Introduction
"Many GNSS failures are not due to satellites - they are due to bad system design."
After years of field experience and forensic analysis of GNSS project failures, a consistent pattern emerges: the satellites are almost never the problem. The atmospheric corrections, the receiver hardware, and the positioning algorithms are rarely to blame. Instead, the vast majority of precision GNSS failures trace back to preventable system design and deployment mistakes. Understanding these common errors - and why they occur - is one of the most practical skills a GNSS engineer can develop.
Mistake 1: Poor Antenna Placement
Antenna placement is the most frequently misunderstood element of GNSS system design, and mistakes here cascade through every measurement made by the system.
Multipath from Nearby Reflectors
When the antenna is placed close to buildings, vehicles, metal fences, water surfaces, or other reflective structures, satellite signals arrive at the antenna via both the direct path and one or more reflected paths. The reflected signal adds constructively or destructively with the direct signal, distorting pseudorange and carrier phase measurements by amounts ranging from a few centimetres to several metres. Multipath cannot be corrected after the fact - it must be avoided through placement.
Obstructed Sky View
Trees, buildings, and terrain features block signals from satellites at low elevations. Fewer visible satellites mean worse geometry (higher DOP) and reduced redundancy for detecting outliers. In RTK, losing even one satellite at a critical geometry moment can cause the ambiguity solution to reset. The antenna must have a clear view of the sky from the elevation mask (typically 10–15 degrees) upward, in all azimuths.
Antenna Too Close to the Ground
Ground reflections are the strongest source of low-angle multipath. An antenna placed on a short tripod or directly on a flat surface is maximally exposed to ground-reflected signals. Elevating the antenna, using a large ground plane, or using a choke ring antenna substantially reduces this effect.
Mistake 2: Incorrect Base Station Setup
Using an Incorrect or Unknown Base Coordinate
This is the single most consequential mistake in precision GNSS. Every rover position in the project is derived relative to the base. If the base's entered coordinate is wrong - even by a small amount - every single rover measurement inherits that error. The error will be consistent (not random), meaning the results will look perfectly precise on the rover's display, but will be systematically wrong in the real world. This is particularly dangerous because the receiver will show no warnings, no flags, and no indication that anything is amiss.
Exceeding Baseline Length Limits
Operating a rover at a distance from the base that exceeds the system's design range - typically 10–20 km for single-base RTK - results in insufficient atmospheric error cancellation. The rover and base experience different ionospheric and tropospheric delays, and the corrections from the base no longer accurately represent the errors at the rover's location. The result is degraded ambiguity resolution and positioning errors that may not trigger obvious quality indicators on the receiver.
Base Antenna in a Multipath Environment
A base station antenna affected by multipath generates biased corrections. These biases are transmitted to all connected rovers and appear in their position solutions. The key symptom is a systematic offset that remains stable over time - the solution looks "fixed" and precise, but is wrong. Selecting a base site with excellent sky view and using a choke ring antenna at the base is non-negotiable for high-accuracy work.
Mistake 3: Weak or Unreliable Communication Links
Assuming the Communication Link Will Work
Engineers frequently assume that NTRIP over a cellular data network, or a UHF radio link, will perform reliably throughout the operational area without testing this in advance. Mobile coverage has dead zones, UHF radio signals are blocked by terrain, and both can be disrupted by interference. A communication link that drops out for 30 seconds is sufficient to break RTK Fixed status and force re-initialisation, losing several minutes of potential productivity.
Excessive Correction Latency
Internet-delivered corrections via NTRIP can experience significant latency spikes when mobile network congestion occurs. A correction age of more than 3–5 seconds degrades RTK performance. Engineers should monitor correction age during operations and configure the receiver to alert or fall back to float solution when the threshold is exceeded.
No Fallback Plan
Operating without raw data logging or a backup communication path means that any communication failure results in a total loss of data for that period. Modern GNSS receivers should always be configured to log raw observations so that PPK (Post-Processed Kinematic) re-processing can recover positioning even if the real-time link failed.
Mistake 4: Misconfigured Receivers
| Misconfiguration | Effect | Prevention |
|---|---|---|
| Wrong antenna height entered | Systematic vertical error equal to the height offset | Measure from correct reference point per manufacturer spec |
| Elevation mask too low | Low-quality satellite observations included, more multipath | Set mask to 10–15 degrees; raise in high-multipath areas |
| Wrong datum or coordinate system | Systematic position error; results incompatible with project data | Verify datum and projection settings match project CRS |
| RTCM message set mismatch | Rover receives incomplete or wrong corrections | Configure base and rover for matching message types |
| Incorrect antenna model selected | Phase center correction errors, especially in vertical | Select exact antenna model from receiver's antenna database |
Antenna Height Errors
Measuring the antenna height incorrectly is one of the most common and damaging mistakes in field surveys. Different manufacturers define the antenna reference point differently - some measure to the bottom of the antenna, others to the centre, others to a specific notch or groove. Applying the wrong measurement span introduces a direct vertical offset into every collected point. Always consult the manufacturer's field guide for the exact measurement procedure for the specific antenna and pole combination being used.
Elevation Mask and Signal Quality Settings
Setting the elevation mask too low allows signals from near the horizon to enter the solution. These signals travel through far more atmosphere than overhead signals, suffer greater multipath risk, and have weaker signal strength. While they add satellite count, they can add more noise than they remove through additional geometry. A mask of 10–15 degrees is the standard starting point, raised to 15–20 degrees in environments with significant low-angle multipath.
Vital Points
- Most GNSS precision failures are preventable - they result from poor system design decisions, not satellite or hardware failures.
- System design is chronically underestimated; engineers often spend more time selecting hardware than planning the deployment and verifying the setup.
- Small mistakes - a 3 cm antenna height error, a base coordinate off by 10 cm, a dropped correction link - lead to errors that are large relative to the centimetre accuracy the system is capable of delivering.
- A pre-deployment checklist covering antenna placement, base coordinates, communication testing, receiver configuration, and power validation will prevent the great majority of field failures.