Bandwidth

Communication channels are classified as analog or digital. Bandwidth refers to the data throughput capacity of any communication channel. As bandwidth increases, more information per unit of time can pass through the channel. A simple analogy compares a communication channel to a water pipe. The larger the pipe, the more water can flow through it at a faster rate, just as a high capacity communication channel allows more data to flow at a higher rate than is possible with a lower capacity channel.

In addition to describing the capacity of a communication channel, the term "bandwidth" is frequently, and somewhat confusingly, applied to information transport requirements. For example, it might be specified that a broadcast signal requires a channel with a bandwidth of six MHz to transmit a television signal without loss or distortion. Bandwidth limitations arise from the physical properties of matter and energy. Every physical transmission medium has a finite bandwidth. The bandwidth of any given medium determines its communications efficiency for voice, data, graphics, or full motion video.

Widespread use of the Internet has increased public awareness of telecommunications bandwidth because both consumers and service providers are interested in optimizing the speed of Internet access and the speed with which web pages appear on computer screens.

Analog Signals

Natural signals such as those associated with voice, music, or vision, are analog in nature. Analog signals are represented by a sine wave, and analog channel capacities are measured in hertz (Hz) or cycles per second. Analog signals vary in amplitude (signal strength) or frequency (signal pitch or tone). Analog bandwidth is calculated by finding the difference between the minimum and maximum amplitudes or frequencies found on the particular communication channel.

For example, the bandwidth allocation of a telephone voice grade channel, which is classified as narrowband, is normally about 4,000 Hz, but the voice channel actually uses frequencies from 300 to 3,400 Hz, yielding a bandwidth that is 3,100 Hz wide. The additional space or guardbands on each side of the voice channel serve to prevent signal overlap with adjacent channels and are also used for transmitting call management information.

Digital Signals

Signals in computing environments are digital. Digital signals are described as discrete, or discontinuous, because they are transmitted in small, separate units called bits. Digital channel capacities are measured in either bits per second (bps) or signal changes per second, which is known as the baud rate. Although these terms are frequently used interchangeably, bits per second and baud rate are technically not the same. Baud rate is an actual measure of the number of signal changes that occur per second rather than the number of bits actually transmitted per second. Prefixes used in the measurement of data transmission speeds include kilo (thousands), mega (millions), giga (thousands of millions), and tera (thousands of giga). To describe digital transmission capabilities in bits per second, notations such as Kbps, Mbps, Gbps, and Tbps are common.

The telephone system has been in a gradual transition from an analog to a digital network. In order to transmit a digital signal over a conventional analog telephone line, a modem is needed to modulate the signal of the sender and demodulate the signal for the receiver. The term modem is an abbreviation of modulate-demodulate. Although the core capacity of the telephone network has experienced an explosion in available bandwidth, local access to homes and businesses, referred to as the local loop in the telephone network, frequently is limited to analog modem connections. Digital transmission is popular because it is a reliable, high-speed service that eliminates the need for modems.

Broadband Communications

Financial and other business activities, software downloads, video conferencing, and distance education have created a need for greater bandwidth. The term broadband is used to refer to hardware and media that can support a wide bandwidth. Coaxial cable and microwave transmission are classified as broadband. Coaxial cable, used for cable television, has a bandwidth of 500,000,000 Hz, or 500 megahertz, and microwave transmission has a bandwidth of 10,000 Hz.

The capacity potential of broadband devices is considerably greater than that of narrowband technology, resulting in greater data transmission speeds and faster download speeds, which are important to Internet users. Data transmission speeds range from a low of 14,400 bps on a low speed modem to more than ten gigabits per second on a fiber optic cable. On the assumption that 50,000 bits represents a page of data, it takes 3.5 seconds to transmit the page at 14,400 bps, but only 8/10 of a second at 64,000 bps. If a page of graphics contains one million bits per page, it takes more than a minute to transmit the page at 14,400 bps, compared to 16 seconds at 64 Kbps. Full motion video requires an enormous bandwidth of 12 Mbps.

Upload versus Download Bandwidth

Among Internet Service Providers (ISPs) and broadband cable or satellite links, there is considerable difference in upstream, or upload, bandwidth and downstream, or download, bandwidth. Upstream transmission occurs when one sends information to an ISP whereas downstream transmission occurs when information is received from an ISP. For example, a broadband cable modem connection might transmit upstream at one Mbps and downstream at ten Mbps.

Typical media used to connect to the Internet, along with upstream and downstream bandwidths include: T3 leased lines, T1 leased lines, cable modems, asymmetric digital subscribe lines (ADSLs), integrated services digital networks (ISDNs), and dial-up modems. As noted in Gary P. Schneider and James T. Perry's book Electronic Commerce, T3 leased lines provide the fastest speeds (44,700 kbps for both upstream and downstream speeds) while the rates for T1 leased lines are 1,544 kbps, ISDNs are 128 kbps, and dial-up modems are 56 kbps. ADSL upstream and downstream speeds are 640 and 9,000 kbps, respectively, while cable modem speeds are 768 kbps upstream and 10,000 kbps downstream.

Each of the connections has advantages and disadvantages. As the speed of the medium increases in the broadband media beginning with T1 lines, costs increase substantially. Although classified as broadband, cable modems are considered optimal in price and performance for the home user.

History of Bandwidth Research

Researchers have studied the effects of bandwidth on network traffic since the 1920s. Research objectives have always focused on the development of encoding techniques and technology enhancements that allow more bits to be transmitted per unit of time. In 1933 Harry Nyquist discovered a fundamental relationship between the bandwidth of a transmission system and the maximum number of bits per second that can be transmitted over that system. The Nyquist Intersymbol Interference Theorem allows one to calculate a theoretical maximum rate at which data can be sent. Nyquist's Theorem encourages data communications professionals to devise innovative coding schemes that will facilitate the maximum transmission of data per unit of time.

In 1948, noting that Nyquist's Theorem establishes an absolute maximum not achievable in practice, Claude Shannon of Bell Labs provided refinements to the theorem to account for the average amount of inherent noise or interference found on the transmission line. Shannon's Theorem can be summarized as saying that the laws of physics limit the speed of data transmission in a system and cannot be overcome by innovative coding schemes.

Thomas A. Pollack

Bibliography

Comer, Douglas E. Computer Networks and Internets. Upper Saddle River, NJ: Prentice Hall, 2001.

Frenzel, Carroll W. Management of Information Technology. New York: Course Technology, 1999.

Lucas, Henry C., Jr. Information Technology for Management. New York: McGraw-Hill, 2000.

Rosenbush, Steve. "Broadband: 'What Happened?'" Business Week, June 11, 2001, pp. 38-41.

Schneider, Gary P., and James T. Perry. Electronic Commerce. Canada: Course Technology, 2001.