Daniel moved apartments. Went from sixty-five square meters to about seventy or eighty, and his single Ubiquiti AP — which beat a three-unit mesh in the old place — now can't reach the bedroom and the storage room. He's surprised, and honestly, I get it. The floor plan barely changed. The walls did.
Concrete.
Concrete. Which means he's got two questions. One, is there such a thing as a fill-in AP — something that covers a small dead zone without blasting so much radio in every direction that it steps on the other units? Or is an AP basically just an AP? And two, beyond the ceiling-mounted smoked-glass disc, are there discreet form factors you can build into walls or blend into the room so Hannah doesn't have to look at them? Those are both real questions, and they're connected in ways that aren't obvious.
They really are. And the thing about Daniel's situation — he's renting, so he can't drill, can't run cable, can't do any of the things we'd tell a homeowner to do. But the questions he's asking are exactly the right ones for someone planning a deployment they actually have to live with. So let's start with why one AP was enough before and isn't now, because that's where the physics lives.
The physics being: concrete eats radio.
It does. And it eats different frequencies differently. Two point four gigahertz penetrates walls reasonably well — it'll go through a couple of concrete walls and still be usable, though degraded. Five gigahertz gets through maybe one wall before it's marginal. Six gigahertz, which is new with Wi-Fi 6E and 7, barely gets through drywall, let alone concrete. The attenuation numbers are stark. A single concrete wall can knock a five gigahertz signal down by ten to twenty decibels. That's the difference between a strong signal and one that barely loads a webpage.
So Daniel's apartment only grew by maybe fifteen square meters, but the walls that separate the bedroom and storage room from wherever the AP lives are probably load-bearing concrete. That's not a coverage area problem. That's an obstacle problem.
And this is the first thing most people get wrong about Wi-Fi planning. They think in square meters. They should be thinking in walls. A hundred and fifty square meter open-plan loft might be fine with one AP. A seventy square meter apartment with three concrete walls can need two or even three.
Which brings us to the pitfall. More APs is not automatically better.
No. And this is where the over-deployment problem Daniel mentioned comes from. When you put two access points too close together on the same channel, they hear each other's beacons and data frames. Wi-Fi is a polite protocol — devices take turns. If two APs are on the same channel and can hear each other, they're sharing airtime. So your client device might show full bars, but the effective throughput is halved, or worse. This is co-channel interference, and it's the single biggest mistake in home deployments.
And the mesh systems that Daniel replaced — the three-unit TP-Link — that was probably part of what was happening. Three units in sixty-five square meters is a lot of radio in a small space. The single Ubiquiti AP beat them because it eliminated the self-interference.
Right. The backhaul in a mesh system — the wireless link between the units — shares the same air as the client traffic unless you've got a tri-band system with a dedicated backhaul radio. Even then, you've still got multiple APs contending for channels. Daniel's single-AP setup was cleaner. One AP, one channel, no self-interference. Now he needs two, but he's right to be cautious about how he adds the second one.
So let's answer his first question directly. Is there a fill-in unit? Something that covers a small dead zone without creating a mess of interference?
Yes. And Ubiquiti makes exactly the product line for this. It's their in-wall series. The U6 In-Wall, and the newer U7 Pro Wall that came out last year. These are not ceiling-mounted omni-directional dishes. They mount into a standard single-gang wall box — the kind you'd put a light switch or Ethernet jack in. They sit flush against the wall, they're powered by PoE, and they usually give you a couple of Ethernet ports on the bottom for wired devices in the room.
And the antenna pattern is different.
That's the key. A ceiling-mounted AP like a U6 Pro or U7 Pro has an omni-directional antenna pattern — it radiates in a roughly doughnut shape, covering the floor below it. An in-wall unit has a forward-firing, semi-directional pattern. It's designed to cover the room it's in and maybe bleed a little into adjacent rooms, but it doesn't throw signal upward or backward nearly as much. This is exactly what Daniel's asking about — it's not blasting general-direction radio everywhere. It's targeted.
