One of the most popular types of communications circuits used for Wide Area Networking applications is the T-1 circuit. T-1 is a high-speed digital line which runs at a rate of 1.544 Mb/s, divided into 24 channels of 64 kb/s each. These lines provide a very high level of speed and flexibility, and can be used in many applications, including data and voice.
This white paper will discuss the basics of how T-1 circuits operate, and will then provide some sample applications showing how to use them. This text is not meant to be an in-depth reference guide to T-1 circuits, rather, it is is designed to provide the reader with a basic understanding of the technology and how to use it.
In order to understand how to plan and use a T-1 line, it is very useful to have a working knowledge of how the line itself works. As mentioned above, T-1 lines consist of 24 channels which are multiplexed together. T-1 lines use a technique called Time Division Multiplexing (TDM) to combine the 24 low-speed channels (also commonly called "DS0"s) over a single high speed circuit. Figure One shows a conceptual view of how TDM works.
Time Division Multiplexing
Figure One is a graph of three digital signals: Input 1, Input 2, and Output. The state of these signals is shown at times T1, T2, T3, T4, T5, and T6. Note that the input signals are running at 1/2 the speed of the output signal.
In TDM, we rapidly sample the inputs to create the output. At time T-1, we sample Input 1. At that moment in time, Input 1 has the binary value of 1, so that appears in the output. At Time T2, we sample Input 2. At that moment, Input 2 has the binary value of 0, so that is passed to the output. At Time T3, we return to Input 1, and pass its current value, 0, to the Output. At Time T4, we again sample Input 2 and pass its value, 1, to the output. Assuming that the particular multiplexor we are modeling has only two inputs and one output, this process of alternating between inputs and passing their current value to the output will repeat continuously.
TDM requires that the sum of all input speeds be less than the output speed. Otherwise, there is no way to scan the inputs fast enough to avoid losing data -- an input could change state before the multiplexor has a chance to sample its value and pass it to the output.
T-1 lines consist of 24 channels which are combined using TDM. The way this is accomplished is to divide the 1.544 Mb/s line into logical units called frames. Each frame consists of 193 bits, divided into 24 time slots of 8 bits each, with an additional one-bit time slot called an "F bit" added for synchronization purposes. Therefore, each channel runs at a rate of 64 kb/s, and 8 kb/s is used for overhead purposes. A T-1 frame is shown in Figure Two
T-1 Frame Format
The frames on a T-1 line are organized into larger structures called super-frames. This is done for purposes of link and system supervision. There are two types of frams which are used:
At the very basic electrical level T-1 uses a format for sending data bits called Alternate Mark Inversion, or AMI. In an AMI encoded signal, digital "ones" are sent as alternating positive and negative voltage levels, and digital "zeros" are sent as a value of zero volts.
In order for the equipment to remain synchronized with the line, we need to ensure that at least one out of every eight bits sent is a "one", as "zeros" do not provide any electrical transition to use for clock recovery.
Luckily, there are work-arounds to this problem. T-1 equipment is designed to use different line codes to replace long strings of "zeros" with a code which ensures at least one "one" will be sent in every group of eight bits. There are two such codes in use, B7ZS and B8ZS.
The exact nature of how these line codes actually operate is extremely complex, and is beyond the scope of this text. As a general rule, simply remember that B7ZS equates to each channel running at a speed of 56 kb/s, and B8ZS equates to each channel running at 64 kb/s. Generally, B8ZS is the preferred technology to use, as it gives a higher effective speed.
Fractional T-1 is a service offered by many providers of T-1 service for users who do not need all 24 channels provided by a standard T-1 line. In actuality, it is a full 24 channel circuit, however the provider will ignore any unused channels and only pass on data on the channels the user is being billed for. Since the provider does not have to forward all 24 channels, it is less costly to provide the service, so the provider can in turn offer the user a less expensive price. Common speeds for Fractional T-1 service are 64, 128, 256, 512, and 768 kb/s. This corresponds to 1, 2, 4, 8, and 12 channels.
Now that we have discussed the basics of how a T-1 line works, we can begin to address some applications for it. Basically, a T-1 line can be thought of as consisting of 24 independant 64 kb/s channels, or as a single high-speed "data pipe" between two points. It all depends on the abilities of the equipment being connected to the line.
The most basic T-1 application is as a point to point high speed leased line. In this configuration, we use a device at each end called a T-1 CSU/DSU with a single T-1 port and a single high-speed serial port (usually V.35). Since all 24 channels on the T-1 are being sent between two points, we effectively &uqot;bundle" them together and encase a high speed stream of data in the T-1 framing structure, and break it back out on the other end. A picture of such an application is shown if Figure Three.
Point To Point High Speed Application
Oftentimes, there is a need to combine multiple applications over a single T-1 circuit. In such a case, instead of using a simple CSU/DSU, a device called a T-1 Multiplexor is used. Such a device will have multiple user interfaces and a single T-1 interface. The user interfaces can be any of various types, such as V.35, RS-232, Voice, etc. An example of this type of application is shown in Figure Four.
T-1 Multiplexing Data & PBX T-1 Tie Line
The application shown in Figure Four is somewhat more complex. Here, we are taking advantage of the chanelized nature of the T-1 circuit and running two separate applications over the line at the same time. Channels 1-12 are used to create a 768 kb/s data circuit, and Channels 13-24 are used as a tie line between two phone systems (PBXs). The device making this application possible is the T-1 Multiplexor, also known as a T-1 Mux. In this case, the T-1 Multiplexors are configured with two V.35 ports and two T-1 ports. The rightmost T-1 port here is being used as the main link between the Multiplexors, and the second T-1 port is being used to create the PBX T-1 Tie Line.
The T-1 Multiplexors are intelligent devices, and are must be programmed to allow this application to work. Here, we would program the unit to route Channels 1-12 to the first V.35 port, Channels 13-24 to the second T-1 port, and no channels to the second V.35 port.
Note that the T-1 Tie Line actually contains 24 channels, however we have programmed the PBXs to only use Channels 13-24. Since a PBX will normally use an entire channel for each simultaneous call, this application allows 12 telephone conversations to be carried on at the same time, along with LAN traffic from the router at 768 kb/s.
The above applications are fairly simple and straight-forward. T-1 setups can be much more complex, as each channel can be assigned to a different application. For example, it is possible to have the provider of the T-1 circuit route certain channels to the Public Switched Telephone Network (PSTN) to provide dial-up circuits, other channels could be tie lines between two PBXs, still other channels can be allocated to data applications. In short, the combinations are really only limited by the imagination.