Beamforming for indoor 5G - practical gains?

Indoor beamforming promises big gains but real rooms have reflections everywhere and users move. What throughput gain did you actually measure indoors versus a well-placed omnidirectional setup?

Indoor beamforming gains are real but almost never match outdoor LOS slides, because the channel is rich and the user moves.

What tends to survive contact with rooms:
1. Modest average throughput / SINR gains vs a well-placed omni are common; peak "hero" MCS only shows up in favourable geometry and collapses with a body or a metal desk in the way.
2. Reflections help and hurt: multipath can fill nulls, but it also confuses sounding if the update rate is slow relative to motion. Tracking latency often eats the theoretical array gain.
3. Placement still wins cheap dB: a ceiling / hallway AP with a sane downtilt often beats a fancy codebook in a corner.

Useful report-out: delta MCS / goodput vs a fixed omni at the *same* AP location, plus how often the beam had to retune when someone walked. Share band (2.4 / 5 / 6 / 60 GHz), antenna count, and whether you logged sounding failures - that usually separates "beamforming helped" from "we just moved the AP".

— rfforum 编辑部 / Editorial Team