CIS 361 Data Communications and Networks

Chapter 8: Data Transmission and Modems



This chapter discusses three kinds of data circuits, and devices used on them. The objectives important to this chapter are:

  • understanding the signaling rate of a circuit
  • understanding the speed of a circuit
  • describe three modes of data transmission
  • understanding synchronous and asynchronous transmission
  • understanding how modems work
The bulk of this chapter concerns methods of sending either analog or digital signals over either analog or digital wires.

Before getting into the nuts and bolts of it, Mr. Rowe gives us another formula on page 270. To understand it, you must understand some vocabulary. Signaling rate is defined as the number of times per second that the signal on a line changes. This is also called the baud rate. The theoretical signaling rate for any medium is two times the bandwidth. This is not practical, however. On a line with a bandwidth of 1 MHz, for instance, the theoretical signaling rate would be 2 MHz. This means that the sender would have to change the signal 2 million times per second. Even harder, the receiver would have to detect changes that lasted less than half of a millionth of a second. Another approach had to be used.

Circuit speed is defined as the number of bits a circuit can transmit in a second (bps). This is different from baud rate. It would be the same, if we only had two unique signals to send on the line. However, systems that can send more unique signals can increase the complexity of the information send with each change.

Transmission modes describe the kinds of conversations that are possible on communications lines. A simplex mode is one that flows one direction, like the remote control for a television. Half duplex mode means that transmission can flow each direction, but the nodes must take turns, like using a CB radio. Full duplex mode is when both parties in a conversation can transmit and receive at any time, like using a telephone. (These terms are used differently in Europe.)

Two types of physical connections are discussed on page 273. Parallel connections typically use several lines or wires simultaneously for transmission, while serial connections typically use one line or wire. The obvious speed advantage is with parallel transmission, but it is only useful for short distances.

The discussion of synchronous and asynchronous transmissions begins on page 273, as well. Asynchronous transmission is described first. In this method, each character that is sent has to be preceded by a start bit, a warning to the receiver that a transmission is about to happen. This adds a lot of overhead to transmissions. Also, some asynchronous methods require that the sender attach stop bits to the characters as well. The advantage to this system is that timing of transmissions is not critical, because the receiver is warned when to pay close attention. In synchronous transmission, he sender and receiver have to synchronize their internal clocks, because the sender will be sending characters on a specific schedule, at least for the duration of the conversation. The receiver will be expecting characters on this set schedule, so it knows exactly when to sample the line.

In asynchronous transmissions, a lot of the bits sent are start and stop bits. These greatly increase the number of bits per message. In synchronous transmission, only a few synchronization bits are sent at the beginning and end of messages. This is more efficient, but requires more expensive hardware on each end of the line.

Regardless of the methods used, signals can start out as analog or digital. The chart on page 277 shows four possible variations.

  • A telephone take an analog sound wave and creates an analog electrical signal with a modulator, to pass along analog lines.
  • A telephone might take the analog sound, and convert to a digital signal with a codec, to pass it along digital lines.
  • A computer creates digital signals, and may use a digital transmitter to pass them along digital lines.
  • A computer creates digital signals, and may pass them through a modem to convert them to analog signals that are passed along analog lines.
In each case, the appropriate kind of signal is passed down lines, either by creation or conversion to the right type.

On page 278, the text illustrates the fact that even if we are talking about digital signals, there are still several ways they might be represented on the communications line. For example, in Unipolar signaling, you are shown that 1s are represented by the presence of a signal, while 0s are represented by the absence of that signal. In Bipolar Nonreturn to Zero, you see that a positive voltage represents 1s, while a negative voltage represents 0 bits. The reference to "Nonreturn to Zero" means that this system sends one signal or the other, never resting in between. The third example show Bipolar Return to Zero, which adds a rest state (zero voltage) between data bit transmissions.

Digital signaling is preferred over analog signaling for the reasons on pages 279 and 280.

  • Better data integrity - less errors.
  • Higher capacity
  • Easier integration of signals carrying different kinds of information.
  • Better security - encryption is better.
  • Lower cost than in the past.

The illustration on page 281 show how an analog signal might become digitized. Essentially, the analog signal is measured, or sampled, many times per second. Each measurement can be converted to a digital signal. The faster the sampling rate, the better the representation of the analog signal. In actual use, the actual value measured may not be the value used. The digital signal generator may only be able to create pulse in integer values, as opposed to the fractional value that the analog wave may have at any given moment. The conversion is called quantization, and the difference between the actual value and the value used is called quantizing noise, or distortion.

On page 284, the text discusses analog transmission of digital signals. This is the most common method for users who dial into an ISP. Note the illustration on page 287 of the four frequencies commonly used in a simple modem, two frequencies for each end of the conversation. Even so, this gives us only two kinds of signal for each modem to send. To improve on the situation, we add phase modulation, as shown on page 288. There are four distinct phases to a wave. Combining these phases with two frequencies gives us eight unique signals.

In the examples on pages 289 and 290, only one frequency is used. On page 289, we see an example of Phase Shift Keying. In this method, one phase of the wave represents a 0 bit, while the other phase used represents a 1 bit. On page 290, there is an example of Differential Phase Shift Keying. Here, we see that a shift from one wave phase to another represents a 1 bit, while the act of not shifting represents a 0 bit. This says, in effect, that we are sending 0s as long as we do not change the wave, but we must change it every time we want to send a 1.

The evolution of modem standards is discussed on page 292. Note that the V.32, V.32bis, V.34 and V.34bis standards all have to do with speed of the modem. On page 295, other standards are mentioned: V.42 is an error correction standard, while V.42bis is a data compression standard. (This is unusual, since the French word "bis" just means "revised".)

Several types of modems are mentioned on page 296. The ones in the pictures are quite old now. Even older, but still usable is the acoustic coupler shown on page 299. This is the way modems had to be used in the beginning, actually putting the receiver of a telephone into a cradle and using the microphone and speaker in the transmission process. A null modem is mentioned on page 298. Usually this is a cable used to connect two computers directly, without using standard transmission media. The most commonly used current type of modem is shown on page 306, the internal modem that fits into a slot in a computer's bus.