Mesh, extender or access points in a concrete flat

You have one router. It sits wherever the line came into the flat, usually the living room. Two rooms away the video stalls, and the call drops when you walk to the kitchen. A shop will sell you an extender. The internet will tell you mesh. Someone else will say you should have run cable when the flat was being done up.

All three are right about something. Which one is right for your flat comes down to things you can check before you spend anything.

Why the far room is weak in the first place

One router covers about 950 sq ft in a concrete Indian home. That is a rule of thumb, not a measurement — construction, layout and where the line enters change it a lot. An indoor router here is allowed 100 mW EIRP — 20 dBm — and the signal still has to arrive above a usable floor at the far end. That leaves roughly 87 dB to spend, and five metres of air plus one reinforced concrete wall spends a large part of it.

−67 dBm is the standard design threshold for voice and video, but strength alone is not the whole story. What decides whether a video plays is how far the signal sits above the noise floor, and in a tower full of neighbouring networks that margin goes twice over — their spill lifts the floor, their traffic takes the airtime — long before the bars on your phone do. Judge the three options below on that margin, not on bars.

What a range extender actually does

A booster, a range extender and a plug-in repeater are the same device under three names. It joins your WiFi as an ordinary client, then re-broadcasts what it hears.

It does that with one radio, on one channel. A WiFi radio cannot transmit and receive at the same moment on the same channel — it takes turns. Every frame on its way to your laptop therefore crosses the air twice, router to extender and extender to laptop, spending the same airtime on the same channel. What you get through it is roughly half of what the extender itself receives. That’s not a defect in cheap hardware; it’s what the arrangement is.

It is usually worse than half. You plug the extender in where the signal is already poor, so its own link to the router runs at a low rate, and slow frames hold the channel longer while everything else waits.

Then there is the second network name. Most extenders create their own SSID, and even one that copies your existing name is a separate network with nothing coordinating the two. Your phone alone decides when to let go, and phones are reluctant. You walk back to the living room and stay on the extender at one bar with the router three metres away. You feel that as a call that goes bad for twenty seconds after you move.

When an extender is genuinely the right answer

Sometimes it is. A small flat, one weak corner, a couple of devices that live there, no drilling because you rent — an extender is inexpensive, plugs into a socket and takes ten minutes. Half of a good signal where you had almost nothing is a real improvement, and buying mesh to fix one room is overbuying.

Two things decide whether it works. Put it halfway, where the signal is still strong, not in the dead room: it can only re-broadcast what it receives, and a repeater fed a weak signal shows bars and does nothing. And accept switching to the second name by hand — fine for a study, not for a route you walk while on calls.

Mesh, and what each hop costs

Mesh fixes the part of the extender problem that annoys people most: one network name across the flat, nodes that know about each other, and 802.11k/v/r roaming, so a node can tell your phone where to go instead of waiting for it to notice. A network that nudges the phone along, instead of leaving the decision entirely to the phone, is most of the value.

What mesh does not change is the airtime arithmetic. On a two-band kit the satellite talks to the router on the same band it serves your devices with, so the hop still spends capacity that would have been yours. Chain a second satellite off the first and you pay again. A node two rooms and two walls past the router is doing the same physics as the extender, with better roaming and the same ceiling above it.

Tri-band mesh keeps the hop off your airtime

A tri-band kit adds a second 5 GHz radio and keeps it for traffic between the nodes. The hop remains, but it no longer competes with your laptop for airtime, so what reaches the far room is much closer to what the node receives. Where cable cannot be pulled and the back bedroom still has to work properly, this is the sensible answer.

It works on one condition that gets skipped: the link between node and router has to be good. Tri-band buys the backhaul a private lane, not a stronger signal, so a node placed where its own link is poor is an expensive extender. Placement is the whole job, which is why I measure signal and interference room by room instead of putting a node wherever there is a free socket.

Wired access points are the reference answer

Run Ethernet to each access point and the hop disappears. Every unit gets the full line over copper, so no capacity is spent repeating. The units sit on different channels at lower transmit power, so their cells have clean edges and your phone has a reason to move. Wired backhaul is the single biggest difference between real coverage and a chain of repeaters fighting each other. It is what I fit when the flat allows: ceiling access points on cable, one name, one controller.

The cost is not the hardware. It is disruption. Cable has to reach the ceiling or wall where an access point belongs — conduit, some chasing or trunking, a day of work in your home, and often written intimation to the society before anyone drills. A flat painted last month is a harder conversation than one being renovated, which is why cabling is worth planning alongside the rest of the interior work. Most 2–4 BHK homes are a single day on site.

When wiring is not practical

Rented flats where permission to drill will not come. Finished marble or panelled walls with no spare conduit. A society that will not allow the work. In those cases tri-band mesh, placed properly, is what I’d fit, and I’ll say so. There is also a middle path: one cable to the far end of the flat, an access point there, mesh beyond it. A single cable to the worst room beats any number of wireless hops.

How to decide

• Flat size. Under about 950 sq ft with a central router, this is often placement rather than more boxes. Larger, or long and narrow, and one radio never had a chance.
• Wall construction. Count what sits between the router and the failing room. A kitchen wall and two bathrooms is a different problem from three partition walls.
• Whether cable can be pulled. This settles it more than anything else. If a route exists, wired access points win.
• Device count. Phones, laptops, TVs, cameras, speakers, every smart bulb. An ISP-supplied box was never built for that load, and an extender does not change that.
• Whether you own or rent. Cable is an improvement to a flat you keep. If you move in a year, the extender or mesh kit goes in the bag with you.

It may not be a coverage problem at all

Rule out the cheaper explanations first. A router shut inside a wooden TV unit gives away signal for nothing. 2.4 GHz in a Mumbai tower is usually congested, shared with the whole building, and a TV clinging to that band is sharing that congestion with the whole building whatever you install, which is why band steering matters more than another box. The headline speed on the box comes from the widest channel, and a wide channel needs a much better margin above the noise to hold together. Two rooms away the link quietly drops to a narrower channel and a lower rate.

None of that is fixed by buying another box, and sometimes it is the line itself. If that’s what’s happening in your flat, I’ll tell you at the survey instead of selling you equipment you don’t need. Message me the size of your flat and the room where it stops working, and I will give you an honest read.