DSLAM Definition: A DSLAM, or Digital-Subscriber-Line-Access-Multiplexer, is a network distribution device that aggregates individual subscriber lines into a high-capacity uplink. These high capacity uplinks, either ATM or Gigabit Ethernet, connect subscribers to their Internet service providers (ISPs). DSLAM units are typically located in telephone exchanges or distribution points. They allow for the high-speed transmission of DSL technology using legacy copper lines. Using advanced multiplexing techniques, these units salvage the utility of the millions of copper lines that were originally deployed for telephone usage in the 1950’s. DSLAMs also come with many advanced traffic management features to separate and prioritize voice, video, and data traffic.
How does a DSLAM connect to customer premise equipment?
DSLAMs are the intermediary units that link end-user equipment to ISP servers located in a central office (CO). ISPs provide end-users with customer premise equipment (CPE) such as routers or modems. These units forward a user’s digital data from their computer or client device to a local cabinet located in the vicinity of a customer’s premise. The data can then travel through a switch, a router, and finally a DSLAM unit.
The DSLAM unit will aggregate or collect individual subscriber lines and transfer data from all their subscribers onto a high-speed capacity uplink that connects to a carrier’s central office using fiber or twisted pair cabling. DSLAMs allow Internet service providers to build Hybrid networks such as fiber to the curb (FTTC) networks. By using fiber for backhaul traffic and twisted copper cables for the last mile of a deployment, ISPs are able to build cost-effective networks that offer high-speed transmission rates.
Once data arrives to a central carrier office, information is routed to a broadband remote access server (B-RAS). These units are responsible for authenticating subscriber credentials, validating user access policies, and routing data to their destinations.
DSLAMs can be classified by the type of xDSL technology they support, by form factor, by architecture, and by deployment location.
By xDSL Type: Single-Service DSLAMs vs. Multiservice DSLAMs
DSLAMs can be either classified as single-service or multiservice units. Single-service DSLAMs only have the capacity of supporting one xDSL technology. Most single-service system DSLAM units will boast backwards compatibility with previous versions of the xDSL type they support. An ADSL2+ DSLAM for example, will boast backwards compatibility with ADSL2 and ADSL, the two previous generations of the ADSL2+.
Multiservice DSLAMs have the capacity of supporting several xDSL technologies. Multiservice DSLAMs allows ISPs or carriers to address the different broadband needs of their customers. For example, a DSLAM chassis that supports VDSL and ADSL line cards gives service providers the advantage of delivering high-speed broadband to customers in short (using VDSL) and long distances ranges (using ADSL). To learn more about the difference between VDSL and ADSL, click here. Multiservice DSLAMs allow ISPs to address scalability, port density and redundant architecture requirements for large-scale deployments.
By Deployment Type
DSLAMs can also be classified by deployment location. A DSLAM designed for outside plant (OSP) deployment such as the VX-M208S, has a smaller subscriber capacity and a smaller form factor in comparison to a DSLAM designed for central office (CO) deployment. OSP DSLAMs are commonly deployed in multi-dwelling units such as apartment complexes or university campuses. These units reside closer to a subscriber’s location and terminate subscriber local loops to achieve faster data transmission rates. “Hardened” OSP DSLAMs provide protection against the elements.
CO DSLAMs are located in distribution points and can support up to 10,000 subscriber lines or more. CO DSLAMs typically reside in distributed shelf architectures. These shelf units can host a number of DSLAMs from different vendors and Internet service providers. CO DSLAMs need to fulfill stringent standards due the large number of subscribers they support. Many CO DSLAMs feature a chassis-type form factor with hot swappable line cards. These units allow ISPs to customize their DSLAMs into multiservice units.
By Form Factor
DSLAMs range in size and interface options. Single-service DSLAMs typically deployed in OSP environments, for example, offer a smaller footprint than CO DSLAMs. These OSP DSLAM units are sometimes referred to as pizza boxes to describe standalone units. CO DSLAMs are typically chassis DSLAMs with swappable line cards and uplink modules. Service providers can oftentimes customize these larger DSLAMs with line cards to support multiple xDSL services. This allows them to fulfill different bandwidth demands and subscribers located at varying distances.
DSLAMs can also be classified by architecture. Centralized models reserve a single central uplink card to perform complex traffic processing. Line cards in centralized models hand-off traffic to the uplink card. In comparison to distributed models, line cards in centralized models offer a more basic function. Centralized architectures are designed to support a high number of subscribers.
