Best Wifi Mesh Network Kits
thewirecutter.com
thewirecutter.com
https://en.wikipedia.org/wiki/Ubiquiti_Networks#U-Boot_confi...
And FWIW, they have recently started shipping a UniFi Mesh product called... UniFi Mesh. I haven't used it so can't really speak for it, but its marketing site is here: https://unifi-mesh.ubnt.com/
Ubiquiti works well in a low RF noise environment.
We have 1k+ unique SSIDs visible throughout our office and I was having horrific latency spikes, lost connections, etc. with top of the line Ubiquiti equipment. Trying tuning them every which way from sunday. No Dice.
Spent the money and switched to ruckus and instantly: perfect wifi. Never looked back.
Be aware with Unifi gear there is no onboard configuration or UI beyond busybox shell via SSH and a basic command line tool. They are intended to be deployed using the Unifi controller which once setup can be shut down, but will be required again to change passwords, etc.
Unifi is definitely at the top of my list for multiple AP deployments, I still prefer MikroTik for edge devices though.
> Be aware with Unifi gear there is no onboard configuration or UI beyond busybox shell via SSH and a basic command line tool. They are intended to be deployed using the Unifi controller which once setup can be shut down, but will be required again to change passwords, etc.
To clarify, this applies to the edge AP/repeaters. I know for sure every router I have purchased from Ubiquiti has had an onboard/hosted web management UI.
That said, you can install the UniFi management portal on a local machine and serve it on demand from a laptop. Or, as you (eltoozero) have done, host it remote.
And finally, you can also buy their "Cloud key" device which acts as a host for the web management portal.
This used to be the case, they now have a app for Android and iOS which will setup and manage APs even without a controller.
I believe you loose some of the functionality, like captive portals on guest networks, but 99.9% of people don't need these things.
I also love the MikroTik router with wifi I have back home though.
My house is 17th century bastide in France which was apparently built to repel canon or nuclear blasts -- my walls are a meter thick in some places and average over a half meter of pure stone. Wifi is a challenge and pulling cable is excruciating.
The software is fine by the big advantage is that it has meraki reliability without the licensing.
The AC-PRO covers my entire house (2 story, 2500+ sqft new construction) and garage with no latency issues or dead spots. There's no handoff issues like having multiple APs/extenders, there's a faster throughput than I was getting with the Nighthawk, and the PoE made it a breeze to install centrally in the house.
I'm only upset that I waited so long to just give up on the purely 'consumer grade' crap.
Also, the linked article is wrong. The iphone app is more than adequate for simple home setups (no crazy java pc installation required while you're up on a ladder / in the attic installing it). It wouldn't surprise me if they have a comparable android app. Didn't check.
Also Ubiquiti are launching a new mesh networking product this month.
If you need a router functionality, get there edgerouter but I'll tell you this, it's super powerful but not something easy to figure unless you have a deeper understanding of network protocols. Great stuff.
You don't need Java to work in the browser. The management tool runs on a Tomcat webserver. That requires Java. I don't know if you need to install Java to get that working, or if Java is included in the tool itself. But you don't need the Java plugin in your browser for the tool to work. To test this, disable or remove the plugin and see if the management tool keeps working!
The Archer C7 the article recommends is solid, too. At least when flashed to OpenWRT, stock firmware is horrible.
I guess I should have said that right away. If interference or false DFS positives aren't an issue, they're excellent radios and with the 34dbi antenna dish UBNT offer you can potentially bridge unbelievable distances.
For LOS Backhauls I'd always go for the AF24HD unless the steep price is an issue. Then again, the license for 18ghz radios alone costs more in Germany and tbh I haven't seen any interference ever with 24ghz, even if it is a free band.
Airfibers are our choice any time there is interference. They handle it better than any other radio we've tested. Have you found another vendor that can handle it better? Obviously there would be no comparison against a licensed setup.
