The house seems very resistant to letting WiFi signals out, especially at the back.
I was wondering about an external aerial or perhaps a complete Access Point to take the WiFi out into the back garden.
My first search was confusing; some very cheap options and some very expensive options.
I assume that I want to have 2.4 GHz and 5 GHz and either a complete unit with Ethernet connection (power over Ethernet?) or a router inside the house with just the aerial outside.
Cabling and connectors at 2.4 and 5GHz tend to have a bit of loss, so once it's got to the antenna you've lost some of the signal and so part of the point of getting the signal there. Additionally modern access points use phased antennas for better reception, which you can't do with one or few randomly placed wired antennas.
I'd suggest going for an external unit with an ethernet connection: either externally mounted or mounted just internally (eg maybe the roof lets signal out where the walls don't). Most units designed for outdoor or ceiling mounting use PoE so you just feed it one cable for both data and power.
You can also use mesh, if you can put the unit somewhere with power that has a whiff of the indoor signal but also a route for its signal to get out.
802.11s is a standard for exactly that. I've tried it on a collection of random routers running OpenWRT. Support in chipsets is a bit mixed so I never got it to work reliably (the wifi chip in the BT Homehub 5a wouldn't play ball, which are most of my pile of scrap routers). But would be doable if the hardware was designed to support it.
I did that, with a router in the roof to ensure a reliable signal from the camera in the wall lamp by the front door. Haven't tested the range for other purposes.
I went with multiple wired WiFi access points networked together with Ethernet cable, with fast roaming.
formatting link
I was confused, thinking I wanted mesh. What I really wanted was to be able to walk about my house and be connected to the right access point, without dropouts. That is called fast roaming.
This Mercusys, kit is nasty, in that admin is done via a cloud service, presumably in China. But... it works reliably. Probably if you pay a bit more you could get something with local admin.
Do you have a room with a window, facing the garden ? That would be a great place to do a test.
You probably own materials that could be used to check how opaque the house is to RF.
1) Cell phone signal. A different frequency than Wifi. Does that penetrate the house ?
2) Take a bluetooth speaker or battery headphone outdoors (2.4GHz). Can the computer in the house, pair with that, or send music to it ?
3) Does your DECT phone work in the back garden, with the DECT base unit sitting in the house ?
4) Use the USB Wifi on your desktop. In Windows 11 there is a Hotspot button, which turns the desktop Wifi into Wifi Direct (an Access Point with its own SSID). The hardest part of using that, is finding the dialog that has the randomly generated password string, so you can enter the password out in the garden, and get up to 54Mbit/sec transfer rate. It's a pitiful rate, but this is just a test of whether a Wifi signal can make it outdoors.
2.4GHz is better at penetrating house walls, than 5GHz is.
5GHz was preferred at one time, because there was relatively less usage of the band, so the channel wasn't as crowded.
If your DECT phone is on 900MHz, that should be really good at getting through a wall. The DECT frequency used in the design, can vary with country.
The router signal is a three dimensional signal. At one time, if your router lived on the house main level, other main level devices worked. The computer in the basement might not see a signal. The upstairs bedroom may have poor reception. (The single antenna on the router back then, sent signal mostly on the same floor it was on.)
On a router with more than one antenna and the antenna is hinged, you can try altering the angle of the antenna, to provide more signal to other parts of the house.
The router signal also has channel numbers. Like Ch 1, 6, 13. You can try changing the channel, if there is too much interference from a neighbours choice for their kit.
If your house walls were filled with foil backed insulation boards, that might be sufficient to cause a problem. Steel door frames. Aluminium window trim. Those could contribute to a poor result.
The router might have a 2dBi isotropic antenna on it. This is intended to spread signal in all directions in a house. The transmitter power (in watts) is limited, by the manufacturers knowledge that only a 2dBi antenna is on the output. If instead, you find a 30dBi dish antenna, and point it at the back garden, that violates the EIRP of the resulting "system", by 28dB. But, that's also a test case. It's not particularly a good investment, an illegal antenna, but it's one way to satisfy your curiosity about whether another device will ever see your SSID. As you were planning to do, I would set up an access point or a mesh routing scheme, before using a dish as my chosen solution. We don't know how these work bidirectionally, and whether one direction works as well as the opposite direction.