DSLAMs with distributed architectures reserve complex traffic processing for smart line cards that are based on programmable network processors such as linecard traffic processors (LTPs). Uplink cards can be in an Ethernet switch if the unit is used in conjunction with Ethernet backhaul or in a full-featured network processor.
What is DSL?
DSL provides internet subscribers with high-speed internet access using the same legacy copper lines originally deployed in the 1950’s by traditional telephone lines. DSL relies on DSLAM’s multiplexing capabilities to transmit digital data or analog signals of several subscriber lines using one uplink. Multiservice DSLAMs can support many DSL technologies, but there are currently no DSLAMs that support all xDSL types. DSLAMs can evenly (or symmetrically) or unevenly (asymmetrically) allocate bandwidth between downstream and upstream speeds. One of the major downsides of DSL is that speeds attenuate the farther away a subscriber is located from a telephone exchange or distribution point. But DSL continues to be a popular deployment option due to its low deployment cost and the option to pair with faster cabling options such as fiber.
ADSL- Asymmetrical Digital Subscriber Line (ADSL)
ADSL prioritizes downstream traffic and allocates only a small portion of bandwidth for upstream traffic. The original ADSL standard could achieve downstream rates of 8.0 Mbps and upstream rates of 1 Mbps. ADSL2/2+ are the improved version of ADSL. ADSL2 can achieve downstream rates of 12 Mbps and upstream rates of approximately 1.3 Mbps. ADSL2+ can achieve even faster downstream rates at around 24 Mbps and comparable upstream rates with ADSL2. The ADSL standard is normally used for distances of up to 18,000 ft.
ADSL deployments originally required professionally installers to install splitters, or microfilters, to separate the DSL data lines from POTS (plain old telephone connection). G.lite is an ADSL substandard that uses different modulation profiles and does not explicitly require the installation of splitters. G.lite can achieve 1.5 Mbps downstream and 512 Kbps upstream rates (at 10,000 ft). With G.lite, splitters are installed locally inside a subscriber’s homes. The asymmetric standard can achieve distances of up to 18,000 ft.
VDSL- Very High Bit-Rate Digital Subscriber Line (VDSL)
VDSL is optimal for shorter distances and signals quickly attenuate after 6,600ft. VDSL can achieve downstream rates of 55 Mbps and upstream rates of 1.5-2.3 Mbps. VDSL2 can achieve downstream rates of 200 Mbps and upstream rates of 200 Mbps up in the first 1,000ft. While VDSL is considered asymmetric, VDSL is both symmetric and asymmetric.
SDSL- Symmetric Digital Subscriber Line (SDSL)
Unlike ADSL which unevenly or asymmetrically allocates bandwidth between downstream and upstream traffic, SDSL evenly or symmetrically allocates bandwidth between downstream and upstream rates. With the ability to reach up to 9,800ft, SDSL typically yields around 1.5 Mbps, depending upon the distance of a customer’s equipment. SDSL is ideal for small businesses with more intensive bandwidth use and offers an ‘always on’ connection.
ISDN- Integrated Services for Digital Network
ISDN was the first protocol to integrate data and voice over copper cables and was traditionally used to carry voice for landline communication purposes. The standard supports data transfer rates of 64 Kbps. B-ISDN is an uncommon version ISDN that utilizes broadband transmission and can achieve rates of 1.5 Mbps with fiber optic cables. Additional ISDN substandards include basic rate interface (BRI), primary rate interface (PRI), and narrowband ISDN (N-ISDN).
IDSL- ISDN Digital Subscriber Line
IDSL is a digital transmission-based technology that eliminates the need to travel to a carrier’s central office. The xDSL standard can achieve 128 Kbps over twisted pair copper. Even though IDSL is a subsidiary of ISDN, IDSL allows for always-on connections and offers a more cost-effective option that eliminates setup delays and per minute fees. Transmission of data occurs over the data network as opposed to the PSTN (public switching telephone network).
HDSL- High-bit-rate Digital Subscriber Line
Even though the HDSL standard was first introduced in 1994, HDSL is still widely used by telephone companies and carriers. Performance is comparable to a T1 line though it is more cost-effective. HDSL can travel up to 12,000 ft and deliver symmetric rates of up to 784 Kbps.
Other High-Speed Alternatives To Connect to The Internet
Besides DSL, high-speed broadband can be accessed via coaxial cables, fiber, or wireless connections. The following will overview the different benefits and drawbacks to different internet connectivity methods.