As for the AF24, they're not an option in Florida for distances over 1 mile due to rain fade. They are great radios, but 24ghz is not a viable frequency if you need it to work when it's humid, foggy, or raining.
Probably unfair to blame it on the AF5X.
As for the weather, I would assume Southern Germany has as much in common with Florida as Switzerland with Hawaii. Snow is an issue but in LoS links it was never a big one. Only on 60/80 Gigahertz links, but those you can kill with a well placed wrapper of tinfoil anyway :)
Our DFS false positives were, since FCC reqs don't matter here, most likely due to congestion or Swiss mobile providers playing with nanocell setups or what not.
All in all I prefered Ubiquity over Mikrotik, Ruckus and Cambium. Mimosa I liked, too, but when it came down to it Ubiquity had the nicest features, the best build quality, the most accessible configuration and the best resulting links.
Not in the WISP biz anymore but I do enjoy talking shop. Did you try the Amplifi the article mentions yet? I have my Unifi setup at home and really don't think I will change it to something else, but I would love to try their first consumer product :)
As a rule of thumb, any time I see "mesh" I run far far away as it usually refers to a bunch of single-radio half-duplex access points and some sort of super "self-healing" magic. One second you get great speed, the next you have horrible speed. But most of the routers mentioned in the article are dual radio with a dedicated "backhaul" thus probably work much better than the mesh networks of the past.
On 24ghz I do have an 18km link running at 250mb/s FDX on a standard set of AF24, but the slightest atmopheric disturbance takes it out. Even my 3km link will drop in a heavy storm. I haven't even bothered with 60/80. I doubt they would make it across the street!
Though the binary firmware blob can still go to hell, it's certainly troublesome or was when I installed about a year ago.
The thread's been going for a while but the problems persist.
Although I think the target is only 2x of 2.4Ghz Wi-Fi N right now, which seems a little disappointing to me. I was expecting something along of 1KM range (I don't think Wi-Fi N actually reaches 500m in the real world).
It may be because the Wi-Fi Alliance is trying to make this an "IoT"-enabling technology as well, rather than allow us to connect to each other's smartphones and notebooks at good speeds and longer ranges.
Anandtech says there's quite a big push for 802.11ah from Wi-Fi silicon vendors.
http://www.anandtech.com/show/9915/wifi-halow-longrange-lowp...
http://www.anandtech.com/show/8253/thread-group-moots-new-ip...
The problem with waiting is: at what point do you stop waiting? Particularly when the vast majority of the time these release dates are pure speculation. If Google releases theirs next month, then word leaks that Amazon has one releasing shortly, should they then wait for that, too? And on and on ad infinitum? I'm sure that at any given moment in time there is _some_ company that is working on a competing product. I'd argue they are absolutely doing the right thing by reviewing what is available regardless of what is on the horizon, otherwise they'd be giving preference to whoever they are waiting on (Google in this case).
For my less technical friends I highly recommend them Eero system.
The results were mixed - they either work really well, or really badly. It seems that the home's wiring and appliances have an overwhelming effect on their efficacy. So, the recommendation is try before you buy if at all possible.
We recently just got a Netgear Orbi and are pretty well satisfied. Turns out the base router/AP is better overall than the other single unit devices we've been using until now and the satellite does seem to make a significant difference in the far top floor room which had problems with both WiFi & Powerline. Lucky for that, too.... our next step was to run cable in the walls.
The Orbi is basically the wireless equivalent of running wires through the walls.
You can have a mesh network without wireless clients at all if you wanted. For example, you can connect two locations together via a mesh of access points by each end plugging into the ethernet port of an AP, then the mesh provides multi-task redundancy.
Having said that, we had reached the point where I was about to call cabling guys to come in and just get it over with... which would have been some multiple of the Orbi number. We had also had various range extenders and bridges, etc. over the years and frankly I wasn't expecting much of the Orbi either.... but we've been pleasantly surprised with how well it's worked. The only complaint I have is that it does seem its configuration UI (web based) is rather sluggish at times... but I rarely use that.