So if you put a U6 In-Wall in the bedroom, it covers the bedroom. The main AP in the living room doesn't hear it as loudly because the in-wall unit's radiation pattern is mostly forward, into the room, and the concrete wall between them provides additional isolation.
Right. And because the in-wall unit has lower transmit power — it's designed for single-room coverage, not whole-floor coverage — the overlap zone is smaller. You still need to do channel planning. You can't just put them on the same channel and hope for the best. But the physics of the antenna pattern works in your favor.
Channel planning. Let's talk about that, because this is where people get lost. Two point four gigahertz has exactly three non-overlapping channels — one, six, and eleven. That's it. If you've got more than three APs within earshot of each other on two point four, someone's sharing.
And in an apartment building, your neighbors are using those channels too. Two point four is a dumpster fire in dense housing. The saving grace is five gigahertz and now six gigahertz. Five gigahertz has twenty-five non-overlapping channels in most regions, and a big chunk of those are DFS channels — Dynamic Frequency Selection — which most consumer gear avoids because it requires the AP to listen for radar before transmitting. But in a planned deployment, DFS channels are free real estate. Nobody's on them. Six gigahertz adds even more channels, and with Wi-Fi 6E and 7, you've got a clean, uncongested band that barely penetrates walls — which, in this case, is a feature, not a bug.
Because if six gigahertz doesn't go through concrete, then the AP in the bedroom on a six gigahertz channel isn't going to interfere with the AP in the living room on a different six gigahertz channel. The wall does the channel isolation for you.
The wall is the filter. And this is the counterintuitive thing about high-frequency Wi-Fi — the thing that makes it worse at coverage is the thing that makes it better in dense deployments. You can put APs closer together without them stepping on each other because the signal doesn't travel as far.
So Daniel's fill-in unit exists. It's an in-wall AP with a directional antenna pattern, lower power, and ideally on a different channel — five gigahertz DFS or six gigahertz if his gear supports it. The interference risk is manageable if you plan the channels.
And Ubiquiti's controller software helps with this. It'll do an RF scan and suggest channel assignments. It's not perfect, but it's better than guessing. The other piece is fast roaming — the 802.11r and 802.11k standards that let clients switch between APs seamlessly. If Daniel's phone is clinging to the living room AP when he's in the bedroom, it doesn't matter how good the bedroom AP is. The in-wall units support all of that.
That's the technical answer. But there's a nuance here that's worth pulling out. Not all APs are the same, and it's not just about the chipset. The antenna pattern, the transmit power, the form factor — these are genuinely different product categories.
The spectrum runs from omni-directional ceiling units — U6 Pro, U7 Pro — which are designed to cover as much area as possible from a central point, to directional in-wall units — U6 In-Wall, U7 Pro Wall — which are designed to cover one or two rooms, to outdoor and point-to-point units which are basically focused beams. Same chipsets in some cases, completely different radiation patterns. The antenna is the product, not the radio.
And the in-wall units have been around for years, but they've gotten much better. The U7 Pro Wall, which dropped in twenty twenty-five, brings Wi-Fi 7 to that form factor — six gigahertz support, four by four MIMO, and it still fits in a single-gang box. That's a serious piece of hardware in something the size of a light switch.
It also has a two and a half gigabit Ethernet uplink and four gigabit Ethernet ports on the bottom. So it's not just an AP — it's a switch. You can plug in a desktop, a printer, whatever. For a home office or a bedroom with wired devices, it's doing double duty.
Which brings us to the second question. Form factor. The ceiling-mounted disc is the default enterprise look, but it's not the only option, and for someone like Hannah — an architect who actually cares what her walls look like — it's a non-starter. What else is out there?
The in-wall units are the obvious alternative, and they're designed to be unobtrusive. The faceplate is white plastic, roughly the size of a double light switch. Ubiquiti sells decorative faceplates, and you can paint them. If you've got white walls, it basically disappears. But that's just the starting point.
