Skip to main content

If You Think You’ve Designed a “Hard” Wi-Fi Network… Try This.

20 Feb 2026.png

https://www.linkedin.com/pulse/you-think-youve-designed-hard-wi-fi-network-try-jarryd-de-oliveira-ox8te 

Every few months someone tells me they’ve just completed a really difficult Wi-Fi deployment. Then I look at it.

Open office. Suspended ceiling. Omni APs. Standard clients. Minimal RF reflection.

That’s not hard.

If you want hard, try automotive manufacturing floors, 20,000+ seat arenas, high-rack warehouses with AMRs, moving assembly lines, overhead-only mounting constraints, heavy IoT on 2.4 GHz, and aggressive user take rates.

That’s where Wi-Fi stops being theoretical and becomes engineering.


The Problem with “Green Heatmaps”

One of the biggest mistakes I still see is the belief that green equals good.

It doesn’t.

High transmit power, massive omni coverage, and wide 80 or 160 MHz channels look fantastic in predictive modelling. But when users show up, reality kicks in.

Now you have co-channel contention, excessive CCI, high channel utilisation, unstable MCS rates, unpredictable roaming behaviour, and APs hearing far more than they should.

In dense or reflective environments, more coverage is often worse.

You don’t need bigger cells. You need controlled cells.


Manufacturing and Warehouse Reality

Industrial and warehouse environments introduce a completely different level of complexity.

You’re dealing with high reflectance from metal and machinery, moving assets such as AGVs and robotics, overlay networks, limited mounting positions, less-than-perfect client radios, heavy 2.4 GHz IoT usage, and voice roaming requirements.

The physics matter.

2.4 GHz behaves very differently to 5 GHz.
6 GHz behaves differently again.

You cannot simply deploy internal omni antennas across a factory floor and expect stability.

You need directionality. You need isolation. You need disciplined transmit power. You need to control what the AP hears as much as what it transmits.

Antenna choice becomes architectural, not aesthetic.


High Density Is Not “More APs”

I’ve seen so-called high-density deployments where omnidirectional APs are five to eight metres apart.

That’s not density. That’s self-inflicted interference.

True high-density design focuses on controlled cell boundaries, reduced listening footprint, balanced transmit power, receiver sensitivity tuning, mandatory data rate strategy, and intelligent channel reuse planning.

In stadium-scale environments, you are not flooding space. You are shaping it.

Beamwidth, orientation, elevation, power limits, and data rate discipline are what separate a network that survives from one that collapses under load.


Lifecycle Discipline Still Wins

No plan survives first contact with the building.

That’s why lifecycle discipline matters more than ever.

Define properly. Design intentionally. Implement accurately. Validate thoroughly. Optimise continuously.

Walk the site. Measure. Adjust. Repeat.

RRM is a tool. It is not a strategy.


Wi-Fi 7 Doesn’t Fix Bad Design

Yes, Wi-Fi 7 introduces 320 MHz support, 4096-QAM, Multi-Link Operation, and expanded 6 GHz capability.

None of that fixes poor RF design.

Wider channels reduce reuse options. Higher modulation requires higher SNR. Multi-Link Operation adds complexity if the RF foundation is unstable.

New technology rewards good fundamentals. It exposes weak ones.


Lessons From Real Deployments

Lower transmit power more often than you raise it.

Use directionality when physics demands it.

Trim minimum data rates intelligently.

Use 20 MHz channels in dense environments.

Control what the AP hears.

Validate with measured data, not assumptions.

Design for airtime efficiency, not marketing throughput numbers.

And above all, adapt.

Buildings change. Mounting changes. Density changes. Expectations change.

Engineering means adjusting without compromising fundamentals.


Final Thoughts

The hardest Wi-Fi environments are not hard because they’re large.

They’re hard because physics, density, mobility, and business expectation collide.

That’s where proper wireless engineering shows up.

If you’re designing warehouse, manufacturing, arena, or high-density enterprise Wi-Fi in 2026, don’t design for coverage.

Design for control.