
“Now that DSL's gone through a transition period, people have said that maybe we should crawl before we walk: provision one line or two lines at a time, keeping it as close to standard POTS [plain old telephone service] as possible,” Benini said.
CVoDSL doesn't replace VoDSL; in fact, Benini sees the two contributing to different markets. “Marketwise and architecture-wise, [VoDSL is] targeting and is optimized for small-business locations eight to 16 [voice] lines,” he said. “We are targeting residential consumers: one to four [voice] lines.”