|
|
|||
|
|
ATM and DSL: A Perfect Combination of Elements
Greg Langdon
06/01/2000
Digital subscriber line and asynchronous transfer mode technologies go together like hydrogen and oxygen. While DSL turns copper phone lines into expressways for digital traffic, ATM (which has become DSL's standard Layer 2 transport protocol) allows different types of traffic to share the road. The union of these two elements has given rise to a third technology that's no infant in the industry--voice over ATM (VoATM). Sometimes associated with VoIP, VoATM is a completely separate entity that possesses some important capabilities its Internet-related cousin has not yet grown into. In fact, VoATM provides the two things most important to any network that carries both data and voice traffic: the superior efficiency associated with data networks and the high-quality voice service users have come to expect whenever they pick up the phone. Sharing the Wealth It's not exactly news that DSL allows the copper twisted pair--which enters virtually every home and business in America--to carry data as well as voice traffic. DSL has effectively bridged the gap between the "haves" of high-speed data services--large companies that could afford dedicated T1/E1 lines--and the "have-nots"--everyone else. The technology allows small to medium-sized businesses and individual subscribers to use existing telephone wiring to establish an always-open, high-speed connection to the Internet or corporate WAN. And unlike cable modems, which require line sharing and may suffer from slowdowns or security breaches, each DSL user has a dedicated connection. However, DSL alone does nothing to expand voice options. A splitter usually serves as a virtual median that keeps the two types of traffic completely separate, which means voice is still carried over the PSTN and reaps none of the efficiency benefits of DSL. If the customer wants more than one voice line, the phone company still has to come out and install another physical line.
Bandwidth Efficiency It quickly becomes apparent why ATM is the perfect complement to DSL. Designed from the ground up to integrate multiple services on one infrastructure, ATM makes it possible for voice and data to truly share the road--a road two dozen lanes wide in this case. In fact, VoATM allows a single twisted pair to carry up to 24 voice lines in addition to data traffic. The ATM Adaptation Layer (AAL) allows for this bandwidth-sharing functionality. In essence, the AAL defines how voice and data traffic is changed into ATM cells and converts higher layers of services (such as TCP/IP) into a format appropriate for the ATM protocol layer. Two flavors of AAL--AAL 1 and AAL 2--are most often used for voice traffic. AAL 1 handles constant bit rate (CBR) traffic, such as voice and video, by allocating fixed amounts of bandwidth to each connection. This prevents interruptions or delays, but stakes out an entire section of bandwidth for each CBR connection, even when the bandwidth is not fully used. AAL 2, on the other hand, slips data traffic into the tiny silences that riddle any phone conversation, ensuring that every bit of available bandwidth is occupied. By identifying different classes of traffic, AAL 2 is able to give voice traffic priority over less time-sensitive data traffic. It then uses statistical multiplexing to allocate bandwidth in real time over a single ATM permanent virtual circuit (PVC). Preserving Infrastructure Just as VoATM prevents existing bandwidth capacity from going to waste, it makes efficient use of existing network infrastructure. The only new piece of equipment required at the customer site is a next-generation IAD. This device, which replaces a standard DSL access box, is connected to the customer's standard telephone equipment and to a PC or LAN. It combines voice and data traffic and sends it over the copper plant to the communications provider, where a voice gateway at the CO splits out the voice and data components. No further retrofitting is required.
Safeguarding Voice Quality If efficiency is the first order of business for data networks, transmission quality has always been of paramount concern on SONETs. While a file transfer is unaffected by a microsecond's delay between data packets, the human ear does not forgive even the tiniest distortion of an audible message. People expect a tele- phone conversation to sound the same as a face-to-face chat. In this regard, VoATM is superior to VoIP. Although often touted as an economical way to offer voice service over existing Internet connections, IP's large, variable-length packets are better at transmitting data than voice. Because the large packets fill relatively slowly, and because IP's standard quality-control method is simply to re-send lost packets, phone conversations often suffer from signal loss and echo. Some corporations with private IP networks have compensated for VoIP's data-oriented design by implementing proprietary protocols. But these solutions are effective only when both ends of the connection use the same hardware and software. The public Internet is a long way from implementing standards that effectively address voice quality issues. In contrast, ATM is an ideal medium for voice as well as data. Because ATM uses small, uniform packets, a VoATM transmission is sent at a rate approximately 10 times faster than a voice transmission on an IP network. Even more important than speed, however, is ATM's excellence in the area of traffic management and QoS. While IP currently treats every transmission the same whether it is an e-mail message or a two-way videoconference, ATM was designed to differentiate between high- and low-priority traffic. Since voice is time sensitive, it is always given higher priority, which means callers don't experience degraded service when the network is busy. Carriers are able to guarantee the "toll- quality" voice connections that users have come to expect. The AAL 2 is actually responsible for VoATM's ability to blend great network efficiency with guaranteed quality. By including content information in each ATM cell, it can prioritize individual packets. This allows voice and data packets to share bandwidth, which is dynamically adjusted based on the traffic flow at any given moment. For instance, when someone takes a breath during a phone conversation, more bandwidth is turned over to data transport; when the conversation resumes, the bandwidth is reclaimed by higher priority voice packets. Greg Langdon is vice president of marketing for Efficient Networks Inc. (www.efficient.com). He can be reached at glangdon@efficient.com.
Share this article: Email,
Slashdot, Digg,
Del.icio.us, Yahoo!MyWeb,
Windows Live Favorites,
Furl
|
|
| Sponsored Links | xchange Announcements |