A Brief About The T1

There are a couple of reasons for spending some time understanding T1:

• T1 is often used as an access circuit technology for services provided to business, government and other organizations. For example, “Dedicated T1s”, Frame Relay and ISDN PRI all use T1 as the access… so whether you like it or not, you may have to end up dealing with T1.
• T1 employs synchronous multiplexing: framing, channels, multiplexers. Once you understand how these things work by using T1 as an example, then you know how the real stuff: SONET works.

T1 History and Current Applications

The most important thing to understand about T1 is its history.  The T1 Carrier System was designed in 1958 by Bell Labs.  The main requirement for this system was to increase the circuit density on existing copper wiring between Central Offices… to increase the density of revenue on existing copper long-distance circuits.  In the old days, this might have been called a pair gain system, because it increases the number of circuits actually carried on each pair of wires. T1 carries 24 digitized voice signals on a single set of copper wires.  It was originally designed and deployed by AT&T for use by AT&T for long-distance voice calls.

Today, T1 is not used much for long-distance circuits.  Fiber optic- and microwave-based carrier systems are used for long-distance circuits.  Instead, T1 is now used more as an access circuit installed from the customer premise to the service provider’s office, to provide “high speed” access to carriers’ voice and data services.

With T1 access, the customer can have connections at T1 speeds (1.5 Mb/s) across the carrier’s network all the time (a “dedicated T1” service), connections for the duration of a communication session (ISDN PRI service), and bandwidth on demand services across the network at T1 access speeds (Frame Relay and IP VPN services).

Basic T1 Circuit Components

A basic T1 system consists of multiplexers, Channel Service Units (CSUs) and the T1 circuit, which is four copper wires with repeaters every mile or so.

The multiplexer (one at each end) is variously referred to as a T1 Multiplexer, a mux, or a channel bank.  Since T1 was designed to carry 24 trunks over 4 copper wires, the mux provides 24 ports, each one running at 64 kb/s. The multiplexer’s aggregate or high-speed output connects to a CSU. The CSU is the interface device connecting the T1 multiplexer and the actual T1 circuit. This device is the one that represents binary digits on the physical T1 circuit.   It performs the same functions as a modem - but since it is a digital device, it is not called a modem.

The T1 circuit is four copper wires, two for each direction.  Binary digits are represented on these copper wires using pulses of voltage.  Repeaters are spaced every 6 kft (1 mile / 1.6 km) along the T1 circuit.  The data rate on the T1 circuit is 1.544 Mb/s, which is a DS1-rate signal.  1.544 Mb/s includes extra overhead bits added in by the multiplexers for framing.

The entire system is two-way simultaneous: both directions at the same time.  When we say “input”, we should really say “input and output”… but it is easier to think about this one direction at a time.

Repeaters are required every 6,000 feet on a T1 circuit, because of attenuation caused by the copper wires.  This is exactly the same problem that we encountered when discussing analog techniques and maximum loop lengths. It is possible to boost up analog signals using an amplifier, to be able to transmit information more than the maximum loop length; however, the amplifiers boost up both the signal and the noise on the line, making analog transmission noisy and distance-limited. The advantage of this digital technique using pulses is that the repeaters do not boost up the incoming signal: repeaters are binary devices that make a decision.  If they decide they detect a pulse coming in, they regenerate a new noiseless square pulse to send down the next cable segment. Using this idea, information can be coded into 1s and 0s, which are in turn represented as pulses on the line and transmitted long distances via regeneration of the pulses at regular intervals.

The result is communicating the information without adding in any noise; so quiet, you could hear a pin drop.

How T1 Is Provided

We do not use T1 for long-distance circuits much at all anymore… we have much better technology in the form of fiber optic transmission systems, capable of radically higher capacity and radically lower per-channel cost.  Synchronous Optical Network (SONET) technology, discussed in upcoming sections, is the most common type of fiber system used for connections between Central Offices, and long-haul connections. In practice, T1 is an access technology, used only for the last mile or two: copper wires used to get a DS1-rate service from the customer premise to these fiber systems. The information is then transported on fiber long-distance.

T1 is a 4-wire copper circuit running from the local phone company’s building (usually a Central Office) to the customer premise.  At the customer side, the wires are terminated on a Channel Service Unit (CSU), which usually provides an interface for the customer’s multiplexer.  At the Central Office, the wires are terminated on an Office Channel Unit, which performs the same functions.  If obtaining service from an Inter-Exchange Carrier (IXC) for a long-distance circuit, the circuit will be carried through the local phone company’s CO to the IXC’s Point of Presence (POP).

Fiber backbone transmission systems carry multiples of DS3-rate signals, not DS1s, and so the information on the T1 will be combined or multiplexed with many other information streams to form DS3-rate streams.  On newer systems, the SONET carrier system is used for synchronous (byte-interleaved) DS3 multiplexing. Older systems use an M13 multiplexer, which takes 28 DS1 inputs and forms them into a single asynchronous (bit-interleaved) DS3 output. These DS3-rate streams are then moved long distance.  At the far end, the reverse process takes place with similar equipment and cabling. If the other end is in Europe, the signal may be delivered as an E1 over the CEPT-1 carrier system.