Wi-Fi 7 and the Width Problem

The same chipset, the same box, two very different throughput ceilings — depending on which country's regulator signed off on the band.

What Wi-Fi 7 Actually Promised

IEEE 802.11be, ratified in 2024, is built around three headline changes: 4096-QAM modulation, multi-link operation (MLO), and 320 MHz channels. The first two are available wherever the hardware ships. The third requires spectrum that not every regulator has released, and that gap between what the standard permits and what local law allows is the core tension of Wi-Fi 7's early deployment.

To understand the channel-width problem, it helps to know where that 320 MHz has to come from. The 6 GHz band, roughly 5.925–7.125 GHz in the United States, is the only place wide enough to fit it. The older 2.4 GHz band tops out at 40 MHz channels before adjacent-channel interference makes it impractical. The 5 GHz band can assemble 160 MHz by bonding non-contiguous segments, which 802.11ac already permitted, but reaching 320 MHz in 5 GHz is geometrically impossible — the band is not wide enough. So Wi-Fi 7's widest channels depend entirely on 6 GHz being open for unlicensed use.

The Regulatory Patchwork

The United States moved first. The FCC opened the full 1,200 MHz of 6 GHz spectrum to unlicensed use in April 2020, creating room not only for 160 MHz Wi-Fi 6E channels but, in principle, for the 320 MHz channels 802.11be would later specify. The UK's Ofcom followed with a similar full-band decision in 2021. Several other markets — including Brazil and Saudi Arabia — have since opened comparable allocations.

Europe moved more cautiously. ETSI and national regulators initially authorized only the lower portion of the 6 GHz band, from 5.925 to 6.425 GHz, a 500 MHz slice. That is enough for three non-overlapping 160 MHz channels or a handful of 80 MHz channels, but it cannot accommodate a single 320 MHz channel. The upper portion, 6.425–7.125 GHz, remains occupied by incumbent licensed services — fixed microwave links, satellite uplinks — and releasing it requires negotiated clearance that most European administrations have not yet completed.

The ITU's regional structure compounds this. The world's spectrum allocations are governed within three ITU regions, and even within a single region, national administrations retain the authority to restrict or withhold secondary allocations. A product legal in Dallas may operate in a degraded mode in Düsseldorf, not because the hardware changed but because the regulator did not finish clearing the band.

The practical consequence is visible in certified product specifications. A Wi-Fi 7 access point sold globally carries the 320 MHz capability in its firmware and silicon, but ships with that capability disabled or automatically downgraded in markets where the upper 6 GHz band is unavailable. The RF front end is built to the full specification; the channel-width table is filtered at initialization against a regulatory domain code. Same product, truncated ceiling.

What the Standard Cannot Fix

802.11be's design was not naive about this. The standard defines the 320 MHz channel as an optional feature in 6 GHz, and the Wi-Fi Alliance's certification program acknowledges regulatory variation by permitting devices to advertise 6 GHz capability while restricting actual channel widths to whatever the deployment country allows. The theoretical peak — roughly 46 Gbit/s aggregate under ideal conditions — is computed assuming 320 MHz channels, 16 spatial streams, and 4096-QAM. Strip the channel width back to 160 MHz and the ceiling roughly halves before any other factor is considered.

Multi-link operation, 802.11be's other distinctive feature, partially compensates. MLO lets a device aggregate simultaneous links across 2.4, 5, and 6 GHz bands, improving both throughput and latency even when the widest single channel is unavailable. But MLO is an architectural improvement, not a substitute for raw spectral width, and spectrum auctions that clear incumbent licensees from the upper 6 GHz band will ultimately determine when European Wi-Fi 7 reaches its rated peak.

The standard was finished on schedule. The spectrum question belongs to a slower process — one where the FCC's April 2020 ruling and its counterparts elsewhere set the pace, not the IEEE ballot.

The third requires spectrum that not every regulator has released, and that gap between what the standard permits and what local law allows is the core tension of Wi-Fi 7's early deployment.

Read next

The rest of in order, with dates

• TimelineGSM goes liveThe first GSM networks opened in 1991 and set the pattern of a standard defined before deployment rather than after it.
• TimelineLTE, and the end of the circuitLTE moved cellular to an all-IP core, which is the change that made voice a service on the data network rather than the other way round.
• TimelineWi-Fi 6 was a rename, not a releaseThe Wi-Fi Alliance applied generation numbers retroactively in 2018, an admission that the lettering had failed as communication.