At this point I think I can fairly say that for us it's been worth the price.
- always plug directly into walls, never into extensions, or multi plugs.
- don't even try to jump phases, performance will be atrocious but...
- if you have to, there are phase converters that you can install into your mains
There's a huge performance gap between cheap and expensive homeplug kits, and protocol implementation varies miserably between manufacturers. Also lowest speed adapter determines network speed.
So ideally you want to buy Devolo or TP-Link Homeplug AV1200 kits with integrated power outlets, stick them into walls directly and you can absolutely get a solid Gigabit without running cable through walls.
I do absolutely agree that troubleshooting is a horrible pain however.
fwiw, MoCA devices, particularly the new 2.0 stuff, have worked much better.
I currently am using a cheap docsis 3 modem (I believe it's a netgear), and an apple airport extreme for the main network, which is extended to the garage through a couple TP Link powerline networking adapters.
I haven't benchmarked the specific powerline performance, but I do know that if I hit up fast.com on my house wifi from my laptop, I get about 100-120 Mbps depending on the day and network conditions. (there's a lot of weird stuff with diff devices and diff routers that I'm eliding here but the Airport Extreme performs very well in my house compared to everything else I've tried)
in the garage, on ethernet to the TP Link powerline adapters, i get about 20 Mbps most of the time. It is pretty reliable, but it's a huge performance hit. Granted, I am using them in an extremely adverse arrangement: old wiring, running through lots of intervening hardware, through the ground, to my garage.
I'm actually happy with them, for my purposes, but they're really not fast.
For any large mesh, using a proper routing protocol is a good idea; the default behavior of ethernet it to construct a spanning tree and prune any links that would result in a cycle, which leads to suboptimal routing if you have more than two routers.
There's been a lot of interesting research in wireless routing. One of the more interesting ideas [1] is to use the probability of packet loss as part of the link cost metric, to discourage routers from always using the path with the fewest number of hops, which might be very unreliable.
It's been about a decade since I looked at any of this stuff, so I don't know what the state of the art is these days.
[1] https://pdos.csail.mit.edu/papers/grid:mobicom03/paper.pdf
http://www.metamesh.org <- the org http://www.metamesh.tech <- the tech & store http://www.pittmesh.net <- the network
[0] https://tools.ietf.org/wg/homenet/
[1] https://tools.ietf.org/html/rfc7788
DD-WRT/tomato/others do have mesh features but it depends on the radio chipset and other things.
You can just buy a 802.11ac repeater for $40.
Use OpenWRT Chaos Calmer, that is the latest stable released and quite good documented at openwrt.org wiki.
This is something that should work fine: https://wiki.openwrt.org/doc/howto/mesh.batman
There was a split and now most of the former OpenWRT devs build LEDE - but they don't have a release yet, however if you need support for newer routers or it's there. It's pretty close to the Linux kernel and most fixes (and also breakage) are timely there. http://www.lede-project.org
for routing: There is olsrdv1 and the new and shiny OLSRv2 that uses IPv6 link-local addresses and has a new metric for optimizing links that takes a lot of stuff into account like packet loss, throughput and so on. https://olsrd.org
There is also babel where lot's of stuff is happening - there is experimental layer3 roaming https://github.com/jech/babeld https://github.com/tcatm/l3roamd
I'm not sure if that's true but I think batman is the best solution for smaller meshes (<50?) in bigger meshes layer2 overhead is a problem.
There are a lot of active communities related to mesh networking:
There is a yearly conference where they benchmark their mesh implementations against each other. Pretty interesting stuff: http://battlemesh.org/
It's really plug'n'play yet if you do it on your own and don't use a mesh firmware i.e. https://libre-mesh.org
https://internetfreedomfestival.org/wiki/index.php/Deploy_an...
Or the software used in a lot of places for freifunk.net meshes in Germany: https://github.com/freifunk-gluon
I'm sure I've forgotten something but these are projects that care about meshes.
