Home Mesh Wi-Fi Coming This Summer From eero

[brossyg]

Can someone please explain why wireless "mesh" in a large house would be better than simply adding an AP with the same SSID/Password as the wireless router?

I do this and it works perfectly, is simple and inexpensive. I placed the AP where the router's wireless signal was the weakest. The mobile device switches seamlessly between the AP and the router.

Thanks.

 

[Nullity]

With Wifi's TDLS, we already kinda have the advantages of both setups. TDLS allows 2 clients to form a direct (Wifi Direct/p2p?) connection in an AP setup.

 

[thiggins]

If they pull it off, eero will be better because it is simpler to set up and use. This system is aimed at people who just want to plug something in and have it work with minimum hassle. APs also normally require Ethernet to connect them together.

You are lucky your client switches between the two APs with no problem. Not all devices do.

Nullity: TDLS isn't mesh. It's more like a modern version of Ad-Hoc networks.

 

[Nullity]

If they pull it off, eero will be better because it is simpler to set up and use. This system is aimed at people who just want to plug something in and have it work with minimum hassle. APs also normally require Ethernet to connect them together.

You are lucky your client switches between the two APs with no problem. Not all devices do.

Nullity: TDLS isn't mesh. It's more like a modern version of Ad-Hoc networks.

Meh. Same difference.


Are you aware of what niche that Mesh is planned to fill? I see lots of research in academia regarding Mesh technology, but I haven't paid enough attention to see if they have a wireless revolution in the making. Perhaps getting internet to 3rd-world countries or something similar?

 

[digitalfreak]

I wonder if you can use one of the ethernet ports instead of wireless for your uplink.

 

[mlg321]

Can additional Eero's be used as access point mode via Ethernet or are additional Eero's strictly mesh capable?

How does this differ to open-mesh or a AirPort Express?

 

[thiggins]

Are you aware of what niche that Mesh is planned to fill? I see lots of research in academia regarding Mesh technology, but I haven't paid enough attention to see if they have a wireless revolution in the making. Perhaps getting internet to 3rd-world countries or something similar?

No. BIG difference.

Ask the average person if they know how to add an access point to their home network and they'll probably ask "what's an access point".

eero is aimed at the mainstream consumer. BIG market that thinks buying one bigger and more expensive router will solve all their Wi-Fi problems.

 

[azazel1024]

No. BIG difference.

Ask the average person if they know how to add an access point to their home network and they'll probably ask "what's an access point".

eero is aimed at the mainstream consumer. BIG market that thinks buying one bigger and more expensive router will solve all their Wi-Fi problems.

Exactly. Aimed at the lowest common denominator with wifi coverage issues.

That happens to be a lot of the market.

How it is being accomplished is no different than a lot of the newest dual band wifi range extenders which have the option of doing a backhaul on one band and extending the other for no loss in performance. The difference is it seems like the Eroo can extend both and dynamically switch which radio is being used to connect to the client and which is doing the back haul, which could be somewhat advantageous.

That and easier setup. Otherwise, the differences are minimal.

It'll still be a lot faster and likely no different experience to just hardwire in access points instead of this mesh networking. As it is I see no mention of any kind of AP directed hand-offs, seamless roaming or anything else like that. So it is really a fancy somewhat better way to connect range extenders.

 

[stevech]

I work with mesh in non-WiFi worlds. Used to engineer mesh for WiFi backhaul.
For consumers, it brings the hope of plug-and-go coverage improvement. Essentially it's self- and auto-installing WDS repeaters that can daisy-chain. Each hop costs half throughput, BUT, if the WiFi link speeds are much faster than the ISP speed, no one will complain, except for "intranet" users doing big file transfers (which shouldn't be on WiFi).

The issue with most meshes is thrashing of automated-route choices when there are 2+ hops.

The Bigger issue is that until the IEEE standard is proven, vendors will jocky for a proprietary mesh solution to force consumers to use all same brand router and mesh routing AP nodes.

This has all been done for years - I think Tropos has/had the best, but proprietary as were Cisco's and Firetide and many others'.
There is/was an open-source mesh for WiFi, a grass roots movement. Not sure it proliferated much.

Most use two radios - one for 802.11 access on 2.4 or 5.8GHz and a second radio that forms a mesh among nearby nodes. Like the infamous attempts by Earthink and others in community WiFi on streetlights and power poles. Failed due to business model, not technology - except for the building-penetration problem yielding speeds lower than people want. But "someone in government" (er, all of us) was going to pay for recurring services, so they solve the so-called Digital Divide.

 

[azazel1024]

Expensive, but one of the best ways to do mesh (I haven't seen it done, but it MIGHT be) is to you multiple radios. I don't simply mean cover in 2.4GHz and backhaul on 5GHz or something, I mean multiple radios.

So you can, say, cover on 2.4GHz channel 1, 20MHz, but then have dual backhaul on 5GHz, a radio on channel 36 and one on channel 141 for example. So that way you can both send packets down the line on one channel/radio and accept packets from further up the mesh on the other channel/radio. That way, other than latency, you have no penalty no matter how large you make the mesh.

I've seen DIY solutions like this for wifi long link where you don't have line of sight. Setup bridge in the middle connecting upstream, wired to a bridge going downstream and having them operate on seperate channels so they don't interfer at all.