Cable originally emerged as a means to deliver access to television programming in mountainous and remote areas. But subscription-based programming did not flourish until the deregulation of the industry in 1984 which spurred carriers to invest “more than $15 billion on the wiring of America” according to this CalCable. But with the widespread adoption of the internet, audiences began to consume content online using popular streaming sites such as Hulu and Netflix.
But carriers were able to salvage coaxial lines using DOCSIS standards (data over cable service interface specification). DOCSIS enables carriers to transmit high-bandwidth data using existing cable coaxial wiring used for cable television. DOCSIS standards have significantly evolved and now offer data speeds that are oftentimes faster than DSL.
DOCSIS 3.0, can achieve downstream speeds of up to 152 Mbps and upstream rates of up to 108 Mbps. But the newest iteration of DOCSIS 3.1 promises to deliver downstream rates of up to 10 G and upstream rates of up to 1 Gbps in laboratory environments. Real world rates tend to dramatically fluctuate, but improvements like these will continue aiding carriers in providing faster services for their customers.
A cable modem termination system (CMTS) in a coaxial network essentially performs the same function as a DSLAM unit in a DSL network. In the same way that a DSLAM feeds subscriber lines to the Internet service provider (ISP), a cable modem termination system (CMTS) feeds the data of hundreds of cable modems and connects users to their ISPs.
Cable relies on a shared line architecture and user speeds can drastically decrease during peak usage. However, cable will typically deliver faster rates than DSL. DSL speeds attenuate the farther away a customer is from a distribution point. With coaxial cable connections, however, the distance from a distribution point does not influence speed.
Many infrastructures already have coaxial cabling and like DSL, it is relatively inexpensive to connect.
Fiber connections offer longer distances and faster transmission speeds in comparison to coaxial cable, wireless, and DSL. Fiber uses light technology to transmit data at up to 1Gbps speeds. Using light technology allows fiber to achieve higher frequencies and data capacities. In comparison to copper-based cabling like DSL and coaxial lines, fiber operates in a near noise-free networking environment with very little interference or energy loss.
Fiber optics is also more costly to deploy than DSL or coaxial cabling. Newly built buildings will include twisted pair copper in their infrastructure making it simple for ISPs to provide connectivity using DSL. But fiber is oftentimes deployed after the construction of a building and represents an additional investment. Fiber is also an intrusive medium to deploy—at times damaging subscriber’s property in the most extreme cases. High deployment costs influence carriers to only deploy fiber in high subscriber density areas such as metropolitan areas. To alleviate the high cost of fiber, carriers will oftentimes build hybrid deployments using fiber and twisted pair copper to create FTTC (fiber-to-the-curb) deployments.
WISP (Wireless Internet Service Providers)
Wireless Internet is supported by radio towers that transmit data in the following ranges: 900 MHz, 2.4 GHz, 4.9, 5.2, 5.4, 5.7, and 5.8 GHz. Wireless Internet service providers (WISP) are carriers responsible for providing Internet connectivity to mobile client devices such as cell phones and wireless hotspots.
Wireless Internet services are the least common types of deployments. Unfortunately, wireless coverage can be spotty and unreliable. Frequent travelers, for example, may note that performance varies by location during the commute of a train. There are several factors that can influence the performance of a wireless connection including altitude or the physical barriers of a building for example.
Why DSL Is Still In Use
Twisted copper pairs is a legacy cabling medium that deteriorates with time and can become a liability without proper maintenance. Verizon has been accused of allowing their DSL copper lines to deteriorate so as to pressure residents into adopting fiber. But broadband providers will continue to rely on DSL technology due to low start-up costs.
Twisted copper pairs can also be used with fiber to build FTTC (fiber to the curb) deployments using DSLAMs. The most expensive portion of fiber deployment occurs in the local subscriber loop where customer premises are located. To avoid some of the high deployment costs of fiber, carriers will oftentimes build hybrid deployments using copper in the local subscriber loop and fiber in the remaining portion of a network. This form of deployment is known as FTTP (fiber to the premises) or FTTH (fiber to the home).
Constant improvements in DSL equipment and chipsets in DSLAMs allow service providers to take advantage of the millions of copper telephone lines that have already been deployed. New chipsets such as G.Fast have been able to achieve up to 1 Gbps at its origin. Improvements such as this will continue to prolong the lifespan of copper pairs.