The thing about in-wall APs is they don't need line-of-sight the way a ceiling unit does. A ceiling AP is designed to radiate downward across an open space. Put it in a closet and you've killed the signal. An in-wall unit is designed for short range and high density — it's meant to cover the room it's in. So you can put it behind a piece of furniture. You can put it in a closet with the door open. The coverage pattern is hemispherical — it radiates forward and to the sides, not backward into the wall.
And that hemispherical pattern has another advantage in multi-story homes. Ceiling units radiate downward, which means they bleed into the floor below. That's great if you want coverage on multiple floors from one AP. It's terrible if you're trying to isolate APs by floor to avoid interference. In-wall units don't penetrate floors nearly as well. They cover their room and maybe the adjacent room on the same level. So in a two-story house, you can put in-wall units on each floor without them stepping on each other.
So the form factor and the RF behavior are linked. The thing that makes it look better also makes it work better in certain deployments.
Right. And then there's the built-in approach. For new construction or major renovations, you can install recessed AP boxes in the wall or ceiling with flush covers that get painted over. Some AV integrators do this with custom faceplates — the AP sits behind a vented cover that matches the wall texture and paint. You literally can't see it unless you know it's there.
There's also the closet trick for renters. Take a standard ceiling AP, mount it in a linen closet or a wardrobe — not ideal for signal, but if the door is thin wood or you leave it open when you need the coverage, it works. Ugly but effective.
I've seen people put APs inside hollowed-out books. Inside air vents — though that's a fire hazard and I can't recommend it. Behind crown molding with a vented cover. The creative solutions people come up with are impressive, but they all trade off some performance. The signal has to get through whatever you're hiding it behind.
And that's the second-order insight here. The best deployment isn't about hiding the AP. It's about placing it where the physics works. A visible AP in the right spot beats a hidden AP in the wrong spot every time. Design should follow RF, not the other way around.
This is the conversation I imagine Daniel and Hannah having. Daniel wants the network to work. Hannah wants the apartment to look like people live in it, not a data center. The in-wall units are the compromise — they're visible, but they look like infrastructure, not tech. They blend in the way a light switch blends in.
And for homeowners planning a renovation, the real move is to run Cat6 to strategic wall locations during the build. Put a single-gang box in every room you might want coverage in, terminate it to a patch panel in a closet or utility room, and you can add in-wall APs later as needed. You don't have to deploy them all at once.
The cable is the hard part. Once the walls are open, running Ethernet is cheap and easy. After the drywall is up, it's a nightmare. And this loops back to Daniel's situation — he's renting, so he can't run cable, can't drill, can't do any of this. His options are basically: live with the dead zone, try to reposition the existing AP, or add a mesh unit that uses wireless backhaul and accept the performance hit.
Which is frustrating, because the problem is solvable. The hardware exists. The deployment strategy is well-understood. He just can't execute it in a rental.
The one thing he could try — and this is a long shot — is powerline Ethernet. Use the electrical wiring as a backhaul to a second AP in the bedroom. Powerline is finicky. It depends on the quality of the wiring, the distance, what else is on the circuit. In some apartments it works great, in others it's useless. But if it works, he could put an in-wall AP in the bedroom with a wired backhaul and solve the problem without drilling.
That's a gamble. But it's the only wired option he's got.
Yeah. And I'd test it with a cheap powerline kit before buying the AP. If the powerline link is garbage, you've only wasted fifty bucks instead of a few hundred.
Let's talk about where this is all heading, because the product landscape is shifting. Wi-Fi 7 is rolling out. Six gigahertz is becoming standard. The in-wall form factor is getting more attention from manufacturers — it's not just Ubiquiti anymore. TP-Link's Omada line has in-wall units. Aruba has them. The trend is toward higher density, lower power, more APs that are designed to blend in.