# Parabolic dish antenna
formatting link
It could be, that the router has degraded with time, and the signal isn't actually as strong as it was when the router was new. You might be able to use your recollection of performance, to judge whether that has happened. Routers can use adaptive transmit power, so even if the claim is its 2 watts, if it detects good signal level, it can turn down the transmit so it's not as hard on any cheap CMOS output.
And we get scattered reports, about "people with weird scan results". In Linux, someone was complaining that half their Wifi SSIDs had gone missing, as if there was some kind of pattern based on the SSID name or something. If there are sun spots or other natural phenomenon, that can temporarily affect radio.
A lot of interference cases, are temporary, and those would be worth listing, if your Wifi "worked sometimes and not others". But your symptom report is, it never works.
TPLink lists TL-WA801N, TL-WA1201, TL-WA901N, TL-WA3001 in increasing price as Access Points. The cheapest is 1/4 the price of the most expensive. The second from the left might be adequate. These are not particularly designed to sit in the rain, they're not outdoor units, but if you had a window facing the garden, might be worth a try. You would want to inspect for evidence of good firmware, good lifespan (no reports of reduced signal after 3 months), rather than just blankly staring at the standard "AC" and assuming you'll get lightning speed and that's it. When there was a review site available for this stuff, and a guy who cared about benching and measurements, the whizzy stuff never seemed to pass about 100MB/sec (and that's on equipment where the WAN side was fast enough to do better). I would probably select "any old thing that wasn't 802.11n".
For example, now that Wifi 7 is out, I'm not "hungry" to buy new kit, just because the stupid thing has 4096 QAM or something. Abusing QAM like that, eventually has to give a diminishing return in the real world, with real world impediments. Putting AI inside the box won't help. Etc :-) And without trustworthy reviews, it's not possible to even remotely guess whether a whizzy thing is worth "spit". One guy had a real test lab, with real lab equipment, anechoic capability, attenuator matting for multipath reduction and so on. But I suspect he's tired of doing test, and is in the back garden in the sun now. If no one can draw us a graph, using a "standard house test", it's pretty hard to make an equipment selection.
One part of the selection process, is downloading the manual to see if the device has a web interface that is decent.
You would likely pay more, for a device packaged for a life in the rain.
More information - I have a TP-Link WDR3600 in the rear living area, about
5 metres from the bifolds.
A wifi camera just outside doesn't get much signal.
{SFX someone wandering around with a phone with a wifi analyser and getting distracted by various things}
I also have a cheapo Netgear WAC104 in the back bedroom fairly near to the window to try and provide a better signal that from the Buffalo router in the office or the above mentioned downstairs router.
[Another diversion - couldn't understand why there was no signal from the downstairs router. Turns out it hadn't been switched back on after our holiday away.]
Also I apparently have WiFi scan throttling enabled on my phone. No idea at the moment how to turn it off.
I am not clear on how/why signal strength seems to vary at one spot about
2 metres from the downstairs router
All I can see so far is that the bifolds significantly reduce the signal when I go just outside.
The rear of the house including internal walls does contain a lot of solid insulation with foil covering.
I think an external AP would probably give a wider scope for WiFi devices in the back garden.
One day I might even get Ethernet to the Mother Of All Sheds.
Depends. 5GHz has less penetration through walls etc, which means the 5GHz spectrum in your house is quieter and less affected by your neighbours. If you are in a populous area there can be so much 2.4GHz mush that sometimes it's hard for devices to get the packets through.
Many IoT devices are based on cheap silicon that only has a 2.4GHz radio and a single antenna, which can be more affected by this than phones and laptops with multiple 2.4 and 5GHz antennas.
I have a router/access point in the loft, second one on ground floor (2.4/5Ghz). I fitted a 2.4/5Ghz extender, outdoor, in my hut, to improve garden access. It works quite well.
Mounting the router indoors, connected to an antenna outdoors, there will be high losses via the coax - so you would need to consider mounting the router very close to the antenna..
There really shouldn't be. Pretty much all RF coax is 50 Ohms, and most small commercial antennas are nominally 50 Ohms.
As long as the antenna is a 50 Ohm type, losses should be small. Something significantly different (e.g. home-made dipole, folded dipole) will need a matching balun transformer to talk to 50 Ohms unbalanced, and a commercial antenna will typically be supplied with one for use with 50 Ohm cable.
Other coax impedances are rare, apart from 75 Ohms for video, which will also work for dipoles, but most RF devices will not be expecting that. Balanced twin feeders will be used for 300 Ohms, it's difficult to make coax of that order of impedance.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.