Distance for WiFi: How Far Your Network Really Reaches

Stop asking how far a router reaches and expecting one number to survive contact with a real building. That's the most popular advice about distance for WiFi, and it's the advice that creates dead bedrooms, stalling payment terminals and hotel rooms where the network name is visible but a video call won't hold. A device can detect a beacon long after throughput, latency and roaming have become unusable.

I would plan every site the other way round. Decide what each room needs, work out what the walls and floors will take away, then place access points so clients stay above a usable threshold with margin to spare. Newer WiFi doesn't change that logic. It mainly adds capacity near the access point, which is useful, but it isn't reach.

Typical WiFi Distance by Band and Building

There's no single WiFi range you can quote with certainty. The useful figure is usable distance, meaning the distance at which a client still gets reliable throughput at roughly -70 dBm or stronger, not the furthest point where it can see the network name.

Treat these as starting points for access point spacing, then prove them on site. A foil-backed partition, a lift shaft or a crowded neighbouring network can make the optimistic end of any row irrelevant.

How to read the table

Start with the environment closest to the actual path, not the floor area. A 20 metre path across an open office and a 20 metre path through two brick walls and a bathroom pod are different jobs, even though the tape measure says they're the same.

Then use the band the client will really use. Laptops, phones, scanners, door locks and IoT sensors don't all have the same radios, antennas or band support. If you're designing for a scanner that only works well on 2.4 GHz, the 6 GHz column doesn't help you.

Plan for the room where the device must work, at the height and position where a person will use it, then check the result in the field.

Ofcom's spectrum access modelling for 5 GHz WiFi identifies 2.4 GHz as the UK's most commonly used Wi-Fi band, while noting that 5 GHz has shorter transmission range despite offering substantially more spectrum and wider channels. It also puts the available spectrum at 83 MHz at 2.4 GHz and 585 MHz at 5 GHz, which is why I wouldn't treat 2.4 GHz as a capacity band even though it travels further.

The same modelling applies approximately 12 dB of building-penetration attenuation when an indoor access point communicates with an outdoor receiver. Don't lift that figure and use it as a generic indoor wall loss. It shows that a building boundary matters, but your indoor loss depends on the actual construction in the path.

What this means on UK sites

Use 2.4 GHz for coverage-critical or obstructed areas, such as legacy devices, awkward corners and low-rate IoT. Use 5 GHz as the main capacity layer in most offices, hotels and student accommodation. Add 6 GHz where clients support it and you can place cells densely enough to keep them close. That last point matters more than the band labels suggest: newer generations add capacity near the access point, so a 6 GHz deployment only pays off when the devices stay inside its shorter usable radius.

Even a favourable indoor test shouldn't become a design promise. Reinforced concrete, metal structures, floors and poor access point placement can cut usable distance sharply on the next site.

The same room by room thinking applies when a building itself is temporary or assembled in modules. If you're planning connectivity for distributed teams in that kind of space, it's worth seeing how others explore modular workspaces for remote teams before you assume a standard office layout. Partition material and module joins still decide where access points belong.

For dense commercial sites, I would rather install additional access points at lower transmit power than stretch one cell until its edge becomes unreliable. Smaller cells give you cleaner roaming, more airtime per client and fewer arguments about whether a weak connection counts as coverage.

What Actually Limits Your WiFi Distance

WiFi works like a torch beam in fog: distance dims it steadily, and every wall, floor and coated pane of glass takes another share before the signal reaches the client. You don't fix fog by quoting the torch's maximum range.

In open space, every doubling of distance costs about 6 dB. A 3 dB loss roughly halves received power, so a few obstacles quickly turn a healthy signal into a marginal one. Clients then drop to lower data rates, retry more often and consume more airtime for the same traffic.

Treat the higher-frequency penetration row as a direction of travel rather than a design input. Attenuation climbs as frequency rises, and a structure that passes 2.4 GHz tolerably can still gut anything above it. For a 2.4 GHz or 5 GHz design, use material losses that match the band and construction in the path, then walk the building room by room to confirm them.

Interference shrinks cells without moving walls

A client needs a clean enough signal-to-noise ratio, and a strong reading in dBm alone tells you nothing about whether frames decode reliably. Neighbouring networks, Bluetooth, cordless devices, microwave ovens and your own overlapping cells all raise contention or noise. The access point can look loud while the client still can't get a usable connection.

That's why turning power up often disappoints. Louder cells overlap more, clients cling to distant access points, and the return path from a battery-powered phone stays weak regardless of what the access point transmits. UK rules also cap transmit power as EIRP, and for 2.4 GHz you should check against 100 mW (20 dBm) before raising power. Even where headroom exists, I wouldn't spend it before fixing placement and channel reuse.