And the six gigahertz band is going to accelerate this. Because six gigahertz doesn't go through walls, the whole deployment model shifts from "one big AP in the middle" to "one small AP per room." The in-wall form factor is perfect for that. You put one in each room that needs coverage, they're all on different channels, and the walls provide natural isolation. It's cleaner than the old model.
The old model was born in an era when Wi-Fi was a luxury and one AP in the hallway was enough for a whole house. That era is over. We've got dozens of devices per home, streaming, gaming, video calls. The demand on the network has gone up, and the physics hasn't changed. Concrete is still concrete.
And this is where I want to push back on something. There's a narrative that mesh systems are the future and discrete APs are legacy. I don't buy it. Mesh has its place — large open spaces, outdoor coverage, situations where you can't run cable. But for a planned deployment in a home you own, wired APs with a switched backhaul will always outperform mesh. The physics of a dedicated cable versus shared wireless spectrum is not a close contest.
The mesh companies have done a brilliant job of marketing, though. "Cover your whole home with these beautiful little pods." They look like air fresheners. And people buy them because they're easy and they don't require thinking about channels or antenna patterns. But Daniel's experience is the reality — a single well-placed wired AP beat a three-unit mesh. That's not an anomaly.
It's not. And the fill-in unit concept is really just an extension of that philosophy. Wired backhaul, targeted coverage, careful channel planning. It's more work up front, but once it's set up, you don't think about it again. The mesh systems I've used require periodic reboots, firmware updates that break things, nodes that randomly decide to connect to the wrong upstream unit. It's a maintenance headache.
So to answer Daniel's questions directly — yes, fill-in units exist. They're in-wall APs with directional antennas and lower power, and they're designed for exactly the scenario he's describing. No, an AP is not just an AP — the antenna pattern and form factor matter enormously. And yes, there are discreet deployment options beyond the ceiling disc. The in-wall units are the main one, and for new construction, recessed flush-mount installations are becoming standard in high-end builds.
The one thing I'd add is that if you're buying today, get something with six gigahertz support. Even if you don't have six gigahertz clients yet, you will. And the channel isolation benefits are real. The U7 Pro Wall is the current pick for in-wall — it's Wi-Fi 7, it's got the two and a half gig uplink, and it's available now.
Daniel's stuck with what he's got until he owns a place, but the lesson for everyone else is: plan the cable runs during renovation, put single-gang boxes in key rooms, and use in-wall APs for fill-in coverage. The hardware is ready. The physics hasn't changed, but the products have caught up to the problem.
Hilbert, you've been quiet. And I know you've got history with this stuff.
Hilbert: Ran low-voltage cable for three years. Late nineties. Commercial offices mostly.
Of course you did.
Hilbert: Had a job in a law firm. Two hundred square meters, open plan with glass partitions. Client insisted on no dead zones. None. So we put in fourteen APs.
Fourteen.
Hilbert: Fourteen. Cisco Aironets. The big ones with the rubber duck antennas. Looked like a porcupine convention on the ceiling.
What happened?
Hilbert: Network was unusable for a week. Lawyers couldn't send email. IT guy flew in from Chicago, walked around with a spectrum analyzer for about four minutes, and told them to pull out nine of them. They kept five. Worked fine after that.
That's a perfect illustration of the over-deployment problem. Fourteen APs in two hundred square meters — they were all hearing each other, all contending for airtime. The clients were probably confused about which one to connect to. It's the same problem at scale.
Hilbert: The senior partner was furious. He'd signed off on the bill for fourteen units plus installation. Tried to make us eat the cost. We didn't.
What was the per-unit cost back then?
Hilbert: The Aironet twelve hundreds were about six hundred dollars each. Plus the controller. Plus our labor. The whole thing was north of fifteen grand and they ended up using five of them.
That's a ten thousand dollar lesson in RF planning.
Hilbert: The concrete walls in that building didn't help. They'd poured the interior walls solid — no conduit, no nothing. We had to surface-mount everything. Looked terrible. The interior designer on that job quit halfway through.
