Bandwidth in signal processing describes the range of frequencies that a signal, system, or channel occupies or can pass. In its simplest form it is the difference between a system's highest and lowest significant frequencies. In radio, audio, optics and electronics this spectral extent determines how much of the frequency domain is used and which signals will be transmitted, filtered, or attenuated.
Definition and measures
Bandwidth is typically measured in hertz (Hz) and for many practical systems is defined by the frequencies at which power falls to a specified fraction (often half-power) of the peak. Two common senses are:
- Passband bandwidth — the width of a band of frequencies that a bandpass system accepts.
- Baseband (or absolute) bandwidth — the highest frequency component present in a lowpass signal.
Other terms such as effective bandwidth, fractional bandwidth, and spectral width are used when signals have non-uniform spectra or when relative measures are more useful.
History and theoretical context
Understanding bandwidth became central as communication systems matured. Information theory links spectral bandwidth to data-carrying capacity: wider spectral bandwidth allows higher information rates for a given noise environment and modulation scheme. Foundational results such as the relationship between bandwidth, signal-to-noise ratio, and maximum achievable data rate are widely cited in engineering.
Practical uses and examples
Bandwidth is a practical design and regulatory parameter. Examples include:
- Radio channel allocation: each station occupies a designated frequency band.
- Audio systems: perceived sound quality depends on the audio bandwidth reproduced.
- Filters and amplifiers: designers specify bandwidth to shape signals.
- Optical communications: laser sources and filters have spectral bandwidth affecting multiplexing.
Distinctions and notable points
In networking contexts "bandwidth" is often used informally to mean data throughput (bits per second). While related, spectral bandwidth (Hz) and data rate (bps) are not identical: modulation, coding, and signal-to-noise ratio determine how many bits can be carried within a given spectral bandwidth. Tools used to measure or visualize bandwidth include spectrum analyzers and signal generators.