This makes the antenna more directional. It increases the range in one direction while decreasing it in every other direction, so I'm not sure how well this would work for most people without relocating the router/AP.
That being said, it's a clever bit of engineering.
Given that it's only somewhat directional, this would actually be useful for the common case that the AP sits in a corner of a one-story house. No sense sending all that RF energy outside, so direct it inside.
When you upgrade from consumer-grade equipment into dedicated APs, they start looking more like something you can put into the center of your living room without being ashamed of it.
The Unifi APs for example look like oversized smoke detectors. They have obnoxious blue LEDs like seemingly all modern equipment, but you can turn them off and then it emits no light at all, and fits quite well into the living room when mounted on the ceiling.
Mikrotik has been moving in this direction as well. This sucker [1], while unfortunately 2.4 GHz only, is an inconspicuous 3.5" white circle you can stick on your ceiling.
My current (rental) place has the router and AP in one corner of the apartment, so signal strength at the other end is terrible. Is there a better (cheap but effective) option than this?
As with the original link, the antenna is directional (and since it's a reflector, probably even more directional), so you lose out on flexibility. It'll be quite good if you don't plan on moving the other point, though.
Better is to get a WDS bridge (wireless repeater). A high gain antenna (higher than in article) won't work well in the "long house" scenario because two clients may not be able to hear each other and may thus collide. (RTS/CTS is supposed to fix this but it relies on clients choosing to use it.) Unfortunately with WDS, the bridge must be the same brand as the AP (which itself must support WDS), so you might be stuck purchasing two new pieces of hardware.
But if the clients are all near each other, you have some money to spend (~$100), and don't mind being stuck with one band (2.4 GHz or 5 GHz), I'd recommend one of these:
Good intuitions. FCC regulations restrict EIRP[0], which is essentially "what is the signal strength in the strongest direction". It's possible that increasing directionality of your antenna puts you over the EIRP limit (given how routers already try to get _really_ close to it to improve performance).
Basically APs on 2.4 GHz are only allowed 36 dBm (4 W) EIRP, no matter what antenna shape. This is usually not a problem though, since stations tend to have weaker transmitters anyway, but will still benefit from the AP's enhanced antenna gain.
The reason this is important is that it prevents private corporations from interfering with public air space (as discussed previously on HN [1]), and prevents landlords from interfering with the rights of tenants to legally place outdoor antennas (such as satellite dishes).
RF energy doesn't quite work like this. If your transmitter has a high EIRP (either due to antenna directionality or high TX power), then its signal is going to travel a long way, even after the signal level has dropped to that of a lower-EIRP transmitter. That is simply the inverse square law at work.
You could place sufficiently absorptive materials that this isn't so (e.g. enclose your house in a Faraday cage). But as hydrogen18 points out you're still limited by the FCC.
I think FCC laws technically apply within your property as well. Hypothetically, what happens if your non-compliant device blocks a visitor on your property from calling 911?
The FCC has stated that the RP-SMA connector (the connector used on nearly all wifi routers that have detachable antennas) shouldn't be allowed on part 15 devices, but also that it will delay enforcement of this indefinitely[1]. So even the FCC is fairly "meh" on enforcing this unless actual interference is reported.
Citation? Plenty of commercial part 15 devices don't even come with antennas, requiring you to "mix and match", unless I misunderstood what you meant by that.
4 W (36 dBm), and the 8.5 dBi gain of this homemade Yagi puts you over that if you're operating at 1 W (30 dBm) max transmit power (as some home routers do).
FCC regulates EIRP, which is output+antenna gain. For this to be strictly legal you would need to adjust the power down until the EIRP is within the legal limit. Actually to be strictly legal I think it would need to be type excepted by the FCC.
That being said if you're using this in your home/office on a legal channel it would be really difficult to even tell from outside you were using it and almost certainly not worth the FCCs time to even try to watch for it. It's not that much extra gain over the stock antenna.
To be fair, the FCC regulates both EIRP (4 W / 36 dBm) and max transmit power (1 W / 30 dBm). E.g. you can't just have a 34 dBm transmitter hooked up to a 2 dBi dipole.
If it's like here in Australia, they do consider EIRP. So a directional antenna would increase the EIRP, but with just 100mw tx at the output, I doubt EIRP would be over the limit (4W for 2.4Ghz ISM in Australia).
No, because no-one wants a home AP that only works if it's pointed at you.
You can buy directional APs if you want, but they're usually used for outdoor point-to-point links, or as sector antennas. e.g. I just installed one of these today: https://routerboard.com/RBSXTG-2HnD in a place that needed a signal far from where the AP was. Usually patch antennas (like this product) are preferred over Yagis for microwave links due to their smaller footprint, and capability of dual polarization in a single package.
