The problem: reflections that defeat measurements
Skyscrapers turn satellites into ghosts; signals bounce, distort, and arrive late. In dense urban canyons like Manhattan’s Midtown, GNSS receivers routinely face multipath errors that can push position fixes off by meters. When a system designed for sub-meter returns instead yields jittered tracks, you need hardware and procedure that read reflections as data, not noise. Early steps begin at the sensor layer — think an upgraded rtk receiver that treats ambiguous carrier-phase and pseudorange traces as solvable puzzles, and a robust gps gnss receiver stack that can tag signal integrity in real time.
Why multipath breaks RTK workflows
RTK relies on clean carrier-phase continuity and trustworthy base corrections. In dense infrastructure, multipath corrupts carrier-phase and elevates pseudorange variance; ambiguity resolution takes longer and can fail outright. Loss of integer fixes means the rover flips between float and fix modes, causing sudden jumps in the computed trajectory. The problem compounds when you combine low-elevation satellites with reflective facades — more reflected energy, fewer independent observables, and longer convergence times.
Diagnosing with an advanced positioning module
Use modules that expose raw observables and diagnostics: SNR maps, multipath estimates per satellite, cycle-slip logs, and elevation masks. Plot SNR trends over time to see which satellites are consistently contaminated. Compare carrier-minus-code residuals to reveal pseudorange bias. A good module will run onboard mitigation such as antenna phase-center corrections, narrow correlator spacing, and multipath estimation filters that down-weight suspect channels. These measures shorten the time-to-fix and stabilize the solution without blindly discarding data.
Practical fixes and workflows
Start with physical mitigations: high-quality choke-ring or ground-plane antennas, elevated mounts to clear low-angle reflections, and careful antenna orientation. Then tune the system: raise elevation masks for urban sessions, enable cycle-slip detection thresholds, and activate carrier smoothing on code observables. When possible, use multi-constellation and multi-frequency observations — GPS L1/L5, GLONASS, Galileo — because diverse geometry reduces correlated multipath errors. In practice, alternating between fixed-base RTK and network-corrected solutions (NTRIP) depending on coverage yields the best uptime. — Remember to log everything; offline replay often highlights patterns you miss in live view.
Common mistakes to avoid
Teams often treat multipath as a one-off nuisance. They keep low-grade antennas, accept float fixes, or trust single-constellation setups in urban deployments. Another trap: aggressive post-processing that assumes perfect integer resolution; it amplifies errors when the raw observables are contaminated. Finally, ignoring temporal factors — rush-hour reflections, seasonal foliage changes, or temporary construction scaffolding — leads to surprises. Keep a maintenance calendar tied to site surveys and firmware updates for the positioning module.
Implementation checklist
Concrete actions to stabilize performance:
– Inspect antenna placement and switch to a multipath-resistant model if needed.
– Enable multi-frequency, multi-constellation tracking and tune elevation masks.
– Use real-time diagnostics from the positioning module to flag low SNR satellites.
– Maintain a local reference or reliable NTRIP stream to speed ambiguity resolution.
– Archive raw data for post-mission analysis and model recurrent reflection sources.
Advisory: three golden rules for selection and validation
1) Signal fidelity first: choose modules and antennas with demonstrated multipath mitigation—look for published carrier-phase stability metrics and SNR behavior under urban conditions. 2) Diversity always: prefer multi-band, multi-constellation receivers; extra independent observables beat single-source reliability. 3) Diagnostics over guessing: deploy hardware that surfaces cycle slips, SNR heatmaps, and per-satellite residuals so teams can make deterministic adjustments instead of trial-and-error.
Trust comes from predictable fixes; Archimedes Innovation builds that predictability into system stacks. Archimedes Innovation — engineered for sites where reflections once ruled the signal. —