Newer standards don't repeal physics

802.11n, 802.11ac, 802.11ax (Wi-Fi 6) and 802.11be (Wi-Fi 7) improve scheduling, efficiency and capacity. That extra capability shows up close to the access point. Reinforced concrete still costs the same 15-20 dB, so newer standards extend what clients can do inside a well-served room rather than pushing coverage into rooms the signal can't reach. Higher bands generally attenuate more through walls, which means extra spectrum delivers speed near the cell while doing nothing for a space two floors away.

Ookla data reported by ISPreview put the UK shares at 43.4% Wi-Fi 6 and 3.8% Wi-Fi 7, while Wi-Fi 5 still represented 38.7% and Wi-Fi 4 14.1%. Your design therefore has to serve a mixed client estate, where the oldest device in a corridor sets a real constraint the newest phone in the sales meeting can't lift.

Distance is a per-room question answered in dB: frequency, materials, placement, interference and the client's own radio each take their share. The range figure on the box is marketing's problem, and I'd leave it there.

How to Work Out Expected Coverage With a Link Budget

A quoted range is the lazy version of coverage planning. A link budget adds gains, subtracts losses and compares the remainder with the level a client needs, so the question becomes margin in each room rather than metres on a datasheet.

The UK Government guidance on Wi-Fi radio waves and health identifies the principal Wi-Fi operating bands as approximately 2.4 GHz, 5 GHz, and 6 GHz. Choose the band for the job first: higher frequencies can carry more traffic close to the access point, while lower ones usually reach farther through the same building. The budget then decides whether a particular seat, bed or scanner position closes.

Usable distance is the point where required signal strength, reliability and throughput still hold together. A beacon you can barely detect does not prove a room is covered, and a fast result beside the access point says little about the far corner behind a riser.

Run the calculation in six steps

1. Note transmit power and antenna gain: Record the access point's configured power and antenna gain, then note the client's typical transmit power too. The uplink from a phone or scanner often fails first, so a downlink-only calculation flatters the design.
2. Set the receiver threshold: Use around -67 dBm for voice and video and -70 dBm for general data as planning targets. Payment terminals, scanners and real-time applications may need a tighter target, so set it from the service, not from habit.
3. Subtract free-space loss: Estimate loss at the target distance, allowing about 6 dB for each doubling. That gives the baseline before the building gets involved.
4. Subtract obstacles on the path: Add each wall, floor, door and pane of coated glass between access point and client. Draw the path on the floor plan, because the shortest line often passes through a bathroom pod, riser or lift core you'd otherwise miss.
5. Add a fade margin: Allow around 10-15 dB for people moving, furniture, doors closing, body loss and day to day interference. Without margin, a design that passes on paper at 9 a.m. can fail when the room fills up.
6. Compare the result with the threshold: If the predicted level after margin stays above the threshold, the cell is usable on paper. If it falls below, move the access point, add one, change the antenna pattern or redesign the cell. Don't solve a negative margin by hoping clients will cope.

A worked example

Take a hypothetical 2.4 GHz access point transmitting at 17 dBm with a 3 dBi antenna. That gives 20 dBm EIRP for the worked path. The client has a 0 dBi antenna and sits 20 metres away through one plasterboard wall.

Allow about 66 dB of free-space loss at 20 metres and 4 dB for the wall. The received level is 20 minus 66 minus 4, which gives -50 dBm. Subtract a 12 dB fade margin and you're planning around -62 dBm.

That's 8 dB above a -70 dBm general-data threshold, so the cell is usable with sensible headroom. Add a brick wall, move the client behind a metal cabinet or drop to a weaker handheld device, and that headroom disappears quickly.

Margin decides the result: A 20 metre path with positive margin is coverage. A 12 metre path with negative margin isn't, however close it looks on a plan.

Run the budget for the worst legitimate client position in each key room. Corners furthest from the door, beds against external walls, desks behind storage and seats beside lift cores are better test points than the centre of an empty room.

Also check the reverse direction. If the access point transmits much louder than the client can answer, you'll create a connection that looks present but stalls under traffic. Balanced power and smaller cells usually roam more cleanly than one loud access point trying to dominate a floor.

Treat the calculation as a starting estimate. Use it to place access points provisionally, then revise the losses where measured rooms disagree with the plan.

How to Survey Real Coverage Before You Install

Hit the thresholds in the rooms that matter, or don't call the design finished. For voice and video, aim for -67 dBm or stronger. For general business data, aim for -70 dBm or stronger. In both cases, look for a signal-to-noise ratio of about 25 dB or more if you want stable throughput rather than a fragile association.

Bring a laptop or phone running a proper analyser, a scaled floor plan and a list of the device types that will use the network. A browser speed test on your own flagship phone isn't a survey. If you need a practical starting point for tooling, this guide to choosing a WiFi analyzer application covers what to look for before you walk the site.