I am getting so tired of this. Why are engineers working for companies that build stuff like this? This is consumer hostile.
The only use case I can think of is transferring files around? Do you think that's so common amongst most users?
You know, everyday stuff, for which modern IoT equivalents somehow need internet connectivity to work.
It seems to me like "troubleshooting why the internet connection is down" is probably a capability you would want to have, no?
I am working on a similar project and is interested to know if anyone have had success with hardware addons like external antenna to build a reliable solution?
The chip capabilities are enough to run any low level to moderate load applications like monitoring, actuating and streaming. It is the hardware that I am doubtful about.
In the end I just used wired network with access points, more expensive but at least it works. I suggest that all these mesh kits should be bought only when return option is available because they may not work in some cases.
I'm amazed... "and usable" implies some of these kits will go down if you lose your internet connection. Not just that they won't be configurable?
For me I'm using AP + a few extenders these days and they worked just fine and are much cheaper.
Says so right on the product page: https://store.amplifi.com/amplifi/amplifi-119.html
It supports "extend your network" as hub spoke leaf model, with both wired-to-wireless bases (hubs) and wireless-to-wireless extenders / relays (spokes), and devices (leaves) will connect to strongest spokes or hubs as you roam.
You can have multiple wired-to-wired and wired-to-wireless bases on the same network each with its own wireless-to-wireless extenders or relays. This has worked and been solid (no reboots needed) for years.
Use the full size tower for hubs and airport expresses for spokes.
Also he says a separate high bandwidth signal coordinates the units, would that limit range? Why not a lower bandwidth signal?
This is a good post about it:
http://www.extremetech.com/computing/184685-what-is-802-11ax...
Their product is their reputation for good reviews. Destroying that is clearly against their interests.
For example, if vendors know that the site is reviewing their products, they can provide support or a higher level of service to the review and influence the rating.
Magazines like Consumer Reports go to great lengths to bake bias out of reviews, and avoid this sort of risk. But their process has its own problems, as it requires that the reviewers really deeply understand the product, or biases ratings towards customer satisfaction.
My own car (my trusty Honda Pilot with 300k miles) is an example. It featured a poorly handled transmission defect and recall, yet CR scored a very high expected reliability rating, due to historical owner satisfaction.
Caveat emptor applies as always but I agree with you the Wirecutter is a generally good review site.
I mean, that's exactly how bias works. In fact, this is exactly what amazon is cracking down on right now, incentivized reviews.
If your revenue model is based on ads, you grow revenue by growing readership which is grown by writing quality content. However if you're revenue model is based on affiliate income, you increase revenue by selling more stuff.
Well... not exactly...
They're cracking down on incentivized reviews on amazon.com product pages. Those onsite reviews being a far stretch away from offsite recommendation blog posts that feature affiliate links, which aren't included in any of their product rating/review metrics.
The review for a particular product isn't an example of pay for play because the Wirecutter gets paid for anything they review that is sold on Amazon. The bias is the push towards Amazon vs Newegg or Walmart and assessments if price based on ephemeral Amazon pricing vs MSRP.
The Amazon crackdown as I understand is about corrupt vendors incentivizing good reviews for shit product.
Is there any evidence to suggest Wirecutter has refused to rate an item due to lack of affiliate links?
If a product isn't on Amazon where affiliate links can point at it, then is it even mainstream enough for The Wirecutter's intended audience?
With that in mind I like the affiliate link model because arguably they can choose any product (in their case the 'best' one) instead of cutting deals with the companies directly (i.e. introducing competition in who pays the most).
They're probably the least biased reviewers out there at the moment and take it very seriously.
Maybe I wont get the best USB power adapter for my car, but I'll be able to choose from a small menu of them, and will probably get something good. And I can skip reading through hundreds of amazon reviews, many of which are totally irrelevant.
They've earned a great reputation in spite of them.
One strategy would be to discount them somehow. If they give something a great review, don't trust it entirely, but remember it if you also see 5 other good reviews.