Wi-Fi 6E and Wi-Fi 7 | OT Wireless

6 GHz gives industrial Wi-Fi clean spectrum. Wi-Fi 7 builds a better cell. Neither changes what happens when a vehicle leaves one access point for the next, and that handoff is still where OT wireless breaks.

What 6 GHz changes

The FCC opened 1,200 MHz from 5.925 to 7.125 GHz¹ to unlicensed use in 2020. Low-power indoor access points can use the whole band indoors, without an automated frequency coordination (AFC) system. Standard-power access points can go outdoors, but only under AFC control and only in the 5.925 to 6.425 GHz and 6.525 to 6.875 GHz sub-bands.

Three things matter on a plant floor.

No radar to dodge. The low-power indoor rules protect incumbents with three restrictions: indoor operation, a contention-based protocol and low power. Radar detection is not one of them. The incumbents are fixed links, broadcast auxiliary services and satellite services, so the DFS channel evacuations that interrupt 5 GHz cells are not part of the 6 GHz rule set.

No legacy clients. Wi-Fi 6E opened spectrum free from legacy device interference. The first amendment to operate there was 802.11ax², so every client in a 6 GHz cell speaks at least Wi-Fi 6.

Less power, especially uplink. Low-power indoor access points are held to 5 dBm/MHz, which reaches 30 dBm only on a 320 MHz channel and 27 dBm on 160 MHz. Clients get 6 dB less. In the EU and UK the low-power indoor limits are 23 dBm and 24 dBm EIRP regardless of channel width. When a vehicle’s control loop depends on its uplink reports, the weak side of the 6 GHz link is the vehicle.

Field note. Bot and AGV density is rising across industries and is saturating channels. 6 GHz is being deployed to relieve utilization.

Wide channels and dense fleets

Wi-Fi 7 adds 320 MHz channels in countries that open 6 GHz³. The U.S. band holds three of them, or seven 160 MHz channels⁴.

Width trades against channel count, and dense OT deployments run out of channels first. To send wide, a station has to find its secondary channels idle, not only its primary. RTS, CTS and acknowledgment frames take the same airtime at any width; only the data portion shrinks. Industrial control messages are small. One industrial evaluation scenario used 48-byte packets⁵. For payloads that size, fixed overhead is a large share of each exchange, and a wider channel does not shorten it.

The cost lands on the neighbors. In simulations of dense WLANs, always taking the widest available channel maximizes a network’s own throughput but often starves the others. Many clients cannot use the width anyway. Access points commonly support 160 or 320 MHz, while many stations operate only on 20, 40 or 80 MHz.

Puncturing helps at the edges. Preamble puncturing lets an access point skip a busy 20 MHz subchannel and keep transmitting on the rest of a wide channel. It works through OFDMA, so its benefit shrinks when few stations are active, which is one reason 802.11be also lets a single station use multiple resource units. Puncturing helps a wide channel survive interference. It does not make a wide channel the right choice for a dense fleet.

OFDMA schedules the cell, not the building

Wi-Fi 6 brought OFDMA, which divides a channel into resource units and gives the access point scheduled uplink access. Trigger-based uplink reduces contention when several stations transmit at once. For many small packets from many clients, that is the right tool.

Two limits remain. Under low-power indoor rules, a station sending on a small resource unit has its total transmit power cut by the narrow bandwidth, which hurts the uplink at the cell edge. The fix, distributed resource units, is an 802.11bn feature. And OFDMA schedules one cell. Between cells, the absence of explicit coordination among access points remains a major obstacle to reliability. Multi-AP coordination is on the 802.11bn feature list, not in Wi-Fi 7.

Multi-link operation: redundancy inside one AP MLD

Multi-link operation (MLO) lets a client use several links, for example one in 5 GHz and one in 6 GHz, through a single association. The access point side is an AP multi-link device (AP MLD): two or more affiliated access points, with each link set up between a pair of affiliated radios. The gains are aggregation, more chances to win the channel, and redundancy across links.

How much of that a client gets depends on its hardware. An enhanced multi-link single-radio (EMLSR) client listens on two or more links but has one fully capable radio, and switches to whichever link it is called on. That is fast link selection, not two copies of every packet. Under high load, the most capable multi-radio mode can block neighboring cells and occasionally produce larger worst-case delays than single-link operation.

The limit that matters most for a vehicle is simpler. Every link in an MLO association belongs to the same AP MLD. When the vehicle drives out of that AP MLD’s coverage, it roams to another one, and transitions between AP MLDs are an 802.11bn mechanism built on MLO⁶, not part of Wi-Fi 7.

Make-before-break roaming is Wi-Fi 8, and it is a draft

Today’s roam is break-before-make: the client leaves its access point, then associates and re-authenticates with the next. IEEE 802.11be, the standard behind Wi-Fi 7, received IEEE Standards Board approval on September 26, 2024⁷, months after Wi-Fi Alliance launched Wi-Fi CERTIFIED 7 on January 8, 2024. It did not change that.

802.11bn, Ultra High Reliability, is the foundation for Wi-Fi 8. It defines a Seamless Mobility Domain (SMD) that groups several AP MLDs into one logical coverage area. Before the roam, the current and target AP MLDs transfer the client’s context. The roam itself is make-before-break: the old AP MLD keeps delivering buffered downlink data while the new one can start, and no re-authentication is needed inside the domain. The project’s authorization sets targets such as 25% reductions in 95th-percentile latency and in MPDU loss, particularly during transitions between overlapping networks.

It is not finished. Draft 2.0 passed its working group letter ballot with 77% approval. At the September 2026 meeting the task group resolved about 1,850 comments, with roaming among the topics, and produced Draft 2.1⁸. It projects an initial IEEE SA ballot in May 2027 and RevCom and Standards Board approval in May 2028. Products are built once the final specification is published. Plan the vehicles you buy now around break-before-make roaming.

Before you buy

The access point is rarely the limit. None of the vehicles in the requirements database that publish their bands lists 6 GHz; each lists 2.4 and 5 GHz only. Ask about the radio on every mobile client:

Does it support 6 GHz, and at which channel widths?
Does it support MLO, and is it single-radio or multi-radio?
Which roaming methods does it implement, and what roam time does its vendor publish, measured how?
Will its hardware support 802.11bn roaming, or will that take new hardware?

Wi-Fi 7 makes each cell better. Your vehicles still have to leave it.