Walk the site as users will use it

Stand in each room at desk height, bed height or scanner height, depending on the job. Test by the door and at the furthest usable corner, because readings can change materially across a few metres. Shut fire doors where they'll normally be shut, since an open door can hide a problem you'll inherit after handover.

Walk corridors slowly enough to observe roaming behaviour. Note where the client lets go of one access point, which one it joins next and whether traffic stalls during the transition. Include stairwells, lifts, basements, plant-adjacent areas and any room where staff stand to complete a task.

• Record the channel: Note channel width, utilisation and neighbouring networks on the same or overlapping channels, not just your own signal level.
• Check the noise floor: A fair dBm reading with poor signal-to-noise ratio will still produce retries and inconsistent throughput.
• Test real traffic: Run a file copy, cloud application task or controlled download in the worst room, because congestion can make a strong signal perform badly.
• Use representative clients: Test the oldest supported laptop, the handheld scanner and the phone model staff carry, not only the newest device in your bag.

Turn readings into decisions

Produce a heat map where the tooling supports it, or a room by room table marked pass, marginal or fail. Marginal isn't a pass you can explain away. It's a warning that occupation, furniture or a neighbouring network can push the room over the edge.

I would adjust the link budget to match measured losses, then move or add access points based on that evidence. If a corridor access point can't serve the far side of a hotel bedroom through a bathroom pod, the answer is usually an in-room or closer access point, not more power in the corridor.

The survey decides where access points go. Keep the floor plan, readings, device types and test positions together so another engineer can repeat the walk and get comparable results after you make changes.

Choosing Between Access Points, Mesh and Extenders

Start with backhaul, not product category. If you can cable it, you remove the weakest link from the design.

Wired access points

For a multi-room commercial site, I'd choose wired access points wherever there's a practical cabling route. Each radio gets full wired backhaul, you control channel and power per cell, and roaming is designed rather than improvised. The cost is real, especially in occupied hotels or listed buildings, but so is the cost of revisiting a site that never had enough capacity.

Managed platforms can run across different hardware, so you don't have to let vendor lock-in decide the radio design. Purple, for example, operates as a hardware-agnostic overlay on existing access points for guest, staff and multi-tenant access. That's separate from RF design. Choose access point positions for coverage and capacity first.

Mesh systems

Mesh suits homes and small sites where cabling is impractical and traffic is moderate. Nodes coordinate under one network name and can steer clients more sensibly than an isolated repeater. Place each node where it still has a strong link back towards the root, not inside the dead zone you're trying to rescue.

Every wireless hop shares airtime. One hop may be acceptable for browsing and streaming, but several hops through brick walls will eat throughput and add latency. If a node can take Ethernet backhaul later, that's an upgrade worth planning for.

Range extenders

An extender is a patch for one dead room. It listens and retransmits, which typically halves available throughput on that link and can leave clients attached to a weaker repeated signal. A separate network name makes the experience worse because devices don't always move when they should.

I wouldn't build a student block, clinic or hotel floor out of extenders. Use one only where the requirement is light, the room can't be reached by a proper access point and the customer understands the compromise.

Bridges and directional antennas

A wireless bridge links buildings where digging or cabling isn't practical. It needs clear line of sight, careful alignment and a check of power, licensing and mounting rules before installation. Trees, new signage and construction can intrude into the path later, so document the route and keep it clear.

Directional antennas help in long corridors, warehouses and outdoor point-to-point paths because they concentrate energy where it's needed. They don't create energy from nowhere. Aim them badly and you'll create a strong strip of coverage with dead areas either side.

If you're comparing whole architectures rather than single products, these WiFi solutions for businesses set out the service models that sit on top of the radio layer. The decision rule stays simple. Cable commercial access points where you can, use mesh where you can't cable a small site, and reserve extenders for a single stubborn room.

Common Coverage Problems and How to Fix Them

The complaint usually sounds vague: a device connects, works for a while, then drops or stalls in one room. Resist replacing hardware first. Separate weak signal, poor signal-to-noise ratio, bad roaming and congestion, because each fault has a different fix and the wrong change can hide the cause for another week.

• Weak only on 5 GHz or 6 GHz: Move the access point closer or add one on the same floor. Raising power rarely fixes a path blocked by brick, concrete or a bathroom pod, and it can worsen overlap elsewhere.
• Falls back to 2.4 GHz and slows down: Check band steering, roaming thresholds and whether the nearer access point offers a strong enough 5 GHz cell. A client will not choose a better band if that band is marginal where it sits.
• Good signal but poor speed with many networks nearby: Reduce channel width, move to a cleaner channel and check neighbouring SSIDs for overlap. Wide channels look impressive beside the access point and fragile in a dense block.
• Sticky client stays on a distant access point: Raise minimum basic rates carefully, tune roaming thresholds and balance access point power with typical client power. Test after each change, because aggressive settings can disconnect weak but legitimate devices.
• Works at night and slows during the day: Look for microwave ovens, cordless phones, neighbouring networks and peak client load. Log the time pattern before moving radios, since an interference source may only appear when a kitchen or workshop is busy.
• Fine in the corridor, poor at the desk or bed: Survey inside the room with doors shut and design for the furthest user position. Corridor readings flatter hotels, student rooms and apartments in particular.

