Bandwidth is the width, in hertz, of the frequency range a signal occupies or that a receiver captures.1 A narrowband voice channel may be ~12.5 kHz wide; an FM broadcast station ~200 kHz; a Wi-Fi channel tens of megahertz. Bandwidth is one of the most consequential numbers in radio: it caps how fast information can flow, it decides how much spectrum a signal consumes, and — on the receive side — it sets how much of the spectrum an SDR must digitise at once.
How it works
Bandwidth arises because information cannot be sent instantaneously. A steady, unchanging carrier is a single spectral line of zero width; modulating it to carry a message spreads its energy into sidebands, and the faster the modulation, the wider those sidebands. The exact width a signal genuinely uses is its occupied bandwidth, usually defined as the span containing 99% of the transmitted power. Regulators allocate each service a channel of a given width and demand that emissions stay within it, leaving a guard band of empty spectrum between channels to prevent interference.
There is a hard link between bandwidth and how much data a channel can carry. The Shannon–Hartley theorem gives the theoretical ceiling as C = B · log₂(1 + SNR), where C is capacity in bits per second, B is bandwidth, and SNR is the signal-to-noise ratio — see Shannon capacity. Capacity scales linearly with bandwidth but only logarithmically with SNR, which is why more bandwidth is the most direct way to raise data rate, and why crowded spectrum is such a valuable resource. Bandwidth cuts the other way too: a wider receiver passband admits more noise (noise power is proportional to bandwidth), so opening the filter beyond what the signal needs only lowers SNR.
In practice
- Channel width sets the standard. Land-mobile digital voice fits inside narrow channels — DMR packs two TDMA slots into 12.5 kHz, P25 Phase 1 uses 12.5 kHz, NXDN offers a 6.25 kHz mode — driving the choice of modulation and symbol rate.
- Match the filter to the signal. A receiver’s channel filter should be about as wide as the occupied bandwidth: wide enough to pass the signal undistorted, narrow enough to reject neighbours and noise.
- Capture versus channel. An SDR’s capture (or instantaneous) bandwidth is far wider than any one channel, letting it see a whole trunked system’s control and voice channels simultaneously.
Relevance to SDR
An SDR’s capture bandwidth is roughly its sample rate: by the Nyquist theorem a complex (IQ) stream sampled at fs represents a band fs wide. To monitor a trunking system, GopherTrunk captures a wide slice — several megahertz — then uses filtering and decimation, and a digital down-converter, to carve out each narrow channel at its own centre frequency and reduce it to the per-channel rate the decoder needs (48 kHz for the 4800-baud C4FM family). The wide capture is what lets one receiver follow a control channel and its voice grants at once; the narrow per-channel bandwidth is what keeps noise out of each individual decode.
Sources
-
Bandwidth (signal processing) — Wikipedia, definition and measures of signal bandwidth. ↩