In short: Airport WiFi fails at peak times because one contended, unlicensed radio network is being asked to carry three very different loads at once: thousands of passenger devices, the retail estate's payment systems, and the operational tools that actually turn aircraft around. The fix is not more access points — it is separating those loads, moving the operational and high-density traffic onto licensed spectrum, and letting WiFi do the one job it is good at.
Key Takeaways
- A single gate can present 400 devices in ten minutes — a wide-body boarding pushes phone, tablet and laptop density past what contended WiFi channels were designed for, which is why the network dies exactly when everyone reaches for it.
- Passenger WiFi and operational traffic share the same congested airwaves — boarding-pass scanners, retail tills and staff devices compete with passenger video streams on unlicensed spectrum, so a busy departure lounge degrades the systems running the terminal.
- UK airports are splitting the load onto licensed spectrum — private 5G and neutral-host networks carry operational and high-density traffic on interference-free frequencies, leaving WiFi as a lightly-loaded passenger amenity rather than critical infrastructure.
In a nutshell
Why is airport WiFi so bad?
Everyone has lived the moment. The gate fills, the delay notice goes up, five hundred people reach for their phones, and the airport WiFi — free, branded, sponsored — slows to nothing. It is one of the most reliable frustrations in modern travel, reliable enough that "why is airport WiFi so bad" is a question travellers genuinely type into search engines while sitting in departure lounges, presumably over mobile data.
The lazy answer is that airports skimp on it. The evidence says otherwise: UK airports have invested repeatedly in WiFi refreshes, and passenger surveys from the likes of SITA and ACI consistently rank connectivity among the top drivers of passenger satisfaction, which airport commercial teams know perfectly well. Gatwick, Heathrow and Manchester have each rebuilt their passenger WiFi more than once in the past decade. The problem is not neglect. It is that the terminal environment breaks the assumptions WiFi is built on, and every refresh runs into the same physics.
Understanding why matters beyond passenger comfort, because at most airports the same wireless estate — or at least the same crowded slice of unlicensed spectrum — is carrying traffic that is far more important than a passenger's boxset download.
What is actually running on the terminal's wireless network?
Peel back the SSID list in any UK terminal and three distinct populations are sharing the air.
The first is passengers: thousands of phones, tablets and laptops, arriving in synchronised waves that follow the flight schedule. A single A380 or 777 boarding presents 300-400 passengers, most carrying two or three WiFi-capable devices, concentrated in a few hundred square metres of gate lounge, all within radio range of the same handful of access points.
The second is the commercial estate. Airport retail is a wireless business: card terminals, tills, stock scanners, digital signage, queue-measurement sensors, and the food-and-beverage units' own systems, many of them running over the airport's WiFi or their own competing access points. Every one of those rogue access points adds interference to the shared channels.
The third — and the one that should concentrate minds — is operations. Boarding-pass scanners at the gate, common-use check-in and boarding platforms, staff devices for turnaround coordination, wheelchair and PRM dispatch, cleaning and security teams, baggage reconciliation. We have written before about the airside half of this problem — stand operations and A-CDM and baggage tracking under Resolution 753 — but the terminal half is just as real. When boarding scanners slow down because the gate lounge is streaming, the aircraft's turnaround absorbs the delay, and delay at a hub propagates across the day.
Putting all three populations on contended, unlicensed spectrum was always a compromise. It made sense when the loads were small. They are not small any more.
Why does device density break WiFi?
WiFi runs in unlicensed spectrum, which means it operates on channels that anyone's equipment may use, with a polite listen-before-talk protocol to share the air. In a home or office this works well. In a terminal, three things defeat it.
Firstly, contention scales badly. When hundreds of devices contend for the same channel, an increasing share of airtime is spent negotiating rather than transmitting; throughput per device collapses non-linearly just as demand peaks. Secondly, the interference environment is uncontrollable: retailers' access points, passengers' personal hotspots, Bluetooth, and the neighbouring gates' own WiFi all overlap, and the airport has no regulatory power to clear the channel, because nobody owns unlicensed spectrum. Thirdly, terminals are hostile radio buildings — vast steel-and-glass halls with mezzanines, hoardings and moving crowds of water-filled human bodies, where coverage that surveyed perfectly in an empty hall behaves quite differently at 6 a.m. in July.
None of this is fixable by adding access points. Past a certain density, more access points on overlapping channels make contention worse, not better. This is the structural reason airport WiFi disappoints: the technology is being asked to do something it was never designed for.
How are UK airports fixing it?
The emerging answer, visible across the UK's transport estate, is to stop treating one radio network as infrastructure for everything, and instead split the load by what it needs.
Licensed cellular spectrum behaves differently from WiFi in exactly the ways a terminal needs: the licensee controls the channel, so there is no rogue interference; cellular scheduling is designed for hundreds of devices per cell rather than tens; and handover between cells is seamless, so a device moving from check-in to gate never drops. Ofcom's shared access licensing framework has made slices of the 3.8-4.2 GHz band available locally at modest cost, which is why private 5G networks have moved from telecoms-giant territory to something a regional airport can own.
In practice this produces a layered design. Operational systems — boarding scanners, staff devices, turnaround coordination, IoT — move onto a private 5G network on licensed spectrum, isolated from the passenger crush and sliced so that safety- and schedule-critical traffic gets guaranteed capacity. High-volume passenger connectivity increasingly rides neutral-host infrastructure, where a single set of in-building cells carries all four UK mobile operators' signal, so passengers simply use their own 4G/5G with full bars — no login page, no SSID, no contention with operations. Neutral-host deployments by the likes of Boldyn and Freshwave across UK transport hubs, including the Elizabeth Line's tunnels, have proven the model at scale. WiFi then remains as an amenity for laptops and long-dwell passengers — a job it can do comfortably once the peak load has been lifted off it.
The interesting point is that the same physical infrastructure — cells, fibre, cabinets — can serve both the private operational network and the neutral host. Built once, it fixes the passenger experience and the operational fragility in the same project.
Does this only make sense for Heathrow?
It is tempting to file all of this under big-hub economics, and certainly the largest deployments are at the largest airports. But the arithmetic arguably favours the mid-sized terminal. A regional airport with two million passengers a year has the same synchronised boarding peaks, the same retail estate on the same WiFi, and the same turnaround sensitivity — with a smaller building to cover and a fraction of the deployment cost. We made the broader case in our piece on regional airports; the terminal wireless estate is a concrete place to start, because the passenger-facing failure is visible in every departure lounge and the operational risk is sitting quietly behind it.
There are trade-offs to weigh honestly. A private network is another system to own, though managed-service models remove most of the burden. Neutral host requires commercial agreements with the mobile operators, which takes time. And WiFi is not going away — laptops, and passengers without roaming, will always want it. The design question is not WiFi versus 5G; it is which traffic belongs on which layer.
What should an airport operations or IT team do next?
Start by finding out what is actually sharing your air. A wireless survey during a real peak — not an empty-hall commissioning test — will show how much of your operational traffic is one busy gate lounge away from degradation, and most airports that run one are surprised. From there, the sequencing is usually: move operational and high-density systems onto licensed spectrum first, because that is where the risk sits; then address passenger connectivity through neutral host as commercial agreements allow; and let the WiFi estate shrink to the amenity role it can actually perform.
We build and manage private 5G networks for operational sites — including the smaller airports and airfields the big integrators overlook — and we are happy to start with the survey rather than the sales pitch. Talk to us about what your terminal's airwaves are really carrying.