Field Guide · term

Also known as: MDS, minimum discernible signal, minimum detectable signal, sensitivity

Receiver sensitivity — often quoted as the minimum discernible signal (MDS) — is the weakest input power a receiver can detect while still meeting a required quality threshold.1 It is not a single mysterious property of the radio but a sum of three legible terms: the thermal noise floor set by kTB, the receiver’s own noise figure, and the signal-to-noise ratio the chosen modulation needs to work. Add them and you have the faintest signal, in dBm, that will decode. Sensitivity defines the bottom of a receiver’s dynamic range and is the number that ultimately decides how far away a transmitter can be heard.

dBm −174 dBm/Hz (kT) + 10·log₁₀(B) → noise floor + noise figure (NF) + required SNR = sensitivity (MDS)
Sensitivity is a stacked budget: the kTB thermal density, widened by bandwidth to the noise floor, raised by the receiver's noise figure, then by the SNR the mode needs to decode.

How it works

The chain from physics to spec is short and exact. Start with thermal noise density, −174 dBm/Hz at 290 K. Spread it over the receiver’s bandwidth B by adding 10·log₁₀(B) in decibels — that gives the input noise floor. Raise it by the receiver’s noise figure, the noise the front end adds on top. Finally add the required SNR the demodulator needs for acceptable quality. The whole thing:

MDS = −174 dBm/Hz + 10·log₁₀(B) + NF + SNR_required

A worked example for a 12.5 kHz land-mobile channel with a 6 dB noise figure and a mode needing 8 dB of SNR:

−174 + 10·log₁₀(12 500) + 6 + 8 ≈ −174 + 41 + 6 + 8 = −119 dBm.

Every term is a lever. Narrower bandwidth lowers the noise floor (but must still pass the signal). A better low-noise amplifier first in the chain lowers NF. A more robust modulation or stronger FEC lowers the required SNR. Sensitivity improves — the MDS number gets more negative — whenever any of these drops.

Variants

  • MDS / MDS at 0 dB SNR. Purists define minimum detectable signal as the level where signal power equals the noise floor (SNR = 0 dB); the practical sensitivity adds the SNR the mode actually needs.
  • SINAD sensitivity. Analog and many commercial receivers are specified by the input level that yields 12 dB SINAD, a listening-quality criterion rather than a raw SNR.
  • BER sensitivity. Digital modes specify the input that achieves a target bit error rate (e.g. 5% BER for P25) — the SNR term is effectively “whatever SNR reaches that BER after FEC.”

Relevance to SDR

Sensitivity puts a floor under what any decoder can do, GopherTrunk included. Two SDRs may claim the same sensitivity on paper yet behave differently in the field because dynamic range, not sensitivity, is the limit in a crowded spectrum — a very sensitive receiver that overloads on a strong neighbour is useless. Still, when a band is quiet and a distant control channel simply will not decode, the sensitivity budget tells you where the margin went: usually the noise figure (add a mast-mounted LNA and shorten feedline) or the required SNR (nothing you can change about the mode).

GopherTrunk sits at the SNR_required end of this equation: its decode thresholds are statements about how much SNR each protocol and vocoder needs. It cannot lower a receiver’s MDS — that is fixed by kTB, bandwidth, and the front-end noise figure — so closing a weak link is an antenna, LNA, and feedline problem first, and a software problem never.

Sources

  1. Sensitivity (electronics) — Wikipedia, receiver sensitivity, MDS, and the noise-floor-plus-NF-plus-SNR budget. ↩

See also