DSLAM Use Cases
Higher capacity central office (CO) DSLAMs are used in distribution points to continue forwarding packets to their destination. But smaller single-card DSLAMs are also used in customer premises in multi-dwelling units (MDU’s) such as campuses, hotels, businesses and enterprise network environments.
DSLAM Deployment Locations:
- rural areas
DSLAMs optimize high-speed transmission by terminating local subscriber loops and transferring traffic into a high capacity uplink. In other words, connecting a series of modems to a DSLAM allows a higher-quality link such as fiber to take over to connect customers to the Internet.
DSLAMs in Rural Areas
Broadband carriers find rural and remote areas unappealing due to low subscriber density. Areas with low subscriber density offer lower returns of investment in comparison to metropolitan areas that boast higher subscriber density per square mile. The Connect America Funds incentivize broadband service providers to bring high-speed connectivity to rural areas. According to the Federal Communications Commission’s (FCC) Connect America Fund (CAF), “approximately 19 million American still lack access” to high-speed broadband. DSL is the preferred type of method in these types of sparsely populated areas due to low startup costs.
Internet service providers (ISPs) are able to provide high-speed broadband to a low volume of subscribers using single card DSLAMs such as the VX-M208S or the VX-M2024S. These units are ideal for smaller scale deployments.
ATM DSLAMs and IP DSLAMs
DSLAMs rely on ATM and IP packet switching technology to transport data. The following will demystify how the different methods transport information.
ATM DSLAMs use the ATM protocol to relay data using permanent virtual circuits (PVC’s) to relay data. These PVC’s require configuration to establish a permanent point to point (PPP) connection to a destination using a virtual circuit.
The ATM protocol splits data into cells made up of 53 bytes. These cells contain very little routing information due to the PPP nature of PVC connections. ATM networks can transport cells at rates of up to 155 Mbps and 622 Mbps.
The ATM protocol establishes a virtual circuit connection from a subscriber to a DSLAM, and then to a B-RAS. The B-RAS then terminates the PPP session and routes traffic to the core network.
As broadband began to add more complex data traffic, ATMs began to incorporate a rudimentary ATM switching fabrics, switched virtual circuits (SVCs), and a variety of other traffic management features.
Broadband now includes many value-added services such as VoIP (voice-over-IP), IPTV (Internet protocol television), VoD (video on demand) and HDTV (high-definition TV). With new concerns for bandwidth, scalability and QoS requirements, IP DSLAMs have managed to consolidate network functions and simplify network deployments. Many IP DSLAMs now have routing capabilities, reducing the number of equipment needed when compared to ATM DSLAM deployments.
IP DSLAMs are a cost-effective alternative to ATM DSLAMs. Many service providers opt to build their networks using Ethernet for their backhaul uplinks. Ethernet, such as Metro Ethernet, can be used for both carrier backbone and access network segments.
Ethernet DSLAMs, or IP DSLAMs, transmit IP-based data known as frames as opposed to ATM-based packets, or cells. Unlike ATM cell relay, frame relay is a packet switching technology that transmits different sized frames. A frame carries more addressing and error handling identifier tags than ATM packets.
Unlike ATM DSLAMs that rely on virtual circuits to relay data to their destinations, IP DSLAMs rely on switches and relay data across constantly-shifting connection paths. However, the frame relay protocol can also be configured to use PVC to forward packets to their destination using permanent pathways as ATM cells do to achieve faster speeds.
The growing complexity of broadband traffic such as Triple Play services known as VoIP, IPTV, and HDTV, have made IP-based DSLAMs and IP-Based architectures popular to do their cost-efficiency and simplified network architecture.
Carrier Ethernet, such as Metro Ethernet, can be used for backbone and access network segments. Ethernet standards are constantly being expanded and improved. In fact, the Ethernet Alliance has recently announced new standards for the backhaul of networks:
- 25 Gbps Ethernet PMD(s) for Single Mode Fiber Study Group
- 50 GBPS Ethernet Over A Single Lane Study Group
- Next Generation 100Gbps and 200Gbps Ethernet Study Group
Be sure to visit the Ethernet Alliance website to learn more about these new standards.
With constantly evolving Ethernet standards, Ethernet has become an integral component that maintains IP-Based networks cost-effective.
Buying a DSLAM
There are several features that DSLAM buyers will need to take into consideration when weighing different DSLAM options. The main differentiating features are subscriber capacity, throughput, packet loss, latency and jitter.