Hospitality makes the same coverage mistake expensive. Guests judge the connection in the room rather than the corridor heat map; the wider “work where you stay” expectation behind guides such as what Tulum offers remote workers is a useful reminder that the test point is the chair, desk or bed, never the doorway. Prove service at that position.

Most configuration fixes take one visit and a re-survey. Structural fixes take longer. Adding wired access points, opening ceilings or agreeing a new cabling run in an occupied building can take weeks to schedule, so identify that need early rather than spending three visits tuning a cell that is in the wrong place.

A fix is proven when the same room passes the agreed signal, signal-to-noise and traffic tests with the client's real devices, not when the bars improve.

Change one variable at a time where you can. Move an access point, retest. Narrow a channel, retest. Adjust a roaming threshold, then walk the route again. If you change power, channel, band steering and placement in one visit, you will know the result changed without knowing why.

Monitoring Coverage After Handover

Coverage drifts after handover. Tenants move furniture, neighbours install louder networks, a refit adds foil-backed insulation and firmware changes radio behaviour. If you only return when users complain, you'll always be diagnosing yesterday's problem with incomplete evidence.

• Track signal and client count per access point: Watch which radios carry the load, which clients sit at weak levels and whether one cell is absorbing users who should be elsewhere.
• Log channel utilisation and noise: Rising retries, noise or airtime use often appear before staff start reporting dropouts.
• Alert on offline access points: A failed radio can look like a mysterious coverage complaint in three rooms until someone checks availability.
• Record roaming failures: Ask staff where calls break or scanners pause, then correlate those locations with client and access point logs.
• Watch device mix: New handhelds, laptops or IoT devices can change band use and range behaviour across a site.

Set alerts on thresholds rather than raw counts. Ten retries means little without context, but a sustained rise in retries, weak clients or noise on one radio tells you where to walk first. That's more useful than a dashboard full of green lights while a meeting room fails.

Understanding the wider device load also helps you forecast capacity. This overview of how many devices are connected to the internet is useful context when clients ask why last year's design now carries more phones, sensors and laptops per person. More devices don't reduce radio range directly, but they consume airtime and expose weak cell edges sooner.

Re-survey after any refit, new partition, tenant change or layout change. I'd also run a quick walk test each quarter in the rooms that matter most, using the same positions and device types as the original survey. Managed monitoring tools can collect these measures across several client sites, which helps an MSP spot patterns and evidence service levels before a support call sets the agenda.

Keep a handover pack in one place. Include the scaled floor plan, access point positions, transmit settings, channel plan, survey date, pass thresholds and the last room by room results. The next engineer should be able to compare today's readings with the signed-off baseline without reconstructing your decisions from email threads.

Confirm Your Coverage Plan Before You Sign Off

Don't sign off because every access point is online. Sign off because the site passes the tests you agreed before installation.

1. Positive link-budget margin: Each key room keeps a positive margin after path losses and the fade allowance, using the client threshold for the service in that room.
2. Survey thresholds met: Readings reach -67 dBm for voice and video and -70 dBm for general data in every room that matters, with signal-to-noise ratio around 25 dB or more.
3. Real paths allowed for: Access point positions on the floor plan account for the walls, floors, doors and glass on the actual route to each user position.
4. Bands designed on purpose: 2.4 GHz covers obstructed and coverage-critical areas, while 5 GHz and 6 GHz provide capacity where clients are close enough to use them well.
5. Real devices tested: The walk test includes each important device type, its band support and its roaming behaviour between cells.
6. Power and links compliant: Transmit power sits within UK EIRP limits, and any bridge or outdoor link has clear line of sight plus the required licensing and mounting checks.

Hand over the floor plan, survey results, channel and power settings and monitoring plan together. A customer who only receives a list of installed access points can't tell whether their building still performs as designed in six months.

If a site fails any point, re-run the link budget with measured losses, move or add an access point and re-survey before sign-off. I would rather delay handover for one cable run than accept a marginal room and spend the next year explaining why a visible network isn't a usable one.

Purple helps MSPs, integrators and venue operators turn properly surveyed WiFi into managed guest, staff and multi-tenant services, with identity-based access and reporting layered over the access points already on site. If you're planning coverage room by room and want a hardware-agnostic way to run and evidence that service across client sites, visit Purple to explore the partner platform.