Better is if your indoor AP can transceive in more of a "pancake" shape, like this: https://routerboard.com/RBOmniTikU-5HnD Note that's almost the same gain as this home-grown Yagi (7.5 dBi vs. 8.5 dBi) but with a more useful directionality.
Antenna gain is a function of directionality and vice versa. Patch antennas are also good because they have often very low "backward gain" but at the same time they are not very directional forward, so they are good to be placed in a corner of a room or near a wall.
It's not illegal per se, if you adjust transmission power accordingly. Router manufacturers generally don't use yagi because it's directional, requiring aiming, unlike omnidirectional dipole antenna. Modern wifi stations have much more sophisticated ways to increase range than passive antenna designs.
It's only illegal if you get net gain out of it[1]; if you turn down the power so that the maximum power transmitted in any direction is the same as before, then it's legal. You can get a small performance improvement by using directional antennas in this way as you reduce the received strength of off-axis signals which can reduce both multipath interference as well as interference from other stations.
It's not all that useful indoors though as 2.4 and 5GHz signals bounce off of walls really well so you lose much the directionality at the very first wall it hits.
1: Actually there is a total ERP requirement, but most routers with external antennas are already close to that requirement, so more than about 3dB gain will often put you over.
So this makes the router sends a stronger signal, farther. But how does this help if my laptop still has the same wifi adapter and can't send a strong signal back?
It also makes the antenna more sensitive in the direction its pointing. The elements make the antenna more directional, sort of focusing the energy. This increases the transmit side by concentrating the energy into a smaller area, but the same effect also helps it collect more energy.
The trade-off is worse performance perpendicular to the extra elements.
Antennas are coupled. Transmitting is receiving, and receiving is transmitting.
When you pull on a chain or a rope, it doesn't matter if the other side is "receiving" the pull, or "sending" it. The two sides are physically coupled.
Similarly, antennas are physically coupled to each other. Increase the gain on one side, the other side still benefits. You're physically connecting a "rope" of electrons over the air.
Not really. It's only really applicable to any router that uses a dipole antenna.
Pragmatically, this means it's only really an option on routers with external whip antennas. Lots of routers have internal antennas, at which point you're going to have problems attaching the thing. Also, if the internal antennas aren't dipoles, it's not an option at all without significant redesign.
You might start at the ARRL's Antennas page [0] and consider the ARRL Antenna Book [1]. Most of that is aimed at amateur radio in particular but the theory is the same.
There's tons of free and/or open-source software (of varying quality) for modeling antennas and such as well.
Take a microwave design class. This would be a second or third course after electrostatic, electrodynamics, then something like microwave theory. At my school we had to petition for the course as a special section. It was great, we built cavity resonators, and final project was an antenna design. All the way from theory, then using modern fea antenna software, and then building it and testing performance. If you have any interest take it. I was not going to but the professor said: "Phil, what's the lower end of the microwave spectrum?", I said "1 GHz", he then asked "How fast is your computer?", I replied "2.8 GHz", He said "I'll see you next semester then?". these days having an understanding of microwave theory will help you no matter your specialization.
1. Find a university course on RF, look up the course textbooks.
In my university we had several courses on "analog electronics" which covered topics like free space path loss and RF concepts (e.g. bandpass filters).
2. Use Google or another search engine (DDG) to find the book in PDF or EPUB version.
3. Enjoy reading the book without forking over hundreds of dollars to a greedy publisher.
if my router has 3 antennas, is there any point to installing it only on one of them? my intuition that's based on exactly no RF experience whatsoever tells me that it'll do more harm than good. (case is weak 5GHz signal in the corner of a house.)
Depends. Some routers have multiple antennas because there are multiple transceivers inside it, and some technologies (i.e. MU-MIMO in 802.11AC) will most likely react strangely having one of their antenna modified. In the end, it would probably make your performance even worse.
theres a two reasons your router probably has 3 antennas:
-beamforming
-split 2.4ghz/5ghz/5ghz-wave2 antennas
if your router is using those three antennas for beamforming, you might trigger some strange behavior if you yagi a single one of them.
if run into issues and can turn beamforming off, you should be able to yagi a single one no problem.
if your router pumps each antenna into a separate radio, then you can reliably pump a single radio into that corner of your house w/ a yagi. however, you won't have seamless roaming between the 2.4ghz and 5ghz networks when you move into and out of that corner of your home.
i recommend getting a mesh network, they're the bees knees and are a huge step above the traditional access point + range extenders model
No, Yagis have a single driven element like a dipole, and you have longer elements behind the driven element (reflectors), and shorter elements in front (directors I think?).
interestingly, the antenna optimizer (AO) software uses genetic algorithm to find the right the antenna topology, therefore, this antenna is not designed, it's evolved. more details can be found here: https://en.wikipedia.org/wiki/Evolved_antenna
That being said, it's a clever bit of engineering.