The term "xDSL" is a catchall, covering a number of similar yet competing forms of digital subscriber line (DSL) technologies. (The "x" in xDSL is actually filled in with other letters, depending on the technology implemented.)
The major xDSL categories are:
 |
High-bit-rate Digital Subscriber Line (HDSL) |
|
|
Asymmetric Digital Subscriber Line (ADSL) |
|
|
Rate Adaptive Digital Subscriber Line (RADSL) (not discussed in this document) |
Each technology offers different speeds, ranges, and operating characteristics. While there may be some overlap in the capabilities of various xDSL technologies, it's likely that they will coexist in a complementary rather than competitive fashion. But that leaves service providers in somewhat of a quandary when it comes to deciding which technology to deploy in the lucrative broadband access market. For that reason its important to know what each implementation offers.
HDSL
Of all the xDSL offerings, HDSL probably has the largest installed base because it the first DSL technology to be invented . The technology arose from carriers' problems in extending broadband speeds -- specifically T1 (1.544 Mbps) and E1 (2.048 Mbps) services -- over long copper loops. Because long copper loops distort signal quality, repeaters or amplifiers are installed on copper pairs at prescribed intervals to restore signal quality. In today's T1/E1 networks, they must be installed about every 3,000 to 4,000 feet -- a time consuming and expensive process.
In the late 1980s, Bellcore began research into a new method of T1 and E1 provisioning that would eliminate repeaters and simplify the overall deployment of high bandwidth networks for the so-called "last mile" into the home or office. The technology, called HDSL, was designed to deliver traditional T1/E1 services over unconditioned wires by placing transceivers on each end of two or three twisted pair.
However, HDSL is also being modified to work over a single copper phone line to provide direct premises connections between a customer and its serving central office. The penalty for operating over one pair is performance; current implementations top out at either 384 Kbps or 768 Kbps symmetrical speeds.
Most predict that single pair HDSL will be an interim development, though, and that it will eventually be supplanted by ADSL since it will provide higher throughput rates.
HDSL's ability to make the most of installed copper has definitely caught the interest of telecom providers interested in extending traditional T1 and E1 services throughout the feeder network plant. But the real demand for multimegabit transports is likely to come from average consumers interested in applications like high speed Internet access and the ability to purchase video movies on demand. Both of these applications are asymmetric in nature, demanding a multimegabit downstream delivery to the consumer, but only a small upstream link into the network to transmit basic commands or the occasional email. Since HDSL cannot provide high enough speeds for these applications, several vendors began work on alternatives to HDSL that could produce fast speeds, albeit in an asymmetric fashion. This is where ADSL, or Asymmetric Digital Subscriber Line originated.
Individual vendor efforts in creating ADSL produced two competing algorithms that both offer downstream rates in excess of 6 Mbps and simultaneous duplex transmissions of 640 Kbps over single pair lines of 12,000 feet or less. One line coding, Discrete Multi-Tone (DMT), is established as the ANSI and ETSI standard; another, Carrierless Amplitude and Phase (CAP) modulation is widely implemented and being pushed by certain proponents as a side-by-side standard with DMT. Just to keep things interesting, some companies are advancing the use of QAM (Quadrature Amplitude Modulation), a line coding scheme similar to CAP.
Simply put, DMT divides the 1 MHz spectrum offered by a phone line into 256 4-KHz channels. It then varies the bit densities on each of these channels to overcome noise and interference that may be present in sections of that spectrum. Proponents argue that DMT is better on noisy lines because of the ability to maximize throughput on good channels and minimize throughput on channels with heavy interference.
In contrast, CAP relies on a single carrier and uses techniques similar to the Quadrature Amplitude Modulation (QAM) used in V.34 (28.8 Kbps) modems to make the most of phone lines.
Despite the ratification of the DMT standard, modems utilizing both methods are being trialed in networks globally in order to gauge cost of deployment and performance attributes. Both encoding algorithms have elicited positive responses from a performance standpoint. If both perform similarly, cost and manageability is likely to be the deciding factor between the two. CAP currently leads DMT in cost and size, but DMT offers greater flexibility. Over time, both will trend towards one another, creating more options for vendors. Which method will be preferred in carrier networks remains to be seen, and most consumers will not know which is implemented in their modems.
From a technical standpoint, both the DMT and CAP approaches place an ADSL modem on each end of a twisted-pair telephone line, creating three information channels: a high speed downstream channel, a medium speed duplex channel, and a POTS (Plain Old Telephone Service) channel (see "Frequency Division in ADSL" Figure). The POTS channel is split off from the digital modem by filters, thus isolating the voice circuit so it can be powered by a traditional phone line. This guarantees uninterrupted POTS connections, even if the ADSL connection or outside power fails.
Frequency Division Multiplexing in xDSL
This frequency division also means service providers can keep voice and data on separate networks; thereby eliminating the congestion on the PSTN that is created by transferring data over circuit-switched rather than packet or cell-switched connections.
Finally, it is expected that the current fixed-rate forms of CAP and DMT will evolve into rate adaptive designs that will enable ADSL connections to overcome loop length limitations of 18,000 or even boost upstream rates to become more symmetrical for applications that require it. DMT designs support rate adaptation in increments of 32 Kbps whereas CAP designs currently do so in 200 Kbps steps. Which approach is favored by service providers and which designs become available in volume first remains to be seen. It's likely that the two line codes will coexist in the near term; this highlights just one deployment issue for service providers: the need for flexible systems.
This information came from various sources (use the search term catchall); the original page I found the information was not found in google.com
Return to previous menu