| Architecture | What it creates | Best for | Strengths | Limits | Typical KPIs |
|---|---|---|---|---|---|
| CATR (Compact Antenna Test Range) | Local plane wave in a quiet zone via reflector | FR2 beamformers, panels, radars, CPE | Pattern accuracy, fast sweeps, moderate footprint | QZ size limited by reflector; edge/diffraction control critical | Gain/efficiency, beams, EIRP/EIS, TRP/TIS |
| Spherical Near-Field (SNF) | Full spherical field sampling → far-field via transform | High-accuracy patterns on any band/DUT | Highest accuracy, full 3D data, polarisation control | Scan time, tight mechanics, probe correction | Gain, patterns, cross-pol, total efficiency |
| Reverberation Chamber (RC) | Statistical isotropic/anisotropic fields by stirring | Throughput/robustness, coexistence, MIMO | Tiny test time, excellent repeatability, production-friendly | Not for classical patterns/beam maps | Throughput/BLER vs SINR, TRP/TIS (stat), desense |
Measurement physics & error mechanisms (the stuff that bites)
CATR
- Plane-wave synthesis: Reflector converts spherical feed to a plane wave in the quiet zone (QZ).
- Errors: amplitude/phase ripple from edge diffraction, reflector surface error, feed cross-pol, mis-focus/tilt, chamber stray.
- Controls: serrated/rolled edges, high-accuracy surface, low-Xpol feed horns, thermal stability, absorber at critical scatter points.
Spherical Near-Field
- Field sampling: Measure tangential fields over a spherical surface; apply NF→FF transform.
- Errors: Probe correction (pattern/phase), sampling/truncation, scanner accuracy, DUT alignment, cable motion.
- Controls: calibrated probes, adequate scan radius, high-precision positioners, flexible de-embedding of fixtures.
Reverberation
- Statistical fields: Create many independent stir states → ensemble average approximates rich multipath.
- Errors: insufficient Q or number of independent states, leakage through ports, non-uniform stirring, and loading sensitivity.
- Controls: verify chamber constant/Q, mode-stir characterisation, reference antenna factor, and loading procedures.
Map DUT classes to the right range
- Handsets/UE, wearables: TRP/TIS and throughput → RC for fast statistical KPIs; CATR/SNF for specific beams/features.
- CPE/routers, gNodeB/small cells, phased arrays (FR1/FR2): Beamshape & EIRP/EIS → CATR (FR2) or SNF (FR1/large apertures).
- Automotive radar (77–81 GHz): Angular/range/Doppler → CATR or compact FF with target simulators.
- IoT modules/sensors: Production-scale radiated sensitivity/robustness → RC; antenna debug → SNF/CATR.
- Large DUTs/odd form factors: Often SNF (with big scanners) or hybrid setups.
KPI alignment (start here, not with hardware)
- Beam/pattern diagnostics, cross-pol, efficiency → SNF or CATR.
- EIRP/EIS, TRP/TIS with acceptance masks → CATR/SNF (deterministic) or RC (statistical).
- User KPIs (throughput/BLER/latency) & coexistence → RC (with interferers/channel emulation).
- MIMO correlation/rank behaviour → RC for fast coverage, MPAC+CATR or SNF when angular detail is needed.
Test time, automation & scalability
- CATR: Fast angular sweeps; good for regression and beam scans.
- SNF: Longer (dense sampling), but richest data; raster reduction and probe-array scanners help.
- RC: Fastest; ideal for production and large test matrices.
- Throughput stacks: Plan for API control (gNB emulator, traffic gen, analysers) regardless of architecture.
Uncertainty & correlation
- Contributors: probe/antenna factor, positioner accuracy, instrumentation linearity/noise, chamber scattering, temperature.
- Good practice: publish expanded uncertainty (k=2) with KPIs; lock golden DUT builds; share seeds/cal files across sites to improve cross-lab correlation.
Facility footprint & infrastructure
- CATR: Reflector + QZ drive size; typical mid-footprint; FR2 absorbers perform at grazing angles.
- SNF: Large scanner + clearance; plan for rigid flooring, vibration control.
- RC: Moderate footprint but robust door/knife-edge seals; mechanical stirrers and RF filtering on penetrations.
- All: HVAC stability, safety interlocks, clean RF feedthroughs, instrument reference distribution.
Upgrade paths (future-proofing)
- CATR: Larger QZ, higher-freq feeds, better edge treatments; add channel emulation for angular spreads.
- SNF: Faster scanners, multi-probe arrays, extended radius for larger DUTs.
- RC: Higher stir bandwidth, more probes/ports, integrated coexistence injectors and fading emulators.
- Hybrid labs: CATR (beams) + RC (throughput) is a powerful combo for Private 5G and device programs.
Decision flow (four questions)
- What KPI matters most? Pattern/beam → CATR/SNF; throughput/robustness → RC.
- Which bands & apertures? FR2 beamformers → CATR; large FR1 arrays → SNF.
- How fast must you test? High-volume → RC; deep debug → SNF/CATR.
- Space & budget? Fit the footprint now, with a clear upgrade path for the next 24–36 months.
Mini case studies
- Private 5G CPE vendor: Needs EIRP/EIS and user throughput. Solution: CATR for deterministic EIRP/EIS + RC for throughput/BLER under coexistence.
- Automotive lab (radar + V2X): Radar beams & Doppler + V2X robustness. Solution: CATR with target simulators + RC for sidelink throughput and coexistence.
- University program: Teaching + research across many devices. Solution: starter RC for labs/production-style KPIs; add SNF for antenna/pattern courses.
Procurement checklist
- Performance masks: QZ amplitude/phase/cross-pol (CATR), probe correction & scan specs (SNF), chamber constant/Q & stir state count (RC).
- Calibration artefacts: reference antennas, alignment tools, uncertainty templates; FAT/SAT procedures.
- Automation: drivers/APIs for positioners, emulators, analysers; data format standards.
- Lifecycle: re-cal kits, drift monitoring, service SLAs, spare parts.
- TCO: capex + installation + yearly calibration/maintenance + operator time.
Pick architecture by KPI, not by fashion. CATR gives you clean plane waves for beams at FR2; SNF gives full-fidelity patterns; RC gives speed and statistical realism for user KPIs. Many high-performing labs run two of the three—and thrive.
Want a neutral second opinion on specs or a blended roadmap (what to buy now vs next)? Novocomms can help draft acceptance masks, uncertainty budgets, and FAT/SAT plans so your range delivers day one.