It would be interesting to see this all in one product (hey, look, XStream already has one 2.4GHz radio and two 5GHz radios...)

Gets pricey, but it would be "ideal".

For maximum expense and wow, have more than just 3 radios, so you can backhaul using two on seperate channels and then extend both 2.4GHz and 5GHz on other channels as well. Sucks spectrum, but it would be cool.

 

[Razor512]

I wish they would do an AP that has 4 radios. Using one band for clients and the other band as a backhaul, will just create bottlenecks.

The pricing seems really steep, and they will need to release a lot more info about the product to even begin having the slightest shred of hope that people will buy the product.

I would also like to know how the management is done, is it all local, or is it relying on any remote servers (e.g., like skydog where they shut the service down, leaving people with useless overpriced devices).

 

[azazel1024]

That would be nice to known on the management.

I am getting tired of "in the cloud". These devices should be quite capable of hosting management locally. Not everyone needs or wants to manage their stuff "over the internet". I don't want my thermostat available over the internet, I'd just like to be able to tinker with it on my own wifi (and if I really need to, I'll port foward so I can access it over the internet).

As for the price, $199 for 3 sounds pretty reasonable, though the non-kickstarter price a wee bit dearer.

 

[L&LD]

The cheapest solution right now for a nominally smaller area is 3x RT-AC56U's with one used as the backhaul on a different 5GHz channel and the other two used on both bands for connection to Clients.

For less than $300 on sale, this offers at a higher reliability what these multi-radio routers only promise, as of now. But the hit to neighboring wireless networks will be huge. As almost all the channels will be used in any given vicinity by this combination.

Either way, multi-radio routers are not my first choice in any situation.

 

[stevech]

two radio APs are good. Get backhaul in a different band. But with the popularity of 5.8GHz increasing, that's a problem.

Reality: Trying to have two same-band radios in the doesn't work. When one is transmitting in band 1 on channel, the other same band radio on some other channel will work poorly due to self-interference. The radios with antennas really close don't have any selectivity (bandpass filters) in that same band.

So radio A will "jam" radio B if they're on the same band but quite separated in frequency, due to the close proximity.

 

[System Error Message]

That would be nice to known on the management.

I am getting tired of "in the cloud". These devices should be quite capable of hosting management locally. Not everyone needs or wants to manage their stuff "over the internet". I don't want my thermostat available over the internet, I'd just like to be able to tinker with it on my own wifi (and if I really need to, I'll port foward so I can access it over the internet).

As for the price, $199 for 3 sounds pretty reasonable, though the non-kickstarter price a wee bit dearer.

Same here, if i need to access anything i do it over VPN and have my router be the VPN server. The only good thing for cloud would be management such as OS configs and such if you have a lot of systems. With VPN i get to use local only services like aerofs remotely just like as if it was dropbox with the space limit of your hard drive.

Mikrotik has had mesh for quite a long time for both wired and wireless but i never tried it nor got it to work for wire. Its the same with openflow, whenever i enable it the router just stops routing for a while till it detects that ive disconnected.

Still if there was something better than WDS and smartconnect to evenly distribute wireless bandwidth that would be a lot better.

 

[eas]

So, I'd guess that the next update to the airport express will included 802.11ac, and it probably won't come out too long after this.

So, I wonder, how will this compare to getting a few Airport Expresses for $100/each and doing an extended network using them.

 

[stevech]

So, I wonder, how will this compare to getting a few Airport Expresses for $100/each and doing an extended network using them.

Do the airport expresses have an access point mode with wireless backhaul to the router in 1 hop? 2 hops?

Airport Express is not in the comps.

 

[azazel1024]

Do the airport expresses have an access point mode with wireless backhaul to the router in 1 hop? 2 hops?

Airport Express is not in the comps.

Yeah, it isn't really comparable. That said, I wouldn't mind if Apple went back to a wall plug format and was AC1200 on them with dual ports for $100.

It would make an awesome access point so long as they didn't muck anything up.

Its one of the things I am looking at doing, for a variety of reasons, through the renovation of my house. I am looking at deploying several small access points running 5GHz only to allow better client bandwidth utilization. Something like the airport express would be rather nice for this, also especially because I am looking to leverage airplay and a distrbuted speaker setup through my house, so 2-4 expresses with only the 5GHz radio active and running airplay to support the speakers in a room would be real spiffy. Then just leverage maybe 2 wireless basestations for 2.4GHz coverage.

 

[sfx2000]

It's not MESH, it's not TDLS, it's not WDS... (or Apple's bastard approach with 'extended networks'), nor is it even WiFi Direct.

Think TDM scheduled links - A lot of this goes back to TDD-LTE technologies... there's a lot of Qualcomm stuff in there from Node to Node. They say as much as the tech is based on Qualcomm..

I'm not an insider here - but a lot of work went into putting Wimax/802.16m into unlicensed space just for this purpose, and I wouldn't be surprised if an TDD-LTE implementation just re-used it.

A scheduled MAC can be very, very, very fast... much faster than the CSMA/CA MAC implementation in 802.11... and LTE/16m have much more granularity with regards to applications flows and Quality of Service.

This is how I would have done it....

sfx

 

[remixedcat]

so its backend is LTE? that's wierd but cool