The 6 GHz coverage holes nobody warned you about

A pattern we have now seen enough times to call it a pattern: an organisation refreshes to Wi-Fi 6E, keeps the existing access point positions because "the coverage was fine", and then spends three months fielding complaints from meeting rooms that worked perfectly last year.

The APs are in the right places for 5 GHz. They are in the wrong places for 6 GHz.

Where the loss comes from

It is not free-space path loss. Going from 5.5 GHz to 6.5 GHz costs about 1.5 dB in free space, which is nothing.

The losses that matter are the lower permitted EIRP under Indoor Low Power operation, and higher material attenuation. Through plasterboard the difference between 5 and 6 GHz is a decibel or two. Through a concrete core or a glass meeting pod with a metallic coating it can be five to ten.

Put together, expect a 6 GHz cell to cover roughly 20 to 30 percent less area than the 5 GHz cell from the same access point.

What that does to a floor plan

A 5 GHz design at -67 dBm with 15 percent overlap has almost no margin. Shrink every cell by a quarter and the overlap disappears first, then the coverage:

• Perimeter meeting rooms fall off 6 GHz and land on 5 GHz. Users notice the change without knowing why.
• Clients become sticky, holding a marginal 6 GHz link because roaming thresholds were set for 5 GHz behaviour.
• Roaming between floors gets choppy, because vertical leakage at 6 GHz is lower than the design assumed.

Change your survey methodology

Survey to a 6 GHz target. If 6 GHz is your primary band, -67 dBm needs to be measured at 6 GHz, not extrapolated.

Calibrate your materials. Default material libraries are usually built around 2.4 and 5 GHz measurements.

Check the client, not the AP. Measure with a real 6 GHz client, not a survey adapter with a better antenna than any laptop in your building.

Test PSC channels specifically. If your 6 GHz BSS is not on a Preferred Scanning Channel, discovery is slow. That reports as a coverage problem and is not one.

The budget conversation

None of this is an argument against 6 GHz. It is an argument for surveying the band you are going to use, and for accepting that a 6 GHz-primary design typically needs 15 to 25 percent more access points than the 5 GHz design it replaces.

Finding that out during design costs a spreadsheet revision. Finding it out after installation costs a second cabling contract.