Wi-Fi 7 in Europe: why 320 MHz stays a data sheet figure for now

Every Wi-Fi 7 data sheet carries a very large number. It comes from 320 MHz channel width, 4096-QAM and the sum of all bands. None of those three describes what a laptop on the third floor actually gets — and in Europe the first is additionally a question of spectrum policy.

The spectrum question Europe has decided

The 6 GHz band splits into two halves. The lower band from 5945 to 6425 MHz is open for wireless LAN across the EU (Commission Implementing Decision (EU) 2021/1067) — it is the band that made Wi-Fi 6E a meaningful generation in the first place. For the upper band from 6425 to 7125 MHz the EU level has now set a direction.

In late October 2025 the responsible federal ministry took a position: Germany will advocate at European level for the entire upper 6 GHz band to be assigned primarily to mobile, arguing that mobile network operators' spectrum demand for future 6G applications is assessed as greater.

Objections from the network industry were sharp. The broadband association and manufacturers — among them Deutsche Giganetz, Deutsche Glasfaser, HPE, Lancom and NetCologne — called the position a misjudgement and warned that fibre connections cannot be exploited without capable wireless inside the building. The European level has since spoken — and in substance followed the German line.

The RSPG decision of November 2025

On 12 November 2025 the Radio Spectrum Policy Group (RSPG), the European Commission's advisory body on spectrum policy, adopted its opinion on the long-term use of the upper 6 GHz band. Its core is a band split with priority: the 540 MHz from 6585 to 7125 MHz are to serve mobile networks with priority at full power. The 160 MHz from 6425 to 6585 MHz stay frozen as a guard band protecting the lower wireless LAN band until the World Radiocommunication Conference WRC-27 — Member States release them neither for mobile nor for wireless LAN. Only afterwards does the RSPG intend to decide on that remainder: if WRC-27 identifies the adjacent 7125–7250 MHz band for mobile, the opinion sees a strong case for wireless LAN in the 160 MHz; if not, for mobile.

Two doors stay ajar. The RSPG recommends studying whether wireless LAN can operate without priority inside the mobile portion — where no mobile network is deployed, such as a large factory in a rural area. And Member States should not be obliged to award spectrum where no mobile demand arises. The technical conditions are being worked out by CEPT under a mandate from the European Commission; the harmonised conditions are scheduled for July 2027.

The channel arithmetic: what 480 MHz yields

Spectrum policy becomes concrete the moment you convert it into channels. The lower 6 GHz band spans roughly 480 MHz. That produces:

A single channel means every access point using it shares it with every neighbour that does the same. In an office building with twenty access points, a channel width without a reuse pattern is not a capacity gain but a shared collision domain. That is why 320 MHz is not a planning value in Europe but at best an option for one isolated access point.

What remains of Wi-Fi 7 then — and it is a lot

The conclusion is explicitly not to skip Wi-Fi 7. The improvements that actually reach users do not depend on channel width at all.

• Multi-link operation. A client uses several bands at once instead of switching between them. The gain lies less in peak throughput than in stability: a briefly disturbed band no longer drops the connection. For voice, video calls and warehouse handhelds that is the real advance.
• Lower latency variance. Perceived quality is decided by the outliers, not the mean — and that is what Wi-Fi 7 addresses.
• Better behaviour in dense environments. The improvements in coordination and resource allocation pay off where many clients share little space: training rooms, manufacturing, event floors.
• A clean starting point. Anyone procuring today procures Wi-Fi 7 anyway. The question is not whether, but with which expectations.

On the hardware side the range is now broad: for the Meraki line Cisco lists the Wi-Fi 7 models CW9171I, CW9172I/H, CW9174I, CW9176I/D1, CW9178I and CW9179F, all of them 6 GHz capable. At Cisco Live US 2026 the Catalyst 9177 was additionally presented as a Wi-Fi 7 access point with 4×4 antennas and support for Ultra-Reliable Wireless Backhaul.

The expensive part is not on the ceiling

The most common miscalculation in a Wi-Fi 7 migration is treating it as an access point purchase. The devices are the smaller item. What gets expensive is everything behind them.

1. 01

   Check uplink speed

   A Wi-Fi 7 access point on a 1 Gbit/s port is the bottleneck in dense environments. Multi-gigabit at the access switch is the prerequisite, not the finishing touch.

2. 02

   Calculate the PoE budget

   What matters is not a single port's rating but the whole switch's budget when fully populated. That is exactly why budgets on new access switches are growing sharply — the Catalyst 9350 presented in 2026 states 4,320 W across 48 ports.

3. 03

   Assess the cabling

   Cat 5e carries multi-gigabit over short runs, but not everywhere and not reliably. Measurement decides, not the 2009 floor plan.

4. 04

   Rethink coverage

   6 GHz has shorter range and worse wall penetration than 5 GHz. Replacing old positions one-for-one produces holes — a survey before mounting saves the second round.

5. 05

   Be honest about the client fleet

   The benefit only materialises with clients that support 6 GHz and MLO. If most of the fleet does not, Wi-Fi 7 is an investment in the next three years — legitimate, but it should be communicated that way.

Wi-Fi 7 planning check: what is missing before procurement?

• Clients: Which models, drivers and applications should benefit? Name a representative pilot client; unknown 6 GHz or MLO support remains a test item.
• PoE: Record the AP model, required operating mode, switch power supplies and total budget at planned occupancy. A spare port does not prove adequate power.
• Uplinks: Document negotiated port speed, cabling measurements and shared uplinks. A data-sheet rating does not verify the existing path.
• Coverage: Record floor plans, materials, working routes and load conditions. RF measurements and application tests define acceptance; existing AP locations are initially assumptions.

When the move pays off — and when it does not

It pays off if one of the following holds: the installed generation is Wi-Fi 5 or older; there are areas with high client density; wireless voice or handhelds are business-critical; or the cabling is being touched anyway.

It does not pay off for its own sake when a well-planned Wi-Fi 6 installation is in place, the client fleet barely supports 6 GHz and the access switches cannot do multi-gigabit. In that case the better order is access layer first, radio second — otherwise you buy access points that wait on their uplink.

How we survey, plan and hand over wireless estates is described under Meraki and cloud management and enterprise networks.