Digital subscriber line access multiplexer
As important as the ADSL terminal unit-remote (ATU-R) and ADSL terminal unit-central office (ATU-C) are for the operation of an ADSL link, the digital subscriber line access multiplexer (DSLAM) is just as important. Without the DSLAM (pronounced dee-slam), the bits have nowhere to go. The DSLAM links the customer side of the network (i.e., the DSLs) with the service side of the network (e.g., the Internet). As the word multiplexer implies, it makes it possible for multiple DSL access circuits to be multiplexed onto higher-speed service circuits.
The graphic to the right depicts the basic architecture of a DSL access multiplexer (DSLAM). On the access side, it is important to note that DSLAMs are used with more than just ADSL. Technically, in the ADSL network architecture, the device that interfaces with the ATU-Cs in the serving exchange is the ADSL access node. An ADSL access node is a very good example of a DSLAM, but the DSLAM is a more generic device, and not just tied to ADSL. In other words, all ADSL access nodes are DSLAMs, but not all DSLAMs are ADSL access nodes. DSLAMs can also terminate HDSL, SDSL, and SHDSL access circuits, to name a few.
On the service side, the DSLAMs commonly implement ATM, Frame Relay, Ethernet, or SONET interfaces. By way of these interfaces they can provide access to the Internet, switched digital video (SDV) servers, or even provide access to ATM, Frame Relay, Carrier Ethernet, or IP-based transport services. Because modern ADSL DSLAMs also incorporate the splitter function and the ATU-C, they typically also have an interface (typically T-1) to the local voice switch.
The DSLAM is a technology typically owned by a local exchange company (LEC). The multiple service operators (MSO) and other cable companies use a different technology that leverages their hybrid fiber/coax (HFC) cable plant.
An Example DSLAM
The graphic to the right is not intended to represent any particular DSLAM product, but rather to convey a general impression of what a DSLAM actually looks like. Physically, the DSLAM shelf is a standard 19" or 21" rack mount with 10" or 12" shelves. Instead of a full 7' rack, the DSLAM is generally a half-height unit about 3½ feet high. However, DSLAM shelves are stacked on top of each other (usually three or four shelves per rack).
Each shelf can have multiple card slots (8, 12, and 16 cards are common configurations), one (or two) of which is typically reserved for a power module and network management module for the shelf. Power and management redundancy is often available, and some shelves allow hot-swappable cards. Each card can typically support multiple access loops, anywhere from 4 to 24.
Access lines enter the DSLAM from the main distribution frame (MDF). These lines might be individual pairs run as customers buy DSL services, but the lines can be gathered in multi-pair cables such as 50 pair or more. The individual pairs are dressed in to the shelves and plugged into the ATU-Cs (these pairs are not illustrated in the graphic to cut down on visual clutter).
DSLAMs can be priced anywhere from $20,000 for a small, bare bones unit all the way up into the hundreds of thousands of dollars, depending on size, redundancy, functionality, and manageability.
The DSLAM can be located in a number of locations including the CO, the customer’s premises, or near the customer’s premises. There are benefits and liabilities to each of these locations.
The DSLAM in the Central Office
When the DSLAM is in the central office (CO), it is in a location physically controlled by the telephone company. This allows the phone company to provide a physically secure location safe from theft, vandalism, and physical hacking. The location can also be controlled for heat, humidity, and power fluctuations, each of which could be detrimental to the equipment. The CO is shielded from radio frequency interference (RFI) that can cause significant impairments to DSL technologies.
The CO location makes DSLAM support easy. Troubleshooting can be done close to the equipment, allowing direct observation of visual alarms, sounds, and other indicators. If maintenance is needed, personnel and spare parts are close by.
The primary liability of having the DSLAM in the CO is that it increases the distance to the customer, which limits the number of customers who qualify for the service. The distance factor also affects the transmission rate that can be supported on the local loop due to the capacitive nature of twisted-pair cables. Local loops to the CO are usually older and can have bridge taps or mixed wire gauges, which degrade or prohibit DSL service.
The DSLAM on Customer Premises
In certain circumstances such as multiple tenant units (MTU), multiple dwelling units (MDU), or campus environments, the DSLAM can be located on the customer premises. There are benefits and liabilities to locating the DSLAM on customer premises.
When the DSLAM is on customer premises the customers are almost always located within the DSL service area (i.e., they are usually within a few hundred feet from the DSLAM), therefore pretty much everyone qualifies for DSL. The local loop in this environment is likely to be free of bridge taps and mixed wire gauges. Wiring is likely to be run through walls or embedded in the earth, which picks up less RFI, resulting in a much higher transmission rate and a lower bit error rate (BER).
On the negative side, when the DSLAM is located on customer premises the service provider loses direct control of the device. This makes the equipment more susceptible to vandalism and environmental damage. While the wiring between the DSLAM and the customers is more likely to be shielded from RFI, the DSLAM location is not likely to have the RFI shielding present in the CO. Troubleshooting and repair are much more difficult in a remote location, particularly if direct access to that location is not available.
Costs to the service provider can be increased if space must be leased or purchased. Providing a temperate, dry, secure space with adequate power protection can also be difficult or very expensive.
The DSLAM near the Customer Premises
There are occasions when the DSLAM is located in an environmentally hardened cabinet to serve a neighborhood or development.
The DSLAM near the customer premises has most of the same advantages as the DSLAM on customer premises; primarily, proximity to the customer location and a local loop with no bridge taps or mixed wire gauges. This means higher speed DSL technologies like ADSL2, ADSL2+, and VDSL can be used. This arrangement is commonly part of a fiber-to-the-node (FTTN) deployment.
This type of DSLAM comes in two basic configurations. Remote access multiplexers (RAM) are usually installed in the digital loop carrier (DLC) cabinet that is served by fiber. The RAM is essentially a mini-DSLAM. The other configuration consists of line cards that perform the DSL function and fit into open slots in the DLC itself. This configuration is only possible with newer DLCs and also requires a fiber connection back to the CO.
The disadvantages of locating the DLSAM near the customer premises are similar to the disadvantages of locating it on customer premises: troubleshooting and repair difficulty, right of way, and the potential for vandalism and environmental damage. Other difficulties with this configuration include heat dissipation, lack of physical space in existing facilities, and the ability to supply adequate power to the location.
|<mp3>http://podcast.hill-vt.com/podsnacks/2008q2/dslam.mp3%7Cdownload</mp3> | Digital subscriber line access multiplexer (DSLAM)|