Sensitivity testing requires a known wanted signal and a setup whose noise and unintended paths do not dominate the result.
Why this matters in the industry
IoT developers need meaningful weak-signal comparisons instead of measurements limited by the laboratory arrangement.
The technical reasoning
Weak-signal testing must distinguish the intended stimulus from noise, leakage and measurement limits. A source display can continue decreasing even when the actual delivered signal is dominated by an unintended contribution. Reference checks and controlled loss changes help establish the useful range of the experiment.
Loss, noise and the position of the first active stage
The effect of loss depends on where it occurs. A passive loss before a low-noise amplifier reduces the wanted signal and adds thermal noise according to its temperature. Under the standard matched model, a passive network at the reference temperature has noise factor equal to its linear loss. In a cascade, later-stage noise contributions are divided by the gains preceding them, so moving the same loss to another position can change the overall noise figure.
How to structure the investigation
Characterize path loss and available signal range. Check background noise, leakage and instrument limitations under the actual settings. Define sufficient observations at each level and identify when the setup can no longer support the intended stimulus.
Draw the sequence of passive and active stages and state the temperatures and reference conditions used by the model. Work in linear factors for cascade calculations, then convert to dB for reporting. Distinguish noise figure from receiver sensitivity, which additionally depends on bandwidth, waveform and the required detection or error criterion. Check that a sensitivity experiment is not limited by source leakage or the measurement setup.
Worked example or engineering scenario
If adding 10 dB loss produces no meaningful change in observed packets near the reported threshold, leakage or an uncontrolled state may be limiting the experiment rather than receiver sensitivity.
Evidence to collect
| Record | Purpose |
|---|---|
| Available level range | Defines the tested state and scope of the comparison. |
| Background noise | Makes the stimulus or route condition reproducible. |
| Leakage | Supports interpretation of variation and possible confounding effects. |
| Observation criterion | Connects the observation with the stated engineering decision. |
Trade-offs and common interpretation errors
Do not apply a simple room-temperature approximation to a different thermal condition without review. A power correction may reconstruct a signal level but cannot undo the signal-to-noise degradation caused by preceding loss. Keep estimates separate from measured noise performance and state the assumptions behind either result.
What the result can support
Establish the setup's low-level capability before reporting a sensitivity threshold or apparent link margin.
A generator display alone does not prove the weak signal reaching the receiver.
Further technical reading
- Keysight: Fundamentals of RF and Microwave Power Measurements
- NIST: Reliable Wireless Systems for Factory Automation
Related industry knowledge
- How to Compare Two IoT Receiver Designs Fairly
- How to Use Dummy Loads During IoT Transmitter Development
Numerical scenarios are illustrative assumptions, not reported measurements of a supplied product or installation.