Wi-Fi Wall Attenuation Calculator

Introduction: why wall material decides your Wi-Fi coverage

A Wi-Fi router radiates a fixed amount of power, and everything that happens after that is subtraction. Air takes its cut through distance spreading. Every wall, floor, door and appliance in the path takes another cut by absorbing and reflecting part of the wave. The number your laptop finally reports, the RSSI in dBm, is what is left. The single biggest lever most people have over that number is not the router; it is what the signal has to pass through to get where it is going. A device three rooms away through stud-and-plasterboard partitions can easily beat a device one room away through a poured concrete shear wall.

This calculator makes that subtraction explicit. You enter the clear-path signal you would expect at the device location, count the barriers that actually sit between the router and the device, choose the band the client connects on, and the page returns the estimated received power in dBm and in absolute power, together with a quality rating against the thresholds practitioners use. The arithmetic is deliberately simple so it runs instantly and so you can see exactly which barrier is responsible for the loss, which is usually the thing you want to know before you start moving hardware around.

Two refinements separate this page from a plain lookup table. The first is frequency. A 5 GHz or 6 GHz link does not lose the same amount in a given wall as a 2.4 GHz link does, and the difference is not a single blanket multiplier: it depends on the material. The second is vertical paths. Floors and ceilings are usually the heaviest single obstacle in a house, and they deserve their own input rather than being faked as extra walls. Both refinements are driven by published ITU-R recommendations rather than rules of thumb, and both are explained below so you can check the working.

How to use the Wi-Fi wall attenuation calculator

Starting signal strength (dBm) is the clear-path reference: the reading you would see at the device's location if nothing were in the way. If you can, walk to the spot with a phone Wi-Fi analyser while standing in line of sight of the router and use that figure. If you cannot, a decent working default for a typical home router is between -35 dBm and -45 dBm a few metres away. The field accepts anything from -120 dBm to 30 dBm; values outside that range are rejected rather than silently used, because they are almost always a typo.

Frequency band selects 2.4 GHz, 5 GHz or 6 GHz. This does not change the starting signal you entered; it changes how lossy each barrier is, using the scaling described in the formula section. Pick the band your client device actually associates on. If you are comparing bands, run the calculation twice and compare the two results.

Wall counts are the number of each material the straight line from router to device crosses. Count a stud partition finished with plasterboard on both faces as one drywall wall, not two. An interior brick or block wall counts as brick. A poured or precast concrete wall, a lift shaft, a stair core or a wall with dense rebar counts as concrete. A window, a glazed partition or a French door counts as glass. Open doorways count as nothing at all, which is exactly why lining the router up with a doorway often beats buying a bigger antenna. Each count is capped at 40, and fractional entries are rounded with a note in the result.

Floors or ceilings crossed is the number of horizontal slabs in the path, and building type tells the calculator which published floor penetration factor to apply. Leave floors at zero for a same-storey path.

Press Calculate signal to run the estimate. The result panel shows the received power in dBm, the same figure as absolute power in picowatts or nanowatts, a full loss breakdown by material, and a quality verdict. The gauge underneath plots that value against the -80, -75 and -67 dBm thresholds. Reset restores the defaults and clears the panel. Copy result puts a one-line plain-text summary on your clipboard.

Formula for path loss in dB through walls and floors

The received power in dBm is the clear-path reference power in dBm minus the sum of every barrier loss in the path, plus the floor penetration term :

Each barrier loss is a 2.4 GHz baseline scaled to the band in use. ITU-R Recommendation P.2040 models a building material with a real relative permittivity and a conductivity , with the frequency in GHz. For concrete, brick, plasterboard and glass the recommendation gives , so the permittivity is flat across the Wi-Fi bands and the attenuation rate in dB per metre, which is proportional to , carries the whole frequency dependence. The loss through a wall of fixed thickness therefore scales as a simple power law:

The exponents come straight from Table 3 of ITU-R P.2040: 0.9395 for plasterboard, 0.16 for brick, 0.7822 for concrete and 1.3394 for glass. That is why the bands behave so differently. Glass nearly triples its loss between 2.4 GHz and 6 GHz, plasterboard roughly doubles, concrete not quite doubles, and brick moves by barely a decibel because its fitted exponent is so small.

The floor penetration term uses the factors published in ITU-R Recommendation P.1238 for floors crossed:

Finally the decibel-milliwatt figure is converted back to absolute power so the scale of the change is visible, using the standard definition in which 0 dBm is exactly 1 mW:

Expressing the answer both ways is worth the extra line. A path that starts at -40 dBm and crosses two concrete walls at 2.4 GHz loses 20 dB and lands at -60 dBm. In dBm that reads like a modest change. In power it is a drop from 100 nW to 1 nW, a factor of one hundred, and that is the number that explains why the video call stutters.

