How to Fix Wi-Fi Dead Zones: A Room-by-Room Troubleshooting Guide

A dead zone is rarely a mystery once you measure it. Radio waves lose energy in predictable ways, and the room where your video call drops is almost always separated from the router by a specific, identifiable obstacle — or is sitting on a channel that four neighbours are also using. Work through the causes in order of how much they cost to fix, and you will usually solve the problem before you reach the part where you spend money.

Step one: put the router in a sensible place

Consumer routers use omnidirectional antennas, which radiate roughly in a doughnut shape around the antenna axis — strong outward, weak along the axis itself. A router lying flat on its back with antennas pointing up puts most of its energy sideways across one floor and comparatively little straight up or down.

Three placement rules follow from that, and all of them are free:

• Central and elevated. Signal fades with distance in every direction, so the router should be in the middle of the coverage area, not in the corner where the fibre or coax happens to enter. A short cable moving it to the middle of the house is the highest-value upgrade most people can make.
• Out of enclosures. A media cabinet with a metal-mesh door, a basement utility room next to the ductwork, or a shelf behind a television all attenuate heavily. Metal is not a partial obstacle; it is a reflector.
• Away from large water and metal masses. Fish tanks, water heaters, fridges and cast-iron radiators absorb or block 2.4 and 5 GHz energy far more than a plasterboard wall does.

Step two: understand what your walls are doing

Attenuation is measured in decibels, and the scale is unforgiving: every 3 dB halves the power. Cisco’s wireless RF reference guide gives typical values of about 3 dB for drywall, 10 dB for a brick wall and 12 dB for concrete with metal content, while cautioning that actual loss depends on moisture, thickness and conductivity.

Glass is where people get caught out. Modern low-emissivity window film carries a thin metallic coating, and it is very effective at blocking radio. A 2024 measurement study by Shakya, Rappaport and colleagues recorded 33.7 dB of penetration loss through a low-emissivity glass wall at 6.75 GHz, rising to 42.3 dB at 16.95 GHz. Thirty-three decibels is a factor of roughly two thousand in power. This is why a modern energy-efficient house often has excellent indoor Wi-Fi and nothing at all on the patio.

Obstacle	Typical loss	Practical effect
Interior drywall partition	~3 dB (Cisco)	Half the power; barely noticeable once or twice
Brick wall	~10 dB (Cisco)	Roughly a tenth of the power; a real barrier
Concrete with metal	~12 dB (Cisco)	Plan an access point on each side
Low-emissivity glass wall	33.7 dB at 6.75 GHz (measured)	Effectively opaque; outdoor coverage needs its own radio

Step three: pick the right band for the distance

Higher frequencies carry more data and travel less far. Free-space path loss alone puts 5 GHz roughly 6 dB below 2.4 GHz over the same distance, and 6 GHz roughly 8 dB below, before any wall is involved. Materials then attenuate higher frequencies more as well.

2.4 GHz

Best range and best wall penetration, worst everything else. In North America the band supports only three non-overlapping 20 MHz channels — 1, 6 and 11 — which every neighbour is also using, alongside Bluetooth, cordless phones, baby monitors, some wireless cameras and microwave ovens, which operate at 2.45 GHz. Keep 2.4 GHz for smart plugs, thermostats and the garage door opener, and stay on channel 1, 6 or 11. Choosing channel 3 or 9 does not avoid interference; it overlaps two of your neighbours instead of sharing one cleanly.

5 GHz

The workhorse. Far more channels, much less legacy interference, moderate range. The best channels in this band are the DFS channels, and most consumer routers avoid them by default, which is why the non-DFS portion is crowded.

6 GHz

Effectively empty, very wide channels, shortest reach. Indoor access points in Canada and the United States operate under low-power indoor rules, so this is a same-floor, same-room-or-next-room band. Excellent for a desk or a media centre, poor for the far end of a basement.

Step four: check the channel and consider DFS

Dynamic Frequency Selection exists because parts of 5 GHz are shared with radar. Under 47 CFR 15.407, devices with any part of their emission bandwidth in 5.25–5.35 GHz or 5.47–5.725 GHz must run DFS. Before transmitting, the device listens for 60 seconds to confirm no radar is present. If radar is detected while in use, all transmissions must cease on that channel within 10 seconds, and the channel is subject to a non-occupancy period of at least 30 minutes. Canada applies an equivalent regime.

The practical upshot: DFS channels are usually far less congested, because so much consumer gear avoids them. If you live away from an airport, a coastline or a weather radar site, enabling DFS on your router often produces an immediate improvement. If you live near one, you will get occasional minute-long dropouts as the radio evacuates a channel — and then you should turn DFS off.

Step five: actually measure, instead of guessing

Two numbers matter. RSSI is received signal strength in dBm, always negative, closer to zero is stronger. SNR is how far that signal sits above the local noise floor, in dB, and it is the better predictor of whether a link will actually work.

Cisco Meraki’s documentation recommends an SNR of at least 20 dB for data and 25 dB for voice, and gives the arithmetic plainly: a client receiving −75 dBm against a −90 dBm noise floor has an SNR of 15 dB. Cisco’s RF reference guide uses roughly −66 dBm as a survey benchmark for client coverage. Those are enterprise targets, but they are the right order of magnitude for a home.

You do not need paid software to collect these:

• macOS — hold Option and click the Wi-Fi menu to see RSSI, noise and channel; Wireless Diagnostics adds a live scanner and a logging monitor.
• Windows — netsh wlan show interfaces reports signal as a percentage, channel and negotiated rate. Percentage is coarser than dBm but adequate for comparing rooms.
• Linux — iw dev wlan0 link gives signal in dBm and the current bitrate.
• Android — several free analyser apps report per-access-point RSSI and channel occupancy.
• iOS — Apple’s AirPort Utility includes a Wi-Fi scanner, enabled from the app’s entry under Settings.

Walk the house at a steady pace and write the numbers down, room by room, for each band. If a room shows −55 dBm and still runs slowly, the signal is fine and the fault is elsewhere.

Step six: if the radio genuinely cannot reach, stop using radio

When a room is behind concrete, behind low-E glass, or two floors down, no amount of channel tuning will help. Move the traffic onto a wire.

MoCA runs Ethernet over the coaxial cable already in most homes. The MoCA 2.5 specification is rated at 2.5 Gbps MAC throughput with average one-way latency under 2.5 ms, using spectrum from 400 to 1675 MHz. If you have a coax outlet near the dead zone and another near the router, this is a genuinely good answer.

Powerline adapters send data over mains wiring. They are the fallback of last resort. Advertised speeds bear little relation to reality, and performance depends entirely on how your house is wired: two outlets on different circuits behind the same breaker panel usually work, two outlets on opposite phases of the panel often do not, and surge protectors between them will kill the link outright. Buy them somewhere with a returns policy.

The practical version

Move the router to the centre of the house and get it out of the cabinet. Put 2.4 GHz on channel 1, 6 or 11 and reserve it for low-bandwidth devices. Try enabling DFS on 5 GHz and see whether the dropouts appear. Walk the house recording RSSI and SNR so you know where the signal actually dies rather than where you think it does. Then, and only then, add hardware — an access point on a cable or a MoCA link if you can, a second radio node if you cannot.