DSLAMs provide a range of subscriber capacity. There are three main metrics that dictate subscriber capacity: line density, subscriber and session capacity. Throughput measurements overview a variety of network environment factors that may influence the overall sustainable throughput of a unit including packet sizes, session volumes, and other network environment features such as IGMP snooping, QoS, AAA, and other related features (depending on the capabilities of a DSLAM).
DSLAMs support anywhere between a single subscriber to tens of thousands, depending on the type of DSLAM and functionality needed. CO DSLAMs can provide sufficient support for thousands upon thousands of subscribers while OSP DSLAMs can provide sufficient support for as little as one subscriber.
Throughput allows carriers to differentiate their service packages from their competitors and is one of the most important factors that carriers take into consideration when deciding which DSLAM to purchase. Though throughput is influenced by a variety of factors, the dominant factor that will determine the performance of a unit will depend on upon the type of xDSL technology used and the location of a customer’s premise. For example, a subscriber that is closer to a central office server of their ISP, will be able to experience faster rates using VDSL2 than a subscriber that lives farther away using the same equipment and xDSL technology. Robust QoS features further improves the accuracy of throughput in real-world settings.
Packet Loss, Latency and Jitter
Broadband has grown in complexity and supports more complex types of traffic such as VoIP, IPTV, and VoD (often known as Triple Play services). These more complex types of traffic are more sensitive to delays or latency and requires more advanced traffic management features to reduce packet loss, latency and jitter. These parameters will influence the performance of a DSLAM. Features such as QoS, Authentication via DHCP Relay, and IGMP Snooping alleviate packet loss. ISPs and network installers can also set the prioritization of voice, video and data traffic to optimize performance. Since voice is more sensitive to delays, incoming and outgoing voice traffic can take priority over data traffic.
Determining the Best DSLAM Units For Your Network
As mentioned before, network installers will need to assess the amount of subscribers they are seeking to serve and the distance ranges they are seeking to cover. DSLAM units come in a variety of sizes with different subscriber capacities. There are a myriad of DSLAM options built for large-scale deployments that can support several thousand subscribers. But there are also single-card DSLAMs that can support a handful of subscribers.
DSL performance rates will depend on upon the distance of a subscriber’s location to the central office (CO). DSL performance is mainly dictated by the type of DSL service a DSLAM supports. Installers will most likely choose VDSL/2 services for distances of up to 6,600ft and ADSL2/2+ for distances greater than 6,600 ft.
As broadband data has grown more complex, DSLAMs have had to account for value-added triple play services placing greater importance on traffic management features.
Common DSLAM features include:
- Traffic management
- Authentication via DHCP Relay
- IGMP Snooping
To demonstrate the range of DSLAM options available, we’ve selected a few examples of different DSLAM equipment types from our product portfolio. These units support varying subscriber capacity and DSL service types.
24 Port VDSL2 IP DSLAM
Models such as the VX-MD4024 24 Port VDSL2 IP DSLAM are also suitable for small scale deployments. As a network grows in size, additional units from different vendors can be added to a network. These DSLAMs are ideal for multi-dwelling units or external cabinets.
48 Port ADSL2+ IP DSLAM
Devices such as the VX-1000HDx provide longer distances and are designed for access networks. These units are heftier in size measuring around 1.5U.
Chassis-type DSLAMs feature hot-swappable line cards. Service providers can customize these DSLAMs to build multi-service units. Units such as these can feature Gigabit Ethernet (GbE) trunk interfaces and SFP ports for fiber connectivity.
Knowing the approximate location of your nearest DSLAM will help you more accurately gauge the expected speed of your Internet service. But DSLAM maps are very rarely found online. If you’re a potential DSL subscriber, and are searching for a DSLAM map to determine potential speeds, you can contact your Internet service provider. They should be able to give you approximate speeds based on your location.
Configuring a DSLAM
With value-added services such as IPTV (Internet protocol television), VoIP (voice over IP), and HDTV (high definition TV) configuring a DSLAM requires setting the traffic prioritization of voice, video and data traffic. Users will need to configure virtual local area networks (VLANs), QoS, and reserve a set amount of bandwidth for voice—the traffic type most sensitive to latency—on their switches. Uplinks will also need to be connected to the DSLAM Ethernet or fiber port.
Users will need to configure data traffic in the following order:
First tier: Voice
Third Tier: Data
To preview how to setup a DSLAM, click on the video below:
To view our full DSLAM product portfolio, click here.
Understanding DSLAM and BRAS Access Devices by Agilent