Field Observation — Wireless & Structured Cabling

The AP got upgraded.
The bottleneck moved into the wall.

Wi-Fi 6 access points ship with 2.5 Gbps uplink ports because 1 Gbps became the constraint. In a lot of buildings, what's actually run to the AP is Cat5e installed before multi-gigabit wireless was a design requirement. The air interface isn't the limit anymore. The cable is.

66,944LinkedIn impressions
133Reactions
69Comments
Since 1992Layer 1 specialists
Cabling running through a ceiling cavity above a commercial site, the kind of installed Cat5e infrastructure a Wi-Fi 6 access point now depends on.

Actual field photograph. Commercial site, Australia. The cabling a new access point inherits is rarely the cabling its uplink port was designed for.

A Wi-Fi 6 access point goes in. Procurement's happy, the AP's rated for gigabit-plus speeds, the project's marked complete. Then performance falls short — and the investigation starts everywhere except the cable in the wall.


Why Cat5e becomes the bottleneck under Wi-Fi 6 load

Wi-Fi 6 access points ship with 2.5 Gbps uplink ports. That's not a spec sheet flourish, it's an admission. The industry already knows 1 Gbps is the bottleneck for an AP aggregating a full floor of concurrent clients. The air interface stopped being the limiting factor years ago. What's limiting it now is whatever's actually run to the AP.

And in a lot of sites, what's run to the AP is Cat5e cabling installed before multi-gigabit wireless was anyone's design requirement, because at the time, it didn't need to be.

The core mismatch

A faster AP doesn't reduce what's demanded of the cabling underneath it. It raises the bar the cabling has to clear. New technologies increase dependence on Layer 1, they don't reduce it.

Aggregate load and PoE+ voltage drop: where Cat5e actually fails

This is worth being precise about, because the loose version of it doesn't hold up — and we know that, because the LinkedIn post this article is based on was tested by exactly the engineers who'd know.

Healthy, short-run Cat5e can do 1 Gbps fine, and sometimes 2.5 Gbps over shorter distances. That's not in dispute. The failure mode isn't "Cat5e is incapable of gigabit." It's two separate things stacking on top of each other.

Failure mode one
Aggregate floor load
A single-client benchmark on a short, line-of-sight run tells you almost nothing about a Cat5e run under a full floor of concurrent Wi-Fi 6 clients hitting one AP. Testing one doesn't validate the other.
Failure mode two
PoE+ voltage drop on long runs
Add distance, add age, add a PoE+ powered AP pulling current down the same cable, and voltage drop works against you quietly. It destabilises the AP before anyone calls it a fault.

Nothing fails outright. It's just degraded enough that nobody's looking at the cable as the cause.

"The bottleneck didn't disappear when the AP was upgraded. It moved into the wall and stopped announcing itself."

The Layer 1 diagnostic blind spot

This is the pattern that shows up site after site: hardware budget gets approved, APs go in, performance falls short of what the spec sheet promised, and the response is always pointed at the active layer. RF planning. Placement. Channel overlap.

The physical layer doesn't get checked because nobody's trained to suspect it first. The AP is new, the configuration is reviewable, the RF environment is measurable. The cable in the wall is none of those things to most of the people troubleshooting the fault, so it's the last thing checked instead of the first.

The pattern that costs you

Risk isn't old technology, it's undocumented, unvalidated infrastructure. If you haven't had the horizontal cabling validated against current standards, the Wi-Fi 6 mismatch is rarely the only thing waiting to surface.

How to run a cable spec audit and fund the fix

If you're running this audit across a multi-closet site, don't try to replace everything at once. That's not a fundable proposal and it won't get approved as one. Treat it as a per-closet sequence instead.

01
Start with the MDF
Build a new optical headend first. That's a far easier business case to fund than "replace all structured cabling," and it gives you the backbone capacity everything downstream depends on.
02
Phase downstream closets against budget cycles
Don't try to do it in one capital cycle. Sequence it. The structure has to survive a CFO who has already signed off on the AP spend once and isn't keen to hear the job wasn't finished.
03
Prioritise front-of-house and revenue-generating areas first
Where the business impact of degraded wireless is most visible and most costly, that's where the cable spec gets checked first.

This is the same diagnostic gap that shows up across most brownfield LAN environments — undocumented changes, untested runs, and infrastructure inherited from a previous design era.

The thesis held up under credentialed pushback.

When this went up on LinkedIn it didn't just get views. It drew principal systems engineers, senior network engineers, and data centre operations leads into a working debate — some confirming the pattern on their own live sites, some pushing back on the specifics. The thesis held up, but it held up because the pushback sharpened it, not because nobody disagreed.

Principal Systems Engineer — 20+ years
The precise point matters: Cat5e runs 1 Gbps fine and sometimes 2.5 Gbps over shorter distances. The real failure mode is degraded, marginal and long runs negotiating down under real-world conditions — not a blanket "Cat5e can't do gigabit."
RF / Wireless Practitioner
A single-client, short-range, line-of-sight benchmark shows gigabit is rarely needed per client. True — and a different layer of the problem to aggregate floor traffic plus PoE+ load on a long run. Different test, different conditions, different result.
Senior Network Engineer
Treat remediation as a per-closet battle. Start with the MDF, build a new optical headend first because it's more fundable than "replace it all," then phase downstream closets aligned to budget cycles, revenue areas first.
Data Centre Operations Lead
Mirrors a live issue of his own — technicians rotating through a different diagnosis on every visit, no root cause ever found. The diagnostic blind spot, playing out in real time on someone else's site.

If a claim about infrastructure risk can survive credentialed engineers testing it in public, that's a stronger signal than a claim nobody's challenged.

The cable spec is the first thing to check, not the last.

We've been validating physical infrastructure since 1992. What changes is the load being placed on cabling that was never designed to carry it — and the cost when someone finally discovers the gap between what the AP needs and what the cable can deliver.

A Layer 1 audit identifies the mismatch before it's misdiagnosed as RF: cable identification, certification testing against current standards, PoE load and run-length assessment, and as-built documentation that reflects what's actually in the wall.

What proper validation looks like

Full physical audit of the installed cabling. Certification testing, run-length and PoE+ load assessment, pathway mapping. Remediation scoped and sequenced against budget cycles. One accountable partner. Nationwide delivery on live, business-critical networks.

Where's the worst infrastructure mismatch you've seen go undiagnosed on a live site — and how long before someone pulled the cable spec?

Find it on your terms, not the network's.

Before the next AP refresh goes in, validate what the cabling underneath it can actually carry.

Book a Layer 1 Assessment