Worked example: reaching the back bedroom at 5 GHz

A router sits in the living room of a flat. Standing in the back bedroom with a clear path, a phone would read about -45 dBm. The real path to that bedroom crosses one poured concrete wall, one interior brick wall and one drywall partition, all on the same storey, and the laptop connects on 5 GHz.

Scale each baseline to 5 GHz with the power law. The frequency ratio is 5 / 2.4 = 2.0833. Concrete: 10 dB × 2.0833^0.7822 = 10 × 1.776 = 17.8 dB. Brick: 6 dB × 2.0833^0.16 = 6 × 1.125 = 6.7 dB. Drywall: 3 dB × 2.0833^0.9395 = 3 × 1.993 = 6.0 dB. The barriers total 30.5 dB, so the estimate is -45 - 30.5 = -75.5 dBm, or about 28 pW of received power. That lands just the wrong side of the -75 dBm usable line, so the calculator rates it Weak: browsing may limp along, a 4K stream will not, and the room is a strong candidate for a mesh node or a wired access point.

Now run the same path on 2.4 GHz. The barriers are simply 10 + 6 + 3 = 19 dB, the estimate is -64.0 dBm, about 398 pW, and the link is comfortably good. Fourteen times more power arrives on the lower band over exactly the same path, which is the whole argument for keeping 2.4 GHz enabled for far-flung devices even when 5 GHz is faster up close. Add one residential floor to the 5 GHz path and the extra 4 dB drops it to -79.5 dBm, below the practical working threshold. Remove the concrete wall instead, by re-routing through an open doorway, and the 5 GHz estimate improves to -57.7 dBm. One concrete wall is the difference between a dead room and a working one.

Typical wall loss table by material and band

These are averages for ordinary construction. Foil-backed insulation, wire mesh in plaster, low-emissivity coated glazing, mirrors and concrete with dense rebar can all attenuate far more than the table suggests, sometimes by 20 dB or more, and behave closer to metal than to the nominal material. Hollow doors, lightweight partitions and large openings let more energy through. The floor rows are held constant across bands because ITU-R P.1238 publishes them as fixed factors rather than as frequency-scaled values.

Sources and standards behind the attenuation values

Sources: the frequency-scaling exponents and the material electrical model come from ITU-R Recommendation P.2040, "Effects of building materials and structures on radiowave propagation above about 100 MHz", whose Table 3 gives the coefficients for the conductivity law used above. The floor penetration loss factors come from ITU-R Recommendation P.1238, "Propagation data and prediction methods for the planning of indoor radiocommunication systems". The 2.4 GHz per-material baselines are consistent with the measured construction-material attenuation reported by the U.S. National Institute of Standards and Technology in NISTIR 6055, "Electromagnetic Signal Attenuation in Construction Materials". The dBm-to-milliwatt conversion follows the standard definition of the decibel-milliwatt, in which 0 dBm equals exactly 1 mW.

Long-form discussion: how buildings really treat radio

Electromagnetic waves at 2.4, 5 and 6 GHz interact with building materials in ways that a single decibel figure can only summarise. The complex permittivity of a material determines how much the wave slows down and how much of its energy is dissipated as heat. Materials with meaningful conductivity, such as reinforced concrete or anything with a metal layer, dissipate strongly and can effectively shield a room. The orientation and spacing of rebar inside concrete produces frequency-dependent behaviour, because a mesh with a spacing comparable to the wavelength acts partly as a screen. The moisture content of a wall matters as well; wet masonry has higher dielectric loss than dry masonry, which is why the same wall can measure differently in winter and summer.

Reflection is the other half of the story, and it is not captured by an absorption figure at all. When a wave meets a wall, part of the energy is transmitted, part is absorbed inside the material, and part is reflected back. The reflected component takes an indirect route to the receiver and arrives with a different phase, so it can add to or cancel the direct component. That is multipath fading, and it is why moving a laptop thirty centimetres can change the reported RSSI by several decibels. The simple subtraction model used here reports an average expected level rather than trying to predict the fine structure of the field. For a proper site survey, engineers walk the space with a calibrated receiver and build a heat map, because no closed-form model reproduces the real pattern of nulls and peaks.

