Why −67 dBm is the WiFi design target

Many enterprise WiFi networks are designed so that devices receive at least −67 dBm from an access point everywhere they are used. It is not a law or a formal standard, but a widely used rule of thumb, drawn from vendor guidance for voice over WiFi, that leaves enough margin for devices to keep clear, reliable connections and roam smoothly.

Where the number comes from

The −67 dBm figure became common through equipment vendors' design guidance for voice over WiFi. Voice is one of the most demanding uses of a wireless network: a call cannot pause and retry the way a web page or email can, and handsets are carried around while in use, so they must move between access points without dropping audio.

Designers found that keeping signal at about −67 dBm or stronger, together with a good signal-to-noise ratio, gave handsets enough headroom to use efficient data rates, recover from brief dips and find a new access point before the old one faded. Because a network designed for voice also handles most other applications well, the figure spread into general enterprise design.

What the target protects

Margin for real-world variation

Signal at any spot moves up and down as people walk past, doors open and close, and devices are held differently. A phone in a pocket or a hand can lose a noticeable amount of signal compared with a survey adapter on a laptop. Designing to −67 dBm leaves room for those losses without falling into the weak range where connections suffer.

Efficient data rates

Stronger signal, with good SNR, lets devices use faster rates. Faster rates mean each transmission takes less airtime, which frees the channel for other users. A network full of devices on slow rates runs out of capacity long before it runs out of coverage.

Smooth roaming

Devices decide when to move to another access point, often when signal drops below their own internal threshold. If coverage is designed so that each area is also covered by a second access point at a usable level, devices have somewhere to go before the call breaks up. See WiFi roaming and handoff for more on how that works.

The target is never just one number

Signal strength alone does not make a good network. −67 dBm is normally paired with other targets, such as:

• Signal-to-noise ratio, commonly around 25 dB for voice. A strong signal in a noisy place still performs badly. See RSSI, SNR and signal strength.
• Secondary coverage, meaning a second access point heard at a usable level, for roaming and redundancy.
• Limited co-channel overlap, so that access points on the same channel do not drown each other out.
• Capacity, so that each access point serves a reasonable number of devices for the applications in use.

When a different target makes sense

−67 dBm is a starting point, not a universal answer. The right target depends on the devices and applications in each area.

Why turning up the power is not the answer

It is tempting to meet the target by setting every access point to full power. That usually backfires. Access points become audible far beyond their intended area, increasing co-channel interference. Devices hear a loud access point but cannot be heard well in return, because their own transmitters are weaker. And devices tend to cling to a distant loud access point instead of roaming to a closer one. Good design meets the target with the right number and placement of access points at moderate power. Access point placement covers how.

Checking a design against the target

A target is only useful if it is checked the same way every time. A practical sequence looks like this:

1. Set the targets per area. Record the primary signal, secondary signal, SNR and overlap targets for each zone, and the device they are based on.
2. Allow for the device. If the survey adapter hears better than the handset, apply an offset agreed in advance, or test with the handset itself.
3. Survey each band. Check the band the critical devices use, not just the one that looks best.
4. Review primary and secondary maps together. An area that meets −67 dBm from one access point but has no second one nearby will still drop calls on the move.
5. Walk it with a call. For voice networks, finish with a live call on the real handset along the routes people actually walk, listening for gaps at the hand-off points.

A worked example

A distribution centre is replacing radios with WiFi handsets for its supervisors. The existing network was designed for laptops in the offices and scanners on the dock, and meets about −70 dBm in most aisles. The voice vendor's guidance calls for stronger signal and a second access point at a usable level throughout.

The survey shows that the aisles meet the primary target only close to each access point, and that secondary coverage disappears between racks. A live call test confirms it: audio breaks up halfway along each aisle. The revised design adds access points at the ends of aisles so each one is covered from two directions, lowers power on the access points over the open dock to reduce overlap, and keeps the voice handsets on 5 GHz. The validation survey checks both primary and secondary coverage against the new targets before the handsets go live.

Common mistakes

• Measuring the target with a survey laptop and assuming phones will see the same.
• Meeting −67 dBm on 2.4 GHz while the devices that matter use 5 GHz.
• Checking primary coverage only and ignoring the second access point.
• Reaching the target with high power rather than good placement, which adds overlap.

How ehZee can help

We set signal, SNR and overlap targets for each area based on the devices and applications you actually use, then measure your network against them with AirMagnet Survey Pro. Our WiFi surveys service shows where those targets are met and where they are not, and what change will close the gap.

Common questions