Speaking of interior designers.
Hilbert: My wife. Retired now. She made me rip out a ceiling AP in our living room about eight years ago. Said it clashed with the Scandinavian minimalism. Her words.
What did you replace it with?
Hilbert: Hollowed out a bookshelf. Took the back panel off, mounted the AP inside, put the books back in front of it. Signal's fine as long as nobody rearranges the Tolstoy.
You're running your home network through Russian literature.
Hilbert: The paperbacks let more signal through than the hardcovers. I tested it.
I want to know how you tested that.
Hilbert: Had a Wi-Fi analyzer app on my phone. Walked around the room with different books in front of the AP. Dostoevsky was about three dB worse than Turgenev. Thicker paper.
This is the most Hilbert thing you've ever said.
Hilbert: The point is, you can hide them. It's not ideal, but it works. My wife's happy, the network works, and nobody knows there's an access point inside the bookshelf unless I tell them.
The heat? Those things generate some warmth.
Hilbert: Left a gap at the top. Convection takes care of it. Been running eight years, no problems.
The Tolstoy access point. That's going to stick with me.
Hilbert: The fourteen-AP office, though — that's the one I still think about. The client thought more hardware meant better coverage. It's the same mistake people make with mesh pods. Just keep adding units until the problem goes away. But the problem doesn't go away, it just changes shape.
The fill-in approach Daniel's asking about is really about discipline. It's not "add an AP wherever the signal is weak." It's "add the minimum AP that solves the specific dead zone, with the right antenna pattern, on a clean channel, with a wired backhaul." That's a lot more thought than just plugging in another mesh node.
That discipline is what separates a network you don't think about from one you're constantly rebooting.
Hilbert: The lawyers learned that the hard way. Cost them ten grand and a week of downtime. Could've been worse. I did a hotel once where they put APs in every other room. Forty-eight units. Same problem. But that's a different story.
We're going to need that story at some point.
Hilbert: It's not that interesting. They pulled out half of them. It's always the same ending.
The office story is the perfect capstone to this whole discussion. Planning matters more than hardware. The fill-in unit concept works, but only if you treat it as part of a plan, not a band-aid. And the aesthetic question — hiding APs — is solvable, but you have to accept the tradeoffs. Hilbert's bookshelf works because he tested it and understood what he was giving up.
Which brings us to the forward-looking question. As Wi-Fi 7 and six gigahertz become standard, is the fill-in unit concept going to become the default deployment model? One AP per room, in-wall, wired backhaul, channels isolated by the walls themselves?
I think it will, for new construction and major renovations. The physics of six gigahertz practically demands it — the signal doesn't go through walls, so you need an AP in every room you want coverage in. The in-wall form factor is the natural fit. And as the hardware gets cheaper — the U7 Pro Wall is already under two hundred dollars — it becomes viable for normal homes, not just commercial installs.
The counter-trend is mesh systems with dedicated backhaul radios getting better. Tri-band and now quad-band mesh units with a separate six gigahertz backhaul channel can get close to wired performance in ideal conditions. But ideal conditions are rare. A cable is a cable.
Cable doesn't care about your neighbor's Wi-Fi. Cable doesn't care about radar on DFS channels. Cable doesn't care about the concrete wall between the AP and the backhaul node. It just works. And until that changes, the wired AP — in whatever form factor — is the right answer for anyone who can run the cable.
Daniel's rental situation is a reminder that not everyone can. The best network is the one you can actually deploy. Sometimes that's a visible AP in the hallway. Sometimes it's a mesh system you accept the compromises on. And sometimes it's a hollowed-out bookshelf full of nineteenth-century Russian novels.
Thanks to our producer Hilbert Flumingtop for keeping this show running and for the single best AP-hiding story I have ever heard.
This has been My Weird Prompts. If you want to send us your own networking disasters or questions, email the show at show at my weird prompts dot com.
We'll be back soon.