The logarithmic decibel scale is the other thing that trips people up. A 10 dB change is a factor of ten in power, and 3 dB is a factor of two. Because Wi-Fi levels are quoted relative to one milliwatt, useful numbers are negative: -30 dBm is a very strong signal found within a metre or two of a router, -67 dBm is the practical floor for smooth video, and -90 dBm sits in the noise. Subtracting three brick walls at 6 GHz removes about 21 dB, which is a factor of more than one hundred in power. That is why construction style, rather than router wattage, dominates real coverage; transmit power is capped by regulation and doubling it only buys 3 dB.

Frequency deserves a longer look because the popular summary, "5 GHz does not go through walls", is only half right. The free-space part of the loss really is worse at higher frequencies for a fixed antenna aperture, and diffraction around corners really is weaker. But the wall-penetration part depends on the material. The ITU-R P.2040 exponents show brick barely changing between 2.4 and 6 GHz, while glass changes by a factor of more than three. In a brick-built home the 5 GHz penalty is mostly a distance and diffraction penalty. In a glass-partitioned office, or a home with low-emissivity double glazing, the penalty is dominated by the glazing. Knowing which of those you live in tells you whether a band-steering change or a second access point is the right fix.

Floors and ceilings are the most under-counted obstacle in home networks. A timber joist floor with a chipboard deck is relatively benign; a concrete slab with in-floor heating pipes and a metal deck is close to a shield. ITU-R P.1238 recognises this by giving residential, office and commercial floors different factors, and by making the loss strongly non-linear in the number of floors: the first office floor costs 15 dB and each additional one only 4 dB, because once a path is diffracting around the building's exterior the extra slab matters less. Placing a router on the middle floor of a three-storey home is almost always better than placing it in the basement, and the calculator makes the size of that difference concrete.

Walls are not the only obstacles that matter. Furniture, appliances and people all contribute. A fridge or a full filing cabinet behaves roughly like a brick wall. A human body is mostly water and costs a few decibels at 2.4 GHz and more at 5 GHz, which is why a crowded room measurably degrades Wi-Fi and why a router placed behind a sofa performs worse than the same router on a shelf. Fish tanks are notoriously bad. Because the model simply sums decibels, all of these can be represented by adding an equivalent wall count in whichever material best matches the obstacle, and the explanation of which count stands for what is worth writing down if you are documenting a deployment.

Understanding barrier loss pays off well beyond home networking. Office fit-outs have to guarantee coverage in glass-walled meeting rooms. Warehouse and factory deployments contend with steel racking and reinforced concrete, where path loss is a first-order design parameter rather than an afterthought. Building-automation radios, smart locks and sensor networks operate with far less link margin than a laptop does, so a 10 dB wall that a phone shrugs off can put a battery-powered sensor permanently offline. The same subtraction, done carefully, scales from a one-bedroom flat to a distribution centre.

The practical takeaway is that signal strength is situational, and the most valuable thing a model like this gives you is a ranked list of what to change. If one concrete wall accounts for 18 of the 30 dB in your path, no amount of antenna tuning will fix it and the answer is a second access point on the far side. If four drywall partitions account for the same 24 dB, moving the router into a hallway that lines up with the doorways may fix it for nothing. Run the numbers, make the change, and then verify with a real measurement from a phone Wi-Fi analyser, because the measurement is the ground truth and the model is only the hypothesis.

Limitations and assumptions behind this estimate

This calculator is a first-pass planning estimate, not a substitute for an on-site survey, and its assumptions are worth stating plainly. It does not model distance at all: the starting signal strength you supply is assumed to already include free-space spreading loss for the path in question, and the calculator only adds the barrier terms. It models straight-line transmission through each barrier and ignores multipath, so it cannot predict the constructive and destructive interference that makes real measurements vary by several decibels over a few centimetres. It ignores diffraction around obstacles and reflection off room surfaces, both of which sometimes deliver more signal than the direct path does. It assumes normal incidence; a wave crossing a wall at a shallow angle travels through more material and loses more. It says nothing about antenna orientation and polarisation mismatch, regulatory transmit-power limits, receiver sensitivity differences between devices, or co-channel interference from neighbouring networks, any of which can dominate the user experience even when the raw signal level is fine. The per-material figures are averages for ordinary construction: metal-backed insulation, wire-reinforced plaster, low-emissivity glazing, wet masonry and densely reinforced concrete can all attenuate far more, while hollow doors and lightweight partitions let more through. The floor factors are published for generic construction classes and cannot know about your particular slab. Treat the output as guidance for where to place a router or a mesh node, then confirm with a real measurement before committing to hardware.

Common Wi-Fi wall attenuation questions