C/N0 is the ratio of received GNSS signal power to noise power density. Receivers commonly report it in dB-Hz for individual tracked signals. It helps describe reception quality, but it is not position accuracy, an interference-suppression rating or proof that a signal is authentic. To interpret a value, identify the signal, the receiver and the observation conditions before comparing numbers.
This guide is for integrators reading receiver logs, antenna comparisons or interference-evaluation results. The practical question is what a C/N0 observation actually establishes—and what additional evidence is needed before changing equipment.
Why the unit is dB-Hz rather than dB
The distinction is the denominator. Signal-to-noise ratio, or SNR, compares signal power with the total noise power in a specified bandwidth. C/N0 instead uses noise power per hertz. Its logarithmic unit is dB-Hz, not dB/Hz. A larger value means more signal power relative to noise density at the stated observation boundary.
For the same signal power and approximately flat noise density across a specified equivalent noise bandwidth B:
SNR (dB) = C/N0 (dB-Hz) − 10 log10(B in Hz).
The bandwidth and processing stage must be stated. A front-end SNR and an SNR after receiver processing are not automatically the same quantity. The original technical explanation by Angelo Joseph in Inside GNSS sets out these definitions and the relationship.
Illustrative calculation, not a receiver specification: take C/N0 = 40 dB-Hz and an equivalent noise bandwidth of 1 MHz. The bandwidth term is 60 dB, giving SNR = −20 dB. With the same C/N0 and a 2 MHz bandwidth, the term is approximately 63 dB and SNR is approximately −23 dB. The different SNR numbers do not imply that the second observation has a lower C/N0. More noise has been included in the specified bandwidth.
Consequently, do not subtract an SNR value in dB from a C/N0 value in dB-Hz and call the difference an antenna improvement. First establish whether the two records describe comparable quantities.
A negative front-end SNR is not the same as an unusable receiver output. GNSS receivers correlate incoming signals with known ranging-code replicas and integrate the correlation results over time. That processing helps extract signals from noise; ESA's correlator explanation describes the mechanism. It does not make reception immune to interference or remove the need to specify the stage at which SNR is measured.
For the mechanism behind reception below the noise floor, see how GNSS correlation and processing gain work, including the limits of longer integration.
Read the signal record, not just the tallest bar
A receiver can track several signals from one satellite and signals from several constellations. A single maximum or average conceals which observations contributed to it. NovAtel's Satellite Tracking documentation illustrates a display separated by antenna, constellation, satellite and frequency, with lock time available separately. These are documented NovAtel interface features, not a claim about ResiNav outputs.
For an antenna comparison, pair observations of the same satellite and signal over comparable epochs and viewing conditions. Keep the receiver configuration and RF path identified. Do not compare the strongest high-elevation satellite in one run with an unrelated signal near the horizon in another and attribute the difference solely to the antenna. ESA's receiver-noise explanation identifies elevation-dependent signal strength as a factor in measurement noise.
An average can also improve when weak signals disappear from the log. A hypothetical record with values 30, 40 and 44 has a simple arithmetic average of 38 dB-Hz. If the 30 dB-Hz signal is no longer tracked, the displayed average of the remaining two becomes 42 dB-Hz, without either retained value improving. This arithmetic illustration is not a recommended averaging method. It shows why a higher summary value must be read alongside the membership of the tracked set and losses of lock.
Separate reception quality, navigation quality and protection performance
| Observation | What it can help establish | What else matters |
|---|---|---|
| C/N0 for a named satellite signal | Signal power relative to estimated noise density in that receiver | Signal identity, geometry, receiver implementation and observation history |
| A change in C/N0 during an event | A reception-quality change worth correlating with other evidence | Obstruction, installation, interference indicators and receiver status; the change alone does not identify a cause |
| Position error against a reference | Navigation error for that setup and time | Reference quality, solution mode, conditions and the required error statistic |
| A specified anti-jam suppression or J/S result | The stated interference-performance metric under its test conditions | Measurement points, waveform, receiver configuration and success criterion |
The last distinction is already developed in the ResiNav data-sheet review guide. Use it when comparing supplier specifications rather than treating every number containing “dB” as a common performance score.
A drop needs context; a high value needs context too
Interference can increase in-band noise and reduce the receiver's signal-to-noise ratio, as NovAtel's interference explanation describes. But signal attenuation and interference are different mechanisms. A falling C/N0 trace alone does not distinguish a blocked view from added RF energy. Correlate the affected bands and times with documented receiver indicators, installation changes and the observed navigation behavior.
Time resolution matters. NovAtel's CNOUPDATE command, for the listed OEM7 platforms, documents approximately 200 ms smoothing by default and a 50 ms option. A smoother plot can therefore reflect a different observation setting rather than a quieter environment. This is an implementation example, not a universal sampling recommendation.
A high C/N0 is not an authenticity check. Receiver signal-quality reporting and spoofing detection are separate functions: NovAtel describes spoofing detection using multiple processing metrics. Equally, a valid-looking coordinate is not independent evidence of accuracy. Assess the output using the documented solution status and the project's reference or integrity criteria.
Use the result to choose the next comparison
If two reports use different signal sets, bandwidth definitions or smoothing settings, first make the observations comparable. If matched observations change with an obstruction or installation change, investigate that reception path. If correlated evidence identifies an interference requirement, evaluate protection under the relevant condition and record both signal observations and navigation outcomes.
There is no single C/N0 pass value supplied by this guide. Set acceptance against the receiver, required operation and defined test conditions—not an arbitrary green bar. The receiver evaluation guide provides the configuration, baseline and recovery framework.
For the protection portion of an integration project, the ResiNav product catalogue provides the published model records and their enquiry paths. Shortlist by the required signals and documented product role, then discuss the relevant reception condition. This article does not establish a ResiNav C/N0 output, threshold or measured improvement.
NEXT STEP
Turn available platform information into an engineering review.
Use the Technology centre to frame the discussion, review relevant application scenarios, then send the available platform and receiver details for confirmation.